If asked to de fine the scope of health eco nom ics, many phy -

Transcription

1 THE CANADIAN JOURNAL OF PSYCHIATRY Volume 46 Ottawa, Canada, August 2001 Number 6 EDITORIAL The Contributions of Health Economics to Mental Health Policy If asked to de fine the scope of health eco nom ics, many phy - si cians would stop af ter men tion ing the eco nomic evalua - tions of health care (that is, cost, cost-ef fec tive ness, and cost-util ity stud ies). In fact, health eco nom ics is a con sid er - ably broader field that draws its the o ret i cal in spi ra tion from other ar eas of eco nom ics. These ar eas in clude in dus trial orga - ni za tion, fi nance and in sur ance, la bour eco nom ics, and pub - lic fi nance. None the less, many econ o mists are en gaged in pol icy-ori ented re search and pub lish pol icy-rel e vant ar ti cles in jour nals that are read by health care pro vid ers and pol icy mak ers. The 2 re view ar ti cles pre sented in this is sue are fine ex am ples of this tra di tion. They ex em plify how eco nomic anal y sis can in form cur rent is sues in men tal health policy. In the first ar ti cle, Dr Car o lyn Dewa and co au thors re view the eco nomic lit er a ture on phy si cians re sponses to fi nan cial in - cen tives (re im burse ment meth ods) and con sider how differ - ent pay ment meth ods ei ther pro mote or fail to promote shared care (the sharing of care and the co or di na tion of care be tween pri mary care phy si cians and men tal health spe cial ists). As a ser vice de liv ery model, shared care arose from the growing rec og ni tion of the ben e fits of col lab o ra tive teams, as well as from the pressing need to ex pand mental health specialists reach to a broader pa tient population. Many prov inces in Canada are con sid er ing re vi sions to physi - cian pay ment meth ods and pro mot ing shared care is only one of sev eral ob jec tives. The rel e vant is sues here are inherently com pli cated and con tro ver sial. In the area of phy si cian re im - burse ment, econ o mists con tri bu tions are a part of a sub stan - tial lit er a ture that seeks to ex plain the vari a tion in pay ment mech a nisms across many oc cu pa tions and in dus tries (1,2). Re cent work in the US sug gests that payment method has a con sid er able im pact on how phy si cians de liver mental health ser vices (3). Dr Dewa and co au thors dis till the sub stan tial the o ret i cal work (and the smaller body of em pir i cal lit er a ture) sur round ing 3 com mon methods of phy si cian re im burse - ment. These are fee-for-ser vice (FFS), pro spec tive pay ment (such as cap i ta tion), and blended meth ods that com bine FFS and pro spec tive pay ment. Their find ings are so ber ing: with - out mod i fi ca tion, nei ther FFS nor pro spec tive payment rep re - sents an at trac tive model to pro mote shared care. The au thors de clare an over all pref er ence for blended pay ment sys tems but are well aware of the chal lenges fac ing any major re vi sion to phy si cian pay ment in the cur rent Ca na dian po lit i cal en vi ron ment. In the sec ond ar ti cle, Dr Eric Latimer pro vides an ex cel lent syn the sis of the ev i dence that ad dresses the eco nomic im - pacts of sup ported em ploy ment pro grams for those with se - vere men tal ill ness. His ar ti cle fea tures a par tic u larly lucid dis cus sion of the pol icy im pli ca tions of his find ings from var - i ous per spec tives. Many read ers of the Jour nal who pro vide care for this pop u la tion are all too fa mil iar with pa tients frus - trated de sires to work in com pet i tive labour set tings. Sup - ported em ploy ment pro grams di rectly place pa tients into com pet i tive em ploy ment and then of fer on go ing and time-un lim ited sup ports to keep them in those po si tions. Studies con sis tently dem on strate that sup ported em ploy ment pro grams lead to higher lev els of com pet i tive em ploy ment, com pared with var i ous al ter na tive pro grams with more in cre - men tal ap proaches. As Dr Latimer notes, the question then be comes how to as sess the costs of sup ported em ploy ment pro grams in re la tion to the ben e fits. Based upon the avail able ev i dence, sup ported em ploy ment costs are mod est. Fur ther, con ver sion of day-treat ment or less ef fec tive vo ca - tional-treat ment pro grams into sup ported em ploy ment pro - grams may be at worst cost-neu tral, and pos si bly cost-sav ing. Does this then sug gest that pro vin cial health sys tems across Can ada should widely im ple ment supported em ploy ment pro grams? Two fac tors sug gest cau tion. First, the rates of com pet i tive em ploy ment achieved by sup ported em ploy ment are modest (around 30%). Those suc cess fully em ployed rep - re sent a self-se lected and likely less dis abled subpopulation of those with se vere men tal ill ness. Pol icy-mak ers must de - cide whether to in vest scarce re sources in this subpopulation or whether they are better al lo cated to in ter ven tions that ben e - fit a wider segment of those with se vere men tal ill ness. Sec - ond, as Dr Lati mer ar gues, the sub stan tial dif fer ences in so cial se cu rity and health care be tween the US and Can ada make ex trap o la tion of the current ev i dence all based in the US ten ta tive at best. US en rollees in sup ported em ploy - ment might ex pect at least mod er ate in creases in in come and Can J Psychiatry, Vol 46, August

2 August 2001 The Contributions of Health Economics to Mental Health Policy 487 ben e fits from com pet i tive em ploy ment. Be cause of uni ver sal health in sur ance, drug ben e fit pro grams, and the high claw - back rates of pro vin cial in come sup port pro grams, how ever, most Ca na dian en rollees would re al ize little ad di tional in - come or other ben e fits from com pet i tive work. Out of ne ces sity, the re views of Dr Latimer and Dr Dewa and col leagues rely al most ex clu sively on US-based re search. Ca - na dian pol i cies for mental health care and other ar eas of health care are better made on the ba sis of solid em pir i cal ev i dence gath ered in this coun try. The re view ar ti cles pre - sented in this is sue of the Jour nal poi gnantly il lus trate the need for greater ex per i men ta tion and eval u a tion within our own health sys tem. References 1. Baker GP, Jensen MC, Murphy KJ. Com pen sa tion and in cen tives: prac tice vs the - ory. Jour nal of Fi nance 1988;43: Pratt JW, Zeckhauser RJ. Prin ci pals and agents: the structure of busi ness. Boston (MA): Har vard Busi ness School Press; Rosenthal MB. Risk sharing and the sup ply of men tal health ser vices. Jour nal of Health Eco nom ics 2000;19: William H Gnam, MSc, MD, FRCPC Guest Editor

3 IN REVIEW Using Financial Incentives to Promote Shared Mental Health Care Car o lyn S Dewa, MPH, PhD 1, Jeffrey S Hoch, PhD 2, Paula Goering, RN, PhD 3 Ob jec tives: To con sider the most com mon pri mary care re im burse ment struc tures, to iden tify in cen tives in her ent in each, and to dis cuss how each could be used to en cour age a shared-care ap proach to treat ing men tal dis or ders at the pri - mary care level. Method: Three ma jor fi nan cial re im burse ment mod els fee-for-ser vice, cap i ta tion, and blended pay ment mech a - nisms are ex am ined. Each is con sid ered in terms of its risk-shar ing el e ments and the conse quent in cen tives. We of fer sev eral sce nar ios to il lus trate how the shared-care prac tice model might be en cour aged un d er each fi nanc ing mech a nism. Re sults: The current fee-for-ser vice sys tem does not en cour age shared care. For wide adoption of the shared-care prac - tice model, there must be a change in the re im burse ment sys tem s in cen tives. While none of the fi nanc ing mech a nisms of fers a per fect so lu tion, each has po ten tial. Each, how ever, must be care fully tai lored to its en vi ron ment. Con clu sions: Fi nan cial con sid er ations are just one as pect to achiev ing shared care. Nev er the less, in de sign ing a sys tem to en cour age col lab o ra tive, co or di nated care for those suf fer ing from mental ill ness, de ci sion mak ers should be wary of cre at ing or main tain ing ob sta cles (fi nan cial or oth er wise) to pro vi sion of ac ces si ble, high-qual ity care. (Can J Psy chi a try 2001;46: ) Key Words: shared mental health care, capitation, fee-for-service, blended payments, financial incentives Over time, the im por tance of the pri mary care phy si cian s role in treating mental dis or ders has in creased (1). Fam - ily phy si cians es ti mate that 20% to 50% of their pa tients have emo tional or psy cho log i cal prob lems (2). In ad di tion, of the 12% of in di vid u als with diagnosable mental dis or ders who use ser vices, more than 50% re ceive care from a pri mary care phy si cian (3). Mean while, a sub stan tial body of lit er a ture sug gests that many in di vid u als re ceiv ing care in the pri mary care set ting do not re ceive ad e quate treat ment for men tal dis or ders. 1 Health Econo mist, Cen tre for Ad dic tion and Men tal Health, Health Sys tems Re search and Con sult ing Unit, To ronto; As sis tant Pro fes sor, Department of Psychiatry, University of Toronto, Toronto, Ontario. 2 Assistant Professor, Department of Epidemiology and Biostatistics; As sis - tant Professor, Department of Family Medicine, University of Western On - tario, London, Ontario. 3 Director, Centre for Addiction and Mental Health, Health Systems Re - search and Consulting Unit, Toronto; Professor, Department of Psychiatry, University of Toronto, Toronto, Ontario. Address for correspondence: Caro lyn S Dewa, Centre for Addiction and Mental Health, Health Systems Research and Consulting Unit, 250 College Street, Toronto, ON M5T 1R8 carolyn_dewa@camh.net De tec tion and treat ment rates by fam ily phy si cians remain rel a tively low for prob lems such as de pres sion and anxiety dis or ders (4 8). In re sponse to the grow ing de pend ence on pri mary care to pro vide men tal health care and to the po ten tial for im proved treat ment, sev eral ini tia tives have emerged both on global and on more lo cal scales to im prove mental health service de liv ery in the pri mary care set ting (9 10). These ini tia tives de fine pri mary care broadly: it com prises all ser vices re lated to pre ven tion, as sess ment, and treatment pro - vided by front-line cli ni cians. In On tario, the Health Ser vices Re struc turing Com mis sion (10) em pha sized col lab o ra tive care with var i ous health care pro fes sion als in volved in the de - liv ery of pri mary care ser vices. Its re port stressed that only an in te grated team can suc cess fully pro vide a com pre hen sive range of ser vices. Pri mary care is the point of first con tact for men tal health prob lems; it is also the level at which most ser - vices are pro vided for ep i sodes of short du ra tion or low to mod er ate se ver ity. From these con cerns about multidisciplinary, com pre hen sive ser vices in the pri mary care set ting has arisen the con cept of shared care. Fundamentally, shared care is a set of prin ci ples and mech a nisms pro mot ing col lab o ra tive and in te grated Can J Psychiatry, Vol 46, August

4 August 2001 Using Financial Incentives to Promote Shared Mental Health Care 489 treat ment ar range ments based in the pri mary care set ting. Since ar ti cles ad vo cat ing shared care of peo ple with mental health prob lems be gan ap pear ing in the sci en tific lit er a ture (11 13), Ca na dian in ter est has been high. This is prob a bly due to the coun try s vast ge og ra phy and com par a tively small and dis persed pop u la tion, which has re sulted in many pockets of un met need and strained psy chi at ric re sources. Shared care is viewed as a re sponse to this sit u a tion: it can po ten tially im - prove the qual ity of mental health care by mak ing specialty re sources available in pri mary care set tings. Al though it is the o ret i cally at trac tive, sev eral bar ri ers pre vent wide spread adoption of shared care. Not least among them is the question of fi nanc ing. In deed, fi nan cial struc tures are con sis tently iden ti fied as ob sta cles to any type of primary care re form (10,14 17). There is lit tle agree ment, how ever, as to which fi nanc ing mech a nism should be im ple mented. This pa per con sid ers the most com mon re im burse ment struc - tures. We ex am ine in cen tives in her ent in each and dis cuss how each can be used to en cour age shared-care treatment of com mon mental dis or ders. Overview of Shared Care Pri mary care phy si cians, psy chi a trists, and con sum ers have long been dis sat is fied with the split-care ser vice model (18). The tra di tional re la tion ship be tween pri mary care phy si cians and psy chi a trists, how ever, is still gen er ally not col lab o ra - tive. These dis sat is fac tions, com bined with prob lems of ac - cess to psy chi at ric ex per tise, have led to sup port for a ser vice-de liv ery concept known as shared-care. Shared care is de fined by a col le gial part ner ship among pri mary care phy si cian and men tal health spe cial ists in which all share re - spon si bil ity for care. A cen tral tenet of shared care is rec og ni - tion on the part of all pro vid ers that no sin gle pro vider can meet all the pa tient s needs (19). There are var i ous models for link ing men tal health spe cial ists with pri mary care prac tices. These range from sharing space in the same prac tice with lit - tle day-to-day con tact be tween pro vid ers to models wherein coun sel lors are fully in te grated into day-to-day care. A large pro por tion of Ca na dian shared mental health care lit - er a ture ex am ines a model known as the con sul ta tion-li ai son model. It is there fore the fo cus of our sub se quent dis cus sions. In this model, mental health spe cial ists serve 4 func tions, of - fer ing con sul ta tion, fol low-up, in di rect ser vices, and educa - tion (20). In On tario, this model has steadily gained at ten tion, mainly in terms of a psy chi a trist li ai son (21 24). Through out the coun try there is also grow ing in ter est in us ing ad - vanced-prac tice psy chi at ric nurses to liaise with pri mary care pro vid ers (25). Reimbursement Mechanisms and Financial Incentives There are cur rently sev eral ob sta cles to shared care s wide - spread adop tion. They in clude pro vider bi ases and prac tice pat terns and so ci etal bar ri ers gen er ated by mental ill ness-re - lated stigma. Sys tems is sues, such as phy si cian dis tri bu tion, in for ma tion ex change, and re im burse ment mech a nisms, also cre ate sig nif i cant bar ri ers. While all are im por tant and worthy of at ten tion, our dis cus sion fo cuses on re im burse ment and fi - nanc ing. We do not imply that phy si cians re spond solely to fi - nan cial in cen tives, but the ory and re search sug gest that they have some in flu ence on phy si cian be hav iour (26 29). One key to any re im burse ment strat egy in volves the el e ment of risk shar ing. Who bears the risk of costs and to what ex tent? The en tity bearing the costs has the in cen tive to con trol them. Risk can be shifted from pro vider to payer or vice versa, de - pend ing on whether payment is made be fore or af ter ser vices are ren dered. Ret ro spec tive and pro spec tive pay ments are re - im burse ment methods that differ ac cord ing to when pay ment for ser vices is made: as their names sug gest, ret ro spec tive pay ment is given af ter ser vices are ren dered and pro spec tive pay ment is given be fore. With ret ro spec tive pay ment, the pro vider is guar an teed re im burse ment for ser vices, and the payer bears all the risk for costs. The payer there fore has a strong in cen tive to con trol costs, whereas pro vid ers have none. Fee-for-service (FFS) re im burse ment is a pop u lar type of ret ro spec tive pay ment. Con versely, un der pro spec tive pay ment, fi nan cial risk shifts to pro vid ers, who receive a set amount for each pa tient be fore know ing what ser vices pa tients will use. If pa tients use few ser vices, the phy si cian makes a profit; if pa tients use many ser vices or high-in ten sity treat ments, the phy si cian faces a loss. Sev eral types of re im burse ment ar range ments fall into this cat e gory, in clud ing cap i ta tion. In the mid dle ground be - tween ret ro spec tive and pro spec tive pay ment falls a mech a - nism re ferred to as blended pay ment (16). It com bines mul ti ple pay ment types and shares the risk be tween pro vider and payer. In this sec tion, we ex am ine FFS, cap i ta tion, and blended pay - ment mech a nisms in re la tion to the pro vi sion of mental health care in a pri mary care set ting. Con sidering each model, we iden tify who bears the risk and in cen tives each en gen ders. Ret ro spec tive Pay ment: FFS In general, FFS fi nanc ing that cov ers costs pro vides incen - tives for phy si cians to in crease the num ber of pa tients they see and the num ber of ser vices they pro vide. At first, it might seem that a wider range of ser vices in di cates higher-quality care. In ter est ingly, how ever, Rob in son notes that US public and pri vate health care pay ers suspect FFS fi nanc ing of con - trib ut ing to costly health sys tem fail ures (30). It is there fore pos si ble that FFS de creases qual ity of care. For ex am ple, when com mu ni ca tion be tween care givers is not sup ported by FFS, they have no fi nan cial in cen tive to co or di nate care with other pro vid ers or to seek and pro vide con sul ta tions with col - leagues. As a re sult, pri mary care phy si cians are not en cour - aged to ask men tal health spe cial ists for ex pert ad vice and

5 490 The Canadian Journal of Psychiatry Vol 46, No 6 coun sel when treat ing their pa tients with men tal disorders. Nor are mental health spe cial ists typ i cally re im bursed for their time spent dis pens ing such ad vice. In deed, Goldberg ob - served that with FFS, it is of ten ex tremely dif fi cult for pri - mary care pro vid ers to find psy chi a trists to whom re fer rals can be made (28). Even when they suc cess fully find psy chi a - trists to pro vide treat ment, phy si cians may not be given refer - ral out comes be cause psy chi a trists have no in cen tive to take ex tra time to com mu ni cate with re fer ring phy si cians, beyond the oblig a tory ini tial con sul ta tion re ports. Thus, with out re - im burse ment for treatment co or di na tion, there is of ten too lit - tle com mu ni ca tion be tween pro vid ers. Pro spec tive Pay ment: Cap i ta tion In capitated health care pay ment plans, a lump sum is paid on the pa tient s be half. It rep re sents pay ment for all ser vices the pa tient uses during a set in ter val of time. Re im burse ment rep - re sents an un der stand ing be tween payer, pro vider, and pa - tient. Usually, the phy si cian re spon si ble for most of the pa tient s care is en ti tled to the capitated fee. The un der stand - ing is that, for a set pe riod of time, the pa tient is as so ci ated with a spe cific pri mary care phy si cian. In con trast to FFS, this means that the pa tient can not nec es sar ily see a dif fer ent phy - si cian at will. Phy si cians in comes start at a fixed level and are re duced by the cost of ser vices pro vided to pa tients un der their care. As treatment costs rise, phy si cian in come declines. The ob vi ous ben e fit for pay ers is that their health care ex - penses are capped at a pre de ter mined level. Be cause phy si cians re ceive a fixed amount for each pa tient, any of the payment not spent on pa tient care be comes profit. Con se quently, they have strong fi nan cial in cen tives to re duce the ser vices to each pa tient (31,32) and to re fer high ser - vice-us ers to other cli ni cians (28,30). In con trast to FFS, where phy si cians have a fi nan cial in cen tive to pro vide more ser vices, re gard less of whether treatment is ben e fi cial, capita - tion en cour ages them to serve more pa tients but to pro vide fewer ser vices for each (33). Such in cen tives either de crease qual ity by dis cour ag ing the pro vi sion of all-en com pass ing ser vices or en hance ef fi ciency by elim i nat ing ser vices of low value. Some ar gue that capita - tion mo ti vates higher-qual ity care be cause phy si cians en - hance ser vices to at tract and main tain a healthy cli ent base and thereby sus tain their de sired in come level (30). This ar - gu ment makes eco nomic sense given 2 con di tions. First, it is as sumed that the capitated rate is de signed to cover a healthy pop u la tion and is not ad justed for the phy si cian s case-mix. This cre ates the in cen tive to re tain only the health i est pa - tients. In this sce nario high-qual ity ser vices are pro vided to healthy pa tients in an ef fort to keep them, be cause the capitated pay ments are greater than their overall ex pected health care costs. The phy si cian has, at the same time, an in - cen tive to re sist treat ing those pa tients with chronic con di - tions, be cause they are a rel a tively higher-cost pop u la tion (34). This could be es pe cially det ri men tal to a pop u la tion with se vere men tal ill ness. The sec ond con di tion re quires some com pe ti tion among phy si cians for health ier pa tients: if the phy si cian has a cap tive pa tient pop u la tion, there is no need to com pete for healthy pa tients by of fer ing high-quality ser vices (35). Here, it should also be noted that, while the the - ory may make sense, cur rent em pir i cal ev i dence is equiv o cal (36); it is un clear how cap i ta tion af fects qual ity. Pauly compares dif fer ences in be hav iours as so ci ated with FFS versus capitated re im burse ment schemes (32). In FFS, phy si cians know for cer tain their max i mum loss; it is the cost in curred if no pa tients are treated (in clud ing, for ex am ple, over head costs such as rent, util i ties, and equip ment). With cap i ta tion, phy si cians know for cer tain their in come; it is the to tal cap i ta tion rev e nue less the op er a tional costs of treating pa tients and run ning the of fice. In con trast to FFS, with pure cap i ta tion (as op posed to soft cap i ta tion, which will be dis - cussed be low) all fi nan cial risk lies with the phy si cian; the max i mum pos si ble loss is un lim ited (for ex am ple, imag ine that all the pa tients ros tered to a par tic u lar pri mary care group be came se ri ously ill). Phy si cians averse to risk might pre fer FFS to cap i ta tion with the same or higher ex pected in come (32). In Can ada, few men tal health ser vices are pro vided on a strictly capitated ba sis. If this mech a nism were in tro duced to pro mote shared care, the first step would be to ad dress the in - cen tive to se lect only healthy pa tients. This could be done by ad just ing capitated pay ments for the risk as so ci ated with the cov ered pop u la tion. For ex am ple, if a pri mary care prac tice were to see many pa tients with serious mental dis or ders, its re im burse ment could be in creased to re flect their treat ment costs. Oth er wise, pri mary care phy si cians and their treat ment teams will be fi nan cially pe nal ized for treating these patients. Nev er the less, it is difficult to im ple ment risk ad just ment suc - cess fully (37). Fur ther, even with ex tremely well-de vised case-mix ad just ment sys tems, pri mary care treat ment teams are still at risk for un usu ally bad years, even though they are fi nan cially pro tected on av er age. Soft cap i ta tion is a mod i fied ver sion of pure cap i ta tion that seeks to ad dress the prac ti cal lim i ta tions of risk ad just ment. Ways of mod i fy ing a pure cap i ta tion agree ment to pro tect pro vid ers from large po ten tial losses in clude payer pro vider shar ing ar range ments for profit and loss, plac ing a limit on in - di vid ual or ag gre gate pa tient ex pense (for ex am ple, a stop-loss pro vi sion), or mak ing ex cep tions for care that can not be pre dicted (for ex am ple, nondiscretionary care) (33). Soft cap i ta tion s goal is to limit the pro vider s ex po sure to risk and en cour age care pro vi sion for pa tients who might be ex pen sive to treat (for ex am ple, those with chronic mental ill ness). In fact, in the US, where cap i ta tion is widely used to re im burse mental health care pro vid ers, the op tion wherein both ser vice pro vider and payer share the risk is more prev a - lent than is pure cap i ta tion (38).

6 August 2001 Using Financial Incentives to Promote Shared Mental Health Care 491 Adapted from Folland and oth ers (38), Fig ure 1 il lus trates the main dif fer ences be tween pure and soft cap i ta tion re im burse - ment methods. On the X-axis is un ad justed phy si cian net in - come (total cap i ta tion pay ments to tal pa tient costs), and on the Y-axis is ad justed phy si cian net in come, in creas ing in the di rec tion in di cated by the axes ar rows. In a pure capitation con tract, the pro vider bears all the risk and does not share any gains or losses with the payer. Thus, re gard less of a year s fi - nan cial out come, there is no ad just ment to the phy si cian s in - come re cord-break ing prof its or losses are the phy si cian s sole re spon si bil ity. This ar range ment is rep re sented as a 45-de gree line (Line A) from the or i gin. It shows a 1-to-1 re - la tion be tween ad justed and un ad justed phy si cian net in - come; the pro vider bears all gains or losses (which are com pletely de ter mined by the costs of caring for ros tered pa tients). Al ter na tively, un der a soft cap i ta tion con tract, pro vider and payer share the risk, and re sult ing wind falls or losses are split on a per cent age ba sis (Note that with pure cap i ta tion, the per - cent age is 100% for the pro vider and 0% for the payer). This is il lus trated by Line B in Fig ure 1. In this case, the Y-axis rep re sents the pro vider s net in come af ter ad just ing for the payer s share of losses or gains (that is, ad justed total rev e - nue to tal costs). As a re sult, the line is tilted to ward the X-axis to re flect the de gree of risk shar ing be tween payer and pro vider: the more Line B bends away from Line A, the more the pro vider is shielded from fluc tu a tions in either gains or losses. For ex am ple, un der a soft cap i ta tion con tract where the payer agrees to ab sorb one-half the risk, Line B would have a slope of ½. If ac tual costs ex ceeded total capitated rev - e nue, the pro vider would only be li a ble for 50% of the short - fall. Con versely, if ac tual costs were less than total capitated rev e nue, one-half the gains would be re turned to the payer. In other words, ev ery dol lar lost or gained is split 50/50 be tween the payer and pro vider. Cap i ta tion can be fur ther al tered to pre vent ex treme vari a - tions in phy si cian in come. Line B s flat por tions rep re sent stop-loss and stop-gain pro vi sions. They oc cur at points where the payer wishes to shield the pro vider from any ad di - tional losses ( stop-loss ) or gains ( stop-gain ). When the pro vider reaches ei ther of these stop points, the payer col - lects or pays 100% of the ad di tional gains or losses. It is im - por tant to note that stop-loss and stop-gain pro vi sions are ad di tional re fine ments that can be in cor po rated into any capi - ta tion re im burse ment scheme. For ex am ple, Line A could be drawn to in clude flat por tions sim i lar to those in Line B.

7 492 The Canadian Journal of Psychiatry Vol 46, No 6 As il lus trated in Fig ure 1, the re la tion among pure and soft cap i ta tion re im burse ment schemes is re lated to the de gree of risk shar ing be tween pro vider and payer. With risk shar ing, the phy si cian gains or loses less for any given level of unad - justed net in come. Fi nan cial in cen tives as so ci ated with soft cap i ta tion are muted, com pared with those as so ci ated with pure cap i ta tion. This is be cause, un der soft cap i ta tion, provid - ers are partly shielded from the fi nan cial re per cus sions of their de ci sions that af fect health care costs. As a re sult, soft cap i ta tion, com pared with pure cap i ta tion, si mul ta neously pro vides less in cen tive for undertreatment and for cost sav ings/ef fi ciency. Pro spec tive Pay ment: Sal ary Sal ary re im burse ment is a spe cial case of cap i ta tion wherein stop-loss and stop-gain pro vi sions are im me di ately en - gaged. This is il lus trated in Fig ure 1 by the hor i zon tal line (Line C). The line is flat be cause there is no re la tion be tween what the phy si cian is paid (ad justed net in come) and the ac - tual gains or losses ex pe ri enced (un ad justed net in come). Eco nomic the ory pre dicts that sal a ried phy si cians will be paid less than they would ex pect to earn with soft cap i ta tion, on av er age. This is be cause the payer as sumes all the risk by guar an tee ing the phy si cian a fixed in come level. In ex change for as sum ing the risk and pro vid ing a risk-free set in come, the payer can ex tract a pre mium from the pro vid ers re lated to their de sire to avoid risk (risk aver sion). With soft cap i ta tion, phy si cians prof its de pend on the in ten sity of the pa tient pop - u la tion s ser vice needs and the ser vices pro vided. If it is a bad year, pa tients may need more ser vices and thereby cost more to treat. As a re sult, prof its will de crease, and the pro - vid ers will make less than ex pected. In con trast, with a sal ary, they are guar an teed a set in come, re gard less of how sick the pa tient pop u la tion is dur ing the year. For that guar an tee, the payer can of fer pro vid ers less. With out bo nuses or in cen tive plans, sal a ried health care pro vid ers are com pletely shielded from the fi nan cial re sults of their actions. In the ory, they have no fi nan cial in cen tives to in no vate and lit tle in cen tive to work hard. Blended Pay ment: Capitiation and FFS Blended pay ment de scribes a re im burse ment method wherein both cap i ta tion and FFS are used to re mu ner ate phy - si cians (31). It of fers a middle ground be tween the opposing in cen tives as so ci ated with ei ther cap i ta tion or FFS alone. Cor rectly bal anced blended pay ment schemes adequately com pen sate for time spent pro vid ing needed ser vices for pa - tients, but they do not com pen sate enough to make un nec es - sary ser vices lu cra tive. Capitated and FFS re im burse ment can be com bined in many dif fer ent ways; the fol low ing 2 ex - am ples il lus trate their po ten tial flexibility. In the first sce nario, pri mary care phy si cians are re im bursed by cap i ta tion plus FFS for a specified set of pro ce dures. In this way, health care ser vices for which cost con tain ment is a pri or ity (for ex am ple, ser vices that tend to be over used, such as lab tests) can be re im bursed un der a fixed bud get, using cap i ta tion methods. For ser vices where undertreatment is a con cern (for ex am ple, ac tiv i ties fre quently not done, such as co or di nat ing care), FFS re im burse ment can be used. The dif - fer ent re im burse ment mech a nisms pro vide dif fer ent fi nan - cial in cen tives that may fa cil i tate di ver gent goals (for ex am ple, in creas ing some ser vices but de creas ing oth ers). When blended, capitated and FFS re im burse ment cre ate fi - nan cial in cen tives re lated to lev els of re im burse ment and risk shar ing. As Casalino notes, how ever, pay ing FFS only for spec i fied ser vices ig nores the im por tance of the fre quency and length of vis its (31). A dif fer ent model pays phy si cians a cap i ta tion fee per pa tient plus FFS at a much re duced rate. Casalino sug gests that ap - prox i mately 70% of phy si cian in come should come from cap - i ta tion, and 30% from FFS but also re marks that ex act pro por tions have to be de ter mined from ex per i ment and ex - pe ri ence (31). Ostbye and Hunskaar re port that these pro por - tions were suc cess fully used in pi lot pro jects in Nor way (16). Al though with this model FFS ac counts for only a small per - cent age of phy si cian in come, its con tri bu tion, in the ory, is large enough to dis cour age undertreatment but too small to en cour age overtreatment. Sce narios Using the Re im burse ment Models to Pro mote Shared Care Cur rently, most psy chi a trists and fam ily phy si cians in Can - ada are paid FFS for coun sel ling and psy chi at ric services. Pres ent fi nan cial ar range ments offer little mon e tary in cen tive to co or di nate care. In di vid ual phy si cians may take time from their busy sched ules to try to in te grate care, but they are not re im bursed for treatment plan ning and co or di na tion. In gen - eral, the fi nan cial dis in cen tive for com mu ni cat ing with an - other care giver equals the re im burse ment phy si cians could have re ceived by treat ing an other pa tient in stead of placing phone calls. Al berta, where col lab o ra tion be tween family phy si cians and psy chi a trists is re im bursed, rep re sents a no ta - ble ex cep tion. Cur rently, there is also lim ited re im burse ment (other than through pri vate in sur ance and sal a ried po si tions in pro grams and clin ics) for other nonphysican mental health specialists, such as psy chol o gists, so cial work ers, psy chi at ric nurses, and oc cu pa tional ther a pists. Yet, re search ev i dence suggests that men tal health spe cial ists from these dis ci plines are ca pa ble of pro vid ing the same psychotherapies as are phy si cians (39 42). Be cause they are not re im bursed by the gov ern ment as in de pend ent prac ti tio ners, it is dif fi cult for them to pro vide ser vices to those in the com mu nity who need men tal health treat ment. From the pa tient per spec tive, there are ad di tional bar ri ers to ac cess be cause out-of-pocket prices significantly de ter them from using nonphysican men tal health spe cial ist vis its (43,44). As with psy chi a trists, full use of these spe cial - ists is not likely to oc cur in the ab sence of a re im burse ment mech a nism for con sul ta tions.

8 August 2001 Using Financial Incentives to Promote Shared Mental Health Care 493 In the next sec tion, we con sider how each of the fi nan cial mech a nisms dis cussed pre vi ously could be used to pro mote shared care in a pri mary care set ting for those with mental ill - ness. Due to this pa per s con straints, we do not exhaustively dis cuss all im ple men ta tion pos si bil i ties. In stead, we offer ex - am ples of how the con sul ta tion-li ai son shared-care model might be en cour aged. We ex am ine each of the sce nar ios with re spect to how they pro mote pa tient care that is ad e quate, co - or di nated, and pro vided by a col lab o ra tive team of primary and mental health spe cialty care pro vid ers. Sce nario 1: Using FFS to Pro mote Shared Care Un der the current FFS sys tem, mental health co or di na tion and col lab o ra tion are not re warded fi nan cially. Al lowing credentialed nonphysician men tal health care pro vid ers to bill for con sul ta tions with fam ily prac tice phy si cians might in - crease the num ber of con sul ta tions. This would not nec es sar - ily change fam ily phy si cians be hav iour, how ever, be cause they are not typ i cally com pen sated for the com par a tively lon - ger visits that men tal health care re quires. Al ter na tively, the pres ent re im burse ment sys tem could be changed to al low credentialed nonphysician men tal health care pro vid ers to bill for men tal health treat ments. Un doubt - edly, more men tal health treat ments would be pro vided; how - ever, there would be lit tle in cen tive to form the col lab o ra tive net works nec es sary for shared care. Nonphysician mental health care pro vid ers might share of fice space with family phy si cians, but there would be little fi nan cial in cen tive to work as a team. Po lit i cal re al i ties also ar gue against the suc - cess of im ple ment ing this funding ar range ment, be cause it ap pears that at least in the near fu ture deep po lit i cal and or ga ni za tional is sues will con tinue to out weigh the pressing need to al le vi ate the treatment shortage for in di vid u als with men tal ill ness. Pros pects for shared care might be en hanced if fam ily physi - cians were al lowed to bill for the coun sel ling and psy cho ther - apy ser vices of nonphysician mental health care pro vid ers as so ci ated with their prac tices. Re im burse ment ar range ments might re sem ble those in place for den tists and hy gien ists work ing in tan dem, where den tists are re im bursed for rou tine clean ing done by dental hy gien ists. In the case of mental health care, fam ily phy si cians would re fer to their nonphysician mental health col leagues those pa tients whom they did not feel they had the time or skill to treat. None the less, be cause of in cen tives in her ent in this model, the schism be tween pri mary mental health care pro vid ers and psy chi a trists might per sist. More over, it seems likely that this fund ing ar range ment might af fect psy chi a trists case mix. It might be come more fea si ble and fi nan cially re ward ing for pri mary care phy si cians to treat the mild, non spe cific mental dis or ders typ i cally pre sent ing in the pri mary care set ting. If these fi nan cial in cen tives were acted upon, psy chi a trists would be left with the most se vere cases, and as a re sult, the se ver ity of their case loads would likely increase. The in trin sic in cen tives of F FS are to in crease the ser vice vol - ume not to co or di nate care. Shared care will seem eco nom i - cally vi a ble only when pri mary care phy si cians can re cover its ad di tional costs through rev e nues based on in creased quan tity. Given the maldistribution of phy si cians in gen eral, and psy chi a trists in par tic u lar (45), re li ance on FFS fi nanc ing in its cur rent form is not en hanc ing op por tu ni ties for shared care. With FFS in cen tives, over worked phy si cians are paid to be too busy to co or di nate care. Sce nario 2: Using Cap i ta tion to Pro mote Shared Care A capitated re im burse ment plan could pro vide strong incen - tives for shared care if it fo cused on the ser vices de liv ered in - stead of on the pro vider de liv er ing them. For in stance, sup pose that a cap i ta tion ar range ment man dates pro vi sion of men tal health ser vices and that cap i ta tion rates are risk-ad - justed for the pop u la tion being served. Fam ily prac ti tio ners might con sider form ing prac tice groups that in clude nonphysician mental health care pro vid ers or psy chi a trists in ei ther full- or part-time ca pac i ties; the pri mary care phy si cian would be re im bursed for the men tal health ser vices re gard less of whether they were pro vided by the fam ily prac ti tio ner, the nonphysician mental health col league, or in col lab o ra tion with a psy chi a trist. Fi nan cial in cen tives would en cour age pri - mary care groups to pro cure psy chi a trist con sul ta tions or use nonphysician spe cial ists to pro vide ser vices they did not feel they had the time or skill to han dle. Pri mary care phy si cians choos ing not to pro vide mental health care ser vices would risk los ing pa tients (and their as so ci ated capitated pay ments) to an other pri mary care group with a shared-care pro gram pro vid ing more ser vices. This im plies that cap i ta tion pay - ments would have to be high enough to equal the cost of refer - ring some pa tient care to a psy chi a trist, psy chol o gist, or nurse prac ti tio ner (who might be paid at a dif fer ent level than is the fam ily practitioner). This sce nario may also lead to fam ily prac ti tio ners hir ing the least ex pen sive mental health cli ni cian (for ex am ple, a bach e - lors level so cial worker or psy chi at ric nurse), rather than a psy chi a trist or psy chol o gist. At is sue from a pol icy per spec - tive is whether the de lay or lack of treat ment by a mental health cli ni cian with ad vanced train ing is worse than timely treat ment of fered by health care pro vid ers with less train ing. This trade-off is fre quently en coun tered in ru ral health care. It should be noted that whether or not cap i ta tion rates are case-mix ad justed, there remain fi nan cial in cen tives to at tract health ier pa tients. These pa tients are safer fi nan cial risks be cause their po ten tial costs vary less. Thus, even when case-mix ad just ment is state-of-the-art, the risk in tro duced with pro spec tive payment fre quently re quires that pure capi - ta tion schemes be mod i fied. Payers ea ger to pro mote shared care might con sider using soft cap i ta tion pay ment plans with

9 494 The Canadian Journal of Psychiatry Vol 46, No 6 risk re duc tion re lated to the de gree to which their ob jec tives are met (for ex am ple, ini ti at ing and fol low ing shared-care mod els). Like wise, sal a ried phy si cians might be given incen - tive bo nuses based on the ex tent to which they par tic i pate in shared-care networks. Sce nario 3: Using Blended Pay ment to Pro mote Shared Care Given that a large, rad i cal change in the phy si cian re im burse - ment sys tem is im prac ti cal, blended pay ment sys tems may rep re sent the most prac ti cal way of creating fi nan cial incen - tives to pro mote shared care. Blended pay ment builds on the al ready ex ist ing FFS sys tem and in tro duces the con cept of shared care through cap i ta tion: it is the com pro mise so lu tion. It po ten tially com pen sates phy si cians in a man ner that pro - duces in cen tives to share care, while re duc ing in cen tives to pro vide un nec es sary services. The blended payment sys tem is a mech a nism en dorsed by the On tario Med i cal As so ci a tion (46) and used in Brit ish Colum - bia s pri mary care dem on stra tion pro ject (47) and in primary care in many parts of Eu rope (35). It is also being used in hos - pi tals and As ser tive Com mu nity Teams, where psy chi a trists are paid ses sional fees for such in di rect ser vices as con sul ta - tion while they con tinue to bill FFS for di rect pa tient care. A blended payment sys tem, where pri mary care phy si cians are paid a capitated fee in ad di tion to FFS at a re duced rate, has much to rec om mend it, if op ti mal fre quency and length of vis - its for pa tients with men tal dis or ders are to be achieved. Guar an teed in come from cap i ta tion pay ments could subsi - dize the use of men tal health spe cial ists in the pri mary care prac tice. These spe cial ists could as sist in, or com pletely pro - vide, mental health ser vices. The cap i ta tion would also free the phy si cian from being wholly ser vice-ori ented. As a re sult, it would en cour age fam ily prac ti tio ners to spend more time co or di nat ing care with other pro vid ers. The abil ity to re ceive FFS re im burse ment in ad di tion to a capitated fee would coun - ter act the fi nan cial in cen tive to undertreat. For ex am ple, the FFS com po nent could off set the ef fort re quired by the family prac ti tio ners to seek spe cial ist con sul ta tions. To be ef fec tive, the FFS re im burse ment rate should be set at a level that dis - cour ages bill ing for un nec es sary con sul ta tions but is high enough to en cour age con sul ta tion. A ma jor lim i ta tion of any sys tem using cap i ta tion, how ever, is that it in tro duces the need for some type of pa tient ros ter - ing: pa tients must be at tached to a sin gle pri mary care prac tice for a set amount of time. This re stric tion may lead to a per - ceived loss of choice on the part of the pa tient. In the end, it may pres ent a sig nif i cant bar rier to this type of model (35), which would need to be im ple mented care fully. Clinical Im pli ca tions Different reimbursement mechanisms offer different incentives. Currently in Canada, the financial incentives to provide shared care are not strong. More empirical research is needed on the best method of financing to promote shared care. Limi ta tions This paper does not exhaustively discuss all po ten tial funding mechanisms. The extent of physicians responses to financial incentives is not dis - cussed. The discussion focuses on the consultation-liaison model of shared care and does not consider other models. Conclusion Re im burse ment schemes have in her ent fi nan cial in cen tives, and we leave the ex tent of phy si cians re sponse to fi nan cial in cen tives for oth ers to de bate. In this ar ti cle, we have ex am - ined FFS, capitated, and blended payment schemes. Based on the fi nan cial in cen tives as so ci ated with these re im burse ment meth ods, we sug gest sce nar ios un der which shared mental health care could be pro moted in a pri mary care set ting. Fi - nan cial con sid er ations are just one as pect of the overall pol icy en vi ron ment. In seek ing to de sign a sys tem that en cour ages col lab o ra tive, co or di nated care for those suf fer ing from men - tal ill ness, de ci sion mak ers should pay par tic u lar at ten tion to ob sta cles (fi nan cial or oth er wise) to the pro vi sion of ac ces si - ble, high-qual ity care. Acknowledgements The authors are grate ful to William Gnam, Wil lard Manning, Sagar Parikh, Da vid Streiner, and Don ald Wasylenki for help ful dis cus - sion and com ments. Dr Hoch grate fully acknowledges fi nan cial support from the Nat u ral Sciences and Engineering Research Coun - cil of Canada (Grant #R3034A01). References 1. Lin E, Goering P. The uti li za tion of phy si cian ser vices for men tal health in On tario TR, To ronto (ON): ICES; Cra ven M, Co hen M, Campbell D, Wil liams J, Kates N. Mental health prac tices of On tario fam ily phy si cians: a study us ing qual i ta tive meth od ol ogy WP, To - ronto (ON): ICES; Parikh S, Lin E, Lesage AD. Men tal health treat ment in On tario: se lected com par i - sons be tween the pri mary care and spe cialty sec tors. Can J Psy chi a try 1997;42: Spitzer RL, Kroenke K, William J, Janet B, Patient Health Ques tion naire Pri mary Care Study Group. Val i da tion and utility of the self-re port ver sion of prime-md: the pa tient health ques tion naire pri mary care study. JAMA 1999;282: Fifer SK, Mathias SD, Pat rick DL, Mazonson PD, Lubeck DP, Buesching DP. Un - treated anxiety among adult pri mary care pa tients in a health main te nance or ga ni - za tion. Arch Gen Psy chi a try 1994;51: Simon GE, Von Korff M. Rec og ni tion, man age ment, and out comes of de pres sion in pri mary care. Arch Fam Med 1995;4: Coyne JC, Schwenk TL, Fechner-Bates S. Non de tec tion of de pres sion by pri mary care phy si cians re con sid ered. Gen Hosp Psy chi a try 1995;17: Perez-Sta ble EJ, Miranda J, Munoz RF, Ying YW. De pres sion in med i cal out pa - tients. Underrecognition and misdiagnosis. Arch In tern Med 1990;150:

10 August 2001 Using Financial Incentives to Promote Shared Mental Health Care World Health Or ga ni za tion. Setting the WHO Agenda for men tal health. Bull World Health Or gan 2000;78: On tario Health Ser vices Re struc turing Com mis sion. Looking back, looking for - ward a leg acy re port. To ronto: On tario Ministry of Health and Long-Term Care; Kates N, Lesser A, Dawson D, Devine J, Wakefield J. psy chi a try and fam ily med i - cine: the McMaster ap proach. Can J Psy chi a try 1987;32: Strathdee G. Pri mary care-psy chi a try in ter ac tion: a Brit ish per spec tive. Gen Hosp Psy chi a try 1987;9: Royal Col lege of Gen eral Prac ti tio ners. Shared care of pa tients with mental health prob lems. Oc ca sional pa per 60, Lon don: Royal Col lege of Gen eral Prac ti tio ners; Leatt P, Pink GH, Guerriere M. To wards a Ca na dian model of in te grated healthcare. Healthcare Pa pers 2000;2: Leggat SG, Walsh M. from the bot tom up and other les sons from Down Under. Healthcare Pa pers 2000;2: Ostbye T, Hunskaar S. a new pri mary care ros ter ing and cap i ta tion sys tem in Nor - way: les sons for Can ada? Can Med Assoc J 1997;157: Col lab o ra tive Working Group on Shared Mental Health Care. Shared mental health care in Can ada: cur rent sta tus and com men tary. Ot tawa and Mississauga: Ca na dian Psy chi at ric As so ci a tion and The Col lege of Fam ily Phy si cians of Can ada; Unger TE, Hoffman BF. Two sol i tudes: psy chi a try and pri mary care fam ily med i - cine: a growing re la tion ship. Health Law Can ada : Kates N, Cra ven M, Crustolo AM, Nikolaou L, Al len C. In te grating mental health ser vices within pri mary care. A Ca na dian pro gram. Gen Hosp Psy chi a try 1997;19: Gask L, Sibbald B, Creed F. Eval u ating mod els of work ing at the in ter face be tween men tal health ser vices and pri mary care. Br J Psy chi a try 1997;170: Kates N, Cra ven MA, Bishop J, Clinton T, Kraftcheck D, LeClair K, and others. Shared men tal health care in canada. Can J Psy chi a try 1997;42 (Suppl):1S 12S. 22. Kates N, Cra ven MA, Crustolo AM, Nikolaou L, Al len C, Farrar S. Sharing care: the psy chi a trist in the fam ily phy si cian s of fice. Can J Psy chi a try 1997;42: Turner T, De Sorkin A. Sharing psy chi at ric care with primary care phy si cians: the To ronto doc tors ex pe ri ence ( ). Can J Psy chi a try 1997;42: Unger TE, Jarmain S. Shared mental healthcare: a col lab o ra tive con sul ta tion re la - tion ship: the North-York Gen eral Hos pi tal ex pe ri ence. Hos pi tal Quar terly 1999;Win ter: Willson B. The role of ad vanced-prac tice psy chi at ric mental health nurses within pri mary care set tings. Van cou ver/rich mond (BC): Van cou ver/rich mond Health Board; Wilkinson BR. Phy si cian pay ment: in cen tives change with sup ply. Can Med Assoc J 1997;157: Kassirer JP. Man aged care and the mo ral ity of the mar ket place. New Engl JMed 1995;333: Goldberg RJ. Fi nan cial in cen tives in flu enc ing the in te gra tion of mental health care and pri mary care. Psychiatr Serv 1999;50: Rice T. The eco nom ics of health health re con sid ered. Chi cago: Health Ad min is tra - tion Press; Rob in son JC. Blended payment meth ods in phy si cian or ga ni za tions un der man - aged care. JAMA 1999;282: Casalino LP. Bal ancing in cen tives how should phy si cians be re im bursed? JAMA 1992;267: Pauly M. Ef fi ciency, in cen tives and re im burse ment for health care. In: Weeks LE, Berman HJ, Bisbee GE, ed i tors. Fi nancing of health care Ann Ar bor: Health Ad - min is tra tion Press; Gold M. Fi nan cial in cen tives cur rent re al i ties and chal lenges for phy si cians. J of Gen Intern Med 1999;14 (Suppl 1):6S 12S. 34. Newhouse JP. Re im bursing health plans and health pro vid ers: ef fi ciency in pro - duc tion ver sus se lec tion. Jour nal of Eco nomic Lit er a ture 1996;34: Hurley J, Hutchison B, Giacomini M, Birch S, Dorland J, Reid R, and others. Pol icy con sid er ations in im ple ment ing cap i ta tion for in te grated health systems. Ot tawa: Ca na dian Health Ser vices Re search Foun da tion; Ar mour BS, Pitts M, Maclean R, Cangialose C, Kishel M, Imai H, and others. The ef fect of ex plicit fi nan cial in cen tives on phy si cian be hav iour. Arch In tern Med 2001;161: Newhouse JP. Risk ad just ment: where are we now? In quiry 1998;35: Folland S, Good man AC, Stano M. The eco nom ics of health and health care. 2nd Ed. To ronto: Prentice-Hall Canada, Inc; Marks I. Con trolled trial of psy chi at ric nurse ther a pists in pri mary care. BMJ 1985;290 (Clin i cal Re search Ed u ca tion 6476): Paykel ES. In no va tions in mental health care in pri mary care sys tem. In: Scott R, Marks I, ed i tors. Mental health care de liv ery. Cam bridge (UK): Cam bridge Uni - ver sity Press; Corney RH. The ef fec tive ness of at tached so cial work ers in the man age ment of de - pressed fe male pa tients in gen eral prac tice. Psychol Med Monogram 1984;6 (Suppl): Knesper DJ, Pagnucco DJ, Wheeler JR. Sim i lar ities and dif fer ences across mental health ser vices pro vid ers and prac tice set tings in the United States. Amer i can Psy - chol o gist 1985;40: McGuire T. Fi nancing and de mand for mental health ser vices. Jour nal of Hu man Re sources 1981;16: Deb P, Holmes AM. Sub sti tu tion of phy si cians and other pro vid ers in out pa tient men tal health care. Health Econ 1998;7: Chan B. Sup ply of phy si cian s ser vices in On tario. To ronto (ON): ICES; Sibbald B. Is fee-for-ser vice on the way out for Ontario FPs? Can Med Assoc J 1999;161: Brit ish Co lum bia Ministry of Health and Min is try Re spon si ble for Se niors. Up date pri mary care dem on stra tion pro ject. Is sue 2. Van cou ver (BC): Brit ish Co lum bia Min is try of Health; Résumé Le recours à des incitatifs financiers pour promouvoir les soins de santé mentale partagés Ob jec tifs : Ex am iner les struc tures de rem bourse ment les plus répan dues dans les soins pri maires, trou ver les in ci ta tifs inhé rents à chacune et pré senter com ment chacune pour rait ser vir à stimuler une ap pro che de soins partagés pour le traite ment des trou bles men taux dans les soins pri maires. Méth ode : Trois prin ci paux modèles rémuné ra tion à l acte, capi ta tion et mé can ismes de ver se ments con fon dus sont ex ami nés, chacun en fonc tion des éléments de part age des risques et des incitatifs qui s en suivent. Nous pré sen - tons plusieurs scé nar ios pour il lus trer com ment le modèle de la pra tique des soins part agés peut être fa vo risé par chaque mé can isme de fi nance ment. Résul tats : Le système ac tuel de rémuné ra tion à l acte n in cite pas aux soins part agés. Pour que le modèle de la pra - tique des soins part agés soit large ment adopté, il faut modi fier les in ci ta tifs du système de rem bourse ment. Bien qu au - cun mé can isme de fi nance ment n offre de so lu tion par faite, chacun offre des pos si bili tés, mais doit être soig neuse ment adapté à son mi lieu. Con clu sions : Les con sidé ra tions fi nan cières ne sont qu un aspect de la réali sa tion des soins part agés. Né an moins, en con ce vant un système qui in cite à des soins co opé ratifs et co or don nés dis pensés aux person nes souf frant de maladie men tale, les dé cideurs doivent pren dre garde de créer ou de main tenir des ob sta cles (financ iers ou autres) à la presta - tion de soins ac ces si bles de grande qual ité.

11 IN REVIEW Economic Impacts of Supported Employment for Persons With Severe Mental Illness Eric A Lati mer, PhD 1 Back ground: Most per sons with se vere mental ill ness pre fer com pet i tive to shel tered vo ca tional set tings. Sup ported em ploy ment (SE) has be come a clearly de fined model for help ing peo ple with se vere mental ill ness to find and main - tain com pet i tive jobs. It in volves in di vid u al ized and rapid place ment, ongoing sup port and as sess ment, and in te gra tion of vo ca tional and men tal health staff within a sin gle clin i cal team. Pre vi ous studies show that SE secures competitive em ploy ment much more ef fec tively than do other ap proaches. This re view fo cuses on its eco nomic im pacts. Methods: Studies re port ing some ser vice use or mon e tary out comes of add ing SE pro grams were identi fied. These out - comes were tab u lated and are dis cussed in nar ra tive form. Re sults: Five nonrandomized and 3 ran dom ized stud ies com pare SE pro grams with day treatment or tran si tional em - ploy ment pro grams. The in tro duc tion of SE ser vices can re sult in any thing from an in crease to a de crease in vo ca tional ser vice costs, de pend ing on the ex tent to which they sub sti tute for pre vi ous vo ca tional or day treat ment ser vices. Over - all ser vice costs tend to be lower, but dif fer ences are not sig nif i cant. Earn ings in crease only slightly on av er age. Con clu sions: Con verting day treat ment or other less ef fec tive vo ca tional pro grams into SE pro grams can be cost-sav - ing or cost-neu tral from the hos pi tal, com mu nity cen tre, and gov ern ment points of view. In vest ments of new money into SE pro grams are un likely to be ma te ri ally off set by re duc tions in other health care costs, by re duc tions in gov ern - ment ben e fit pay ments, or by in creased tax rev e nues. Such in vest ments must be mo ti vated by the value of in creas ing the com mu nity in te gra tion of per sons with se vere men tal ill ness. (Can J Psy chi a try 2001;46: ) Key Words: vocational rehabilitation, supported employment, individual placement and support, mental illness, psychiatric disability, costs, cost-effectiveness, economics In creas ingly so phis ti cated ef forts are being made to sup port the com mu nity in te gra tion of per sons with se vere mental ill ness. Work is an es sen tial el e ment of that in te gra tion. Sur - veys in di cate that most in di vid u als with se vere mental ill ness want to work in reg u lar em ploy ment set tings, earning the min i mum wage or better (1). More over, stud ies show that work in creases their self-es teem (2,3) and qual ity of life (4). Yet US data in di cate that 10 years ago, at least, more than three-quar ters of per sons with se vere men tal ill ness re mained un em ployed (5). Tra di tional ap proaches to pro vid ing this Manuscript received and accepted June Research Scientist, Douglas Hospital Research Centre, Verdun, Quebec; Assistant Professor, Department of Psychiatry, McGill University, Mont - real, Quebec; Associate Member, Department of Economics, McGill Uni - versity, Montreal, Quebec. Address for correspondence: Dr E Latimer, Douglas Hospital Research Centre, 6875 LaSalle Boulevard, Verdun, QC H4H 1R3 lateri@douglas.mcgill.ca pop u la tion with mean ing ful work have in volved shel tered or tran si tional em ploy ment, which rarely lead to com pet i tive em ploy ment (6,7). Over the past 15 years, a rad i cally dif fer ent ap proach called sup ported em ploy ment (SE) has emerged (8,9). Ini tially ap - plied to vo ca tional re ha bil i ta tion of per sons with phys i cal and de vel op men tal dis abil i ties (8), SE in volves plac ing cli ents di - rectly into com pet i tive jobs and then of fer ing them ongoing sup ports place-train, rather than train-place. During the 1990s, US re search ers for mal ized the SE model for peo - ple with psy chi at ric dis abil i ties, un der the la bel In di vid ual Place ment and Sup port (IPS) (10,11). This for mu la tion of SE bor rows from the As ser tive Com mu nity Treat ment (ACT) model (12) the prin ci ple that treatment and re ha bil i ta tion ser - vices should be in te grated within a sin gle clin i cal team. The term sup ported em ploy ment, ap plied to per sons with se - vere mental ill ness, is in creas ingly be ing con strued in a man - ner con sis tent with the prin ci ples of IPS (13). That is the sense in which it will be used through out this re view. Can J Psychiatry, Vol 46, August

12 August 2001 Economic Impacts of Supported Employment for Persons With Severe Men tal Illness 497 SE prin ci ples have been de scribed ex ten sively else where (14) and will only be listed here: the search for com pet i tive rather than shel tered em ploy ment; no, or min i mal, prevocational train ing; in te gra tion of vo ca tional and men tal health treat - ment within the same clin i cal team; at ten tion to cli ent pref er - ences; com pre hen sive, con tin u ous, work-based as sess ment; and time-un lim ited sup port. The em pir i cal ba sis for these prin ci ples has also been ad dressed (14). In ad di tion to these prin ci ples, screen ing for work-readi ness is avoided, so that all cli ents who ex press a de sire for work may have that op por tu - nity (15). Var i ous studies of SE in volv ing per sons with se vere mental ill ness were car ried out in the US during the 1990s, sev eral of them rig or ous in de sign. They have con sis tently found that SE yields much higher rates of com pet i tive em ploy ment than do var i ous al ter na tive strat e gies in volv ing more grad ual ap - proaches (6,16 21). A recent metaanalysis sum ma riz ing 5 well-con ducted tri als com par ing SE with vo ca tional training con cludes that sub jects in SE pro grams are more likely to be in com pet i tive em ploy ment than are those re ceiv ing prevocational train ing, up to 18 months (for ex am ple, 34% vs 12% at 12 months); in that re view, the dif fer ence at 24 months, based on 2 tri als only, fa vours SE but is mar gin ally sig nif i cant (7). Fur ther, ev i dence that pro grams which score higher on a SE fidelity scale (22) achieve higher com pet i tive em ploy ment rates has also re cently been re ported (23). Sur pris ingly, these studies have gen er ally not re ported that SE ser vices have any ef fects on symp toms, self-es teem, or qual ity of life. Re cent ev i dence sug gests, how ever, that there is an as so ci a tion be tween the ex pe ri ence of work ing and im - prove ments in symp toms; in sat is fac tion with vo ca tional ser - vices, lei sure, and fi nances; and in self-es teem (3,24). Being of fered SE ser vices does not in prac tice guarantee steady em - ploy ment: as in di cated above, Crow ther and oth ers cal cu lated a mean em ploy ment rate of 34% at 12 months (7). Thus, the ac tual ef fects of work ing on these do mains could be di luted by the pres ence of many sub jects who work lit tle or not at all, even though they are re ceiv ing SE ser vices. Whether work it - self does af fect these do mains, or whether some other causal mech a nism is op er at ing, re mains un clear at pres ent. None the less, there ex ists a widely shared be lief that in te gra - tion of those with se vere mental ill ness into reg u lar set tings in the com mu nity is pref er a ble to main tain ing them in seg re - gated settings. On this ground, the finding that SE more ef fec - tively pro motes in te gra tion into reg u lar work set tings, with at worst no dis cern ible neg a tive out come for cli ents, sug gests that more SE pro grams should be de vel oped. The prac ti cal ques tion then arises as to the eco nomic fea si bil - ity of im ple ment ing SE ser vices for per sons with se vere men tal ill ness. This ar ti cle re views the ev i dence on the is sue. Pre vi ous re views have been con ducted (25 29). Most have in volved peo ple with de vel op men tal dis abil i ties. This re view in cor po rates more recent stud ies (20,21). Methods Po ten tially rel e vant studies were iden ti fied through searching the lit er a ture, iden ti fy ing ref er ences noted in lit er a ture re - views, and com mu ni cat ing with ex perts. Studies were sought that 1) de scribed SE pro grams set up in par al lel with other com mu nity sup port ser vices; 2) were de signed to es ti mate the SE pro gram s ef fects in isolation from those of other commu - nity sup port ser vices; 3) re ported at least some eco nomic data other than per cent age or time com pet i tively em ployed; and 4) tar geted a broad cross-sec tion of per sons with se vere men - tal illness. This led to ex clud ing stud ies on sev eral cri te ria: first, stud ies were ex cluded that dealt pri mar ily or wholly with SE for per - sons with de vel op men tal or phys i cal dis abil i ties, be cause per sons with se vere men tal ill ness have lower base line rates of em ploy ment and in te grat ing them into the reg u lar workforce pres ents dis tinct chal lenges. Sec ond, studies were ex cluded that re ported on in te grat ing vo ca tional re ha bil i ta - tion ei ther into ACT pro grams (30,31) or into com pre hen sive pro grams other than ACT (32), be cause the sep a rate ef fects of the vo ca tional com po nent on eco nomic outcomes cannot be iso lated. Fur ther, ACT pro grams target only a small sub group of per sons with se vere mental ill ness, es ti mated at 0.07% to 0.1% of the gen eral pop u la tion (33), whereas per sons with se - vere mental ill ness rep re sent about 2.6% of the gen eral popu - la tion (34). The eco nomic im pli ca tions of in clud ing a fo cus on vo ca tional re ha bil i ta tion in ACT pro gram ming are not dis - cussed ex plic itly in this re view. Third, stud ies that re ported only clin i cal out comes and workforce par tic i pa tion rates, with no data on earn ings, ser vice use, or service costs, were also ex cluded (for ex am ple, [6] and some pre lim i nary re sults of on go ing studies cited in [13]). As a re sult, in ci den tally, this meant that the only Ca na dian study iden ti fied in this search, car ried out in Brit ish Co lum bia (35), was ex cluded. Finally, 2 un pub lished re ports and 1 un pub lished Mas ter s the sis, cited in (25), as well as an other un pub lished re port cited in (13), which might not have been ex cluded ac cord ing to the pre vi - ous cri te ria, were not ob tained in time for in clu sion in this re view. To in crease com pa ra bil ity across stud ies, where nec es sary, data were re-ex pressed in terms of units per cli ent per year (for ex am ple, dol lars per client per year). In one case, stan - dard errors of the mean were con verted into stan dard de vi a - tions for the same rea son (20). Costs are re ported in the same units as in the orig i nal study.

13 498 The Canadian Journal of Psychiatry Vol 46, No 6 Table 1. Description of supported employment (SE) studies reviewed Study and location % a Previous work experience (entire sample) Groups (n) Duration Fidelity of SE program Nonrandomized studies Rogers and oth ers (16), Bos ton, MA 37 Mean of 12.4 months full-time in previous 5 years SE (19) b 1 year Low to medium c Drake and oth ers (18); Clark and others (38), Ru ral New Hampshire % in paid competitive employment at study start SE (71) DT (112) d,e 1½ years High Drake and oth ers (19); Clark and others (38), Ru ral New Hampshire % in paid competitive employment at study start SE (112) f 2 years High Bai ley and oth ers (50) 77.4 g 14.6% in competitive employment at study start SE (31) CSP (31) Becker and oth ers (21) % employed in any competitive job in previous 5 years SE (73) DT (41) 1 year High 2 years High Randomized studies Bond and oth ers (17, 39), In di an apo lis, IN and area 66 70% employed continuously 1 year or longer in competitive job SE (43) GE (43) 1 year High at one site, medium at the other h Drake and oth ers (40); Clark and others (41), New Hamp shire cit ies 46.9 Relatively good employment histories ; 0% in competitive employment at study start IPS (74) GST (69) 1½ years High at one site, low to medium at the other i Drake and oth ers (20), Wash ing ton, DC months paid work in previous 5 years SE (76) EVR (76) 1½ years High CSP = com mu nity sup port pro gram; DT = day treatment; EVR = en hanced vo ca tional rehabilit ation; GE = grad ual en try; GST = group skills training a Per centage with schizo phre nia or schi zoaf fec tive dis or der in en tire sam ple unless oth er wise speci fied. b Serv ices prior to in tro duc tion of SE are not clearly speci fied; they in clude day treatment and vo ca tional re ha bili ta tion (see Ta ble 2). c Some screening for work readi ness ap pears likely. A struc tured 7- week pre em ploy ment phase was in tended to pre cede ac tive job search (51). Vo ca tional re ha bili ta tion staff did not ap pear to work closely with the clini cal team. d DT in clud es skills training groups, so ciali za tion groups, and shel tered work within the men tal health cen tre; CSP in volves case man age ment with no day treat ment; GE clients had to re ceive at least 4 months of pre vo ca tional prepa ra tion be fore be com ing eli gi ble for the SE pro gram; GST in volves pre em ploy ment training to choose, get, keep a job, fol - lowed by in di vidu al ized sup port; EVR in volves sev eral agencies en dors ing the goal of com peti tive em ploy ment but re ly ing on paid work- adjustment training in shel tered work - shop set tings. En hanced re fers to the presence of a vo ca tional coun sellor who fa cili tated cli ent in volve ment with agencies. e Quasi- experimental de sign with com pari son group. f In this study, the day treat ment pro gram that served as com pari son group in the Drake and oth ers (18) study con verted to In di vid ual Place ment and Sup port (IPS). The study com pares out comes be fore and af ter this con ver sion. g Per cent age of sam ple with a psy chotic dis or der. h Sepa rate vo ca tional and mental health pro grams at one site. i At one site, em ploy ment spe cial ists em pha sized shel tered work. The in for ma tion is syn the sized in nar ra tive form. Metaanalytic methods were not used, due to im por tant differ - ences in the pro grams being com pared, and in other factors, which imply that for any given di men sion, such as em ploy - ment earnings, there is un likely to be in fact a sin gle un der ly - ing true ef fect size to be es ti mated. Con trolling for dif fer ences in pro grams, cli en teles, and other fac tors, using mul ti ple re gres sion meth ods as in (36), was not pos si ble, due to the small num ber of avail able stud ies. Results Ta ble 1 de scribes the studies in cluded in this re view. Five nonrandomized and 3 ran dom ized stud ies were iden ti fied. Fi - del ity to the SE model is gen er ally high, with the ex cep tion of the study by Rog ers and oth ers, while com par i son ser vices range from day treatment with rel a tively little em pha sis on vo ca tional re ha bil i ta tion (the com par i son pro gram in [18]) to state-of-the-art step wise pro grams. Data on em ploy ment his tory the single base line char ac ter is tic that con sis tently pre dicts em ploy ment out comes (18,20,21) are not re ported con sis tently, and it is dif fi cult to in fer to what ex tent this im - por tant char ac ter is tic ac tu ally var ies across stud ies. Ta ble 2 sum ma rizes the ef fects of SE on use of the fol low ing ser vices: hos pi tal iza tions, emer gency room (ER) visits or cri - sis ser vices use (high lighted as a ba rom e ter of in sta bil ity and likely sub se quent re source use), and a gen eral other cat e - gory that mostly in cludes case man age ment and out pa tient costs. In gen eral, con sis tent with other re ports (for ex am ple, [13]), SE does not ap pear to af fect hos pi tal iza tion rates ma te - ri ally or sig nif i cantly. The single pos si ble ex cep tion is the study by Rog ers and oth ers (16), where a re duc tion in hos pi tal days of more than 50% was ob served be tween the base line and fol low-up years. The sam ple size of 19 is very small, and

14 August 2001 Economic Impacts of Supported Employment for Persons With Severe Men tal Illness 499 Table 2. Supported employment impacts on use of services (per client per year unless otherwise specified) Study Hospitalization Emergency room or crisis services use Other service use Rogers and others (16) Drake and others (18); Clark and others (38) Drake and others (19); Clark and others (38) Bailey and others (50) Becker and others (21) Before SE: 6.5 days After SE: 2.9 days (not tested) 28.2% of IPS group rehospitalized during follow-up period; 25% of comparison group (not significant) (18). Mean length of stay for sample of n = 27 regular day treatment users at experimental site, from 2.63 to 2.00 (not significant); comparison site (n = 31 regular users), from to (not significant) (38) 25% hospitalized during baseline period (that is, follow-up period for comparison group of Drake and others [18]), 14.3% during follow-up period (P < 0.025) (19) for n = 31 regular users, from to (not significant) (38) Days of hospitalizations did not change (no numbers reported) SE: from 16.4% admitted in past year at baseline to 9.6% at follow-up. DT: from 26.8% to 22% (program by time interaction not significant) Crisis services (hours): before, 1.5 after, 0.2 (not tested) Crisis services: experimental site, 1.24 (SD 2.07) to 1.41 (SD 4.09) mean hours per client; comparison site: 1.71 (SD 3.83) to 1.67 (SD 3.4) hours From 1.67 (SD 3.40) hours to 1.86 hours (SD 3.44) (not significant) Days of crisis housing did not change (no numbers reported) Not specified Day treatment (days): before, 9.2; after, 3.5 Supported residential (days): before, 33.1, after, 5.5 Other services (hours): before, 141; after, a (no differences tested) Experimental site (community services crisis services, case management, outpatient, partial hospitalization, vocational): (SD ) to (SD 69.02) Control site: (SD ) to (SD ) (site by time interaction not significant at 0.05 level: P < 0.09) From (SD ) hours to (SD ) (not tested) Outpatient mental health service use did not change (no numbers reported) Not specified Bond and others (17,39) Accelerated: 5.8 days. Gradual: 5.3 days (not significant) Not specified Accelerated: 85.8 employment specialist hours (n = 25). Gradual: 40.8 hours (not tested) Drake and others (19); Clark and others (41) Similar large decreases for both groups (numbers not specified) Drake and others (20) SE: 20.9 (baseline rate 30.3). EVR: 12.1 (baseline rate 17.4) (difference not significant when adjusted for difference in baseline rate) Not specified Not specified Over 18 months: SE: 61.6 (SD 37.1) direct contact hours. GST: 74.1 (SD 59.2) (difference not significant) Not specified a In de creas ing or der of im por tance at base line: ther apy or medi ca tion evalua tion, non- SE vo ca tional serv ices, case man age ment, so cial or rec rea tional ac tivi ties, sup ported educ a - tion. Table 3. Supported employment impacts on vocational rehabilitation and health and social services costs (on annualized, per client basis) a Study Alternative program SE program Inpatient costs Other service costs Overall increase (decrease) or difference (SE other) in service costs Rogers and others (16) $620 $7128 new program + $607 usual vocational services Before: $2220 After: $1011 (not tested) Before: $8319 After: $5053 (not tested) $2639 (not tested) Drake and others (18); Clark and others (38) SE site baseline: $7686 b $1595 (SD$ 1261) SE: before, $1130 (SD$ 3227); after, $864 (SD$ 2476) (not significant) DT: before, $9300 (SD$ ); after, $ (SD$ ) (not significant) SE: before, $6349; after, $8089 DT: before, $6325; after, $5897 c SE: ($5670) DT: ($482) Drake and others (19); Clark and others (38) $6597 (SD$ 8315) $1878 (SD$ 2003) Before: $ (SD$ ) After: $ (SD$ ) Before: $5897 After: $7958 c ($2080) (not significant) d Bailey and others (50) Not specified Not specified Not specified Not specified Service costs did not change (no numbers reported) Bond and others (17,39) $6103 ($4667 day treatment + $1436 SE) $4463 ($1443 day treatment + $3020 SE) Not specified Accelerated entry: $1713 Gradual entry: $2371 e (not tested) ($2298) (not tested) Drake and others (40); Clark and others (41) $3757 $3688 SE: $4095 GST: $4457 (not significant) Outpatient costs: SE: $5525 GST: $6685 (not significant) ($1453) (not significant) a Becker and oth ers (21) and Drake and oth ers (20) do not re port any cost data and are not in cluded in this table. b Weighted av er age of both sites at base line and com pari son site at follow- up. Since this aver age was cal cu lated by the author, no stan dard deviation or test is re ported. c Case man age ment, out pa tient, emer gency, and (af ter pe riod in SE) par tial hos pi tali za tion costs. d Count ing only com mu nity costs, dif fer ence is ($2657) and is sta tis ti cally sig nifi cant using a 1- tailed test (P < 0.05). e Other serv ice costs include, in de creas ing order of im por tance: drop- in cen tre, outpatient serv ices, medi ca tion clinic, club house, psy chia trist, sub stance abuse coun sel ling.

15 500 The Canadian Journal of Psychiatry Vol 46, No 6 Table 4. Supported employment impacts on earnings, benefits, and taxes paid (per client per year) Study Employment Earnings Benefits Taxes Rogers and others (16) Increase in earnings of $1846 (not tested) Reduction in transfer payments of $933 (not tested) Increase in taxes of $425 (not tested) Drake and others (18,19); Clark and others (38) Statistically significant difference in favour of SE, counting all clients (amounts not reported) Not specified Not specified Bailey and others (50) Significant improvement in total wages for SE group (P = 0.001) (amounts not reported) Not specified Not specified Becker and others (21) SE: $340. DT: $91 (not significant) Not specified Not specified Bond and others (17,39) Accelerated entry: $1525 Gradual entry: $574 (not tested) Not specified Not specified Drake and others (40); Clark and others (41) SE: $2897 (3882). GST: $1783 (1918) SE: $6661. GST: $6912 (not significant) SE: $471. GST: $277 (P < 0.05) (P < 0.05) a Drake and others (20) SE: $1333 (2254). EVR: $1337 (1967) (not significant) Not specified Not specified a Dif fer ence is sta tis ti cally sig nifi cant as shown when pre vi ous work is used as a covaria te (40). no con fi dence in ter val or sta tis ti cal tests are re ported, so this find ing could be the re sult of sam pling vari a tion. If it is real, how ever, the fact that em ploy ment spe cial ists worked very in ten sively with cli ents (ra tio 10:1), and to some ex tent as - sumed a case-man age ment role, could have meant that they had the same kind of ef fect on hos pi tal iza tion rates as is usual with in ten sive case man age ment or ACT (36,37). There are fewer re sults on ER vis its or cri sis ser vices than on hos pi tal iza tions, but those avail able sug gest the same pattern as found with hos pi tal iza tions: no ef fect, or per haps an ef fect if em ploy ment spe cial ists also serve as case man ag ers. In terms of other ser vice use, Rog ers and oth ers (16) at trib ute an in crease in sup ported ed u ca tion hours to 2 cli ents who re - turned to school. They also at trib ute a sig nif i cant in crease in use of so cial and rec re ational ac tiv i ties to strong efforts by pro gram staff to help cli ents strengthen their so cial sup port net works. In some of the New Hamp shire day treat ment con - ver sion studies (18,19,38), the ap par ent re duc tion in commu - nity-care hours fol low ing con ver sion of day treat ment into SE is due to a dra matic re duc tion in partial hos pi tal iza tion (clo sure of day treat ment) that more than off sets smaller in - creases in vo ca tional, out pa tient coun sel ling, and case-man - age ment hours. (As re flected in the cost data pre sented be low, how ever, the day treat ment ser vices were de liv ered in group for mat and cost less per cli ent per hour than the more in ten - sive SE ser vices that re placed them.) Finally, in the study by Bond and oth ers, con trol group sub jects could re ceive SE, but only af ter a de lay of 4 months or more, dur ing which they re - ceived prevocational prep a ra tion (17,39). The em ploy - ment-spe cial ist hours are higher in the SE group. Ta ble 3 brings to gether find ings on vo ca tional re ha bil i ta tion costs and health care and so cial ser vices costs. The im pact of SE on over all costs of vo ca tional re ha bil i ta tion var ies dra mat - i cally from one study to the next. At one ex treme, in the study by Rog ers and oth ers (16), a high-in ten sity form of SE was sim ply added to ex ist ing ser vices in a set ting where vo ca - tional re ha bil i ta tion ser vices were min i mal. At the other ex - treme, in the New Hamp shire day treat ment con ver sion stud ies for which de tailed data are available (18,19,38), an ex pen sive day treat ment pro gram was closed and re placed with a less costly SE pro gram. In be tween, in the New Hamp - shire ran dom ized study (40,41), 2 sim i larly ex pen sive vo ca - tional pro grams were com pared. The study by Bond and oth ers (17,39) pres ents yet a dif fer ent in ter me di ate case, where cli ents in the ac cel er ated-en try con di tion spent less time in day treat ment, re sult ing in some what lower overall net costs for vo ca tional re ha bil i ta tion services. In pa tient costs closely re flect find ings on hos pi tal days re - ported above: the eco nomic sig nif i cance of any SE ef fects on hos pi tal iza tions ap pears neg li gi ble, ex cept in the pos si ble case of the Rog ers and oth ers study (16). The pattern ob served for other service costs is con sis tent with the above ob ser va tions. In the study by Rog ers and oth ers (16), the costs of other ser vices are re duced be cause the SE spe cial ists sub sti tute for part of those ser vices. In the New Hamp shire day treat ment con ver sion stud ies for which we have cost data, day treat ment clo sure and re duced re ha bil i ta - tion staff ing ap pear to be as so ci ated with in creased case-man age ment and out pa tient costs. The 2 other stud ies for which such costs are avail able show some what lower other costs (by < 30% in both cases) as so ci ated with SE, due mostly to some what lower out pa tient costs.

16 August 2001 Economic Impacts of Supported Employment for Persons With Severe Men tal Illness 501 The last col umn of Ta ble 3 reports that in ev ery study ex cept the one by Rog ers and oth ers (16), SE is as so ci ated with nom - i nally lower ser vice costs, tak ing vo ca tional, in pa tient, and as sorted out pa tient costs into ac count. (In the study by Bond and oth ers [17,39] hos pi tal days were slightly fewer in the SE con di tion; had they been costed, the dif fer ence in fa vour of SE would have been slightly greater.) Ta ble 4 shows dif fer ences in em ploy ment earn ings, benefits, and taxes as so ci ated with SE. As ex pected given the greater suc cess of SE at in creas ing em ploy ment rates, em ploy ment earn ings are, in ev ery case ex cept one, much greater with SE than with the al ter na tives (though still very mod est when av - er aged over all cli ents). The ex cep tion is a ran dom ized study by Drake and oth ers (20), where total earn ings are al most iden ti cal. In that study, 94% of wages earned in the SE con di - tion came from com pet i tive em ploy ment (av er age hourly wage, USD5.82), com pared with 8% in the en hanced vo ca - tional re ha bil i ta tion (EVR) con di tion. Two-thirds of EVR wages came from shel tered em ploy ment, at an av er age hourly rate of USD1.00. Im pacts on gov ern ment ben e fits and on taxes are re ported in only 2 stud ies and ap pear neg li gi ble, con sis tent with the low av er age earn ings. None of the studies re viewed re port any mea sures of care giver time or ex pense, nor do they mea sure jus tice-re lated costs. Discussion Per haps the most striking ob ser va tion to emerge from this re - view, con sis tent with Clark s pre vi ous ob ser va tion made on the ba sis of the New Hamp shire studies alone (27), is the ex - tent to which the net ef fect on vo ca tional re ha bil i ta tion costs of pro vid ing SE ser vices de pends on the con text. In a setting where no, or hardly any, vo ca tional re ha bil i ta tion ser vices are pro vided at the out set, in tro duc ing an SE ser vice is likely to in crease vo ca tional re ha bil i ta tion costs, simply be cause there is no op por tu nity for sub sti tu tion. At the other ex treme, con - vert ing ex ist ing vo ca tional pro grams into sup ported em ploy - ment ap pears to al low a sig nif i cant cost re duc tion. The find ings sum ma rized here offer little hope for a sig nif i - cant re duc tion in other health care costs (such as in pa tient care) fol low ing the in tro duc tion of SE un less, per haps, the SE spe cial ist as sumes the role of case man ager in the ab sence of such ser vices. In that case, how ever, the em ploy ment spe cial - ist s ef fec tive ness is likely to be com pro mised: in a re cent study, ded i ca tion of SE spe cial ists to a vo ca tional role (as op - posed to their as sum ing a more gen er a list role that in cludes vo ca tional re ha bil i ta tion) has been found to be strongly as so - ci ated with the success of an SE pro gram (23). The ab sence of sig nif i cant re duc tions in other health care costs could, how ever, be partly at trib ut able to the way some of the pro grams were funded. In the New Hamp shire ran dom - ized study, SE ser vices were paid out of a re search grant, and there was lit tle in cen tive for either of the 2 men tal health cen ters to limit out pa tient treat ment costs (27). It is per haps also at trib ut able to the rel a tively short fol low-up pe riod of the studies. Thus, from the point of view of a hos pi tal or com mu nity cen - tre, fi nanc ing an SE pro gram without new funds ap pears un - likely un less the use of other vo ca tional pro grams can be cur tailed. The study by Bond and oth ers (17,39) found that this could be ac com plished with out con ver sion of existing ser vices. That find ing, how ever, may be spe cific to the state s fi nanc ing sys tem at that time, wherein pro vid ers billed Medicaid on a fee-for-ser vice ba sis for day treat ment and other ser vices. In Ca na dian set tings, hos pi tals and commu - nity cen tres op er ate out of global bud gets; there is un likely to be any off set ting re duc tion in vo ca tional re ha bil i ta tion costs un less, as in the New Hamp shire and Rhode Is land experi - ments, ex ist ing pro grams are con verted into SE pro grams. (Un der man aged care in the US, sim i lar in cen tives also apply [27]). The find ings re ported in Ta ble 4 sug gest that the cost of SE ser vices ranges roughly from USD2000 to USD4000 per cli - ent yearly, al low ing for in fla tion. Using a con ver sion factor that takes rel a tive pur chas ing power into ac count rather than cur rency ex change rates, this cor re sponds to about CAD2500 to CAD5000 per cli ent yearly. The wages paid to em ploy - ment spe cial ists, the amount of over head, and the ra tio of cli - ents to em ploy ment spe cial ists (which nor mally should range be tween 20:1 and 25:1 once the pro gram has reached ma tu - rity) are the main fac tors that would in flu ence where in that range the cost per cli ent is likely to fall. As suming a fairly pessimistic sce nario, if an em ploy ment spe cial ist costs CAD per year in clud ing ben e fits, with over head and other ex penses add ing an other 50%, and if the cli ent-to-staff ra tio is 20:1, the cost per cli ent works out to CAD4500. In many Ca na dian set tings, a lower fig ure may be achiev able. Given this rel a tively modest cost, hos pi tals that have day treat ment cen tres that do not serve as a sub sti tute for acute hos pi tal iza tion, but that have be come long-term re ha bil i ta - tion or day-care fa cil i ties, are likely to find con ver sions to SE fea si ble. The au thor has pre vi ously es ti mated the di rect costs of day pro grams (ex clud ing day hos pi tal iza tion and in clud ing shel tered em ploy ment) at 2 psy chi at ric hos pi - tals in Mon treal to be in the range of CAD110 to CAD140 daily (doc u ment avail able from the au thor). A suf fi ciently large such pro gram, serv ing an av er age of about 30

17 502 The Canadian Journal of Psychiatry Vol 46, No 6 par tic i pants each week day, could be con verted into an SE pro gram serv ing sev eral times that num ber of cli ents. The foregoing anal y sis ig nores pos si ble changes in costs over time, be gin ning with start-up costs. In the con text of a hos pi - tal or com mu nity cen tre with a global bud get, con vert ing hu - man re sources from one pro gram to an other will in cur mostly out side train ing costs and, pos si bly, some space conversion costs; the re duced ef fi ciency during the team de vel op ment pe riod will re sult in fewer ser vices be ing given but not in in - creased costs as such. Later, it is also pos si ble that some cli - ents will be come able to re main em ployed without fur ther vo ca tional sup port. Al though anecdotally this does oc cur, the ev i dence re mains lim ited and some what con tra dic tory (42,43). The foregoing anal y sis has pur posely adopted the budgetary per spec tive of a hos pi tal or com mu nity cen tre be cause this is the en tity most likely to de cide whether to im ple ment an SE pro gram. Nev er the less, other per spec tives are of in ter est as well. Among these, the one of greatest prac ti cal im por tance is that of the cli ents them selves, who will be more or less in ter ested in par tic i pat ing in SE. From their point of view, aside from the per sonal value of as sum ing a more use ful, so cially in te grated role, ef fects on health ben e fits and on total earn ings are par a - mount. Al though av er age im pacts noted in Ta ble 4 are small, for an in di vid ual cli ent the po ten tial in creases in earn ings, and con se quent loss of ben e fits, may be con sid er able. In the US, re cent fed eral leg is la tion (44) is ex pected to help many in di - vid u als with psy chi at ric dis abil i ties re turn to work while main tain ing their gov ern ment-spon sored health in sur ance (to the ex tent that states ad just their reg u la tions ac cord ingly). To date, how ever, con cerns that the in crease in earn ings will be short-lived but sufficient to trig ger a loss of ben e fits that will be dif fi cult to re es tab lish re main a major rea son why many cli ents in US set tings hes i tate to reg is ter in SE pro grams. Used to en joy ing phy si cian and hos pi tal in sur ance ben e fits that are in de pend ent of em ploy ment or wel fare sta tus, Cana - dian cli ents may have less rea son to fear in creases in earn ings than have their US coun ter parts. Uni ver sal pharmacare pro - grams avail able in some prov inces (no ta bly Que bec) fur ther re duce such con cerns. At the same time, the greater the share of their earn ings that cli ents can keep without losing their in - come-sup port ben e fits, the greater will be their in cen tive to work. In Que bec, ev ery ad di tional dol lar earned above a thresh old of CAD100 re sults in an al most equal re duc tion in in come sup port. With such a small gain in earn ings, virtually the only ben e fit to the cli ent of work ing more hours is the in - trin sic sat is fac tion of doing so. A de tailed dis cus sion of opti - mal pol icy in this re gard is beyond the scope of this pa per. Closely re lated to the cli ents per spec tive is that of their fam - ily mem bers. A sec ond ary anal y sis car ried out on the New Hamp shire day treat ment con ver sion study pro vides the only ev i dence on how fam ily mem bers view SE in the con text ad - dressed here: it suggests that fam ily re ac tions are on bal ance pos i tive (45). Data on care giver time and ex pense are needed to better char ac ter ize the costs and ben e fits of SE from the fam ily mem bers point of view. A fourth per spec tive of in ter est is that of the gov ern ment, which can ex er cise var i ous le vers to pro mote the de vel op - ment of SE pro grams. This is a difficult per spec tive to de fine. From a purely bud get ary point of view, any cost-neu tral or cost-re duc ing (from the point of view of the hos pi tal or com - mu nity cen tre) con ver sion of vo ca tional pro grams to SE is ir - rel e vant to the gov ern ment: it will have no di rect ef fect on bud gets al lo cated to hos pi tals or com mu nity cen tres. The lim - ited ev i dence re viewed here sug gests that re duc tions in wel - fare pay ments and in creases in tax rev e nues are likely to be im ma te rial from the point of view of gov ern ment decision mak ers. Fur ther, in creases in earn ings and, hence, in creases in taxes paid by, and re duc tions in ben e fits paid to, some in di - vid u als en abled to en ter the la bour force may re sult in di rectly in tax re duc tions and ben e fit in creases for other individuals, as some of a fi nite num ber of job va can cies are filled in a con - text of less-than-full em ploy ment. In other words, if an SE cli - ent finds a job, an other job-seeker may have to look lon ger and per haps stay on un em ploy ment lon ger. Such in di rect ef - fects, which will some times oc cur, will at ten u ate the already neg li gi ble bud get ary ben e fits to gov ern ment. The pos si ble im pacts of SE on jus tice-re lated costs have not been in ves ti - gated but, given the low costs for the av er age client (46), they are also likely to be im ma te rial. Thus, from a purely bud get - ary per spec tive, in the Ca na dian con text the de vel op ment of SE ser vices, fi nanced with out new money, is es sen tially ir rel - e vant to the gov ern ment (whether re gional, pro vin cial, or fed eral). Of course, no gov ern ment s per spec tive is so lim ited. The ben e fits to cli ents from in creased com mu nity in te gra tion are of some value, and it is up to con sumer ad vo cates to make that value ap pear greater to the rel e vant gov ern ment decision mak ers: for gov ern ments, in flu enced by po lit i cal con sid er - ations, there is no unique or ob jec tive way of as sign ing value to such in tan gi ble benefits. The most com pre hen sive per spec tive is that of so ci ety. This is the only com monly adopted per spec tive that is di rectly rel e - vant to no stake holder. But, be ing the most com pre hen sive, it is, econ o mists ar gue, the one that should guide re source al lo - ca tion de ci sions. In the pres ent con text, it differs from the gov ern ment per spec tive in 5 prin ci pal ways. First, it in cludes care giver costs. Sec ond, changes in costs over time be come rel e vant: as ex plained above, they may have no ef fect on a hos pi tal or com mu nity cen tre s bud get ary out lays, but they af fect the value of ser vices ac tu ally pro vided. As we have

18 August 2001 Economic Impacts of Supported Employment for Persons With Severe Men tal Illness 503 seen, little ev i dence is avail able on which to es ti mate these 2 dif fer ences. Third, earn ings, which are ir rel e vant from a gov ern ment bud - get ary per spec tive ex cept in so far as they af fect ben e fit pay - ments and tax rev e nues, need to be taken into ac count. How to do so re mains an un set tled ques tion among econ o mists. Many would ar gue that the wage paid to work ers must be equal (at least as long as the prod uct and labour mar kets are com pet i tive) to the value of what they have pro duced. Thus, wages mea sure the value of la bour force par tic i pa tion: that is the tra di tional hu man cap i tal ap proach. But at least one al ter - na tive point of view has been force fully ar tic u lated in re cent years. If the real gain from a per son s em ploy ment is mea - sured by the con se quent in crease in Gross Na tional Prod uct (GNP), and if the job filled is, as sug gested ear lier, one of a fi - nite num ber in a con text of less-than-full em ploy ment, the ac - tual ef fect on GNP of an SE cli ent taking a job may well be less than his or her wage (25). That is the basic con sid er ation un der ly ing what is called the fric tion-cost method for valuing changes in la bour force par tic i pa tion (47), a method that has been ap plied in at least one pre vi ous ar ti cle in this jour nal (48). Thus, in creases in earn ings are val ued at their nom i nal value or less, de pend ing on which of the 2 ap proaches is used. A fourth dif fer ence be tween the gov ern ment and so ci etal per - spec tives is that any changes in ben e fits or taxes paid are ir rel - e vant from a so ci etal per spec tive, ex cept in so far as they af fect ad min is tra tive ex penses: they in volve no net gain or loss to so ci ety, but merely a trans fer of re sources. Given that ef fects on av er age earn ings and changes in ben e fits and taxes ap pear min i mal, this dif fer ence, as well as the third, be tween the per - spec tive of the gov ern ment and that of so ci ety is un im por tant. Finally, a fifth dif fer ence be tween the gov ern ment and so ci - etal per spec tives is that the latter per spec tive seeks to assign value to in tan gi ble ben e fits in an ob jec tive and im par tial way, so that ben e fits ex pected from al ter na tive pro grams can be com pared. There are 3 ba sic ap proaches to do ing this. The first, and the only one used to date in anal y ses of SE pro - grams, is to mea sure ben e fits in terms of an out come ap pro - pri ate to the pro gram being in ves ti gated (for ex am ple, in the pres ent con text, weeks in com pet i tive em ploy ment). The sec - ond ap proach is to use a com mon met ric to eval u ate health care ben e fits, so that com par i sons of cost-ef fec tive ness across dif fer ent health care pro grams (for ex am ple, or tho pe - dic sur gery vs vo ca tional re ha bil i ta tion) can be made. The qual ity ad justed life year (QALY) is the best-known such met ric (49). QALYs have not of ten been used in psy chi at ric re search, and they have not, to the au thor s knowl edge, been used to value the ben e fits of SE. It is not known to what ex tent they would capture vari a tions in well-be ing as so ci ated with changes in em ploy ment sta tus among per sons with mental ill ness. The third method that could be used to value in tan gi ble ben e - fits is called will ing ness-to-pay. In con texts such as the pres - ent one, this in volves asking stake holders to in di cate how much they would be will ing to pay to achieve a cer tain ben e - fit: not only cli ents but also their fam ily mem bers, and even the pub lic at large, could be asked what value they would place on them selves, or on their loved one, or on more in di - vid u als with se vere men tal ill ness being in te grated into the workforce. Again, this has not yet been done, to the au thor s knowl edge. More re search needs to be done be fore the dif fer ence be tween the gov ern ment and so ci etal per spec tives can be fully as - sessed. Ex cept for the val u a tion of the ben e fit of in creased com mu nity in te gra tion, how ever, the dif fer ences ap pear likely to be small. In con clu sion, the ev i dence to date sug gests that conversion of day treat ment or step wise vo ca tional pro grams into SE pro grams would be cost-saving or cost-neu tral from the points of view of hos pi tals, com mu nity cen tres, and gov ern - ments and that such con ver sions are de sir able on the grounds of in creas ing the com mu nity in te gra tion of the population with se vere men tal ill ness. Avail able ev i dence fur ther sug - gests that in vest ments of new money into SE pro grams will not be ma te ri ally off set by re duc tions in other health care costs or in gov ern ment ben e fit pay ments or by in creased tax rev e nues. In the pres ent state of knowl edge, such in vest ments must there fore be mo ti vated by the value of in creas ing the com mu nity in te gra tion of per sons with men tal ill ness. These in fer ences, how ever, are based on only a few, lim ited cost-ben e fit studies and there fore remain pro vi sional. Fur - ther, be cause all the studies re viewed have been carried out in the US, and be cause there are ma te rial dif fer ences be tween the US and Can ada in ser vice fi nanc ing, in ben e fit rules, in leg is la tion af fect ing em ploy ers ac tions to ward the dis abled, and in other fac tors likely to in flu ence the costs and ben e fits of SE pro grams, the in fer ences are even more pro vi sional in the Ca na dian con text. Ad di tional in ves ti ga tions, sev eral of which are cur rently un der way in the US and Can ada, will help to clarify the costs and ben e fits of SE. Acknowledgements Sup port from the Agence d Évaluation des Technologies et Modes d Intervention en Santé (AETMIS) is grate fully acknowledged. The au thor also thanks Deborah Becker, Gary Bond, Robin Clark, Rob - ert Drake and William Gnam for their many helpful comments and sug ges tions, and Youcef Ouadahi for his assistance. References 1. Lehman AF. Vo ca tional re ha bil i ta tion in schizo phre nia. Schizophr Bull 1995; 21: VanDongen C. Qual ity of life and self-es teem in work ing and nonworking per sons with mental ill ness. Com mu nity Ment Health J 1996;32: Mueser KT, Becker DR, Torrey WC, Xie H, Bond GR, Drake RE, and oth ers. Work and nonvocational do mains of func tion ing in per sons with se vere mental ill ness: a lon gi tu di nal anal y sis. J Nerv Ment Dis 1997;185:

19 504 The Canadian Journal of Psychiatry Vol 46, No 6 4. Rosenfield S. Factors con trib ut ing to the sub jec tive qual ity of life of the chron i cally men tally ill. Jour nal of Health and So cial Be hav ior 1992;33: Jacobs H, Wissusik D, Collier R, Stackman D, Burkeman D. Cor re la tions be tween psy chi at ric dis abil i ties and vo ca tional out come. Hos pi tal and Com mu nity Psy chi a - try 1992;43: Gervey R, Bedell JR. Sup ported em ploy ment in vo ca tional re ha bil i ta tion. In: Bedell JR, ed i tor. Psy cho log i cal as sess ment and treat ment of per sons with severe men tal dis or ders Wash ing ton (DC): Tay lor and Fran cis; p Crow ther RE, Mar shall M, Bond GR, Huxley P. Helping peo ple with se vere mental ill ness to ob tain work: sys tem atic re view. BMJ 2001;322: Wehman P. Sup ported com pet i tive em ploy ment for per sons with se vere dis abil i - ties. Jour nal of Ap plied Re ha bil i ta tion Coun seling 1986;17: Drake RE. A brief his tory of the in di vid ual place ment and sup port model. Psy chi at - ric Re ha bil i ta tion Jour nal 1998;22(1): Becker D, Drake R. A working life: the in di vid ual place ment and sup port pro gram (IPS). Con cord (NH): Dartmouth Psy chi at ric Re search Cen ter; Becker D, Drake R. In di vid ual placement and sup port: a com mu nity mental health cen ter ap proach to vo ca tional re ha bil i ta tion. Com mu nity Ment Health J 1994;30: Stein LI, Santos AB. As ser tive com mu nity treat ment of persons with se vere mental ill ness. New York: WW Norton; p Bond GR, Becker DR, Drake RE, Rapp CA, Meisler N, Lehman AF, and others. Im ple menting sup ported em ploy ment as an ev i dence-based prac tice. Psychiatr Serv 2001;52: Bond G. Prin ci ples of the in di vid ual place ment and sup port model: em pir i cal sup - port. Psy chi at ric Re ha bil i ta tion Jour nal 1998;22(2): Wehman P. Sup ported em ploy ment: to ward zero ex clu sion of persons with severe dis abil i ties. In: Wehman P, Moon M, ed i tors. Vo ca tional re ha bil i ta tion and sup - ported em ploy ment. Bal ti more (MD): Paull Brookes; p Rog ers SE, Sciarappa K, Mac Don ald-wil son K, Danley K. A ben e fit-cost analysis of a sup ported em ploy ment model for per sons with psy chi at ric dis abil i ties. Eval u a - tion and Program Planning 1995;18: Bond GR, Dietzen LL, McGrew JH, Miller LD. Ac cel er ating en try into sup ported em ploy ment for per sons with se vere psy chi at ric dis abil i ties. Re ha bil i ta tion Psy - chol ogy 1995;40(2): Drake RE, Becker DR, Biesanz JC, Torrey WC, McHugo GJ, Wyzik PF. Re ha bil i - ta tive day treat ment vs sup ported em ploy ment: I. vo ca tional out comes. Com mu - nity Ment Health J 1994;30: Drake R, Becker DR, Biesanz JC, Wyzik PF, Torrey WC. Day treatment vs sup - ported em ploy ment for per sons with se vere men tal ill ness: a rep li ca tion study. Psychiatr Serv 1996;47: Drake RE, McHugo G, Bebout R, Becker D, Harris M, Bond G, and others. A ran - dom ized clin i cal trial of sup ported em ploy ment for in ner-city pa tients with severe men tal dis or ders. Arch Gen Psy chi a try 1999;56: Becker DR, Bond GR, Mc Car thy D, Thomp son D, Xie H, McHugo GJ, and others. Con verting day treat ment cen ters to sup ported em ploy ment pro grams in Rhode Is - land. Psychiatr Serv2001;52: Bond GR, Becker DR, Drake RE, Vogler KM. A fi del ity scale for the in di vid ual place ment and sup port model of sup ported em ploy ment. Re ha bil i ta tion Coun - seling Bulletin 1997;40: Becker DR, Smith J, Tanzman B, Drake RE, Tremblay T. Fi del ity of sup ported em - ploy ment pro grams and em ploy ment out comes. Psychiatr Serv 2001;52: Bond G, Resnick S, Drake R, Xie H, McHugo G, Bebout R. Does com pet i tive em - ploy ment im prove nonvocational out comes for peo ple with se vere mental ill ness? Jour nal of Clin i cal and Com mu nity Psy chol ogy. Forth com ing. 25. Clark RE, Bond GR. Costs and ben e fits of vo ca tional pro grams for peo ple with se - ri ous men tal ill ness. In: Moscarelli M, Rupp A, Sar to rius N, ed i tors. Handbook of men tal health eco nom ics and health pol icy. Sus sex (UK): John Wiley and Sons Ltd; p Rog ers SE. Cost-ben e fit stud ies in vo ca tional ser vices. Psy chi at ric Re ha bil i ta tion Jour nal 1997;20(3): Clark RE. Sup ported em ploy ment and man aged care: can they co ex ist? Psy chi at ric Re ha bil i ta tion Jour nal 1998;22(1): Clark RE, Dain BJ, Xie H, Becker DR, Drake RE. The economic ben e fits of sup - ported em ploy ment for persons with mental ill ness. The Jour nal of Mental Health Pol icy and Eco nom ics 1998;1: McCaughrin WB, Ellis WK, Rusch FR, Heal L. Cost-ef fec tive ness of sup ported em ploy ment. Mental Re tar da tion 1993;31(1): Weisbrod BA, Test MA, Stein LI. Al ter na tive to mental hos pi tal treat ment: II. eco - nomic ben e fit-cost anal y sis. Arch Gen Psy chi a try 1980;37: McFarlane WR, Dushay RA, Deakins SM, Stastny P, Lukens EP, Toran J, and oth - ers. em ploy ment out comes in fam ily-aided as ser tive com mu nity treat ment. Am J Orthopsychiatry 2000;70: Chandler D, Meisel J, Hu T-W, McGowen M, Mad i son K. A capitated model for a cross-sec tion of se verely men tally ill cli ents: em ploy ment out comes. Com mu nity Men Health J 1997;33: Conseil d Évaluation des Tech nol ogies de la Santé. Le suivi intensif en équipe dans la communauté pour personnes atteintes de trou bles mentaux graves. Montréal: Ministère de la Santé et des Ser vices sociaux; Clinical Implications Conversion of ex ist ing day treatment or less effective vocational re - habilitation programs into supported employment (SE) programs can be cost-neutral or cost-saving from a budgetary point of view and should be carried out in such cases. Where such conversions are not possible, development of SE pro - grams can be justified on the grounds that they promote community integration of persons with severe mental illness more effectively than do other methods currently available. Limi ta tions Results are based on a limited number of studies. Cost estimates from some of these studies may be significantly bi - ased. Results are based on US experience and may not generalize to Can - ada. 34. Kessler R, Berglund P, Zbao S, Leaf P, Kouzis A, Bruce M, and others. The 12-month cor re lates and prev a lence of se ri ous mental ill ness in mental health. Rockville (MD): De part ment of Health and Hu man Services Public Health Ser vice Sub stance Abuse and Mental Health Ser vices Ad min is tra tion Cen ter for Mental Health Ser vices; DHHS Pub li ca tion Num ber (SMA) Block L. The em ploy ment con nec tion: The ap pli ca tion of an in di vid ual sup ported pro gram for persons with chronic mental health prob lems. Ca na dian Jour nal of Com mu nity Mental Health 1992;11(2): Lati mer E. Economic im pacts of as ser tive com mu nity treat ment: a re view of the lit - er a ture. Can J Psy chi a try 1999;44: Mueser KT, Bond GR, Drake RE, Resnick S. Models of com mu nity care for severe men tal ill ness: A re view of re search on case man age ment. Schizophr Bull 1998;24(1): Clark RE, Bush PW, Becker DR, Drake RE. A cost-ef fec tive ness com par i son of sup ported em ploy ment and re ha bil i ta tive day treat ment. Ad min is tra tion and Pol icy in Men tal Health 1996;24(1): Bond GR, Dietzen LL, Vogler K, Katuin CH, McGrew JH, Miller LD. To ward a frame work for eval u at ing cost and ben e fits of psy chi at ric re ha bil i ta tion: 3 case ex - am ples. Jour nal of Vo ca tional Re ha bil i ta tion 1995;5: Drake RE, McHugo GJ, Becker DR, An thony WA, Clark RE. The New Hampshire study of sup ported em ploy ment for peo ple with se vere mental ill ness: vo ca tional out comes. J Con sult Clin Psychol 1996;64: Clark RE, Xie H, Becker DR, Drake RE. Ben e fits and costs of sup ported em ploy - ment from three per spec tives. The Jour nal of Be hav ioral Health Ser vices and Re - search 1998;25(1): McHugo GJ, Drake RE, Becker, DR. The du ra bil ity of sup ported em ploy ment ef - fects. Psy chi at ric Re ha bil i ta tion Jour nal 1998;22(1): Bond GR, Drake RE, Mueser KT, Becker D. An Up date on sup ported em ploy ment for peo ple with se vere men tal ill ness. Psychiatr Serv 1997;48: Golden TP, O Mara S, Ferrell C, and oth ers. A the o ret i cal con struct for benefits plan ning and as sis tance, in the ticket to work and work in cen tives im prove ment act. Jour nal of Vo ca tional Re ha bil i ta tion 2000;14: Torrey W, Becker D, Drake, R. Re ha bil i ta tive day treatment ver sus sup ported em - ploy ment: II. cli ent, fam ily, and staff re ac tions to a program change. Psychosocial Re ha bil i ta tion Jour nal 1995;18(3): Wolff N, Helminiak TW, Di a mond RJ. Es ti mated so ci etal costs of as ser tive com - mu nity mental health care. Psychiatr Serv 1995;46: Koopmanschap MA, Rutten FF, Ineveld BMv, Roijen Lv. The fric tion cost method for mea sur ing in di rect costs of dis ease. Jour nal of Health Eco nom ics 1995;14: Goeree R, O Brien BJ, Blackhouse G, Agro K, Goering P. The Val u a tion of pro - duc tiv ity costs due to pre ma ture mor tal ity: a com par i son of the hu man-cap i tal and fric tion-cost meth ods for schizo phre nia. Can J Psy chi a try 1999;44: Drummond MF, O Brien B, Stoddart GL, Torrance GW. Methods for the eco nomic eval u a tion of health care programmes, Sec ond Ed. To ronto: Ox ford Uni ver sity Press; Bailey EL, Ricketts SK, Becker DR, Xie H, Drake RE. Do long-term day treat ment cli ents ben e fit from sup ported em ploy ment? Psy chi at ric Re ha bil i ta tion Jour nal 1998;22(1): Danley KS, An thony WA. The choose-get-keep model. Amer i can Re ha bil i ta tion 1987;6 9:27 9.

20 August 2001 Economic Impacts of Supported Employment for Persons With Severe Men tal Illness 505 Résumé Répercussions économiques de l em ploi assisté pour les personnes souffrant de grave maladie mentale Con texte : La plupart des per son nes souf frant de grave mala die men tale préfèrent un mi lieu pro fes sion nel con cur ren - tiel plutôt que pro tégé. L em ploi as sisté (EA) est de venu un modèle bien défini pour aider les gens souf frant de grave mala die men tale à trou ver et à con server des em plois con cur ren tiels. Il fait ap pel à un place ment in di vidu al isé rap ide, à une évalua tion et à un soutien per ma nents ainsi qu à l inté gra tion de tra vail leurs de l em ploi et de la santé men tale au sein d une même équipe clin ique. Les études pré céden tes in diquent que l EA ob ti ent des em plois con cur ren tiels de façon beau coup plus ef fi cace que les autres ap pro ches. Cette étude se penche sur les réper cus sions économiques. Méth odes : Les études por tant sur l u tili sa tion des serv ices ou les résul tats fin an ciers de l a jout des pro grammes d EA ont été repé rées. Ces résul tats ont été to tal isés et sont pré sentés sous forme nar ra tive. Résul tats : Cinq études non aléa toires et 3 études aléa toires com par ent les pro grammes d EA aux pro grammes de traite ment de jour ou d em ploi de tran si tion. L in stau ra tion de pro grammes d EA peut avoir n im porte quel résul tat, d une aug men ta tion à une dimi nu tion des coûts des serv ices d em ploi, dans la me sure où ils se sub stitu ent aux services d em ploi ou de traite ment de jour anté rieurs. Les coûts glo baux des services ten dent à être infé rieurs, mais les diffé - rences ne sont pas sig ni fi ca tives. La rémuné ra tion n aug mente que lé gère ment en moy enne. Con clu sions : La con ver sion des pro grammes de traite ment de jour ou d autres pro grammes d em ploi moins ef fi caces en pro grammes d EA peut être rent able ou ne pas en gager de coûts pour l hôpi tal, le cen tre com munautaire et le gou - ver ne ment. L in ves tisse ment d ar gent frais dans les pro grammes d EA ne sera prob able ment pas sen si ble ment com - pensé par des coûts réduits de soins de santé ail leurs, par des ver se ments de presta tions réduits du gou ver ne ment ou par des re cet tes fis cales ac crues. Ces in ves tisse ments doivent être mo tivés par la va leur d une meil leure inté gra tion à la com munauté des per son nes souf frant de mala die men tale.

21 ORIGINAL RESEARCH Mild Dementia or Cognitive Impairment: The Modified Mini-Mental State Examination (3MS) as a Screen for Dementia Roger C Bland, MB, ChB, FRCPC 1, Ste phen C Newman, MSc, MD 1 Ob jec tive: To ex am ine the Mod ified Mini-Men tal State Ex am i na tion (3MS) as a screen for de men tia. Method: A group of 1092 el derly Ed mon ton com mu nity res i dents com pleted the 3MS and the Ge ri at ric Mental State Ex am i na tion (GMS). 3MS sen si tiv ity and spec i fic ity were de ter mined by com par ing posi tive 3MS screens (score 77) with those clas si fied as GMS or ganic (se ver ity level 3, equiv a lent to a clin i cal di ag no sis). In the Ca na dian Study of Health and Aging (CSHA), 2914 subjects re ceived the 3MS and a clin i cal ex am i na tion. A group de scribed as hav ing cog ni tive im pair ment but no de men tia (CIND) was identified. Re sults: In Ed mon ton, the 3MS showed 88% sen si tiv ity, 90% spec i fic ity, 29% pos i tive pre dic tive value (PPV), and 99% neg a tive pre dic tive value (NPV). In the CSHA, 30% of sub jects re ceiv ing both the 3MS and a clin i cal ex am i na tion were clas si fied as CIND. One-half of these were clas si fied as having age as so ci ated mem ory im pair ment (AAMI) or as un spec i fied. Con clu sions: The 3MS with a cut ting score of 77/78 proved a rea son able screening in stru ment; 1 case in 3 screen ing pos i tive has de men tia, but few (0.64%) will be missed by screening neg a tive. CIND, ac count ing for 2 out of 3 cases screened pos i tive by the 3MS in the Ed mon ton study, is a sub stan tial, het er o ge neous group that is not nec es sar ily predementia but that in many cases merits fur ther in ves ti ga tion. (Can J Psy chi a try 2001;46: ) Key Words: dementia screening, dementia epidemiology, Modified Mini-Mental State Examination (3MS), Geriatric Mental State Examination (GMS) Dementias are pri mar ily dis eases af fect ing older age groups; they are likely to have the most im pact in the es - tab lished mar ket econ o mies, where life ex pec tancy is the great est, ap proach ing 80 years. In Can ada, as sum ing con stant prev a lence, ac tual cases are pro jected to in crease more than 3-fold, from in 1991 to in 2031, an in crease en tirely due to de mo graphic change (1). To ex am ine the Mod ified Mini-Men tal State Examination (3MS) ef fec tive ness (2) as a screen ing test for de men tia, we used a com mu nity-based ran dom sam ple of el derly sub jects. The gold stan dard for the di ag no sis of de men tia was a Manuscript received October 2000, revised, and accepted April Professor, Department of Psychiatry, University of Alberta, Edmonton, Al - berta. Address for correspondence: Dr RC Bland, Department of Psychiatry, 1E7.07 Mackenzie Centre, 8440B112 Street, Edmonton, AB T6G 2B7 Roger.Bland@ualberta.ca di ag no sis of or ganic on the Ge ri at ric Mental State Ex am i - na tion (GMS) (3 5) with a se ver ity level of 3, equiv a lent to a clin i cal di ag no sis. Classifying Dementia While little dif fi culty is ex pe ri enced in de tect ing se vere and ad vanced de men tia cases, the prob lems of early de tec tion re - main un re solved and arise at sev eral lev els (6 8). The DSM-III-R, ICD-10, and DSM-IV clas si fi ca tion sys tems are not con sis tent. DSM-III-R de scribes 2 cat e go ries of mild cog - ni tive im pair ment: type 1, short- and long-term mem ory im - pair ment with no func tional dis abil ity; and type 2, short-and lon ger-term mem ory im pair ment with no func tional dis abil - ity but with at least 1 of sev eral other def i cits. ICD-10 de - scribes 3 cat e go ries, also called mild cog ni tive im pair ment: type 1, short- or lon ger-term mem ory im pair ment with no func tional dis abil ity; type 2, short- or long-term mem ory im - pair ment and in tel lec tual de cline with no func tional dis abil - ity; and type 3, short-or long-term mem ory im pair ment with Can J Psychiatry, Vol 46, August

22 August 2001 Mild Dementia or Cognitive Impairment 507 in tel lec tual de cline and per son al ity change but with no func - tional dis abil ity. DSM-IV in tro duces the terms mild neurocognitive dis or der, which re quires a neu ro log i cal or gen eral med i cal eti ol ogy, and age-re lated cog ni tive de - cline, a de cline in func tion ing that is within nor mal lim its for age and is not due to a gen eral med i cal con di tion or other men tal dis or der. The term age-as so ci ated mem ory im pair ment (AAMI) has es tab lished cri te ria spec i fy ing per for mance 1 SD be low the young adult mean on 1 or more tests. Other terms used in - clude age-con sis tent mem ory im pair ment, late-life for get - ful ness, and age-as so ci ated cog ni tive de cline (9). The prob lems with these clas si fi ca tions and their cri te ria in clude com plex ity, re li ance on tests with out ad e quate norms, vague def i ni tions of mem ory im pair ment, and in con sis tent ap pli ca - tion in re search all lead ing to poor com pa ra bil ity of re sults. Prognostic Significance Table 1. Canadian study of health and aging subcategories of Cognitive Impairment But No Dementia (CIND) a If early cog ni tive de cline is be nign and those af fected do not pro ceed to de men tia, this is clearly im por tant. Con versely, if early cog ni tive de cline ac cu rately de tects early de men tia, those af fected would be suit able subjects for early in ter ven - tions, when op por tu ni ties arise. In pro spec tive stud ies, sev - eral reports have shown that sub jects iden ti fied with mild cog ni tive im pair ment proceed in most cases to de men tia af ter a 2- to 7-year fol low-up (10 14). Nev er the less, even in those stud ies that show a high per cent age of cases pro gress ing from mild cog ni tive im pair ment to de men tia, at least one-third do not prog ress. Al though a re cent study found 18.5% of the pop u la tion over age 50 years to have AAMI, it would seem be yond the realm of pos si bil ity that such a high pro por tion of this pop u la tion is in the stage of predementia, nor do the au - thors sug gest it (15). Förstl and oth ers re ported a pro spec tive study show ing that the dis tinc tion be tween AAMI and Alz - hei mer s dis ease (AD) did not pre dict cog ni tive de te ri o ra tion dur ing a 2-year fol low-up (16). n % Age-associated memory impairment (AAMI) Cerebrovascular Depression General vascular Psychiatric or delirium Alcohol or drug abuse Other specified Unspecified a Adapted from (17) Cognitive Impairment But No Dementia (CIND) Dur ing the course of the Ca na dian Study on Health and Aging (CSHA), 2914 sub jects re ceived both the 3MS and a clin i cal ex am i na tion (1). Of this group, 30% were iden ti fied as having CIND and were fur ther ex am ined to de ter mine the cause of cog ni tive im - pair ment (17) (Ta ble 1). One-half of this group were clas si fied ei ther as hav ing AAMI or as un spec i fied. CIND was de ter mined as fol lows: de - men tia cri te ria were not met, a cli ni - cian iden ti fied a mem ory def i cit, there was a def i cit in 1 other area of cog ni - tion, the def i cit did not se ri ously in ter fere with daily liv ing, and there was a clin i cal im pres sion that some thing was go - ing on. Methods Sub jects Trained lay in ter view ers in ter viewed a stratified ran dom sam ple of 1092 com mu nity res i dents age 65 years or over from the city of Ed mon ton, Al berta, (pop u la tion over ). The sam ple was ob tained using pro vin cial health care reg is tra tion data that in clude more than 99% of the popu - la tion. The re sponse rate was 82.3% of the el i gi ble sub jects. Strat i fi ca tion was by age to oversample the old est groups. In stru ments All sub jects were ad min is tered the 3MS (2) and the GMS (3 5). The 3MS has dem on strated su pe ri or ity to the more pop u lar Mini-Men tal State Ex am i na tion (MMSE); it shows better va - lid ity in iden ti fy ing de men tia and all lev els of cog ni tive im - pair ment. The ar eas un der the re ceiver op er at ing char ac ter is tic (ROC) curve were sig nif i cantly greater at 0.94 for the 3MS, com pared with 0.89 for the MMSE. The 3MS in - cor po rates ad di tional ques tions in 4 ar eas (personal in for ma - tion, ver bal flu ency, ab strac tion, and long-term re call) and more im por tant, uses an ex panded scor ing system. The cut - ting score used on the 3MS was 77/78. This was the cutting score de ter mined in the CSHA to pro vide op ti mum sen si tiv - ity (87%) and spec i fic ity (89%) for de tect ing de men tia (18). The GMS is an in stru ment that has un der gone pro gres sive de - vel op ment over many years and has been widely used in both clin i cal and com mu nity stud ies. The ver sion used (version A3) was com puter-ad min is tered by trained in ter view ers. The di ag no sis is de rived using the com put er ized di ag nos tic sys - tem, AGECAT, not clin i cal judg ment. Current symp toms are col lected into 8 syn drome clus ters, 1 of which is the or ganic

23 508 The Canadian Journal of Psychiatry Vol 46, No 6 years and 3.5% in those age 75 to 84 years to 14.3% in those age 85 years and over. The prevalence of CIND for the comparable age groups was 3.8%, 10.4%, and 22.0%. Thus, it may be seen that the increase with age of both GMS organic and CIND is almost logarithmic, but with a steeper slope for the GMS organic classification. At all ages, the prevalence of CIND is 2 to 3 times greater than the prevalence of GMS organic (Figure 1). Figure 1. GMS Organic and CIND by age, age-specific prevalences, both sexes (n = 1092). Figure 2 shows the statistics comparing the 3MS as a screening test with GMS organic as the gold standard. The sensitivity is 88%, specificity 90%, positive predictive value (PPV) 29%, and negative predictive value (NPV) 99%; the chance of having the illness (dementia) if the 3MS test is negative is 0.64%. Discussion cluster. This cluster includes the dementias but does not differentiate the various etiologies. Each cluster is assigned a level of confidence, a measure of diagnostic certainty on a scale of 0 to 5, where 3 is the level of clinical significance that would ordinarily require intervention. A second level of analysis with the AGECAT introduces hierarchies of diagnosis, but this was not used in our study. Diagnoses The diagnosis of dementia was based on a GMS AGECAT classification of organic at a severity level of 3, the level that is considered to be the equivalent of a clinical diagnosis. For this study, the diagnosis of CIND was defined as scoring 77 or less on the 3MS and not being classified as GMS organic. Note that this is somewhat simpler than the definition used in the CSHA, wherein there was direct clinical examination. Analyses The prevalence of GMS organic disorders and of CIND as defined above was calculated for each elderly age group. The test characteristics of the 3MS compared with the GMS were also calculated. Results The response rate from the eligible subjects was 82.3%; 59.6% were women, 40.4% were men, and 53.0% of the respondents were married. Further descriptions of the methods are available in earlier communications (19,20). The overall weighted prevalence of GMS organic in the population age 65 years and older was 2.9% (men 2.5%, women 3.2%); this varied from a low of 1.4% in those age 65 to 74 At 2.9%, the prevalence of dementia (GMS organic) in this study is slightly lower than the prevalence found in the CSHA community sample from the Prairies, which was 3.7% (4.8% for men, 2.9% for women). This CSHA sample included subjects from 4 centres, not just Edmonton. Neither the prevalence found using the GMS in this study (2.9%) nor that found in the CSHA represents the population prevalence, because institutionalized people were not part of these samples. It is obvious that a considerable number of the elderly show cognitive impairment but fail to meet criteria for dementia. Follow-up studies have demonstrated that a proportion of these subjects will likely develop dementia. In all studies, however, a considerable number do not seem to progress to dementia. The designation of CIND for those who show cognitive impairment without any additional evidence of dementia seems a useful concept: it may indicate the need for further clinical investigation, and certainly, in some cases, it predicts future dementia. It should be noted that the term CIND was also used in the CSHA. Although the concept is similar in both studies, the operational definitions (as given above) differ. Direct comparisons of prevalences of CIND between the studies must therefore be made cautiously. Like many other studies, the Edmonton study clearly demonstrated that the prevalence of dementia increases almost exponentially with age, but so does CIND. The prevalence of CIND is at all ages greater than that of dementia by a factor of 2 or 3. As did the CSHA, the Edmonton study illustrated that the 3MS (using the cutting score of 77/78) is a reasonable

24 August 2001 Mild Dementia or Cognitive Impairment 509 Clinical Implications The Modified Mini-Mental State Examination (3MS) is a highly sensitive instrument for the detection of dementia. The 3MS detects cases with cognitive impairment but no dementia (CIND) that warrant further investigation. The 3MS is easily used as a screen ing instrument. Limitations The 3MS does not differentiate the causes of dementia. Only one-third of those screening positive for cognitive impairment on the 3MS will actually have dementia. The 3MS takes slightly longer to administer than the Mini- Mental State Examination (MMSE). mem ory com plaints should be eval u ated and pa tients fol - lowed up to as sess pro gres sion when care givers or in for mants de scribe cog ni tive de cline, these ob ser va tions should be taken se ri ously, with cog ni - tive as sess ment indicated. Figure 2. Edmonton Community Study 3MS 77 as the screening test, GMS organic as the gold standard. screen ing in stru ment for de tect ing cog ni tive im pair ment and de men tia. That this in stru ment is readily avail able, takes only a few min utes to ad min is ter, and can be ad min is tered by trained nonclinicians all in di cate that it is sat is fac tory for screen ing gen eral el derly pop u la tions. It can eas ily be admin - is tered in gen eral prac tice set tings, in pub lic health clin ics, or in clinics for the el derly. A screening in stru ment for the popu - la tion should be sen si tive in the de tec tion of the mild or early cases that may be found out side clin i cal set tings; it should be noted that the Ed mon ton study used a ran dom sample of com - mu nity-res i dent el derly and not a clin i cal sam ple, which would have in tro duced a bias to ward more se vere cases. The results of the Ca na dian Con sen sus Con fer ence on De - men tia were pub lished in 1999 (21) and, with ref er ence to screen ing and case finding for de men tia, con cluded that there was in suf fi cient ev i dence to rec om mend for or against screening in the ab sence of symp toms there was in suf fi cient ev i dence for or against screening with short mental sta tus ques tion naires in un se lected older peo ple given the bur den of de men tia for older people and their care givers, fam ily phy si cians should have a high in dex of sus pi cion for de men tia and fol low up con cerns re gard ing func tional de cline and mem ory loss These com ments are largely based on ex pe ri ence using the MMSE in gen eral pop u la tions of 65- to 74-year-old peo ple. The con sen sus group stated that the MMSE has an av er age sen si tiv ity of 83% and an av er age spec i fic ity of 82%, with a false-pos i tive rate of 93%. Ob vi ously, these results are less sat is fac tory than those which we ob tained using the 3MS. Nev er the less, we found a false-pos i tive rate of 71%, and this could cre ate un nec es sary anx i ety if ran dom pop u la tion screen ing were to be un der taken, not to mention pos si bly ex - pen sive and time-con sum ing un nec es sary further in ves ti ga - tion. Against this must be weighed the very low pro por tion of false neg a tives (0.64%) and the con sid er ation that must be given to causes of cog ni tive im pair ment other than de men tia; these may need in ves ti ga tion. It should be noted that the Con - sen sus Con fer ence group ques tioned the use ful ness of identi - fy ing CIND, seem ingly on the ba sis that its natural his tory is un clear. The group states, how ever, that an nu ally 5% to 6% of CIND sur vi vors prog ress to de men tia. More over, Ebly, Ho - gan, and Parhad have clearly shown that about one-half of CIND cases are due to causes that re quire further in ves ti ga - tion and treat ment (17). Thus, we ar gue that it is both prac ti cal and useful to de tect cases of CIND. We also rec om mend that in rou tine clin i cal prac tice there are good rea sons to re place the orig i nal MMSE with the 3MS. We dem on strated that the 3MS has a high sen si tiv ity in de - tect ing de men tia cases in the el derly pop u la tion, with only a 0.64% chance of the ill ness being present if the test is neg a - tive. Of those screening pos i tive on the 3MS, how ever, only 1 in 3 in the gen eral el derly pop u la tion will ac tu ally have an ill - ness clas si fi able as a de men tia (the false-pos i tive rate was 71%). This con sid er ably nar rows the field of those who may need to be in ves ti gated fur ther. Par tic u lar at ten tion should be paid to pos si ble vas cu lar causes, to psy chi at ric ill ness

25 510 The Canadian Journal of Psychiatry Vol 46, No 6 in clud ing de pres sion or de lir ium, and to sub stance use in those clas si fied as CIND; that is, those falling be low the cut-off on the 3MS but failing to meet cri te ria for a di ag no sis of GMS or ganic. Acknowledgment This study was funded in part by the National Health Research De - vel op ment Pro gram. References 1. CSHA Working Group. Ca na dian study of health and ag ing. Study meth ods and prev a lence of de men tia. Can Med Assoc J 1994;150: Teng EL, Chui HC. The mod i fied mini mental state (3MS) ex am i na tion. J Clin Psy - chi a try 1987;48: Cope land JR, Kelleher MJ, Kellett JM, Gourlay AJ, Gurland BJ, Fleiss JL, and oth - ers. A semi-struc tured clin i cal in ter view for the as sess ment of di ag no sis and men tal state in the el derly: the ge ri at ric mental state sched ule. I. De vel op ment and re li abil - ity. Psychol Med 1976;6: Cope land JR, Dewey ME, Griffiths-Jones HM. A com put er ized psy chi at ric di ag - nos tic sys tem and case no men cla ture for el derly sub jects: GMS and AGECAT. Psychol Med 1986;6: Cope land JR, Dewey ME, Wood N, Searle R, Davidson IA, McWilliam C. Range of men tal ill ness among the el derly in the com mu nity: prev a lence in Liv er pool us - ing the GMS-AGECAT pack age. Br J Psy chi a try 1987;150: Henderson AS, Huppert FA. The problem of mild de men tia. Psychol Med 1984;14: Kay DWK, Henderson AS, Scott R, Wil son J, Rickwood D, Grayson DA. De men tia and de pres sion among the el derly living in the Hobart com mu nity: the ef fect of the di ag nos tic cri te ria on the prev a lence rates. Psychol Med 1985;15: Mowry BJ, Burvill PW. A study of mild de men tia in the com mu nity using a wide range of di ag nos tic cri te ria. Br J Psy chi a try 1988;153: Blackford RC, LaRue A. Cri te ria for di ag nos ing age-as so ci ated memory im pair - ment: pro posed im prove ments from the field. De vel op men tal Neurosychology 1989;5: Katzman R, Aronson M, Fuld P, Kawas C, Brown T, Morgenstern H, and others. De vel op ment of de ment ing ill nesses in an 80-year old vol un teer co hort. Ann Neurol 1989;25: Ru bin EH, Morris JC, Grant EA, Vendegna T. Very mild se nile de men tia of the Alz hei mer type. Arch Neurol 1989;46: Flicker C, Ferris SH, Reisberg B. Mild cog ni tive im pair ment in the el derly. Neu rol - ogy 1991;41: Pe ter son RC, Smith GE, Tangalos EG, Kokmen E, Ivnik J. Lon gi tu di nal out come of pa tients with a mild cog ni tive im pair ment. Ann Neu rol ogy 1993;34: O Connor DW, Pollitt PA, Jones BJ, Hyde JB, Fellowes JL, Miller ND. Con tinued clin i cal val i da tion of de men tia di ag nosed in the com mu nity us ing the Cam bridge men tal dis or ders of the el derly ex am i na tion. Acta Psychiatr Scand 1991;83: Barker A, Jones R, Jennison C. A prev a lence study of age as so ci ated mem ory im - pair ment. Br J Psy chi a try 1995;167: Förstl H, Hentschel F, Sattel H, Gei ger-kabisch C, Besthorn C, Czech C, and oth - ers. Age as so ci ated mem ory im pair ment and early Alz hei mer s dis ease: only time will tell the dif fer ence. Arzneimittel-Forschung 1995;45: Ebly EM, Ho gan DB, Parhad IM. Cog ni tive im pair ment in the nondemented el - derly: re sults from the Ca na dian study of health and ag ing. Arch Neurol 1995;52: McDowell I, Kristjansson B, Hill GB, Hebert R. Com mu nity screening for de men - tia: The mini mental state exam (MMS) and mod i fied mini men tal state exam (3MS) com pared. J. Clin Epidemiol 1997;50: Newman SC, Shel don CT, Bland RC. Prev a lence of de pres sion in an el derly com - mu nity sam ple: a com par i son of GMS-AGECAT and DSM-IV di ag nos tic criteria. Psychol Med 1998;8: Newman SC, Bland RC, Orn HT. The prev a lence of mental dis or ders in the el derly in Ed mon ton: a com mu nity sur vey us ing the GMS-AGECAT. Can J Psy chi a try 1998;43: Patterson CJS, Gauthier S, Berg man H, Co hen CA, Feighner JW, Feldman H, and oth ers. The rec og ni tion, as sess ment and man age ment of de ment ing dis or ders: con - clu sions from the Ca na dian con sen sus con fer ence on de men tia. Can Med Assoc J 1999;160 (Suppl 12): 1S 15S. Résumé Démence légère ou déficit cognitif : le mini-examen modifié de l état mental (3MS) pour le dépistage de la démence Ob jec tif : Étud ier le mini- examen modifié de l état men tal (3MS) comme in stru ment de dépistage de la démence. Méth ode : Un groupe de per son nes âgées de la ré gion d Ed mon ton ont répondu au 3MS et à l ex amen de l état men tal gé ri atri que (GMS). La sen si bil ité et la spé ci fic ité du 3MS ont été dé ter minées en com par ant les scores po si tifs du 3MS (score 77) avec ceux clas sés comme étant «or ganiques» du GMS (ni veau de gravité 3, l équiva lent d un diag - nos tic clin ique). Dans l Étude sur la santé et le vieillis se ment au Can ada (ESVC), su jets ont répondu au 3MS et subi un ex amen clin ique. On y a dis tin gué un groupe dé crit comme ayant «un déficit cog ni t if mais sans démence (DCSD)». Résul tats : À Ed mon ton, le 3MS a démontré une sen si bil ité de 88 %, une spé ci fic ité de 90 %, une va leur prédic tive posi - tive (VPP) de 29 % et une va leur prédic tive né ga tive (VPN) de 99 %. Dans l ESVC, 30 % des su jets ayant répondu au 3MS et subi un ex amen clin ique ont été clas sés DCSD. La moitié de ces der ni ers ont été clas sés comme ayant un trou ble de la mémoire lié à l âge (TMLA) ou «non spé ci fié». Con clu sions : Avec une note tron quée de 77/78, le 3MS s est révélé un in stru ment de dépistage rai son na ble. Un cas sur 3 ayant une note «posi tive» souf fre de démence, mais quelques- uns (0,64 %) ne sont pas dé tectés à cause d un DCSD «né ga tif», ce qui re pré sente 2 cas sur 3 qui ont eu un dépistage po si tif par le 3MS dans l étude d Ed mon ton. Il s agis - sait d un groupe im por tant et hé té rogène qui n est pas né ces saire ment au stade de la «pré- démence» mais qui, dans bien des cas, mé rite un ex amen ap pro fondi.

26 ARTICLE DE SYNTHÈSE Schizophrénie et psychothérapies cognitivo-comportementales Amal Abdel-Baki, MD, FRCPC 1, Luc Nicole, MD, FRCPC 2 Objectif : Différencier les approches de psychothérapie cognitivo-comportementale pour la schizophrénie selon leurs buts, objectifs et tech niques, puis ex poser les études d efficacité. Méthode : Synthèse de l information recueillie au moyen de recherches électroniques ( MEDLINE, PSYchlit) et bibliographiques. Résultats : Les thérapies cognitivo-comportementales ont pour objectif global une meilleure ad ap ta tion à l expérience psychotique au point de vue cognitif, comportemental et affectif, en proposant au pa tient un nou veau modèle explicatif de la psychose, soit le modèle vulnérabilité-stress. Ces approches comportent différents niveaux et visent différents buts. Certaines sont axées sur la cor rec tion des déficits cognitifs de base ou sur la mod i fi ca tion des symptômes psychotiques ou de la détresse associée. À l autre extrémité, les approches métacognitives visent la mod i fi ca tion et la restructuration des schèmes du soi et de l environnement dysfonctionnels pour faciliter l acquisition de stratégies cognitives plus adaptées et généralisables. L efficacité de ces thérapies a été démontrée par quelques études dont la puis sance et la méthodologie sont limitées. Con clu sion : Les thérapies cognitivo-comportementales se révèlent un traitement adjuvant prometteur dont l efficacité a été démontrée en ce qui concerne l amélioration de l ajustement so cial, la qualité de vie ainsi que la dim i - nu tion des symptômes psychotiques et de la détresse entraînée par l expérience de la psychose. Elles visent à la fois les symptômes positifs, négatifs, cognitifs, comportementaux, et affectifs en ten ant compte de la phase de la maladie et des besoins particuliers du pa tient. Des recherches fu tures devront permettre de préciser la durée, la fréquence idéale d administration et la spécificité de ces approches. (Rev can psychiatrie 2001;46 : ) Mots clés : Schizophrénie, psychose, trouble psychotique, psychothérapie, psychothérapie cognitivocomportementale, psychothérapie cognitivo-béhaviorale, psychothérapie cognitive, traitement Introduction Globalement, les thérapies cognitivo-comportementales ont pour objectif une meilleure ad ap ta tion à l expérience psychotique, au point de vue cognitif, comportemental et affectif. Ces thérapies reposent en bonne partie sur l adoption par le pa tient d un nou veau modèle explicatif de la psychose, soit le modèle vulnérabilité-stress. Le constructivisme est une théorie selon laquelle l humain construit activement sa représentation du monde à travers ses expériences de vie. Cette représentation détermine la Manuscrit réçu en juin 1999; révisé, accepté en juin Psychiatre, CHUM, Chargé d enseignement clinique, Université de Montréal, Montréal (Québec). 2 Psychiatre, Hôtel-Dieu de Lévis, Lévis (Québec), et professeur de clinique, Université Laval, Qué bec (Québec). Adresse de correspondance : Dr Luc Nicole, Hôtel-Dieu de Lévis, 143, rue Wolfe, Lévis (Québec) G6V 3Z1 Courriel : amalbaki@hotmail.com sig ni fi ca tion qu il accordera à ses expériences. C est le cadre conceptuel sous-jacent à toute démarche cog ni tive. Perris souligne que les thérapies cognitives comportent différents niveaux (Tab leau 1) et visent différents buts (1). Il décrit un con tin uum qui contient, à une extrémité, les approches axées sur la cor rec tion des déficits cognitifs de base ou la mod i fi ca - tion des symptômes spécifiques. À l autre extrémité, Perris décrit les approches métacognitives qui visent la mod i fi ca - tion et la restructuration des schèmes du soi et de l environnement dysfonctionnels (Fig. 1). Un pre mier type de démarche est axé sur une meilleure maîtrise des déficits cognitifs et l amélioration de la gestion des émotions provoquées par des facteurs de stress internes et externes. Ces approches misent sur une restauration des apti - tudes à traiter l information. Le programme intégratif de thérapies psychologiques (IPT) de Brenner (2,3) et son équipe et l Emotional man age ment ther apy (EMT) de Hodel (4) axent donc le tra vail thérapeutique sur le processus de la pensée. Rev can psychiatrie, vol 46, août

27 512 La Revue canadienne de psychiatrie vol 46, no 6 Tableau 1. Les trois niveaux où peuvent se situer les troubles de la cognition Niveau cognitif Définition Événement cognitif Expé ri ence sub jec tive des évé ne ments. C est le con tenu de la pensée con sti tué en tre autres des pensées auto ma tiques et des dé li - res. Processus de la pensée Mé can isme par le quel on perçoit les évé ne - ments. C est la manière dont l in for ma tion est traitée, le siège des er reurs de la ges tion de l in for ma tion ainsi que des dis tor sions cogni - tives. Plusieurs études ont de mon tré des défi cits de la vigi lance et de la mémoire ver bale chez les pa tients at te ints de schizo phré nie (5). Schéma cognitif (ou structure cognitive) Système de clas si fi ca tion des stimuli ap pris au fil de l expé ri ence de vie. C est la base des croy ances selon lesquelles on donne une sig - ni fi ca tion aux évé ne ments et qui dé ter mi nent la per cep tion qu on a de soi- même et des autres. Un deuxième type de thérapie vise plus particulièrement à diminuer l intensité et la sévérité de la symptomatologie psychotique (délires, hal lu ci na tions) ou la détresse, l anxiété et la dépression qui y sont liées. Ces thérapies tentent ainsi de mod i fier le contenu dysfonctionnel de la pensée (événements cognitifs) ou la réponse af fec tive ou comportementale associée. Les auteurs en sont principalement les équipes britanniques de Kingdon et Turkington (6), Chadwick et Birchwood (7), et Tarrier (8). Finalement, un troisième type de thérapie met l accent sur un tra vail psychothérapeutique ciblant les struc tures ou schèmes cognitifs ainsi que sur l image et l estime de soi en rap port avec l expérience psychotique. Ce tra vail portant sur l identité vise à diminuer la vulnérabilité personnelle au stress et à permettre d acquérir une meilleure capacité de gestion des émotions. Ces thérapies sont la COPE (cognitively ori ented psy cho ther apy for early psy cho sis) mise au point par l équipe de McGorry (9) et la «thérapie personnalisée» élaborée par l équipe de Hogarty (10). L équipe de Fowler (11) a créé une approche plus intégrative qui s adresse tant aux schémas cognitifs (estime et im age personnelle) qu aux événements cognitifs (symptômes psychotiques et réponse associée). L efficacité de ces thérapies, en as so ci a tion avec un traitement antipsychotique, a été démontrée par quelques études dont la puis sance et la méthodologie sont limitées. En effet, la durée du suivi, les di ag nos tics des pa tients et les étapes de la maladie sont vari ables d une étude à l autre, et aucune ne com pare deux thérapies cognitivo-comportementale en tre elles. Néanmoins, ces études con stit u ent un pre mier pas dans l avancement de cette prometteuse approche adjuvante au traitement actuel. Le texte qui suit résume les différentes approches et ex pose brièvement les études cas-témoin qui s y rattachent, en s attardant uniquement aux résultats significatifs sans les critiquer. Démarches axées sur le processus de la pensée Le programme IPT a été élaboré par Brenner (3,4) et son équipe en Suisse, en Répandu en Eu rope, il est utilisé en Californie et depuis 1996, dans la région de Québec (12). L hypothèse de base en est que l amélioration de déficits dans les processus cognitifs, qui contribuent à la vulnérabilité du pa tient, permettra l acquisition d aptitudes plus com plexes. L objectif est donc de permettre au pa tient une meilleure per - cep tion et maîtrise des déficits cognitifs (p. ex., difficulté d attention), des facteurs de stress internes et externes de même que des dysfonctions comportementales qui ont aussi un effet sur la cog ni tion. Ainsi, le pa tient développe des stratégies de sta bili sa tion du soi. Le but est d interrompre les cercles vicieux établis d une part à l intérieur des processus cognitifs perturbés, et d autre part, dans les systèmes de rétroaction pos i tive en tre les dysfonctions cognitives et les facteurs de stress psychosociaux. La thérapie est donc centrée sur les dysfonctions du traitement de l information et leurs répercussions sur le fonctionnement so cial. Le pa tient est intégré dans un programme d entraînement hautement structuré selon un ordre hiérarchique de 5 sous-programmes. La programme se fait selon des critères de complexité cog ni tive et d intégration de facteurs émotionnels et sociaux (Fig. 2). Un sixième sous-programme, «Gestion des émotions», a été modifié et est maintenant considéré par Hodel comme une approche psychothérapeutique distincte. Les 5 sous-programmes de l IPT s enchaînent selon la séquence suivante : 1. Différenciation cog ni tive : le pa tient reçoit un entraînement qui tou che la pensée conceptuelle (for ma tion de classes et con cepts), la capacité d abstraction et la mod u la tion de con cepts. La dis crim i na tion cri tique en tre l essentiel et l accessoire est abordée. L exercice se fait en trois rubriques : le classe ment de cartes selon des critères pour exercer l attention et la mémorisation; l élaboration de systèmes de con cepts verbaux (p. ex., synonymes, antonymes, anal o gies, définition de mots, hiérarchie sémantique, con cepts à sig ni fi ca tions différentes selon le contexte); l utilisation de stratégies de re cher che de con cepts (par un jeu de ques tions). 2. Per cep tion sociale : exercice portant sur la dis crim i na tion et l interprétation des stim uli (à l aide de diapositives). 3. Com mu ni ca tion verbale : vise l altération de la com mu ni - ca tion. Le but est de re specter la con tri bu tion de chacun des par tic i pants et d essayer de saisir le raisonnement d autrui.

28 Août 2001 Schizophrénie et psychothérapies cognitivo-comportementales 513 Ac cent Focus sur sur les leschèmes cognitifs centraux Approches métacognitives Modification et restructuration des schèmes dysfonctionnels de soi et des autres Réhabilitation cognitive et promotion de la compétence interpersonnelle Modification et réduction des expériences hallucinatoires et délirantes Approches moléculaires Détection et correction des déficits cognitifs de base Accent Focussur surles lesévé événements cognitifs de base et et sur le pro ces sus de la pensée sur le processus de la pensée Figure 1. Psychothérapies cognitives pour les patients atteints de schizophrénie (Adapté de Perris, 1994) 4. Entraînement aux compétences sociales : dans différents contextes, par des jeux de rôle préparés en groupe par les pa tients. 5. Résolution de problèmes interactionnels : afin de faciliter et d optimiser la gestion des facteurs potentiels de stress. Suite à différentes études d efficacité, les auteurs ont précisé les in di ca tions suivantes : les 3 pre miers sous-programmes (différenciation cog ni tive, per cep tion sociale, com mu ni ca - tion verbale) s adressent davantage aux pa tients présentant des trou bles cognitifs prononcés, des peurs sociales considérables, une symptomatologie négative, une mo ti va - tion thérapeutique fai ble et une longue durée d hospitalisation. Les trois derniers sous-programmes, qui visent l amélioration des compétences sociales (com mu ni ca - tion verbale, compétences sociales et résolution de problème), sont suggérés pour les jeunes pa tients caractérisés par un man que d habileté à com poser avec les sit u a tions sociales, une mo ti va tion plus grande et plusieurs hos pi ta li sa - tions de brève durée. L IPT a été évalué par plusieurs études cas-témoin(3,13). Les résultats sont résumés au Tab leau 3. La thérapie EMT fut mise au point en Suisse, en 1995 par Hodel et son équipe (5). Cette thérapie est is sue du tra vail effectué à partir de l IPT. L hypothèse de base sup pose que les personnes atteintes de schizophrénie sont plus lentes et moins précises dans l identification des stim uli émotionnels, surtout si les stim uli environnants sont com plexes ou stressants. Cette incapacité à détecter ou décoder les in di ces essentiels pour reconnaître l intensité des af fects (surtouts négatifs) serait reliée à une dysfonction du traitement de l information. Comme la vulnérabilité des jeunes gens atteints de schizophrénie mène indirectement au retrait so cial et à l isolement, les oc ca sions d acquérir des compétences sociales sont moindres, ce qui provoque le cercle vicieux du retrait so cial. Le but est donc de raffiner la re con nais sance efficace des émotions et les habiletés à les gérer, pour réduire l effet de la détresse et de l anxiété sur le traitement de l information chez les personnes atteintes de psychose à un stade précoce. L EMT comporte 2 sous-programmes : 1. L apprentissage de tech niques de re lax ation : s effectue en thérapie individuelle afin d éviter l anxiété provoquée par le

29 514 La Revue canadienne de psychiatrie vol 46, no 6 Résolution de problèmes interactionnels Gestion des émotions Charge émotionnelle Interaction de groupe Compétences sociales Communication verbale Perception sociale Différentiation cognitive Figure 2. Représentation schématique du programme IPT (adapté de Bren ner et coll., 1990) groupe et de permettre l introduction pru dente de la prochaine modalité. 2. Le sous-programme «Gestion des émotions» est administré en très pe tits groupes de 2 ou 3 personnes. En 8 étapes, il vise le développement d habiletés à lutter contre le stress. Au début, le pa tient doit faire une de scrip tion verbale d émotions démontrées à l aide de matériel visuel. Chacun rapporte ensuite sa propre expérience de ces émotions dans le passé. À partir de cette expérience, le pa tient précise les conséquences de sa cog ni tion et de son comportement sur la gestion de ces émotions. Des stratégies d adaptation de rechange sont ensuite ajoutées à celles rapportées par les personnes. La constructivité et la praticabilité de ces stratégies d adaptation sont analysées. Une stratégie particulière est sélectionnée pour un jeu de rôle à partir d une sit u a tion définie. Les auteurs attirent l attention sur le fait que la plupart des personnes atteintes de psychose à un stade précoce, comparativement à celles présentant des mal a dies de longue évolution, sont en core stimulées par le tra vail ou l environnement so cial. Ils soulignent que les sujets utilisés en psychothérapie de groupe devraient être proposés par les pa tients. Le Tab leau 3 résume l étude d efficacité sur l EMT (4). Démarches axées sur le contenu de la pensée (événements cognitifs) L équipe de Kingdon et Turkington (6) en Angleterre a décrit au début des années 90 une approche de thérapie individuelle, la «réponse rationnelle». Dérivée des travaux de Beck et Ellis, elle a pour objectif l explication et la déstigmatisation de la maladie psychotique. Le tra vail thérapeutique est centré sur une compréhension des symptômes dans le cadre du modèle vulnérabilité-stress et repose sur l hypothèse selon laquelle les hal lu ci na tions et les délires peuvent être accessibles au raisonnement par une approche de col lab o ra - tion qui évite la con fron ta tion. Le tra vail sur les délires se divise en 4 étapes : Les symptômes sont repérés dans le temps jusqu à la date de leur ap pa ri tion. Le thérapeute amène le pa tient à l identification des cognitions erronées liées à cette période en utilisant le questionnement inductif. Les idées délirantes sont explorées avec le pa tient dans le but de comprendre la rai son pour laquelle une telle sig ni fi ca tion a été attribuée aux événements et expériences. L exploration d autres ex pli ca tions est entreprise en col lab o ra tion avec le pa tient. Le cas échéant, on uti lise une interprétation normalisante dans le cadre du modèle

30 Août 2001 Schizophrénie et psychothérapies cognitivo-comportementales 515 Tableau 2. Les différentes thérapies cognitivo-comportementales pour les troubles psychotiques Thérapie Niveau cognitif visé Nombre de séances Individuel/groupe But principal de la thérapie Programme intégratif de thérapies psychologiques (IPT) (2) Processus cognitif 2 3 séances de 90 minutes par semaines 90 minutes par séance (8 10 mois) Groupe de 5 à 7 patients, 2 à 3 cothérapeutes Acquérir une meilleure perception et maîtrise des déficits cognitifs, des facteurs de stress internes et externes et de leur lien avec les dysfonctions comportementales Gestion des émotions (EMT) (5) Processus cognitif 11 séances, 3 4 semaines Individuel + groupe de 2 à 3 patients Raffiner la reconnaissance efficace des émotions et les habiletés à les gérer afin de réduire la vulnérabilité aux stresseurs psychosociaux Réponse rationnelle (6) Événements cognitifs (contenu de la pensée) Non spécifié Individuel Obtenir une meilleure compréhension des symptômes psychotiques dans le cadre du modèle vulnérabilité-stress afin d'amoindrir la stigmatisation Adaptation des stratégies d adaptation (ASA) (8) Événements cognitifs 10 séances (5 semaines) Individuel Augmenter le répertoire de techniques d'adaptation Modifier les facteurs perpétuants Modification des croyances (7) Événements cognitifs 3 heures / semaines quotidien (12 semaines) Individuel ± groupe Modifier la signification donnée aux hallucinations Thérapie personnalisée (10) Schéma cognitif Jusqu à 3 ans Individuel et groupe Sensibiliser davantage à la vulnérabilité personnelle en améliorant la capacité à détecter de façon plus précise ses propres états affectifs Cog ni tively ori ented psy cho ther apy for early psy cho sis (COPE) (9) Schéma cognitif séances, Jusqu à 9 mois Individuel Faciliter l adaptation à l épisode psychotique et amoindrir ou prévenir la morbidité secondaire lors de la résolution du 1 er épisode Cognitive behavioral therapy for psychosis (11) Événements cognitifs et schéma cognitif séances (6 mois) Individuel Améliorer l adaptation du patient suite à un épisode psychotique en ciblant les symptômes positifs, négatifs et anxiodépressifs vulnérabilité-stress pour aider le pa tient à surmonter les stigmates associés à la maladie psychotique. Aucune étude contrôlée n a été publiée concernant cette approche. L équipe anglaise de Chadwick et Birchwood, en 1994, a élaboré une autre approche cognitivo-comportementale : la mod i fi ca tion des croyances, dont le but est de rendre l expérience psychotique moins pénible en modifiant la sig - ni fi ca tion donnée aux hal lu ci na tions et aux délires associés plutôt qu en diminuant les symptômes (7). La démarche est similaire aux thérapies cognitives de la dépression ou des trou bles anxieux. Appliquée dans une atmosphère de col lab o - ra tion, cette approche individuelle comporte 3 grandes par ties : 1. La phase d ouverture consiste en l engagement du pa tient, l établissement d une re la tion thérapeute-pa tient et un volet psychoéducatif. Le pa tient définit ses hal lu ci na tions et les croyances qui y sont reliées : leur na ture, leur identité, leur pouvoir, leur sig ni fi ca tion et l effet de l observance (ou de la non-ob ser vance) des in ci ta tions émises par les voix. Puis, les faits à la base de ces croyances délirantes (incluant le contenu des hal lu ci na tions) sont examinés. L étape suivante est la clar i fi ca tion du lien en tre les croyances, les réponses affectives ou comportementales, et l intensité de ces dernières. Le pa tient est ensuite amené à considérer les avantages et désavantages de l hypothèse de la fausseté de ses croyances. Durant cette phase, le pa tient est mis en con tact avec d autres personnes ayant eu des hal lu ci na tions, de façon directe ou par des enregistrements vidéo de pa tients discutant de leurs hal lu ci na tions. 2. La dis cus sion de la véracité des croyances fait ap pel à deux tech niques cognitives : la con tra dic tion hypothétique qui mesure l ouverture des pa tients à des preuves contredisant leurs croyances et la re mise en ques tion verbale, qui exige du pa tient d évaluer les faits sur lesquels se fondent ses croyances, puis de trouver d autres interprétations plausibles. La tâche du thérapeute est ensuite de faire ressortir l incohérence et l irrationalité du raisonnement, puis d offrir d autres ex pli ca tions du phénomène hallucinatoire (p. ex., les voix sont possiblement générées par le pa tient et les croyances qui y sont reliées sont une ten ta tive de la part du pa - tient de les expliquer et de leur donner un sens). 3. L expérimentation empirique des croyances, par une ten ta - tive de contrôle sur les hal lu ci na tions et la vérification de l hypothèse in verse, peut se faire. Pour les hal lu ci na tions, une évaluation cog ni tive des in di ces déclencheurs des voix est entreprise, puis la tech nique de ver bali sa tion concomitante

31 516 La Revue canadienne de psychiatrie vol 46, no 6 Tableau 3. Études d efficacité des TCC pour la schizophrénie axées sur le processus de la pensée Thérapie Type d étude, population étudiée Résultats IPT (3) 5 études cas-témoin com binées n = 165 pa tients Schizo phré nie (DSM- III / CIM-9) Âge moyen 30 à 36 ans Du rée to tale d hos pi tali sa tion : 6 mois à 15 ans symptômes psychotiques fonc tions cog ni tives (at ten tion) Pas de change ment du fonc tion ne ment cog ni tif supérieur Nor mali sa tion com por te men tale n est pas né ces saire ment as so - ciée à amé lio ra tion cog ni tive EMT et réadaptation réguliere (4) Thé ra pie 4 se maines Suivi 8 mois Cas-témoin n = 19 pa tients hos pi tal isés en psy chose ai guë à stade précoce de la mala die (6 mois d évo lu tion) Dans une 2 e étape com parai son avec 16 pa tients at te ints de schizo phré nie chronique (6 ans) Phase précoce de la maladie ha biletés cog ni tives Meil leure inté gra tion so ci ale et des re chutes (à 8 mois) Pas de change ment du fonc tion ne ment so cial et du bien-être émo tion nel à court terme Phase chronique fonc tion ne ment so cial et bien- être émotionnel ha biletés cog ni tives (pré coce = chronique) Tableau 4. Études d efficacité des TCC pour la schizophrénie axées sur les événements cognitifs Thérapie Type d étude, population étudiée Résultats Amélioration des stratégies d adaptation (ASA) et résolution de problèmes (RP) et liste d attente (8,16) Thérapie 5 semaines Modification des croyances (individuelle + groupe; 1 heure/jour) et thérapie de soutien (1heure/jour) (14, 15) Thérapie 12 semaines Suivi 9 mois 27 pa tients at te ints de schizo phré nie ran domisés, réfrac - taires au traitement Du rée de la mala die 12 ans Âge moyen 43 ans 4 hos pi tali sa tions (moy enne) Der nière ad mis sion 3,7 ans 44 pa tients hos pi tal isés pour psy chose ai guë (dont 66% = 1 er ou 2 e épi sode et 70% ont moins de 5 ans évo lu tion) Âge moyen 30 ans 6 ans d évo lu tion 3 épi sodes antérieurs Di ag nos tics : Schizo phré nie (60%) et trouble schizoaffectif 40%) symptômes psy cho tiques Meil leure ad ap ta tion variant pro por tion nel le ment aux symptômes psy cho tiques Straté gies d a dap ta tion posi tives ef fi caces avec ASA mais avec RP Pas de change ment des symptômes né ga tifs, ou du fonc tion - ne ment so cial Rémis sion des symptômes po si tifs accé lé rée (25% 50%) et plus mar quée dès la 7 e se maine de thé ra pie Pas d ef fet sur les symptômes né ga tifs, désor gani sa tion con vic tion dé li rante TCC plus que con trôles dès la 7 e se - maine symptômes psy cho tiques rési du els (jusqu à 9 mois) Dé lai et de gré de rémis sion iden tiques chez 1 er épi sode et mul ti ples épi sodes antérieurs Rémis sion plus rap ide si sexe féminin, courte du rée de la mala die et courte du rée de psy chose non- traitée est utilisée pour éliminer les voix. Pour les autres croyances, le test empirique peut être élaboré avec le pa tient. Dans l étude d efficacité menée en 1996 par ces mêmes auteurs (14), le traitement comportait un deuxième volet de thérapie cognitive en groupe où les pa tients devaient critiquer les délires des autres par tic i pants ainsi que les faits à la base des croyances, puis analyser leurs stratégies d adaptation. Drury et Bichwood ont publié une étude cas-témoin sur l efficacité de la thérapie cog ni tive (individuelle et de groupe) dans la rémission de la psychose aiguë chez des pa - tients hospitalisés pour décompensation (14,15). Les résultats sont résumés dans le Tab leau 4. L amélioration des stratégies d adaptation (ASA) mise au point par l équipe de Tarrier en Angleterre, en 1993, est une thérapie individuelle qui re pose sur l hypothèse selon laquelle deux facteurs augmentent la probabilité de l apparition des symptômes psychotiques : la présence de précipitants environnementaux et la réaction cog ni tive, comportementale ou physiologique aux expériences hallucinatoires ou délirantes (8). L objectif de cette thérapie est donc d atténuer les symptômes en apprenant au pa tient à maîtriser les facteurs précipitants et ses réactions aux symptômes psychotiques. Le pre mier but visé est de for mer une batterie de tech niques d adaptation, en optimisant le répertoire des stratégies habituelles du pa tient. L autre but consiste à mod i fier les facteurs de perpétuation des symptômes psychotiques pour réduire à la fois les symptômes et les émotions négatives associées. La première étape de la thérapie est l identification des stratégies d adaptation habituellement utilisées par le pa tient. Puis, le pa tient apprend à reconnaître ses symptômes (hal lu ci na tions, contenu bi zarre ou inhabituel des pensées délirantes, etc.) comme on l enseigne au pa tient déprimé ou anxieux pour les pensées erronées. Les symptômes cibles sont ensuite classés en ordre hiérarchique, du plus fac ile au plus difficile, afin

32 Août 2001 Schizophrénie et psychothérapies cognitivo-comportementales 517 Tableau 5. Études d efficacité des TCC pour la schizophrénie axées sur les schémas cognitifs Thérapie Type d étude/ population étudiée Résultats Thérapie personnalisée (TP) et thérapie de soutien et thérapie familiale et combinaison soutien + familiale (18,19) Suivi et thérapie 3 ans n = 151 pa tients ran domisés Di ag nos tic RDC (schizo phré nie 70% et trou ble schi zoaf fec tif 30%) Âge moyen 28 33ans Du rée de la mala die 6 10 ans Meil leur ajus te ment so cial avec TP Pas d ef fets sur les symptômes Pa tients plus anxieux avec TP qu avec les autres thé ra pies Patients vivant avec leur famille de la non- observance re chutes psy cho tiques et af fec tives avec TP thérapies soutien ou famili ale Meil leure per form ance globale avec TP qu avec les autres Patients vivant sans leur famille Amé lio ra tion du ren de ment au travail et des re la tions sociales Plus de re chutes psy cho tiques avec TP qu avec thé ra pie soutien Cognitive behavioral therapy For psychosis (London East Anglia) (CBT) et approche «Case management» (21, 22, 23) n = 60 Dx schizo phré nie, trou ble schi zoaf fec tif, trou ble dé li rant résis tants Évo lu tion de la mala die 13 ans Âge moyen 40 ans 25% BPRS à 9 mois avec CBT (items sus pi cion, dé li res, hal lu ci na tions) Taux réponse 50% contre 31% con trôle Rythme d amé lio ra tion simi laire à la clo zap ine 80% des pa tients sont satisfaits Autoévalua tion : con vic tion dé li rante, détresse associée Fréquence des hal lu ci na tions Main tien de la réponse à 18 mois ( 29% BPRS contre 2 % pour les témoins) Nou velles amé lio ra tions attrib uta bles à la thé ra pie : détresse re liée aux dé li res et fréquence des hal lu ci na tions COPE vs contrôles (thérapie pas offerte) vs ceux qui ont refusé la thérapie (17) n = 80 Maladie psychotique Durée psychose jours Durée d hospitalisation jours Âge au début de la maladie 21 ans échelle QOL et BDI Meil leure inté gra tion de l expé ri ence psy cho tique et modèle ex pli ca tif plus adapté Pas de diffé rence au BPRS, GSI, taux de re chute et de sui cide Patients traités avec COPE vs ceux ayant refusé BDI moins élevé chez ceux ay ant refusé la thérapie Avec COPE, meil leure inté gra tion de l expé ri ence psy cho tique et du Modèle ex - pli ca tif de la psychose BPRS = Brief Psy chi at ric Rating Scale; BDI = Beck De pres sion In ven tory; GSI = Global Symptom In ven tory Rat ing Scale; QOL = Qual ity of Life Scale d entreprendre une approche graduelle. Par contre, si la réduction rapide d un autre symptôme était prioritaire à cause de la détresse ou du dysfonctionnement im por tant associé, ce symptôme pourrait avoir préséance sur d autres plus faciles. Enfin, c est la mise à l épreuve des stratégies d adaptation : cognitives (p. ex., dis trac tion, centrage de l attention, auto-in struc tion); comportementales (p. ex., aug men ta tion des activités sol i taires et des in ter ac tions sociales, test comportemental); physiologiques (p. ex., re lax ation, exercices de respira - tion). La stratégie reconnue la mieux appropriée pour un symptôme est fragmentée en étapes. On uti lise cette stratégie d abord du rant les séances de sim u la tion d une sit u a tion et par la suite, in vivo. Des exercices à faire à la maison sont proposés afin d encourager la généralisation des habiletés à d autres situa - tions. Le pa tient cible un symptôme à la fois jusqu à amélioration. Un max i mum de 2 à 3 séances par symptôme est fixé. S il n y a aucune amélioration, le symptôme suivant est ciblé, afin d éviter le découragement du pa tient. Thérapies axées sur les structures ou schèmes cognitifs L équipe australienne de McGorry et Jack son ont créé la COPE (Cog ni tively Ori ented Psy cho ther apy for Early Psy - cho sis) dont l objectif est de faciliter l adaptation de la personne et de prévenir ou amoindrir la morbidité secondaire lors de la résolution du pre mier épisode psychotique (9). C est la seule thérapie créée spécifiquement pour l intervention précoce auprès de pa tients souffrant de psychose. Cette approche s appuie sur plusieurs hypothèses concernant l expérience psychotique. Elle se fonde sur la métathéorie du constructivisme (voir l introduction). L accent est mis sur le sen ti ment d efficacité personnelle, c est-à-dire l impression d avoir la capacité de s adapter ou de gérer une sit u a tion. Ces auteurs considèrent la psychose comme étant une expérience traumatique menaçant l intégrité du soi ou l identité du jeune adulte, ce qui peut l amener à penser que plusieurs sit u a tions dépassent sa capacité d adaptation. L intervention plus précoce au de but de maladie pourrait arrêter ou prévenir une détérioration de la per cep tion (con science) de soi et du sen ti ment d efficacité personnelle. Le résultat souhaité est de réduire le ris que de sui cide ainsi que de prévenir l incapacité et la concrétisation des déficits sévères (symptomatologie négative).

33 518 La Revue canadienne de psychiatrie Vol 46, No 6 Tableau 6. Facteurs prédicteurs de réponse à la TCC pour les délires selon l étude de London East Anglia (22) Facteurs prédicteurs de réponse à la TCC pour diminuer les délires La flexi bil ité cog ni tive (me surable au Maud sley Scale à l item «pos si bil ité de se tromper» en ce qui con cerne les croy ances dé li ran tes Taux élevé d hospitalisations ré cen tes Facteurs étudiés mais non prédicteurs de réponse Fac teurs démo graphiques His toire évo lu tive de la mala die In tel li gence Flu ence ver bale Dépres sion, anxiété Es time de soi Score au Brief Psy chi at ric Rat ing Scale De gré de con vic tion des dé li res In ten sité des préoc cu pa tions dé li ran tes Détresse re liée aux délires Cette démarche vise à agir sur la stig ma ti sa tion et les stéréotypes reliés à la maladie, à aider la personne à comprendre sa maladie, à cibler la morbidité secondaire dès son émergence (dépression, anxiété sociale, etc.) et à promouvoir l adaptation en encourageant la poursuite d objectifs de vie. La notion de l identité est au cœur de cette approche. Cette thérapie individuelle souple s ajuste au rythme de l évolution du pa tient. Les moyens utilisés sont la psychoéducation, des tech niques cognitives pour le tra vail du soi (stig ma ti sa tion, stéréotypes, etc.) et des tech niques cognitivo-comportementales pour cibler la morbidité secondaire (dépression, anxiété, etc.). Cette approche n inclut pas de traitement des symptômes positifs car ces auteurs observent que les pa tients présentent habituellement une bonne réponse au traitement pharmacologique, lors du pre mier épisode. La thérapie se divise en 4 volets : L évaluation s étend à 3 ou 4 séances. Elle a pour but d identifier le modèle explicatif du pa tient en ce qui a trait à sa maladie, de déterminer le niveau d autocritique, les mécanismes et le niveau d adaptation à la psychose. L effet de la maladie sur la perception qu a le pa tient de lui-même et sur son estime personnelle de même que la présence de morbidité secondaire sont également évalués. C est durant cette étape que se forme l alliance thérapeutique. L engagement s établit à compter de la quatrième séance, où un contrat et un plan de tra vail sont élaborés en col lab o ra tion avec le pa tient. Le plan inclut habituellement la psychoéducation et une approche cognitivo-comportementale pour cibler les problèmes de retrait so cial, de stig ma ti sa tion et de mo ti va tion. On uti lise l approche cognitivo-comportementale pour la morbidité secondaire (dépression et anxiété). La thérapie visant à promouvoir l adaptation comporte 4 phases. Les deux premières visent à maintenir l espoir par l exploration du répertoire de mécanismes d adaptation du pa tient et par une prise de con science de son potentiel et de sa capacité à se réaliser. Les deux dernières phases con sis tent en l ajustement des cognitions d adaptation et des comportements. L amélioration de la con science de soi (sense of self ) et la distanciation par rap port aux as pects négatifs de l environnement (p. ex., stig ma ti sa tion) sont également ciblées. Le Tab leau 5 résume une étude menée par l équipe de Jack - son et McGorry dans la région de Mel bourne, en Australie (17). Quatre-vingt pa tients ont été divisés en trois groupes : ceux qui ont accepté la thérapie COPE, ceux qui l ont refusée et ceux à qui la thérapie n a pas été offerte (témoins). La thérapie personnalisée à été élaborée en 1995 par l équipe de Hogarty aux États-Unis (10). Ce groupe travaillait déjà depuis 20 ans à des programmes psychoéducatifs chez des pa - tients atteints de schizophrénie et leur famille. Cette nou velle thérapie est administrée en complémentarité avec leurs autres programmes. Elle se dis tingue des autres formes de thérapies psychosociales par l accent mis sur les réponses reliées spécifiquement aux af fects, indépendamment des stresseurs. Trois prémisses sous-tendent cette démarche : L altération des af fects joue un rôle cen tral dans la schizophrénie et dans l exacerbation psychotique en plus de compromettre l ajustement per son nel et so cial, d où l importance de stratégies de contrôle des comportements déterminés par le système limbique. La nécessité d une thérapie centrée sur la maladie plutôt que sur la personne. Il faut tenir compte des as pects neuropsychologiques de la diathèse sous-jacente. Les in ter ven tions sont graduellement plus complexes selon le rythme et le niveau de rémission du pa tient, afin d éviter qu elles deviennent une source de stress supplémentaire risquant de précipiter une décompensation.

34 Août 2001 Schizophrénie et psychothérapies cognitivo-comportementales 519 L objectif prin ci pal de la thérapie est de développer une sensibilité ac crue à la vulnérabilité personnelle. On améliore la capacité de détecter de façon précise ses propres états affectifs (en tre autres, par les in di ces internes de mauvaise régulation des af fects), pour ensuite en cour ager l expression appropriée et contrôlée des émotions, et l évaluation des réponses réciproques d autrui. Les moyens utilisés sont des tech niques comportementales et la psychoéducation, dont le contenu est propre à chacune des trois phases. Durant la phase de base qui dure de 3 à 6 mois, on vise la sta - bili sa tion clinique, la for ma tion d une al li ance thérapeutique et l établissement d un contrat thérapeutique où l on souligne la no tion de responsabilité, de dose minimale efficace de médication et d acquisition de stratégies d adaptation. Par ailleurs, le pa tient participe, avec 4 ou 5 personnes, à un ate - lier de groupe psychoéducatif sur la maladie. Le modèle présenté se cen tre sur la vulnérabilité, la nécessité du traitement pharmacologique et l abstinence de drogues (3 séances). La phase intermédiaire s étale sur 6 à 18 mois et vise le développement d une prise de con science des états affectifs, cognitifs et comportementaux, et à la fois, l augmentation des compétences personnelles d auto-régulation et de gestion des émotions. Le pa tient participe à un ate lier où l on aborde les symptômes précurseurs de rechute et l on discute de ces symptômes avec chacun des par tic i pants. On discute également du rôle de la mauvaise régulation af fec tive et des in di ces de détresse associés, de même que du répertoire de stratégies utilisées par le pa tient. Un deuxième ate lier aborde les nouvelles stratégies à apprendre : re lax ation, entraînement aux habiletés sociales, aux tech niques de per cep tion sociale et à la résolution de problème. Au cours du troisième ate lier, on aborde les questions relatives au travail : habiletés nécessaires, expériences passées, problèmes éventuels. Durant la dernière phase (avancée), le pa tient se tourne vers l extérieur en travaillant sur son ori en ta tion occupationnelle et la socia li sa tion. Il est alors dans un processus de réinsertion au travail. On analyse également les conséquences sur son état affectif de l expression des af fects et des changements dans l environnement. L adaptation interne est renforcée par les sit u a tions in vivo que le pa tient rencontre; on cible particulièrement les sit u a tions où de fortes émotions sont exprimées. L équipe de Hogarty à Pitts burgh, aux États-Unis, a effectué une étude randomisée de 3 ans de traitement comparant la thérapie personnalisée à la thérapie de soutien et/ou familiale (18,19), (Tab leau 5). Les auteurs concluent que l application de la thérapie personnalisée ne devrait idéalement commencer qu après l atteinte de la sta bili sa tion des symptômes et l installation au lieu de résidence puisque dans leur étude, les pa tients vi vant sans leur famille font plus de rechutes psychotiques lorsqu ils sont traités avec la thérapie personnalisée (TP) plutôt qu avec la thérapie de soutien, contrairement aux pa tients vi vant avec leur famille. Psychothérapie intégrative En 1989, l équipe de Fowler et Morley en Grande-Bretagne a conçu une approche intégrative décrite dans un manuel intitulé Cog ni tive Be hav ioral Ther apy for Psy cho sis (11,20). Cette thérapie agit notamment sur l estime, l image personnelle (schéma cognitif) et les événements cognitifs (symptômes psychotiques ou réponse qui y est associée). L hypothèse théorique veut que les délires, qui ne sont ni fixes ni immuables, peuvent être compris en tant que processus psychologiques. Ces auteurs soulignent qu une pro por tion substantielle de pa tients souffrant de schizophrénie présentent des symptômes psychotiques résistant à la médication, une per tur ba tion émotionnelle grave et un hand i cap so cial marqué. Considérant l hétérogénéité qui caractérise ces problèmes, la thérapie est individualisée et le choix des tech niques est guidé par la for - mu la tion des problèmes et les besoins du pa tient. Cette thérapie globale, structurée et de durée limitée a pour con cept cen tral l attitude empathique du thérapeute et comporte 6 étapes : 1. L évaluation et l engagement durant lesquels s établit l alliance thérapeutique. Les problèmes du pa tient sont identifiés de même que les facteurs qui sont associés à leur développement et leur maintien. 2. L application de stratégies d adaptation cognitivo-comportementales, dans le but d inspirer un senti - ment de contrôle, de maintenir l espoir et de contenir les sen - ti ments envahissants de terreur ou d incurabilité. Différentes stratégies d autorégulation des symptômes cibles sont suggérées au pa tient afin de l aider à gérer les expériences psychotiques, les réactions émotionnelles graves ou les pas - sages à l acte impulsifs. Les stratégies cognitives consistant à maîtriser les pensées dysfonctionnelles identifiées, les stratégies comportementales utilisant l exposition et les stratégies physiologiques, comme la re lax ation, font l objet d un entraînement. 3. Une dis cus sion est engagée à propos d un nou veau modèle explicatif de l expérience personnelle du pa tient à partir de for mu la tions cognitives des symptômes psychotiques, à l aide du questionnement socratique et du modèle vulnérabilité-stress. 4. La restructuration cog ni tive s attarde aux délires et aux hal - lu ci na tions qui sont source de problème d intégration sociale, d incapacité ou de souffrance pour le pa tient. L attention est centrée sur les délires, les interprétations paranoïdes, les croyances à propos des voix, etc. dans le but de fournir des ex - pli ca tions basées sur la réalité, impliquant donc que l expérience psychotique est le produit d un dysfonctionnement biologique. Si le pa tient n est pas prêt à considérer d autres so lu tions de rechange, le thérapeute travaillera à l intérieur du système délirant afin de diminuer la détresse et de promouvoir une évaluation de la réalité mieux adaptée. 5. L analyse à l aide d une grille cognitivo-comportementale des présomptions erronées à propos de soi et des autres (p.

35 520 La Revue canadienne de psychiatrie Vol 46, No 6 ex., im pres sion d être sans valeur, diabolique, dangereux, que les autres sont indignes de confiance etc.). Les stratégies utilisées sont similaires à celles décrites chez les pa tients déprimés ou ayant des trou bles de personnalité. 6. La con sol i da tion des nouvelles per spec tives sur les problèmes de la personne; l objectif est de concevoir un plan pour maîtriser d éventuelles rechutes psychotiques et contrer l inaptitude sociale à moyen et long terme. Le Tableau 6 résume l étude randomisée de Lon don East Anglia (21 23). La thérapie cognitivo-comportementale, selon le manuel écrit par l équipe de Fowler (11) est comparée au traitement habituel à l aide d une approche de case man age ment ou agent de suivi communautaire. Pour le groupe traité avec TCC, les délires étaient les symptômes démontrant le plus d amélioration. Conclusion «Nous devons prendre du recul et amorcer une réflexion sur ce que peut signifier cette distorsion biologique de l âme humaine pour la personne qui en souffre et sur les in ter ven - tions requises» (Coursey [24]). L élaboration d interventions psychologiques cohérentes avec no tre con - cep tion actuelle de la maladie, notamment les causes et traitements biologiques, a connu un progrès considérable au cours des 10 dernières années. Ce progrès est concret. Il se définit par les in ter ven tions précises et rigoureuses que nous avons révisées. Cette précision et cette ri gueur doivent continuer d accompagner l ouverture et l humanisme in dis pen sa bles au traitement des personnes atteintes de schizophrénie et de leur famille. Ce sont également cette précision et cette ri gueur qui rendent pos si ble la réalisation de recherches. Celles complétées à ce jour permettent de jeter un re gard cri tique sur ces in ter ven tions et contribuent à définir les cor rec tions nécessaires. Celles à venir pourraient tout d abord nous aider à raffiner l évaluation du pa tient et des buts poursuivis par l intervention psychologique. Il est clair que no tre con cep tion de la schizophrénie ne permet que peu de généralisation au chapitre des in ter ven tions psychologiques comme ailleurs. La schizophrénie demeure une maladie complexe, à caractère évolutif où la démarche thérapeutique repose essentiellement sur une question : «Quelle est la combinaison spécifique d interventions qui sera la plus utile pour ce pa tient avec ce type particulier de schizophrénie, à cette phase précise de la maladie?» (Fenton, McGlashan, [25]). Bibliographie 1. Perris C, Skagerlind L. Cog ni tive ther apy with schizo phrenic pa tients. Acta Psychiatr Scand 1994;89 (suppl. 382) : Pomini V, Brenner HD, Hodel B, Roder V. Thérapies psychologiques des schizophrénies. Primont : Pi erre Mardaga; Konen A, Neis L, Hodel B, Brenner HD. À propos des thérapies cognitivo-comportementales de la schizophrénie: le programme intégratif de thérapies psychologiques (IPT). L Encéphale 1993;XIX : Hodel B, Brenner HD, Merlo MCG, Teuber JF. Emo tional man age ment ther apy in early psy cho sis. Br J Psy chi a try 1998;172 (suppl 33) : Im pli ca tions clin iques Les psychothérapies cognitivo-comportementales peuvent être bé - néfiques dans l intervention auprès de patients souffrant de schizo - phrénie tant en phase aiguë que chronique. Ces approches devraient faire partie de programmes d intervention précoce pour les patients atteints de schizophrénie afin de prévenir la morbidité secondaire et de faciliter l adaptation psychologique et so - ciale de la personne à l expérience traumatique que peut représenter la psychose. L alliance thérapeutique et un climat de collaboration empirique sont essentiels au succès de ces démarches. Limitations La méthodologie des études d efficacité de ces thérapies et les me - sures utilisées dans ce but sont très variables. L hétérogénéité des populations étudiées (diagnostics, âge, sévérité et phase de la maladie) rend difficile les comparaisons en tre ces études. Les études ne permettent pas de conclure quant à l efficacité relative de ces méthodes selon le type de patient ou de symptomatologie ou quant à la durée optimale du traitement. 5. Green MF. What are functionnal con se quences of neurocognitive def i cits in schizo - phre nia? Am J Psy chi a try 1996, 153 (3). 6. Kingdon D, Turkington D, John C. Cog ni tive be hav iour ther apy of schizo phre nia: the amenability of de lu sions and hal lu ci na tions to rea son ing. Br J Psy chi a try 1994;164 : Chadwick P, Birchwood M. The om nip o tence of voices: a cog ni tive ap proach to au - di tory hal lu ci na tions. Br J Psy chi a try 1994;164 : Tarrier N, Beckett R, Har wood S, Baker A, YusupoffL, Ugarteburu I. A trial of two cog ni tive-be hav ioural meth ods of treating drug-re sis tant re sid ual psy chotic symp - toms in schizo phrenic pa tients: I. Out come. Br J Psy chi a try 1993;162 : Jack son H, Mc Gorry P, Ed wards J, Hulbert C. Cog ni tively ori ented psy cho ther apy for early psy cho sis (COPE). Dans : Cotton P, Jack son H, eds. Early in ter ven tion and pre ven tive ap pli ca tion of clin i cal psy chol ogy. Mel bourne (Aus tra lia): Ac a - demic Press; Hogarty GE, Kornblith SJ, Green wald D, DiBarry AL, Cooley S, Flesher S, et coll. Per sonal ther apy: a dis or der-rel e vant psy cho ther apy for schizo phre nia. Schizophr Bull 1995;21 (3): Fowler DG, Garety P, Kuipers L. Cognitive be hav iour ther apy for psy cho sis: the - ory and prac tice. Chichester (G-B): John Wiley and Sons; Nicole L, Pires A, Routhier G, Bélanger R, Buissières G, Gingras N, et coll. Schizophrénie, approche spécialisée et continuité de soins: le programme spécifique in ter ven tion «pre mier épisode» de l Hôtel-Dieu de Lévis. Santé mentale au Qué bec. 1999;24 (1). 13. Brenner HD, Hodel B, Kube G, Roder V. Kognitive therapie bei schizophrenen: Problemanalyse und empirische ergebnisse,. Nervenarzt 1987;58: Drury V, Birchwood M, Cochrane R, MacMillan F. Cog ni tive ther apy and re cov - ery from acute psy cho sis: a con trolled trial. I. Im pact on psy chotic symp toms. Br J Psy chi a try 1996;169 : Drury V, Birchwood M, Cochrane R, MacMillan F. Cog ni tive ther apy and re cov - ery from acute psy cho sis: a con trolled trial. II. Im pact on re cov ery time. Br J Psy - chi a try 1996;169 : Tarrier N, Sharpe L, Beckett R, Har wood S, Baker A, Yusupoff L. A trial of two cog ni tive-be hav ioural meth ods of treating drug-re sis tant re sid ual psy chotic symp - toms in schizo phrenic pa tients: II. Treat ment specific changes in coping and prob - lem solving skills. Soc Psy chi a try Psychiatr Epidemiol 1993;28 : Jack son H, Mc Gorry P, Ed wards J, Hulbert C, Henry L, Francey S, et coll. Cog ni - tively ori ented psy cho ther apy for early psy cho sis (COPE): Pre lim i nary re sults. Br J Psy chi a try 1998;172 (suppl 33) : Hogarty GE, Kornblith SJ, Green wald D, DiBarry AL, Cooley S, Ulrich RF, et coll. Three-year trials of per sonal ther apy among schizo phrenic pa tients liv ing with or in de pend ent of fam ily. I: De scrip tion of study and ef fects on re lapse rates. Am J Psy chi a try 1997;154(11): Hogarty GE, Green wald D, Ulrich RF, Kornblith SJ, DiBarry AL, Cooley S, et coll. Three-year trials of personnal ther apy among schizo phrenic pa tients living with or in de pend ent of fam ily. II: Ef fects on ad just ment of pa tients. Am J Psy chi a try 1997;154 (11) : Fowler D, Garety P, Kuipers E. Un der stand ing the in ex pli ca ble: an in di vid u ally for mu lated cog ni tive ap proach to de lu sional be liefs. Dans : Perris C, McGorry P,

36 Août 2001 Schizophrénie et psychothérapies cognitivo-comportementales 521 Abstract Schizophrenia and Cognitive-Behavioural Therapy (CBT) Ob jec tive: To dis tin guish be tween dif fer ent ap proaches of cog ni tive-be hav ioural ther apy (CBT) for schizo phre nia de - pend ing on the goals, ob jec tives and meth ods of these ap proaches, then to dis cuss ef fi cacy studies. Method: A sum mary of in for ma tion col lected through elec tronic (MED LINE, PSYchlit) and bib lio graphic re search. Re sults: CBTs all broadly at tempt to bring a bet ter cog ni tive, be hav ioural and emo tional ad just ment to the psy chotic ex - pe ri ence by sug gest ing to the pa tient a new ex plana tory model of psy cho sis: the vulnerability-stress model. These ap proaches involve dif fer ent levels and goals. Some fo cus on cor rect ing ba sic cogni tive deficits or modifying the psy chotic symp toms and the re lated dis tress. At the other end of the spectrum, me ta cog ni tive thera pies aim to modi fy and re struc tur e dys func tional self and en vi ron ment sche mas to en able the de vel op ment of bet ter-ad justed and gen er ally ap - plied cog ni tive strate gies. A few stud ies with lim ited power and methods have shown the effi ciency of those therapies. Con clu sion: CBTs prove to be a prom is ing ad di tive treat ment. They have been shown to im prove so cial adj ust ment and qual ity of life, and to di min ish psy chotic symp toms and the related dis tress. They ad dress all posi tive, nega tive, cog ni - tive, be hav ioural, and emo tional symp toms while con sid er ing the stage of the dis ease and the pa ti ent s spe cial needs. Fur ther re search is needed to es tab lish the du ra tion, the best pro vi sion fre quency, and the specificity of these ap proaches. eds. Cog ni tive psy cho ther apy of psy chotic and per son al ity dis or ders: Handbook of the ory and prac tice. John Wiley and Sons Ltd. Chichester (UK): 1998; 464p. 21. Kuipers E, Garety P, Fowler D, Dunn G, Bebbington P, Free man D, et coll. Lon don East-Anglia ran dom ised con trolled trial of cog ni tive- be hav ioural ther apy for psy - cho sis. I: Ef fects of the treat ment phase. Br J Psy chi a try 1997;171 : Garety P, Fowler D, Kuipers E, et coll. London East-Anglia ran dom ised con trolled trial of cog ni tive-be hav ioural ther apy for psy cho sis. II: Pre dic tors of out come. Br J Psy chi a try 1997;171 : Kuipers E, Fowler D, Garety P, Chishlom D, Free man D, Dunn G, et coll. Lon don East-Anglia ran dom ised con trolled trial of cog ni tive- be hav ioural ther apy for psy - cho sis III: Fol low-up and economic eval u a tion at 18 months. Br J Psy chi a try 1998; Coursey RD. Psy cho ther apy with per sons suf fer ing from schizo phre nia: the need for a new agenda. Schizophr Bull 1989;15 : Fenton WS, McGlashan TH. We can talk: in di vid ual psy cho ther apy for schizo - phre nia. Am J Psy chi a try 1997;154(11) :

37 ORIGINAL RESEARCH Polydipsia, Psychosis, and Familial Psychopathology AG Ahmed, MB 1, Li ana M Heigh, BA, BSW 2, KV Ramachandran MB, BS, MRCPsych 3 Ob jec tive: To com pare the de mo graphic and clin i cal fac tors and fa mil ial psychopathology of chronic psy chi at ric in pa - tients with, and with out, polydipsia. Method: We un der took a case-con trol study of chronic psy chi at ric in pa tients both with, and with out, polydipsia. Clin i - cal and de mo graphic data were gath ered using a predesigned ques tion naire, the Pos i tive and Neg a tive Syn drome Scale (PANSS), the Ab nor mal In vol un tary Move ment Scale (AIMS), the Mini-Mental State Ex am i na tion Scale (MMSE), and the Fam ily His tory Re search Di ag nos tic Cri te ria (FH-RDC). Re sults: The prev a lence rate of polydipsia was 20.2%. The group with polydipsia was sig nif i cantly youn ger, both at the time of their first-ever psy chi at ric and current psy chi at ric ad mis sions, com pared with the group with out polydipsia. The 2 groups were sim i lar in terms of their ill ness char ac ter is tics and psy chi at ric di ag no ses. In the group with polydipsia, al co hol abuse pre dated the psy chotic ill ness by a mean of 10.5 (SD 4.4) years, com pared with 4.8 (SD 1.6) years for the same pe riod in the un af fected group. The 2 groups did not differ sig nif i cantly re gard ing the antipsychotic med i ca tion dos age, the pro por tion on con com i tant anticholinergic medicaton, the doc u mented pre vi ous re sponse to antipsychotic med i ca tion, or past treat ment with electroconvulsive ther apy (ECT). First-degree rel a tives of pa tients with polydipsia were found to have sig nif i cantly higher rates of al co hol de pend ence. Con clu sion: This study pro vides further ev i dence for the higher rate of polydipsia among chronic psychia t ric pa tient pop u la tions and for high rates of al co hol-re lated prob lems among their first-de gree rel a t ives. (Can J Psy chi a try 2001;46: ) Key Words: polydipsia, psychiatric inpatients, first-degree relatives, psychopathology, alcohol abuse Ex ces sive fluid con sump tion is not an un com mon finding among chronic psy chi at ric pa tients, par tic u larly those with schizo phre nia. The prev a lence of polydipsia ranges from 2.5% to 17.5% (1,2). The di ag no sis is of ten missed un til se vere dilutional hyponatremia causes life-threat en ing gen er - al ized sei zures. A sig nif i cant pro por tion (29%) of pa tients with polydipsia may de velop wa ter in tox i ca tion, and this sub - group is said to have a less fa vour able prog no sis, com pared with pa tients who have schizo phre nia without polydipsia or schizo phre nia with polydipsia but without wa ter intoxication (3). The pathogenesis of polydipsia re mains poorly un der - stood and may in volve sev eral mech a nisms, in clud ing hypo - Manuscript received October 2000, revised, and accepted May Assistant Professor, University of Ottawa; Staff Psychiatrist, Royal Ottawa Health Care Group, Brockville Cam pus, Brockville, On tario. 2 Social Worker, Saskatchewan Hospital, North Battleford, Saskatchewan. 3 Chief Psychiatrist, Saskatchewan Hos pi tal, North Battleford, Saskatche - wan. Address for correspondence: Dr AG Ahmed, University of Ottawa and Royal Ottawa Health Care Group, Brockville Campus, PO Box 1050, Brockville, ON K6V 5W7 vkahmed@hotmail.com tha lamic and hippocampal dis tur bance. Raskind and oth ers sug gest that both the psy chotic ill ness and polydipsia may be man i fes ta tions of the same cen tral ner vous sys tem pa thol ogy (4). Psychotropic med i ca tion has been pro posed as hav ing a po ten tial role in its eti ol ogy. Polydipsia can not, how ever, be fully ex plained as a side ef fect of psychotropic med i ca tion, be cause the phe nom e non has been doc u mented in pa tients not tak ing psychotropic med i ca tions (3). The well-known com pli ca tion of wa ter in tox i ca tion may re - sult from the se ver ity of polydipsia, from kid ney dys func tion, or from in creased antidiuretic hor mone (ADH) ac tiv ity and the presence of the syn drome of in ap pro pri ate se cre tion of antidiuretic hor mone (SIADH). It may also re sult from os - motic dysregulation as so ci ated with chronic psy cho sis. It has been sug gested that brain sen si tiv ity to hyponatremia is an - other probable patho genic fac tor that war rants fur ther study (3). Just as the pathogenesis of polydipsia re mains poorly un der - stood, so does its man age ment re main a clin i cal chal lenge. Var i ous treat ment mo dal i ties have been tried, with varying suc cess. Case re ports of de creased fluid con sump tion and Can J Psychiatry, Vol 46, August

38 August 2001 Polydipsia, Psychosis, and Familial Psychopathology 523 Table 1. Characteristics of patients with, and without, polydipsia Characteristics Group with poly dip sia n (%) Group with out poly dip sia n (%) χ 2 P Sex Male Fe male 27 (79.4) 7 (20.6) 27 (79.4) 7 (20.6) 0 ns Ethnic Background European Ca na dian Na tive Ca na dian 28 (82.4) 6 (17.6) 30 (88.2) 4 (11.8) 0.46 ns Education In com plete high school Com plete high school Post high school 28 (82.4) 4 (11.8) 2 (5.9) 21 (61.8) 7 (20.6) 2 (17.6) 3.80 ns Employment status before admission Em ployed Un em ployed 21 (61.8) 13 (38.2) 20 (58.8) 14 (41.2) 0.06 ns Marital Status Sin gle Mar ried Sepa rated or di vorced 30 (88.2) 1 (2.9) 3 (8.8) 23 (67.6) 4 (11.8) 7 (20.6) 4.30 ns Financial Status Com pe tent In com pe tent 25 (73.5) 9 (26.5) 27 (79.4) 7 (20.6) 0.32 ns nor mal iza tion of se rum so dium with clozapine use in pa tients with polydipsia or hyponatremia sug gest that se ro to nin-do - pa mine an tag o nists may be valu able in treating this con di - tion. Data on the ef fects of other atyp i cal antipsychotics (par tic u larly risperidone) on polydipsia remain equiv o cal, how ever (5 8). This study com pared the de mo graphic and clin i cal char ac ter - is tics of 2 groups of chronic psy chi at ric in pa tients: one group with polydipsia and the other, without polydipsia. In ad di tion, we also com pared the rates of psychopathology in the first-de gree rel a tives of the groups. Method This study was con ducted at the Sas katch e wan Hos pi tal in North Battleford, Sas katch e wan, a long-stay psy chi at ric hos - pi tal with a total pa tient pop u la tion of about 170. The hos pi tal pro vides psy chi at ric re ha bil i ta tion pro grams for pa tients with chronic mental dis or ders and fo ren sic psy chi at ric as sess - ments for the en tire prov ince of Sas katch e wan. Most pa tients in the psy chi at ric re ha bil i ta tion pro gram have schizo phre nia. All pa tients with the clin i cal di ag no sis of polydipsia as identi - fied by the multidisciplinary clin i cal team and meet ing the fol low ing cri te ria were in cluded in the study: at least one doc u mented ep i sode of hyponatremia (se rum so dium less than 135 mmol/l) a doc u mented his tory of ex ces sive fluid con sump tion re - sult ing in rapid weight gain (greater than 4 kg for men or greater for 3.5 kg for women or in ex cess of 5% of base line weight) on more than 3 oc ca sions since ad mis sion to hos pi - tal past or cur rent res i dence on the spe cial ized polydipsia unit will ing ness to par tic i pate in the study ab sence of a doc u mented phys i cal health con di tion that might ac count for the polydipsia (for ex am ple, diabetes, chronic re nal fail ure, or con ges tive heart fail ure). All pa tients who met the in clu sion cri te ria were matched with con trol sub jects for age (plus or mi nus 5 years), sex, and race. All in ter views were con ducted in clear con scious ness. The Battleford Health Dis trict Ethics Com mit tee ap proved the re - search pro to col, and all pa tients gave in formed con sent to par tic i pate in the study. The Pos i tive and Neg a tive Syn drome Scale (PANSS) (9) was used to measure ill ness se ver ity. The Mini-Mental State Ex - am i na tion Scale (MMSE) (10) was used to screen for cog ni - tive im pair ment. The se ver ity of dyskinetic move ments was mea sured us ing the Ab nor mal In vol un tary Move ment Scale (AIMS) (11). Pa tients charts were re viewed to ob tain in for - ma tion on de mo graphic char ac ter is tics, past psy chi at ric ill - ness, and med i ca tion his tory. Ei ther the pa tient or a rel a tive went through the Fam ily His tory Re search Di ag nos tic Cri te - ria (FH-RDC) (12) in ter view with the sec ond au thor to es tab - lish psy chi at ric di ag no ses in their first-de gree rel a tives. All but the first au thor were blind to the pa tient group ings. The fol low ing bi o log i cal pa ram e ters were ob tained on all pa tients

39 524 The Canadian Journal of Psychiatry Vol 46, No 6 Table 2. Illness characteristics of patients with, and without, polydipsia Characteristics Group with poly dip sia (SD) Group without polydipsia (SD) t P Mean age At first psy chi at ric con tact (years) At first psy chi at ric ad mis sion (years) 18.3 (3.9) 19.6 (4.2) (6.4) 24.4 (6.3) ns Current Admission Mean age on ad mis sion (years) Mean du ra tion (years) 31.5 (9.9) 11.8 (8.8) 37.4 (8.7) 7.5 (8.6) ns Mean number of previous admissions 10.6 (8.7) 9.9 (7.8) 0.35 ns Table 3. Pharmacotherapy history Medication variables Group with poly dip sia n = 34 (%) Group with out polydipsia n = 34 (%) χ 2 P Current antipsychotic medication Typi cal Atypi cal Com bi na tion Concomitant medication Ben zo diap zepine An ti de pres sant An ti manic (ex clud ing lith ium) An ti cho liner gic Documented Response to previous antipsychotics Full re sponse Par tial re sponse Poor re sponse Un known Past treatment with electroconvulsive therapy 11 (32.3) 9 (26.5) 14 (41.2) 5 (14.7) 4 (11.8) 12 (35.3) 15 (44.1) 4 (11.8) 11 (32.4) 10 (29.4) 9 (26.4) 4 (11.8) 15 (44.1) 8 (23.5) 10 (29.4) 4 (11.8) 8 (23.5) 14 (41.2) 13 (38.2) 8 (23.5) 18 (52.9) 4 (11.8) 4 (11.8) 5 (14.7) ns ns ns ns 0.02 ns ns 0.02 ns ns ns ns dur ing the data col lec tion: 72-hour, twice-daily, urine spe - cific grav ity and se rum so dium level. Data were en tered into Epi Info Ver sion 6 (13), and ba sic anal y ses, in clud ing fre quency dis tri bu tion and tests of sta tis - ti cal sig nif i cance, were car ried out at the 5% level; we used Pearson s chi-square test to study as so ci a tion be tween cat e - gor i cal vari ables and the 2-tailed t-test to com pare means. Anal y sis was done us ing a 95%CI. Ob served sig nif i cant lev - els were cor rected by ad just ing the num ber of com par i sons made. Results The multidisciplinary clin i cal team iden ti fied 34 pa tients (27 men and 7 women) as having polydipsia and meet ing the in - clu sion cri te ria. This gave a polydipsia prev a lence rate of 20.2% at the time of data col lec tion. The group with polydipsia was matched with 34 pa tients with out polydipsia. The 2 groups were matched for age (with polydipsia, [SD 7.5] years; with out polydipsia, [SD 11.0] years), sex, and eth nic ity. A com par i son of the 2 groups re vealed no sta tis ti cally sig nif i cant dif fer ences in premorbid level of edu - ca tion, mar i tal sta tus, preadmission em ploy ment sta tus, cur rent fi nan cial com pe tence sta tus, and le gal sta tus on ad - mis sion (Ta ble 1). Ta ble 2 shows the ill ness char ac ter is tics of the 2 groups. The group with polydipsia was sig nif i cantly youn ger at the time of their first-ever psy chi at ric and current psy chi at ric ad mis - sions, com pared with the group with out polydipsia. The group with polydipsia also had lon ger ad mis sions, com pared with their un af fected coun ter parts. Over all, the 2 groups were sim i lar in terms of their psy chi at ric di ag no ses, with 85.3% (29) of the group with polydipsia and over 73.5% (25) of the un af fected group meeting the DSM-IV di ag nos tic cri te ria for schizo phre nia and schizoaffective dis or ders (14). Bi po lar dis or der was di ag nosed in 1 (2.9%) of the pa tients with polydipsia and 5 (14.7%) of the pa tients without polydipsia. Four (11.7%) pa tients in each group met the di ag nos tic cri te - ria of other psy chi at ric dis or ders, mainly psy chotic dis or der not oth er wise specified (NOS) and chronic anx i ety dis or der. Twelve pa tients (35.3%) in the group with polydipsia and 11 pa tients (32.4%) in the un af fected group met the DSM-IV cri - te ria for comorbid sub stance-re lated dis or der (χ 2 = 0.07, P = 0.79). Seven (20.5%) and 5 (14.7%) spe cif i cally met the diag - nos tic cri te ria for al co hol abuse in the group with, and with out polydipsia, re spec tively. Among those with polydipsia and

40 August 2001 Polydipsia, Psychosis, and Familial Psychopathology 525 Table 4. Psychopathology profile of patients with, and without, polydipsia, suffering from schizophrenia and schizoaffective disorder Psychopathology profile Group with poly dip sia (n = 29) Group with out poly dip sia (n = 25) t P Positive subscale ns Negative subscale ns Positive and negative symptom scale (PANSS) composite score ns General psychopathology subscale ns Anergia ns Thought disturbance ns Activation ns Paranoid belligerence ns Depression ns Mini-mental state examination (MMSE) score ns Table 5. Characteristics of first-degree relatives of patients with, and without, polydipsia Characteristics First-de gree rela tives of pa tients with polydipsia (n = 204) (%) First-de gree rela tives of pa tients with out polydipsia (n = 172) (%) χ 2 P Sex Male Fe male 98 (48.0) 106 (52.0) 93 (54.0) 79 (46) ns ns Relationship Par ents Sib lings Off spring 68 (33.3) 136 (66.7) 0 (0.0) 59 (34.3) 112 (65.1) 1 (0.6) ns ns FH-RDC Diagnoses Schizo phre nia De pres sion Bi po lar dis or der Drug mis use Al co hol de pend ence Other psy chi at ric dis or der 6 (2.9) 17 (8.3) 0 (0.0) 13 (6.4) 47 (23.0) 3 (1.5) 3 (1.7) 10 (5.8) 3 (1.7) 4 (2.3) 23 (13.4) 6 (3.5) ns ns ns ns ns FH-RDC = Fam ily his tory re search di ag nos tic criteria with an al co hol-re lated prob lem, al co hol abuse pre dated the psy chotic ill ness by a mean of 10.5 (SD = 3.4) years, com - pared with 4.8 (SD = 1.6) years for the same pe riod in the group with out polydipsia. As far as the pharmacotherapy his tory was con cerned, all the pa tients in volved in this study were on antipsychotic med i ca - tion at the time of data col lec tion, but the group with polydipsia had had on av er age a trial of 4 dif fer ent antipsychotics, com pared with 3 in the un af fected group (t = 2.3, P = 0.02, (95%CI, to 0.104). Sim i lar pro por tions were on typ i cal or atyp i cal antipsychotics, or com bi na tions of both, at the time of data col lec tion. The 2 groups did not dif fer sig nif i cantly re gard ing the antipsychotic med i ca tion mean dos age (with polydipsia,700 mg [SD = 350 mg]; with out polydipsia, 650 mg [SD = 200 mg] chlorpromazine equiv a lents); the pro por tion on con com i tant anticholinergic med i ca tion; or past treat ment with electroconvulsive therapy (ECT). Sig nif i cantly fewer pa tients with polydipsia had doc - u mented full re sponse to med i ca tion in the past (χ 2 = 1.62, P = 0.02). None of the pa tients in volved in the study were on lith - ium ther apy. A sig nif i cantly higher pro por tion of the un af - fected group was on con com i tant an ti de pres sant med i ca tion, com pared with the group with polydipsia (χ 2 = 1.62, P = 0.02). All the pa tients on con com i tant an ti de pres sants had a di ag no sis of either schizoaffective disorder or bi po lar disor - der with psy chotic fea tures (Ta ble 3). Ta ble 4 shows the se ver ity of psychopathology in pa tients with schizo phre nia or schizoaffective dis or der in both groups. The group with polydipsia had higher mean scores on the pos i tive, neg a tive, and gen eral psychopathology

41 526 The Canadian Journal of Psychiatry Vol 46, No 6 subscales of PANSS, as well as symp tom clus ters of anergia and thought dis tur bance. The dif fer ences did not reach sta tis - ti cally sig nif i cant lev els, how ever. The MMSE scale also re - vealed no sig nif i cant dif fer ence be tween the 2 groups. The 2 groups were sim i lar in terms of the se ver ity of dyskinetic move ments as mea sured on the AIMS (t = 1.04, P = 0.30). The group with polydipsia had sig nif i cantly less con cen trated urine (1.011), com pared with the group with out polydipsia (1.016) (t = 3.49, P = 0.001). The se rum so dium lev els were, how ever, sim i lar in the 2 groups (139 mmol/l and 140 mmol/l, re spec tively). Ta ble 5 shows the sex, re la tion ship, and di ag nos tic cat e go ries of the first-de gree rel a tives. For the 68 probands in volved in the study, de tails were avail able on 376 (87% of to tal sam ple) first-de gree rel a tives. Over all, the first-de gree rel a tives of the 2 groups were sim i lar, al though the group with out polydipsia had fewer living par ents com pared with the group having polydipsia. Across-group com par i son re vealed a sta tis ti cally sig nif i cant higher prev a lence rate of al co hol de pend ence among the first-de gree rel a tives of those with polydipsia com pared with the un af fected group. Al though there were al - most equal num bers of sui cide at tempt ers in the 2 groups, the first-de gree rel a tives of pa tients with polydipsia had more sui cide attempts. Dis cus sion In this study, the polydipsia prev a lence rate of 20.2% com - pares with 17.5% re ported in US psy chi at ric in sti tu tions (2). The higher rate may re flect the study de sign, be cause our def - i ni tion of polydipsia was based on nurses ob ser va tions of ac - tual reg u lar ex ces sive fluid con sump tion or be hav iour sug ges tive of reg u lar ex ces sive fluid in take. This study, like many oth ers of its kind, ex cluded pa tients with known causes of polydipsia, in clud ing pa tients on lithium ther apy. The higher rate of polydipsia in this study may also be ac counted for by the fact that the sam ple in cluded all pa tients iden ti fied with life time polydipsia, even though they may not neces - sar ily have had acute polydipsia when the data were col - lected. The fo cus of this study was, how ever, not to mea sure the ab so lute prev a lence of polydipsia but in stead to com pare groups with, and without polydipsia. De mo graphically, the 2 groups were sim i lar. As far as di ag nos tic and ill ness char ac ter is tics were con - cerned, the find ings in this study are in keep ing with pre vi ous re ports. Slightly over 85% of the group with polydipsia met the di ag nos tic cri te ria for schizo phre nia or schizoaffective dis or der, which is sim i lar to the findings of Jose and oth ers (1) and Blum and oth ers (2). Most pa tients in the 2 groups had schizo phre nia or schizoaffective dis or der that was mod er - ately se vere but rel a tively sta ble. The ex clu sion of pa tients on lith ium from both groups might have re duced the rate of bipo - lar dis or ders. Al though sim i larly high rates of sub stance-re - lated comorbidity were found in the 2 groups, first-de gree Clinical Im pli ca tions Chronic-care psychiatric patients need regu lar monitoring for evi - dence of polydipsia. Familial alcohol-related problems may be a risk factor for the devel - opment of polydipsia in chronic-care psychiatric inpatients. Families should be involved in assessing and treat ing psychiatric pa - tients with polydipsia. Limi ta tions The study design was nonprospective. The sample was small. Investigation was limited to a few biological correlates of psycho - genic polydipsia. rel a tives of pa tients with polydipsia had sig nif i cantly higher rates of al co hol de pend ence, based on the FH-RDC in ter - view. Ripley and oth ers (15) re ported that 8 (47.0%) of their 17 pa tients with polydipsia had first-de gree rel a tives who met one or more Feighner cri te ria for al co hol ism, com pared with 2 (11.8%) in the control group. These find ings, and the fact that al co hol abuse started ear lier among the group with polydipsia, may sup port the hy poth e sis, pro posed by Millson and oth ers (7), that self-in duced wa ter in tox i ca tion may, like al co hol ism be ad dic tive. This may sug gest that polydipsia and al co hol ism share a sim i lar pathogenesis and may ben e fit from sim i lar treat ment in ter ven tions. The find ings in this study of sim i lar PANSS scores in both groups were, like those by Emsley and oth ers (16), con trary to the find ings of Law son and oth ers (17) and Kirch and oth ers (18), who re ported more neg a tive symptomatology in the group with dis or dered wa ter reg u la tion. This may sug gest that the 2 groups are not eas ily dis tin guish able clin i cally. We also found our group with polydipsia to be sim i lar to our un - af fected group on cog ni tive screen ing un like the pre vi ous stud ies, which reported cog ni tive im pair ment and ev i dence of struc tural brain dam age (16 18). This dif fer ence in find - ing may be ac counted for by the fact that the ear lier stud ies used more elab o rate cog ni tive as sess ment tools, while our study used only a screening tool. This study, like the one by Umbricht and oth ers (19), found no ev i dence sup port ing a pathophysiological link be tween the man i fes ta tion of ab nor - mal movement and polydipsia in chronic psy chi at ric pa tients. The ex act mech a nisms of ex ces sive fluid con sump tion in the psy chi at ric pa tients with polydipsia re main un clear; how - ever, med i ca tion ef fects like dry mouth and ex cess cen tral dopaminergic neurotransmision have been pos tu lated as pos - si ble mech a nisms (20). The find ings in this study re vealed that ex po sure to var i ous classes of antipsychotic med i ca tion and con com i tant use of anticholinergic drugs were in sig nif i - cant in dif fer en ti at ing pa tients with polydipsia from un af - fected pa tients. Con trary to the findings of Siegler and oth ers (21), who re ported a sig nif i cant as so ci a tion be tween the

42 August 2001 Polydipsia, Psychosis, and Familial Psychopathology 527 development of hyponatremia in psy chi at ric in pa tients and the use of fluoxetine and tricyclic an ti de pres sants, a sig nif i - cantly higher pro por tion of pa tients with out polydipsia, com - pared with pa tients in this study who had polydipsia, were on con com i tant an ti de pres sants. There fore, the use of an an ti de - pres sant did not ap pear to be a risk factor for the de vel op ment of polydipsia. This finding re mains valid even when the data on the 6 pa tients with polydipsia and a past his tory of hyponatremic seizures were con sid ered. The find ings in this study do no more than do pre vi ous studies to sup port or re fute cur rent hy poth e ses of fered to explain polydipsia in chronic psy chi at ric pa tients. This study, how - ever, pro vides fur ther ev i dence for its high prev a lence rate among chronic psy chi at ric in pa tient pop u la tions, par tic u larly those with schizo phre nia or schizoaffective dis or ders. Aside from the ap par ent ear lier onset of ill ness in the pa tients with polydipsia, their ill ness char ac ter is tics and se ver ity are simi - lar to those of un af fected pa tients. The first-de gree rel a tives of those with polydipsia also have a higher rate of al co hol de - pend ence, which may sug gest fa mil ial vul ner a bil ity. These find ings may sug gest the use of some of the ther a peu tic in ter - ven tions em ployed in ad dic tions treatment and fam ily in - volve ment in the care of pa tients with polydipsia, be cause re turn to the fam ily with out psychoeducation may con sti tute a high-risk sit u a tion. The eti o log i cal sig nif i cance of these find ings needs fur ther ex plo ra tion in a pro spec tively de - signed study with a larger sam ple size. References 1. Jose CJ, Perez-Cruet J. In ci dence and mor bid ity of self-in duced wa ter in tox i ca tion in state men tal hos pi tal pa tients. Am J Psy chi a try 1979;136: Blum A, Tempey FW, Lynch WJ. So matic find ings in pa tients with psy cho genic polydipsia. J Clin Psy chi a try 1983;44: de Leon J, Verghese C, Tracy JL,Josiassen, RC, Simpson GM. Polydipsia and wa ter in tox i ca tion in psy chi at ric pa tients: a re view of the ep i de mi o log i cal literature. Biol Psy chi a try 1994;35: Raskind MA, Orestein H, Chris to pher TG. Acute psy cho sis, in creased water in ges - tion and in ap pro pri ate antidiuretic hor mone se cre tion. Am J Psy chi a try 1975;132: Khreis IY, Slaugh ter JR. Risperidone for pri mary psy cho genic polydipsia. Poster presentated at the 148th an nual meeting of the Amer i can Psy chi at ric As so ci a - tion;may 1995;Miami (FL). 6. Lan dry P. Ef fect of risperidone on polydipsia and hyponatremia in schizo phre nia [let ter]. Can J Psy chi a try 1995;40: Millson RC, Emes CE, Glackman WG. Self-in duced water in tox i ca tion treated with risperidone. Can J Psy chi a try 1996;41: Leadbetter RA, Shutty Jr MS, Pavalonis D. Risperidone treat ment in polydipsia and in ter mit tent hyponatremia. Paper pre sented at the 6th in ter na tional congress on schizo phre nia re search; April 12 16, 1997; Col o rado. 9. Kay SR, Fiszbein A, Opler LA. The pos i tive and neg a tive syn drome scale (PANSS) for schizo phre nia. North Touawanda (NY): Multi-Health Sys tems, Inc; Folstein MF, Folstein SE, McHugh PR. The mini mental state ex am i na tion. J Psychychiatr Res 1975;12: Camp bell M, Palji M. Mea sure ment of side ef fects in clud ing tardive dyskinesia. Psychopharmacol Bull 1985;21: Endicott J, Andreasen N, Spitzer RL. Fam ily his tory re search di ag nos tic criteria. New York: NIMH; Dean AG, Dean J, Bur ton AH. Epi info, ver sion 6: a word pro cess ing, da ta base and sta tis ti cal pro gram for ep i de mi ol ogy on mi cro com puter. Stone Mountain (GA): USD Inc; Amer i can Psy chi at ric As so ci a tion. Di ag nos tic and sta tis ti cal man ual of men tal dis - or der. 4th ed. Wash ing ton (DC): Amer i can Psy chi at ric As so ci a tion Ripley TL, Millson RC, Koczapski AB. Self in duced water in tox i ca tion and al co - hol abuse. Am J Psy chi a try 1989;146: Emsley RA, Spangenberg JJ, Rob erts MC, Taljaard FJ, Chalton DO. Dis or dered wa ter ho meo sta sis and cog ni tive im pair ment in schizo phre nia. Biol Psychiatry1993;34: Law son WB, Karson CN, Bigelow LB. In creased urine vol ume in chronic schizo - phrenic pa tients. Psy chi a try Ref er ence 1985;14: Kirch DG, Bigelow LB, Weinberger DR, Law son WB, Wyatt RJ. Polydipsia and chronic hyponatremia in schizo phrenic in pa tients. J Clin Psy chi a try 1985;46: Umbricht DSG, Saltz B, Pol lack S, Woerner M, Kane JM, Lieberman JA. Polydipsia and tardive dyskinesia in chronic psy chi at ric pa tients re lated dis or - ders. Am J Psy chi a try 1993;150: Dourish CT, Jones RSG. Do pa mine ag o nist-in duced res to ra tion of drinking in re - sponse to hypertonic sa line in adipsic do pa mine denervated rats. Brain Res Bull 1982;8: Siegler EL, Tamres D, Berlin JA, Al len-tay lor L, Strom BL. Risk fac tors for the de vel op ment of hyponatremia in psy chi at ric in pa tients. Arch In tern Med 1995;155: Résumé Polydipsie, psychose et psychopathologie familiale Ob jec tif : Com parer les facteurs démo graphiques et clin iques ainsi que la psy cho pa tholo gie famili ale des pa tients psy - chia tri ques chroniques hos pi tal isés souf frant ou non de poly dip sie. Méth ode : Nous avons en tre pris une étude cas- témoin de pa tients psy chia tri ques chroniques hos pi tal isés dont cer tains étaient at te ints de poly dip sie et d autres, pas. Les données clin iques et démo graphiques ont été re cueil lies à l aide d un ques tion naire prédé ter miné, de l é chelle des syn dromes po si tifs et né ga tifs (PANSS), de l é chelle des mou ve ments in vo - lon taires anor maux (AIMS), du mini- examen de l état men tal (MMSE) et des critères di ag nos tiques de re cher che des anté cédents fa mili aux (FH- RDC). Résul tats : Le taux de préva lence de la poly dip sie était de 20,2 %. Le groupe comptant pa tients at te i nts de poly dip sie était sig ni fi ca tive ment plus je une, tant à l époque de la pre mière hos pi tali sa tion psyc hia tri que qu à celle en cours, com para tive ment au groupe sans poly dip sie. Les 2 groupes étaient sem bla bles en ce qui con cerne les caracté ris tiques de leur mala die et les di ag nos tics psy chia tri ques. Dans le groupe souf frant de poly dip sie, la con som ma tion ex ces sive d al cool pré cédait la mala die psy cho tique d une moy enne de 10,5 (écart- type 4,4) ans, par rap port à 4,8 (écart-type 1,6) ans pour la même période chez le groupe non at te int. Les 2 groupes ne pré sen taient pas de diffé rence sig ni fi ca tive sur le plan du dos age des an tipsy cho tiques, de la pro por tion des médica ments an ti cho liner giques con comi tants et de la pré cédente réac tion docu mentée aux an tipsy cho tiques ou au traite ment passé d é lec tro choc. Les par ents au premier de gré des pa tients souf frant de poly dip sie ont révélé des taux sig ni fi ca tive ment plus élevés d al cool isme. Con clu sion : Cette étude ap porte d autres preu ves du taux plus élevé de poly dip sie chez la popu la tion des pa tients psy - chia tri ques chroniques et du taux élevé de pro blèmes liés à l al cool chez leurs par ents au pre mier de gré.

43 ORIGINAL RESEARCH Alternative Medicine Use by Individuals With Major Depression JianLi Wang, PhD 1, Scott B Pat ten, MD, PhD, FRCPC 2, Mar ga ret L Rus sell, MD, PhD 3 Ob jec tive: To de scribe the use of al ter na tive med i cine (AM) by per sons with major de pres sion and to ex am ine the fac - tors as so ci ated with AM use among these in di vid u als. Methods: We used data from the and Na tional Pop u la tion Health Sur veys. We se lected sub - jects who had ma jor de pres sion ac cord ing to the Com pos ite In ter na tional Di ag nos tic In ter view Short Form for Ma jor De pres sion (CIDI-SFMD). The prev a lence of AM and con ven tional health ser vice use by the sub jects was cal cu lated for each sur vey and was strat i fied by prov ince. We em ployed lo gis tic re gres sion to exami ne the fac tors as so ci ated with AM use. Re sults: There was a tem po ral trend to ward in creas ing use of AM among per sons with ma jor depression. The prev a - lence of AM use among sub jects with major de pres sion was 7.8% in and 12.9% in Fe male sex, hav ing more than 12 years ed u ca tion, and hav ing 1 or more long-term med i cal con di tions were as so ci ated with an in - creased like li hood of using AM. The sex dif fer ence in AM use de pended on sub jects age in Con clu sion: Gen eral prac ti tio ners, mental health spe cial ists, and AM pro vid ers should be aware of their pa tients use of both con ven tional med i cal ser vices and AM be cause there may be in ter ac tions be tween con ven tional and alternative treat ments. Com mu ni ca tion and, if pos si ble, co op er a tion may lead to im proved out comes in the man age ment of de pres - sive disorders. (Can J Psy chi a try 2001;46: ) Key Words: major depression, alternative medicine, general practitioner Al ter na tive med i cine (AM) in cludes such prac tices as chiropractic, ther a peu tic mas sage, acu punc ture, herbalism, ho me op a thy, and naturopathy (1). AM use has been stud ied in many pa tient groups, in clud ing those with can cer (2), with rheu ma toid dis ease (3), with gas tro in tes ti nal dis or ders (4), and with dermatological prob lems (5). Few stud ies, how ever, have ex am ined the ex tent to which AM is used by in di vid u als with major de pres sion. A na tional sur vey in the US re ported that 40.9% of the sub - jects who had de pres sion used AM ther a pies in the past 12 months (6). De pres sion in this study, how ever, was Manuscript received January 2001, revised, and accepted June Post-Doctoral Fellow, Department of Community Health and Epidemiol - ogy, Faculty of Medicine, Dalhousie University, Halifax, Nova Sco tia. 2 Associate Professor, Departments of Community Health Sciences and Psy - chiatry, Faculty of Medicine, University of Calgary, Calgary, Alberta. 3 Associate Professor, Department of Community Health Sciences, Faculty of Medicine, University of Calgary, Calgary, Alberta. Address for correspondence: JianLi Wang, Department of Community Health and Epidemiology, Dalhousie University, 5849 University Avenue, Halifax, NS B3H 4H7 Jwang@tupdean2.med.dal.ca self-re ported, and self-re ported de pres sion can differ from that eval u ated using struc tured di ag nos tic in stru ments. In a re cent US na tional sur vey, 22.4% of sub jects who had major de pres sion, mea sured by the WHO s Com pos ite In ter na - tional Di ag nos tic In ter view Short Form for Major De pres sion (CIDI-SFMD), reported us ing AM in the past 12 months (7). Fe male sex, mid dle age, and a higher level of ed u ca tion were found to be as so ci ated with AM use (6 8). Ca na dian data have not pre vi ously been avail able. Most pa tients with de pres sion who seek treat ment are seen by gen eral prac ti tio ners (GP) or fam ily doc tors in a pri mary care set ting (9,10). These pa tients may use both AM and con ven - tional med i cine, and con ven tional prac ti tio ners are of ten un - aware of their pa tients use of AM (11). Eisenberg and oth ers re ported that about 60% of peo ple who used AM did not dis - close their AM use to phy si cians (6). Ad di tionally, AM ap - proaches have not been tested suf fi ciently (12). Some stud ies as sert that AM may be as ef fec tive as con ven tional an ti de - pres sant treat ment among pa tients with mild and mod er ate de pres sion (13 15). A re cent ran dom ized, dou ble-blind, pla - cebo-con trolled clin i cal trial found that a widely used herbal prod uct, St John s wort (Hypericum perforatum), was not ef - fec tive for treat ment of ma jor de pres sion (16). This study Can J Psychiatry, Vol 46, August

44 August 2001 Alternative Medicine Use by Individuals With Major Depression 529 Table 1. National prevalence of alternative medicine (AM) and conventional health service in and National Population Health Sur veys (NPHS) among persons with major depression, by type of practitioner % 95%CI % 95%CI Chiropractor , , 18.1 Other AM provider , , 15.8 Any AM provider , , 26.9 GP or family doctor a , , 92.6 Medical doctor , , 94.8 Any conventional health professional b , , 94.9 a GP or family doc tors and other medi cal doctors in clud ing psy chia trists. b GP or family doc tors, other medi cal doctors in clud ing psy chia trists, nurses, so cial worke rs and coun sel lors, and psy cholo gists. used a Ham il ton De pres sion Rating Scale (HDRS) score of 20 or greater as an el i gi bil ity cri te rion and may have evalu - ated a group of pa tients with more se vere de pres sion than did pre vi ous stud ies. There are also con cerns about the pos si ble neg a tive ef fect of AM on con ven tional man age ment of mood dis or ders (17). Some stud ies re port that St John s wort com - bined with trazodone, sertraline, or nefazodone may cause se - ro to nin syn drome (18 20). It may also potently af fect the ac tiv i ties of ma jor hu man drug-me tab o liz ing en zymes (21,22). Thus, with out ef fec tive com mu ni ca tion be tween con ven tional prac ti tio ners and AM pro vid ers, the ef fi cacy of con ven tional an ti de pres sant treat ment could be af fected by AM use. Given the pos si ble in ter ac tion be tween AM prod ucts and an - ti de pres sants, it is im por tant to un der stand both the fac tors re - lated to AM use and its pat tern among per sons with major de pres sion in Can ada. The ob jec tive of this anal y sis was to es ti mate the prev a lence of AM use in Canada among in di vid - u als with major de pres sion and to ex am ine the de mo graphic and clin i cal fac tors as so ci ated with AM use. Methods Data from the and Ca na dian Na tional Pop u la tion Health Sur vey (NPHS) were used in this anal y sis. The tar get pop u la tion of the NPHS house hold com po nent was house hold res i dents in all prov inces, ex clud ing those on In dian re serves or mil i tary bases, in the Yu kon or North west Ter ri tories, in some re mote ar eas of Que bec and On tario, and in long-term in sti tu tions. The NPHS meth od ol ogy has been de scribed in detail else where (23 26). Only the sub jects with ma jor de pres sion in one or both sur veys were in cluded in this anal y sis. In the NPHS, major de pres sion was mea sured by the (CIDI-SFMD). The CIDI-SFMD was de rived from the full ver sion of the CIDI and val i dated by Kessler and col leagues (27). In the NPHS, 1043 ( ) and 3133 ( ) in di vid u als were iden ti fied as hav ing had an ep i sode of major de pres sion in the pre ced ing 12 months, based on the CIDI-SFMD. Re spon dents to the NPHS were asked, In the past 12 months, have you seen or talked to an al ter na tive health care pro vider such as an acu punc tur ist, ho meo path or mas sage ther a pist about your phys i cal, emo tional or mental health? If the sub - jects an swered yes to this ques tion, they were also asked whether they had vis ited the fol low ing AM pro vid ers in the past 12 months: a mas sage ther a pist, an acu punc tur ist, a ho - meo path or na tu ro path, a feldenkrais or al ex an der teacher, a re lax ation ther a pist, a bio feed back teacher, a rolfer, a herb al - ist, a reflexologist, a spir i tual healer, a re li gious healer, or any other AM pro vider. In the NPHS, in for ma tion about chiro - prac tor visits in the past 12 months was col lected in the con - ven tional health care use com po nent. In our anal y sis, chiropractic was con sid ered a cat e gory of AM. For compari - son pur poses, a sep a rate cat e gory of AM ( other AM ) ex - clud ing chiropractic was also cre ated. The rates of AM vis its in dif fer ent AM cat e go ries were cal cu lated. In the NPHS, the sub jects were also asked, In the past 12 months, how many times have you seen or talked on the tele - phone with a fam ily doctor or gen eral prac ti tio ner about your phys i cal, emotional or men tal health? In our anal y sis, an an - swer of 1 or more times to this ques tion was de fined as visit - ing GPs or fam ily doc tors. The NPHS also in quired about the fre quen cies of talk ing to or vis it ing other med i cal doc tors (in - clud ing psy chi a trists), nurses, so cial work ers or counsellors, and psy chol o gists about phys i cal, emo tional, or men tal health in the past 12 months. We cal cu lated sep a rately the rates of vis it ing each of the fol low ing: GP or fam ily doctors, GP or fam ily doc tors and other med i cal doc tors, and any health pro fes sion als. We es ti mated the prev a lence of AM and con ven tional medi - cal care-pro vider vis its in each of the 2 sur veys. The rates

45 530 The Canadian Journal of Psychiatry Vol 46, No 6 Table 2. The prevalence of AM visits by persons with major depression in each survey by region (95%CI) At lan tic prov inces Que bec On tario Prai rie prov inces Brit ish Co lum bia % 95%CI % 95%CI % 95%CI % 95%CI % 95%CI Chi ro prac tors , 18.4 a ,16.9 a , 27.0 a , 34.8 a Other AM providers , 9.2 a , 22.7 a , 9.1 a , 17.0 a Any AM pro vid ers , 9.8 a , , , , Chi ro prac tors , 18.8 a , 17.3 a , 27.9 a , 35.9 a Other AM providers , , 21.0 a , , , 25.7 a Any AM pro vid ers , 14.1 a , 30.5 a , 26.3 a , , 48.6 a In di cates that the co ef fi cient of sampling vari ance of the es ti mate is be tween 16.3% and 33.3%; indicates that the co ef fi cient of sam pling vari ance of the es ti mate is greater than 33.3%, thus, not pre sented. were fur ther bro ken down for pro vin cial com par i son, where pos si ble. The es ti mates were weighted to ac count for the com plex de sign of the NPHS (24). Lo gis tic re gres sion mod - el ling was used to in ves ti gate the de mo graphic and clin i cal fac tors as so ci ated with AM use. These fac tors in cluded sex, age, mar i tal sta tus, ed u ca tion (12 years or less school educa - tion vs more than 12 years ed u ca tion), in come, em ploy ment, re gion (ur ban vs ru ral), chronic con di tions (1 or more long-term con di tions vs no long-term med i cal con di tions), and an ti de pres sant use (used an ti de pres sants vs did not use an ti de pres sants) in the past 12 months. This anal y sis was per - formed using STATA 6.0 (28). Results Ta ble 1 pres ents the prev a lence of vis it ing AM and con ven - tional care pro vid ers among the sub jects with major de pres - sion in each sur vey. Most of these subjects re ported making 1 or more vis its to con ven tional health care pro vid ers. The rates of con ven tional care pro vider use in the 2 sur veys did not change dra mat i cally be tween and The rates of vis it ing chi ro prac tors in the past 12 months did not differ be tween the sur veys. In con trast, the prev a lence of vis it ing other AM pro vid ers (ex clud ing chi ro prac tors) in was higher than that in On a pro vin cial level, sev eral es ti mates of AM use for the At - lan tic prov inces and On tario achieved in ad e quate pre ci sion in ; there fore, they are not pre sented. As Ta ble 2 shows, there was a trend of de creas ing AM use from Western to East ern prov inces in ; in , this trend was not clearly ev i dent. The rates of vis it ing AM pro vid ers were com pa ra ble in the At lan tic prov inces and in Que bec. In Brit ish Co lum bia, the prev a lence of vis it ing other AM pro - vid ers in ap peared to be higher than that in Sam pling vari abil ity cannot be ex cluded as an al - ter na tive ex pla na tion for this dif fer ence. In On tario and the Prai rie prov inces, there was an in creased rate of vis it ing other AM pro vid ers in , com pared with the sur vey. Cor re lates of AM use were ex plored for Lo gis tic re gres sion in di cated that sub jects who had 1 or more chronic con di tions (ad justed odds ra tio = 1.98; 95%CI, 1.22 to 3.23) and those who were be tween age 30 and 44 years (ad justed odds ra tio = 2.25; 95%CI, 1.11 to 4.53) were more likely to visit any AM pro vider in When chi ro prac tors were ex cluded, how ever, age and hav ing 1 or more chronic con di tions were no lon ger as so ci ated with AM vis its. Fe male sex (ad justed odds ra tio = 3.86; 95%CI, 1.77 to 8.51) and hav - ing higher ed u ca tion (ad justed odds ra tio = 2.30; 95%CI, 1.28 to 4.12) were, how ever, found to be as so ci ated with AM vis - its. None of the re main ing vari ables acted as con found ers or as ef fect mod i fi ers of these as so ci a tions. In , per sons who had 1 or more chronic con di tions (ad justed odds ra tio = 2.71; 95%CI, 1.77 to 4.17) and those who had higher ed u ca tion (ad justed odds ra tio = 1.64; 95%CI, 1.10 to 2.44) were more likely to use any AM ser vice. When chi ro prac tors were ex cluded, lo gis tic re gres sion mod - el ling in di cated a sex age in ter ac tion (like li hood ra tio test, χ 2 = 5.98, df 1, P = 0.01), which sug gested that the sex dif fer ence in AM use de pended on age. The sex age-spe cific rates of other AM use in are pre sented in Fig ure 1. The rate of other AM use among women in creased with age un til the age of 44 years, then de creased slightly. This pattern was not ob served among men. Women who were aged 12 to 19 years and those who were aged 65 years or older had a higher rate of other AM use than did men in the same age groups. Sim i larly, women who were be tween age 30 and 44 years were more likely to visit other AM pro vid ers than were men. Lo gis tic re gres sion also showed that hav ing more than 12 years ed u ca tion (ad justed odds ra tio = 2.39; 95%CI, 1.42 to 4.04) and having 1 or more chronic con di tions (ad justed odds ra tio = 2.63; 95%CI, 1.48 to 4.68) con tin ued to be as so ci ated

46 August 2001 Alternative Medicine Use by Individuals With Major Depression 531 Men with MDE Women with MDE Age Figure 1. Sex age-specific rates of AM use among subjects with major depression in (excluding chiropractor visits). with AM use in models in clud ing age, sex, and age sex in ter - ac tion terms. Discussion The re sults of this analysis con firmed that most subjects who had ma jor de pres sion were seen by GP or fam ily doc tors in the pe ri ods and There was no dra - matic change in vis its to chi ro prac tors among these sub jects. In creasing use of other AM pro vid ers by these sub jects was found, how ever. There was also re gional variation in vis its to AM pro vid ers, with higher AM use in the west ern prov inces than was found in the east ern prov inces in Can ada. The data in di cated that, among these sub jects, the ef fects of sex, age, ed u ca tion, and chronic con di tions on rates of visiting AM pro vid ers de pended on how AM was de fined in this anal y sis (that is, whether chiropractic was in cluded or ex cluded). The in creas ing use of AM (ex clud ing chiropractic) among per sons with ma jor de pres sion may be at trib ut able to many fac tors. First, me dia in ter est in this topic may be par tially re - spon si ble (29). Sec ond, those with more ed u ca tion may be more ex posed to var i ous in for ma tion sources (for ex am ple, books, jour nal ar ti cles, or the Internet), which may prompt their in ter est in un con ven tional ther a pies (30). This as so ci a - tion has been con sis tently ob served in pre vi ous stud ies (6,7). Hav ing long-term med i cal con di tions may be an other factor lead ing to AM use in per sons with major de pres sion. Re - search sug gests that in di vid u als who have chronic phys i cal ill nesses and who have less suc cess in the con ven tional health care sys tem tend to ex per i ment with dif fer ent types of health care pro vider (31). We found that 73.1% of the sub jects in - cluded in this anal y sis had 1 or more long-term med i cal con - di tions and that having 1 or more chronic con di tions was in de pend ently as so ci ated with AM use. Astin re ported that chronic pain, fa tigue, and head aches are the ma jor health prob lems leading to AM use in the US (32). These health prob lems, how ever, may also be the so matic man i fes ta tions of major de pres sion. It is pos si ble that some per sons with ma - jor de pres sion may use AM ther a pies pri mar ily to re lieve their de pres sive symp toms. In Astin s study, it was found that de pres sion was also a major health prob lem prompting peo - ple to use AM (32). Some stud ies have dem on strated that St John s wort ef fec - tively treats mi nor and mild de pres sion and some times has fewer side ef fects than an ti de pres sant treatment (13 15). Per - ceived ef fi cacy may be a mo ti va tion for vis it ing AM

47 532 The Canadian Journal of Psychiatry Vol 46, No 6 pro vid ers. It is also pos si ble that per sons with major de pres - sion per ceive AM to be a sup ple ment to con ven tional medi - cine or a method to pro mote their over all health, rather than a treat ment for spe cific ill nesses. Finally, some may think that AM treat ment is more nat u ral than pharmaceuticals. The mech a nism un der ly ing the sex-spe cific change in the age-re lated AM use pattern is prob a bly com plex. Astin sug - gested that the de ci sion to use AM was highly con text- or situ - a tion-de pend ent (32). Its use among chil dren and ad o les cents may be largely due to par ents in flu ence. For ex am ple, an Al - berta study found strong within-fam ily clus ter ing of AM use (33). Sim i larly, an other study, con ducted in a pe di at ric clinic in Que bec, also in di cated that chil dren s AM use was strongly as so ci ated with par ents AM use (30). The ex ist ing lit er a ture can not, how ever, clearly explain why sex dif fer ence in AM use var ies with age, as ob served in this analysis. The re gional variation of AM use that we ob served is con sis - tent with a pre vi ous Ca na dian re port, sug gest ing that this dif - fer ence may be re lated to the funding for chiropractic ser vices pro vided un der pro vin cial health in sur ance plans in British Co lum bia, On tario, and the Prai rie prov inces (34). Chiropractic ser vices are not pub licly funded in Quebec and the At lan tic prov inces. Dif fer ences in pop u la tion com po si - tion may also con trib ute to this vari a tion. Most new im mi - grants re side in Brit ish Co lum bia, the Prai rie provinces, On tario, and Que bec (35). Eu ro pean im mi grants are more likely than Asian im mi grants to stay in Que bec, and most Asian im mi grants are in Brit ish Co lum bia, On tario, and Al - berta (35). Various cul tural or i gins may af fect AM use in these regions. In this anal y sis, the fre quency of AM use could have been un - der es ti mated, partly be cause of the dis crep an cies in its ex act def i ni tion and its narrow def i ni tion in the NPHS. This analy - sis did not ex am ine the pro por tions of spe cific AM use among its subjects be cause the num ber of sub jects in each cat e gory was small. Ad di tionally, ma jor de pres sion mea sured in the NPHS re ferred to the past 12 months. The chronicity of major de pres sion among the sub jects was un known. Fu ture stud ies are nec es sary to ex am ine whether the chronicity of ma jor de - pres sion af fects the AM use. GP or fam ily doc tors and psy chi a trists should be aware that use of var i ous AM ther a pies among their pa tients is in creas - ing. Given the pos si ble in ter ac tions be tween some AM prod - ucts and con ven tional an ti de pres sants, GP or fam ily doc tors and psy chi a trists should ask their pa tients about their AM use so that the chance of dan gers and mis un der stand ings can be min i mized. References 1. Crellin JK, Andersen RR, and Connor JTH. Gen eral in tro duc tion: a pot pour ris of is - sues. In: Crellin JK, Andersen RR, and Connor JTH, ed i tors. Al ter na tive health care in Can ada: nine teenth and twen ti eth century per spec tives. To ronto: Ca na dian Scholar s Press Inc; p 3. Clinical Implications Individuals with major depression are increasingly using alternative medicine (AM) serv ices. General practitioners and mental health specialists should be aware of AM use among their pa tients. Despite occasional animosity between traditional and alternative medicine providers, cooperation and communication may become an imperative. Limi ta tions The prevalence of AM use may be underestimated in this analy sis because a standard definition for alternative medicine (AM) is lack - ing. Specific types of AM used by the subjects, and their motivation for it, could not be examined. The data are self-reported; therefore, reporting bias is possible. 2. Cam pion EW. Why un con ven tional med i cine? [ed i to rial] N Engl J Med 1993;328: Fahner JB. Phy si cians and heal ers: un wit ting part ners in health care. N Engl J Med 1992;326: Furnham A, Smith C. Choosing al ter na tive med i cine: a com par i son of the be liefs of pa tients vis it ing a gen eral prac ti tio ner and a ho meo path. Soc Sci Med 1988;26: Rousseau N, Saillant F, Desjardins D. Socio-de mo graphic and socio-pro fes sional pro file of ho lis tic theraphists in Que bec. Can J Pub lic Health 1991;82: Eisenberg DM, Da vis RB, Ettner SL, Ap pel S, Wikey S, Van Rompay M, and oth - ers. Trends in al ter na tive med i cine use in the United States, JAMA 1998;280: Unutzer J, Klap R, Sturm R, Young AS, Marmon T, Shatkin J, and oth ers. Mental dis or ders and the use of al ter na tive med i cine: re sults from a na tional sur vey. Am J Psy chi a try 2000;157: Kelner M, Wellman B. Who seeks al ter na tive health care? A pro file of the us ers of five modes of treat ment. Jour nal of Al ter na tive and Com ple men tary Med i cine 1997;3: Si mon GE, VonKorff M, Barlow W. Health care costs of pri mary care pa tients with rec og nized de pres sion. Arch Gen Psy chi a try 1995;52: Unutzer J, Patrick DL, Si mon G, Grembowski D, Walker E, Rutter C, Katon W. De pres sive symp toms and the costs of health ser vices in HMO pa tients aged 65 years and older: a 4-year pro spec tive study. JAMA 1997;277: Thomas KJ, Carr J, West lake L, Wil liams BT. Use of non-or tho dox and con ven - tional health care in Great Britain. BMJ 1991;302: Askster CW. Con cepts in al ter na tive med i cine. Soc Sci Med 1986;22: Woelk H. Com par i son of St John s wort and imipramine for treat ing de pres sion: ran dom ised con trolled trial. BMJ 2000;321(7260): Philipp M, Kohnen R, Hiller KO. Hypericum extract ver sus imipramine or pla cebo in pa tients with mod er ate de pres sion: ran dom ised multicentre study of treat ment for eight weeks. BMJ 1999;319(7224): Wheatley D. LI 160, an ex tract of St John s wort, ver sus amitriptyline in mildly to mod er ately de pressed out pa tients a con trolled 6 week clin i cal trail. Pharmocopsychiatry 1997;30 (Suppl 2): Shelton RC, Keller MB, Gelenberg A, Dunner DL, Hirschfeld R, Thase ME, and oth ers. Ef fec tive ness of St John s Wort in ma jor de pres sion: a ran dom ized con - trolled trial. JAMA 2001;285: Sov er eign AE. Na tu ro pathic ser vices. Can Med Assoc J 1990;142: Demott K. St John s wort tied to se ro to nin syn drome. Clin i cal Psy chi a try News 1998;26: Lantz MS, Buchalter E, Giambanco V. St John s wort and an ti de pres sant drug in - ter ac tions in the el derly. J Geriatr Psy chi a try Neurol 1999;12: Martin TG. Se ro to nin syn drome. Ann Emerg Med 1996;28: Obach RS. In hi bi tion of hu man cytochrome P450 en zymes by con stit u ents of St John s wort, an herbal prep a ra tion used in the treatment of de pres sion. J Pharmacol Exp Ther 2000;294: Du.rr D, Stieger B, Kullak-Ublick GA, Rentsch KM, Steinert HC, Meier PJ, and oth ers. St John s Wort in duces in tes ti nal P-glycoprotein/MDRI and in tes ti nal and hepatic CYP3A4. Clin Pharmacol Ther 2000;68: Pat ten SB, Charney DA. Al co hol con sump tion and ma jor de pres sion in the Ca na - dian pop u la tion. Can J Psy chi a try 1998;43: Sta tis tics Canada. Na tional pop u la tion health sur vey over view, 1994/95. Ot tawa: Min is try of In dus try; Cat a logue No Beaudet MP. De pres sion. Health Re port 1996;7(4):11 24.

48 August 2001 Alternative Medicine Use by Individuals With Major Depression Wade TJ, Cairney J. Age and de pres sion in a na tion ally rep re sen ta tive sam ple of Ca na di ans: A pre lim i nary look at the na tional pop u la tion health sur vey. Can J Pub - lic Health 1997;88: Kessler RC, An drews G, Mroczek D, Ustun B, Wittchen H. The World Health Or - ga ni za tion com pos ite in ter na tional di ag nos tic in ter view short form (CIDI-SF). In - ter na tional Jour nal of Methods in Psy chi at ric Re search 1998;7: STATA 6.0. Col lege Sta tion, Texas: Stata Corportion; Klymchuck K. Club price. Info Club 1992;5: Spigelblatt L, Laine-Ammara G, Pless IB, Guyver A. The use of al ter na tive med i - cine by chil dren. Pae di at rics 1994;94: Lipowski ZJ. Somatization: the con cept and its clin i cal ap pli ca tion. Am J Psy chi a - try 1988;145: Astin JA. Why pa tients use al ter na tive med i cine: re sults of a na tional study. JAMA 1998;279: Verhoef MJ, Rus sell ML, Love EJ. Al ter na tive med i cine use in ru ral Al berta. Can J Pub lic Health 1994;85: Millar WJ. Use of al ter na tive health care prac ti tio ners by Ca na dian. Can J Public Health 1997;88: De part ment of Cit i zen ship and Im mi gra tion. Im mi gra tion sta tis tics, Ot tawa: Min is ter of Public Works and Gov ern ment Ser vices Can ada; Cat a logue No. MP22-1/1996. Résumé Recours à la médecine douce par les personnes souffrant de dépression majeure Ob jec tif : Dé crire l u tili sa tion de la médecine douce (MD) par les per son nes souf frant de dépression ma jeure et exam - iner les fac teurs as so ciés au re cours à la MD chez ces per son nes. Méth odes : Nous avons utilisé les données des En quêtes na tion ales sur la santé de la popu la tion de et Nous avons choisi des sujets souf frant de dépres sion ma jeure selon la forme abrégée pour la dépres sion ma jeure du Com pos ite In ter na tional Di ag nos tic In ter view (CIDI- SFMD). La préva lence du re cours à la MD et aux serv ices de santé tra di tion nels chez les su jets a été cal culée pour chaque en quête et strati fiée par prov ince. Nous avons em ployé la ré gres sion lo gis tique pour ex am iner les fac teurs as so ciés à l u tili sa tion de la MD. Résul tats : Il y avait une ten dance tem porelle à l ac crois se ment de l u tili sa tion de la MD chez les per son nes souf frant de dépression ma jeure. La préva lence du re cours à la MD chez ces sujets était de 7,8 % en et de 12,9 % en Les femmes ayant plus de 12 ans de sco larité et 1 trou ble médi cal à long terme ou plus étaient as so ciées à une prob abil ité ac crue d u tili sa tion de la MD. La diffé rence selon le sexe d u tili sa tion de la MD re po sait sur l âge des su jets, en Con clu sion : Les gé né ral istes, les spé ci al istes de la santé men tale et les fournis seurs de MD doivent être au cou rant de l u tili sa tion que font leurs pa tients tant des services médi caux tra di tion nels que de la MD, parce qu il peut y avoir des in ter ac tions en tre les traite ments tra di tion nels et par allèles. La com mu ni ca tion et, si pos si ble, la col labo ra tion peu vent me ner à de meilleurs résul tats du traite ment du trou ble dépressif.

49 ORIGINAL RESEARCH Risperidone Treatment of Outpatients With Schizophrenia: No Evidence of Sex Differences in Treatment Response Alain Labelle, MD, FRCPC 1, Margaret Light, PhD 2, Fiona Dunbar, MB, BCh 3 Ob jec tive: Given the re newed in ter est in the role of sex dif fer ences in schizo phre nia, we un der took a post hoc analysis to de ter mine whether sex dif fer ences in treatment re sponse were pres ent among out pa tients with schizo phre nia who re - ceived risperidone in an 8-week, open-label, Phase IV clin i cal study. Method: We eval u ated 330 adult pa tients (232 men, 98 women) with a DSM-III-R di ag no sis of schizo phrenia for safety and 292 (206 men, 86 women) for ef fi cacy. Antipsychotic and antiparkinsonian med i ca tions were dis con tin ued at study en try. Treat ment with risperidone was ini ti ated at a dos age of 2 mg daily, in creased to the tar get dos age of 6 mg daily by day 3, and main tained at 6 mg daily until day 14. The dos age was then main tained at 6 mg daily, in creased or de - creased by 2 mg daily each week, based on the pa tient s re sponse. Risperidone treat ment was given for 8 weeks; the per - mit ted dos age range was 4 mg to 10 mg daily. Re sults: Both male and fe male par tic i pants re sponded well to risperidone treat ment; by the fi nal as sess ment day, they had ex pe ri enced de creases from base line in their total Pos i tive and Neg a tive Syndrome Scale (PANSS) scores of 41.0% and 36.5%, re spec tively. Most male (77%) and fe male (78%) par tic i pants were con sid ered to be PANSS re spond ers: risperidone was ef fec tive against both the pos i tive and neg a tive symp toms of schizo phre nia. Both sexes showed im - prove ments over base line in the in ci dence and se ver ity of parkinsonism, dystonia, and dyskinesia. No sig nif i cant (P > 0.05) sex dif fer ences in treatment re sponse were ob served for any of the ef fi cacy out comes or in the in ci dence and se - ver ity of extrapyramidal symp toms (EPS). Con clu sions: In this pop u la tion of out pa tients with chronic schizo phre nia, both men and women re sponded well to flex i ble doses of risperidone. No sig nif i cant sex dif fer ences were evident either in treat ment re sponse or in neu ro log i cal side ef fects. The ab sence of sex dif fer ences in re sponse to risperidone treat ment may ob vi a te the need for a sex-based dif fer en tial dosing in schizo phre nia man age ment. (Can J Psy chi a try 2001;46: ) Key Words: risperidone, schizophrenia, gender, sex, antipsychotics, serotonin, dopamine, extrapyramidal symptoms Sex dif fer ences in schizo phre nia were first doc u mented in the early 1900s (1). Over the past 2 de cades, there has been a re newed in ter est in the role of sex in schizo phre nia, as re search ers at tempt to gain a greater un der stand ing of the het - er o ge ne ity, pathophysiology, and treat ment of this dis ease (1,2). One hy poth e sis, which has been widely studied, is that men and women are prone to dif fer ent sub types of this disor - Manuscript received October 2000, revised, and accepted June Director, Schizophrenia Program, Royal Ottawa Hospital, Ottawa, Ontario. 2 Director, Clinical Research, Janssen-Ortho Inc, Toronto, Ontario. 3 Vice-President, Clinical Affairs, Janssen-Ortho Inc, Toronto, Ontario. Address for correspondence: Dr A Labelle, Di rec tor, Schizophrenia Pro - gram, Royal Ottawa Hospital, 1145 Carling Avenue, Ottawa, ON K1Z 7K4 alabelle@rohcg.on.ca der. If so, it is probable that treat ment de ci sions in future will be based, at least in part, on a pa tient s sex. Based on early ep i de mi o log i cal data, it was thought that schizo phre nia af fects men and women with equal in ci dence; how ever, more re cent data sug gest that men may have a higher in ci dence of this ill ness than do women (1,3). Most stud ies have found that, com pared with women, men experi - ence schizo phre nia onset at an ear lier age, poorer premorbid so cial and in tel lec tual func tion ing, poorer so cial functioning over the course of the ill ness, and greater struc tural brain ab - nor mal i ties (1,2). In ad di tion, men ex hibit both a poorer dis - ease course and a poorer med i ca tion re sponse than do women (1). Al though women and men have com pa ra ble rat ings of psychopathology for acute psy chotic ep i sodes, women s re - sponse to med i ca tion is faster, oc curs at a lower dose, and is more ex ten sive than men s re sponse (4 14). Can J Psychiatry, Vol 46, August

50 August 2001 Risperidone Treatment of Outpatients With Schizophrenia 535 There is some ev i dence that more women than men experi - ence late-on set schizo phre nia and that postmenopausal women may re quire higher doses of antipsychotics than do premenopausal women (14,15). In at least 1 other study, how - ever, no con sis tent in creases in daily use of antipsychotics were seen after meno pause (16). Fur ther, sex-re lated treat - ment dif fer ences have not been found in all stud ies (17 20). For ex am ple, in a sub group of 69 antipsychotic nonresponders treated with clozapine, women did not ex hibit a ther a peu tic re sponse su pe rior to men s (19). Sim i larly, in a smaller study, no sig nif i cant dif fer ences in treat ment re - sponse were ob served be tween 24 male and 20 fe male pa - tients who had been well matched for clin i cal, de mo graphic, and treat ment char ac ter is tics (20). It should be noted that most of these stud ies in ves ti gat ing sex and treat ment re sponse were con ducted with older antipsychotic med i ca tions that work pri mar ily through dopa - mine block ade. Be cause es tro gen also ex hib its antidopaminergic prop er ties, it has been ar gued that it ex erts antipsychotic ef fects sim i lar to those of antipsychotic drugs and, thus, may confer a pro tec tive effect in women with schizo phre nia (1). It may be that at least some of the sex dif - fer ences seen in cases of schizo phre nia are a re sult of differ - ences in es tro gen dis tri bu tion. For the newer atyp i cal antipsychotic agents that do not act solely through dopaminergic mech a nisms, lim ited data are avail able on sex-re lated dif fer ences in treatment re sponse. Risperidone is a novel benzisoxazole de riv a tive, with po tent se ro to nin 5-HT 2 blocking prop er ties at low doses and po tent do pa mine D 2 re cep tor-block ing prop er ties at higher doses (21,22). It has been proven safe and ef fec tive in the treat ment of schizo phre nia and other psy chotic dis or ders, com pared with pla cebo, con ven tional antipsychotics (23 29), and atyp - i cal antipsychotics (30,31). In adults with chronic schizo - phre nia, risperidone re duced both pos i tive and neg a tive symp toms ef fec tively. More over, its use was as so ci ated with de creased se ver ity of dyskinesia and was not as so ci ated with in creased EPS. Be tween June 1, 1993 and April 30, 1995, a Phase IV, multicentre, open-la bel clin i cal study was con ducted to as - sess the ef fi cacy and safety of risperidone in a pop u la tion of out pa tients with schizo phre nia (32). A to tal of 330 patients, en rolled at 61 sites across Can ada, par tic i pated. In that study, risperidone was found to be su pe rior to pa tients pre vi ous antipsychotic ther apy: it was more ef fi ca cious and EPS were less servere. In our in ves ti ga tion, we con ducted a post hoc anal y sis of the clin i cal data from this Phase IV study to de ter - mine whether the treatment of schizo phre nia with risperidone was as so ci ated with the sex dif fer ences in re sponse and tolerability. Patients and Methods De tails of the study de sign, pa tient se lec tion, and treat ment sched ule for the Phase IV study have been out lined pre vi - ously (32). In brief, this was an 8-week, multicentre, open-la - bel, flex i ble-dose study de signed to as sess risperidone s ef fi cacy and safety in the treat ment of out pa tients with schizo phre nia. A group of 330 men and women, aged 18 to 65 years, who met the DSM-III-R cri te ria for subchronic or chronic schizo - phre nia (33) and who were being treated as out pa tients, were el i gi ble for the study. Par tic i pants had a to tal score ranging from 60 to 120 on the Pos i tive and Neg a tive Syn drome Scale (PANSS) for schizo phre nia (34) and had been able to dis con - tinue current antipsychotic and antiparkinsonian med i ca - tions. No an ti de pres sants, mood sta bi liz ers, lith ium, or other antipsychotics were al lowed at any time through out the study. When a sed a tive or hypnotic med i ca tion was in di cated, chlo ral hy drate was the pre ferred agent. Written in formed con sent was ob tained from all pa tients or their le gal rep re sen - ta tives, when re quired. Pa tients were screened for the study at base line and seen at days 7, 14, 28, and 56 (end of treat ment). For pa tients who dis con tin ued prior to day 56 but were oth er wise con sid ered evaluable for ef fi cacy, we used the re sults of the last as sess - ment visit to cre ate a Last Avail able Visit (LAV) end point. There fore, data from pa tients who dis con tin ued early due to in ad e quate re sponse, ad verse ex pe ri ences, and other rea sons were in cluded in the ef fi cacy eval u a tion. At study ini ti a tion, pa tients dis con tin ued their antipsychotics and, if ap pli ca ble, their antiparkinsonian agents. Treat ment with risperidone was ini ti ated at a dos age of 2 mg daily. The dos age was in - creased by 2 mg daily on each of the fol low ing 2 days; a dos - age of 6 mg daily was achieved on day 3 and was main tained un til day 14. The in ves ti ga tor then main tained the dos age at 6 mg daily or in creased or de creased it by 2 mg daily (at most once each week), based on the pa tient s clin i cal re sponse. The per mit ted dose range for risperidone was 4 to 10 mg daily. Treat ment com pli ance was mon i tored at each visit. The pri mary ef fi cacy pa ram e ters were the percent change from base line in total PANSS score and the per cent age of PANSS re spond ers. Re sponders were de fined as those in whom the to tal PANSS score de creased by 20% or more. Other ef fi cacy end points in cluded changes from base line in the Clin i cal Global Im pres sion (CGI) scale of se ver ity; CGI im prove ment; PANSS total score, pos i tive, neg a tive, and gen eral psychopathology subscales; and the Global As sess - ment of Func tioning (GAF) scale. The Extrapyramidal Symp tom Rat ing Scale (ESRS) (35) and the CGI scales for se ver ity of dyskinesia, parkinsonism, and dystonia were used to as sess the oc cur rence and se ver ity of EPS. Fur ther, risperidone s safety and tolerability were eval u ated on the ba - sis of vi tal signs, phys i cal ex am i na tion, and ad verse events

51 536 The Canadian Journal of Psychiatry Vol 46, No 6 Table 1. Baseline demographic and psychiatric characteristics of the efficacy analyses population by sex a Men Women Number (% of total population) 206 (70.5%) 86 (29.5%) Age (years) 34.9, SD 9.7 (17, 63) 40.1, SD 10.3 (20, 62) Weight (kg) 78.9, SD , SD 17.9 Height (cm) 175.3, SD 8.6 (132, 195) 160.3, SD 7.1 (137, 181) Schizophrenia diagnosis (Axis I) Para noid 133 (64.6%) 57 (66.3%) Un dif fer en ti ated 35 (17.0%) 19 (22.1%) Re sid ual 23 (11.2%) 7 (8.1%) Dis or gan ized 13 (6.3%) 3 (3.5%) Cata tonic 2 (1.0%) 0 (0%) Any hospitalization history 191 (92.7%) 81 (94.2%) Any hospitalization in past 6 months 48 (23.4%) 18 (20.9%) Age at onset (years) 22.4, SD 6.4 (9, 51) 24.5, SD 6.2 (11, 40) Age at first hospitalization (years) 24.0, SD 6.7 (13, 51) 26.3, SD 6.7 (11, 44) Duration of illness (years) 12.5, SD 8.2 (0, 37) 15.6, SD 9.4 (0, 34) Total Positive and Negative Syndrome Scale (PANSS) score 85.7, SD 16.6 (60, 120) 87.7, SD 16.9 (61, 119) PANSS positive subscale score 19.3, SD 6.3 (7, 34) 20.2, SD 5.6 (9, 37) PANSS negative subscale score 24.6, SD 6.0 (11, 41) 24.1, SD 7.0 (9, 38) PANSS General Psychopathology Subscale (GPS) score 41.8, SD 8.4 (21, 65) 43.5, SD 8.7 (24, 64) Clinical Global Impression (CGI) severity score 4.4, SD 0.8 (3, 6) 4.4, SD 0.8 (3, 6) Global Assessment of Functioning (GAF) score 45.5, SD 12.7 (4, 90) 46.8, SD 13.1 (21, 85) a Un less oth er wise speci fied, data are ex pressed as ei ther number, (% of male or fe male popu la tion) or as mean, stan dard de via tion (SD) (mini mum, maximum) re ported in re sponse to gen eral ques tions about prob lems or con cerns. In this in ves ti ga tion, we per formed t-tests to as sess whether sig nif i cant sex dif fer ences in pa tient de mo graphic and psy - chi at ric char ac ter is tics were pres ent at base line. The Wilcoxon Rank Sum test was used to test for pos si ble sex dif - fer ences in the use of pre vi ous antipsychotic ther a pies, con - verted to chlorpromazine equivalents. We used t-tests to test for sex dif fer ences in the fi nal ti trated dose of risperidone and in the time needed to achieve the fi nal dose; a chi-square test was used to as sess dif fer ences in dos - age changes (that is, no change, in crease, or de crease). We used 2-tailed Fisher s ex act tests to as sess sex dif fer ences in the num ber of pa tients com plet ing the study and in those who dis con tin ued due to an ad verse event. We used the Kaplan-Meier log-rank test to de ter mine whether sig nif i cant sex dif fer ences af fected the study-dis con tinu a tion time. We also per formed t-tests to ex am ine the relation be tween sex and treatment re sponse for most of the ef fi cacy pa ram e - ters, in clud ing all PANSS end points and the CGI severity score. We used 2-tailed Fisher s ex act tests to as sess sex dif fer ences in the per cent age of PANSS re spond ers, chi-square tests for CGI im prove ment from base line (cat e go - ries were im proved, no change, and worse ), and Wilcoxon Rank Sum tests for GAF scores. Sex dif fer ences in re sponse to risperidone, com pared with the pa tient s pre vi ous antipsychotic med i ca tion, were as sessed us ing the chi-square test (cat e go ries were better, the same, or worse ). We also per formed t-tests to ex am ine the relation be tween sex and treat ment re sponse for the safety pa ram e ters, in clud - ing all ESRS subscales and the CGI dyskinesia se ver ity, parkinsonism, and dystonia subscales. Ad verse events expe - ri enced by male and fe male pa tients over the course of the study were tab u lated and re ported as per cent ages. Results Base line Char ac ter is tics A total of 330 pa tients (232 men and 98 women) were en - rolled in the study; all 330 were in cluded in the safety analy - sis. The dis po si tion of these pa tients has been de scribed pre vi ously (32). Briefly, 38 (11.5%) pa tients dis con tin ued prior to the day 14 visit and were there fore ex cluded from the

52 August 2001 Risperidone Treatment of Outpatients With Schizophrenia 537 Table 2. Baseline medication history in chlorpromazine equivalent dosages (mg) by sex Previous neuroleptic therapy Men Women P Oral Mean, SD Mini mum, maxi mum n = , , 4000 n = , , Depot Mean, SD Mini mum, maxi mum n = , , 3040 n = , , Combined Mean, SD Mini mum, maxi mum n = , , 4140 n = , , ef fi cacy anal y sis. There were no sex dif fer ences in the num - ber of par tic i pants dis con tinu ing (26 men and 12 women, P = 0.851) or in those dis con tinu ing due to an ad verse event (14 men and 8 women, P = 0.504). A total of 292 (88.5%) of the 330 pa tients were evaluable for ef fi cacy, and of these, an ad di tional 48 pa tients with drew prior to the pro to col-spec i fied 56 days. There fore, 244 (73.9%) pa tients com pleted the en tire 8-week study. No sex dif fer ences were de tected in the num ber of pa tients com plet - ing the study (P = 0.386) or in the study-dis con tinu a tion time (P = 0.364), when early with drawal oc curred. In ad di tion, no sex dif fer ences were de tected among the pa tients who with - drew due to an ad verse event (n = 17), lack of ef fi cacy (n = 14), or for other rea sons (n = 17). More than twice as many men as women par tic i pated in this study, and they were, on av er age, 5 years youn ger than the women (Ta ble 1). Male par tic i pants weighed more than fe male par tic i pants and were taller. Most pa tients (ap prox i - mately 65%) had a di ag no sis of para noid schizo phre nia; over 90% of them had been hos pi tal ized for their ill ness. Male par - tic i pants had a sig nif i cantly ear lier age of dis ease on set than did fe male par tic i pants (22.4 vs 24.5 years, P = 0.011); how - ever, fe male par tic i pants had the dis ease for a lon ger time (15.6 vs 12.5 years). Men and women were sim i lar at base line in terms of to tal PANSS scores; the pos i tive, neg a tive, and gen eral psychopathology PANSS subscale scores; CGI se - ver ity score; and GAF score. No sig nif i cant sex dif fer ences were ev i dent in the patients pre vi ous use of oral, de pot, or oral and depot com bined antipsychotic med i ca tions (Ta ble 2), when con verted to chlorpromazine equiv a lents. Youn ger women (age 40 years or less) re ceived lower dos ages of pre vi ous antipsychotics (chlorpromazine equiv a lents) than did the men (535 mg vs

53 538 The Canadian Journal of Psychiatry Vol 46, No 6 Table 3. Summary of efficacy analyses by sex: differences from baseline at last available visit Men (n = 206) Women (n = 86) Efficacy Outcomes Mean SD Mean SD P Primary % change in total PANSS score % PANSS re spond ers 76.7 n/a 77.9 n/a Other CGI se ver ity score CGI im prove ment % 85.0 n/a 82.6 n/a PANSS to tal score PANSS posi tive subscale PANSS negative subscale PANSS GPS GAF score Table 4. Summary of safety analyses by sex: differences from baseline at last available visit Men (n = 206) Women (n = 86) Extrapyramidal Symptom Rating Scale (ESRS) Outcomes Mean SD Mean SD P Parkinsonism Dystonia Dyskinetic movements CGI: severity of dyskinesia CGI: severity of Parkinsonism CGI: severity of dystonia mg daily); how ever, this dif fer ence did not reach sta tis ti - cal sig nif i cance. Older women (> 40 years) re ceived a mean of 602 mg daily of pre vi ous antipsychotic med i ca tion. Ef fi cacy Out comes Male par tic i pants ex hib ited a slightly greater per cent age change from base line in to tal PANSS score, com pared with fe male par tic i pants, at each visit (Fig ure 1): by the final as - sess ment day, male par tic i pants had ex pe ri enced a 41.0% de - crease in their to tal PANSS score; fe male par tic i pants ex pe ri enced a de crease of 36.5%. These sex dif fer ences were not sta tis ti cally sig nif i cant (P = 0.287) (Table 3). A sim i lar num ber of male and fe male par tic i pants were as - sessed as PANSS re spond ers to risperidone treat ment (Ta ble 3). The num ber of re spond ers steadily in creased during treat - ment (Fig ure 2). Compared with day 7, the num ber of pa tients iden ti fied as re spond ers had more than dou bled by the fi nal as sess ment day: from 37.9% to 76.7% of men and from 34.1% to 77.9% of women. Based on data from the last avail able visit, there were no sig nif i cant sex dif fer ences in the per cent age of PANSS re spond ers (P = 0.880) (Ta ble 3). No sig nif i cant sex dif fer ences were ev i dent on the fi nal as - sess ment day for any of the other ef fi cacy mea sures (P > 0.05) (Ta ble 3). Im prove ments from base line in the mean total PANSS scores were ev i dent for both men and women by day 7 (Fig ure 1). Scores im proved steadily during treat ment. By the fi nal as sess ment day, male par tic i pants had ex pe ri enced a de crease of 23.1 points in their mean total PANSS score; fe - male par tic i pants ex pe ri enced a de crease of 21.6 points. The sex dif fer ence in mean change in total PANSS score on the last avail able visit was not sta tis ti cally sig nif i cant (P = 0.539) (Ta ble 3). Both men and women showed sim i lar im prove ments from base line in their scores on each of the PANSS pos i tive, neg a - tive, and gen eral psychopathology subscales; the CGI se ver - ity scale; the CGI im prove ment scale; and the GAF scale (Ta ble 3). Fur ther more, no sex dif fer ences were evident when com par i sons were made be tween re sponse to risperidone

54 August 2001 Risperidone Treatment of Outpatients With Schizophrenia 539 treat ment and the pa tient s pre vi ous antipsychotic (P = 0.359): at the last available visit, 77.5% of men and 69.6% of women re ported im prove ment with risperidone. At the same time, 13% of men and 19% of women re ported no change, and 9.5% of men and 11.4% of women re ported that their re - sponse to risperidone was poorer than to their pre vi ous antipsychotic. No sig nif i cant sex dif fer ences were ev i dent in the mean fi nal dos age of risperidone or the time to reach it. Risperidone was ti trated to a mean fi nal dos age of 6.1 mg daily in male partici - pants and 6.2 mg daily in fe male par tic i pants (P = 0.699). The mean time to reach fi nal dosage was 15.6 days for male par - tic i pants and 15.1 days for fe male par tic i pants (P = 0.825). Fur ther more, no sig nif i cant sex dif fer ences were ev i dent when com par i sons were made be tween re sponse to risperidone treat ment and the pa tient s pre vi ous antipsychotic (P = 0.359). At the last avail able visit, 77.5% of men and 69.6% of women re ported im prove ment with risperidone, com pared with their pre vi ous antipsychotic; 13% of men and 19% of women re ported no change; and 9.5% of men and 11.4% of women re ported that their re - sponse to risperidone was poorer. Safety and Tolerability Out comes Based on ESRS, both male and fe male par tic i pants ex hib ited sig nif i cant re duc tions from base line in both the in ci dence and se ver ity of EPS over the course of the study (32). There were no sig nif i cant sex dif fer ences in any of the out comes as sessed (Ta ble 4). Both men and women showed im prove ments over base line in the in ci dence of parkinsonism, dystonia, and dyskinetic move ments. Sim i larly, based on the CGI physi - cian rat ing scales, both men and women showed im prove - ments over base line in the se ver ity of parkinsonism, dystonia, and dyskinesia. Sex dif fer ences were ev i dent in some, but not all, of the most com monly re ported side ef fects. The most com monly re - ported (5% of pa tients) events were in som nia (22.8% M, 22.5% F), headache (12.1% M, 23.5% F), nau sea (13.4% M, 22.5% F), som no lence (15.1% M, 13.3% F), and fa tigue (11.2% M, 11.2% F). Weight gain was reported as an ad verse event only by male par tic i pants (4%); how ever, no sta tis ti - cally sig nif i cant sex dif fer ences were ev i dent in the per cent - age change from base line weight (in crease, 2.3% M; in crease, 1.5% F; P = 0.085). Some male par tic i pants (13.8%) also re - ported ejac u la tion fail ure or dis or der. Comments No sex dif fer ences in re sponse to risperidone ther apy were ev i dent either among the 292 out pa tients with chronic schizo - phre nia in cluded in the ef fi cacy anal y ses or among the 330 in - cluded in the safety anal y ses. This was true not only for all ef fi cacy end points ex am ined, but also for the EPS of parkinsonism, dystonia, and dyskinesia. The im prove ment in re sponse and the less ened se ver ity of EPS were achieved at com pa ra ble doses of risperidone for male and fe male pa tients. Al though the present data are based on a post hoc anal y sis and as such should be viewed with cau tion, some general ob ser va - tions can be made. Con sis tent with the pre vi ous re ports, the fe male pa tients in this study were older at the onset of their ill - ness. The youn ger fe male pa tients ( 40 years) en tered the trial with a lower dose (chlorpromazine equiv a lents) of antipsychotic med i ca tion and were underrepresented (only 30%) in this group of pa tients with chronic schizo phre nia. These re sults are not con sis tent, how ever, with other stud ies show ing that men and women re spond dif fer ently to antipsychotic med i ca tions more spe cif i cally, that women have a better re sponse to antipsychotic ther apy. This ap par ent ab sence of sex dif fer ences in re sponse to risperidone therapy could be at trib uted to the se lected pa tient pop u la tion. As men - tioned, there were ap prox i mately twice as many male pa tients in this co hort. In ad di tion, the 2 groups were not dif fer ent with re spect to their dosage of pre vi ous antipsychotics and se ver - ity of psychopathology, sug gest ing that the groups may have been equally re spon sive ness to antipsychotics at the on set of the study. The ab sence of sex dif fer ences could also have been sec ond - ary to the ti tra tion method used in this trial. All pa tients were ti trated to 6 mg daily and main tained at this dos age for 2 weeks, when dosage ad just ments were made based on the de - gree of re sponse or ev i dence of side ef fects or both. Al though the pat tern of dosage ad just ments was not sta tis ti cally differ - ent, slightly more women than men (49% F; 42% M) were main tained at 6 mg daily, and slightly fewer women re quired a dose in crease (21% F; 25% M). This fixed-dose ti tra tion, how ever, did not al low de tec tion of pos si ble sex dif fer ences in the low est ef fec tive dos age levels or in the tim ing of re - sponse on set. The cur rent risperidone pre scrib ing in for ma - tion rec om mends a start ing dos age of 1 to 2 mg either once or twice daily, ad justed grad u ally over sev eral days, based on clin i cal re sponse, to a tar get dosage of 4 to 6 mg daily. Some pa tients ben e fit from a lower ini tial dosage and a slower ti tra tion. An other pos si ble and in trigu ing hy poth e sis for the ap par ent lack of sex dif fer ence in antipsychotic treat ment with risperidone may be drawn from the fact that the previously ob served sex-based re sponse dif fer ence was based pri mar ily on studies using con ven tional antipsychotics, which ex ert their ef fects pri mar ily through do pa mine re cep tor block ade. Es tro gen, which it self ex erts antidopaminergic ef fects, may have con ferred ad di tional antipsychotic ef fects on con ven - tional ther apy, thus pro vid ing fe male pa tients with the ob - served pro file of better re sponse to antipsychotic ther apy. This may have been the ba sis for the slight dif fer ence in the pre vi ous antipsychotic dose at base line ob served in the youn - ger women in this trial. Risperidone s dual mech a nism of

55 540 The Canadian Journal of Psychiatry Vol 46, No 6 ac tion, which is me di ated through se ro to nin 5-HT 2 an tag o - nism and do pa mine D 2 an tag o nism, may make it more equally ef fec tive and tol er a ble in both sexes. Risperidone has been shown to be equally ef fec tive and tol er - a ble in male and fe male pa tients with chronic schizo phre nia. The ab sence of sex dif fer ences with this agent may ob vi ate the need for a sex-based dif fer en tial dosing in the manage - ment of schizo phre nia. Acknowledgements This study was supported by a grant from Janssen-Ortho Inc., To - ronto, On tario. We also wish to acknowledge the institutions and in - ves ti ga tors that participated in the orig i nal RIS-CAN-3 study. References 1. Salem JE, Kring AM. The role of gender dif fer ences in the re duc tion of etiologic het er o ge ne ity in schizo phre nia. Clin Psychol Rev 1998;18: Tamminga CA. Gen der and schizo phre nia. J Clin Psy chi a try 1997;58 (Suppl 15): Iacono WG, Beiser M. Where are the women in first-ep i sode stud ies of schizo phre - nia? Schizophr Bull 1992;18: Andia AM, Zisook S, Heaton RK, Hesselink J, Jernigan T, Kuck J, and oth ers. Gen - der dif fer ences in schizo phre nia. J Nerv Ment Dis 1995;183: Szymanski S, Lieberman JA, Alvir JM, Mayerhoff D, Loebel A, Geisler S, and oth - ers. Gender dif fer ences in onset of ill ness, treat ment re sponse, course, and bi o logic in dexes in first-ep i sode schizo phrenic pa tients. Am J Psy chi a try 1995;152: Hogarty GE, Goldberg SC, Schooler NR, The Col lab o ra tive Study Group. Drug and sociotherapy in the af ter care of schizo phrenic pa tients. III. Ad just ment of nonrelapsed pa tients. Arch Gen Psy chi a try 1974;31: Hogarty GE, Goldberg SC, Schooler NR, Ulrich RF, The Col lab o ra tive Study Group. Drug and sociotherapy in the af ter care of schizo phrenic pa tients. II. Two-year re lapse rates. Arch Gen Psy chi a try 1974;31: Dworkin RJ, Ad ams GL. Pharmacotherapy of the chronic pa tient: gender and di ag - nos tic fac tors. Com mu nity Ment Health 1984;20: Young MA, Meltzer HY. The re la tion ship of de mo graphic, clin i cal and out come vari ables to neuroleptic treat ment re quire ments. Schizophr Bull 1980;6: Goldstein MJ, Rodnick EH, Ev ans JR, May PR, Steinberg MR. Drug and fam ily ther apy in the af ter care of acute schizo phren ics. Arch Gen Psy chi a try 1978;35: Goldberg SC, School er NR, Davidson EM, Kayce MM. Sex and race dif fer ences in re sponse to drug treat ment among schizo phren ics. Psychopharmacologie 1966;9: Chouinard G, Annable L. Pimozide in the treat ment of newly ad mit ted schizo - phrenic pa tients. Psychopharmacology 1982;76: Kolakowska T, Wil liams AO, Ardern M, Reveley MA, Jambor K, Gelder MG, and oth ers. Schizo phre nia with good and poor out come. I: Early clin i cal fea tures, re - sponse to neuroleptics and signs of or ganic dys func tion. Br J Psy chi a try 1985;146: Seeman MV. In ter ac tion of sex, age and neuroleptic dose. Compr Psy chi a try 1983;24: Lindamer LA, Lohr JB, Harris MJ, McAdams LA, Jeste DV. Gen der-re lated clin i - cal dif fer ences in older pa tients with schizo phre nia. J Clin Psy chi a try 1999;60: Salokangas RKR. Gen der and the use of neuroleptics in schizo phre nia. Fur ther test - ing of the oes tro gen hy poth e sis. Schizophr Res 1995;16: Szymanski S, Lieberman J, Pol lack S, Kane JM, Safferman A, Munne R, and oth - ers. Gen der dif fer ences in neuroleptic nonresponsive clozapine-treated schizo - phren ics. Biol Psy chi a try 1996;39: Pinals DA, Malhotra AK, Missar CD, Pickar D, Breier A. Lack of gen der dif fer - ences in neuroleptic re sponse in pa tients with schizo phre nia. Schizo phre nia Re - search 1996;22: Jeste DV, Kleinman JE, Potkin SG, Luchins DJ, Weinberger Dr Ex uno multi: subtyping the schizo phrenic syn drome. Biol Psy chi a try 1982;17: Glick M, Mazure CM, Bowers MB, Zigler E. Premorbid so cial com pe tence and the ef fec tive ness of early neuroleptic treat ment. Compr Psy chi a try 1993;34: Janssen PAJ, Niemegeers CJE, Awouters FHL, Schellekens KHL, Megens AAHP, Meert TF. Phar ma col ogy of risperidone (R64766), a new antipsychotic with se ro - Clinical Im pli ca tions In patients with chronic schizophrenia, sex differences are not seen in short-term treatment response to risperidone. Given the overall improvements in psychological rating scales and in the incidence and severity of extrapyramidal symptoms, switch - ing patients to risperidone ther apy will likely benefit those with chronic schizophrenia. This analy sis does not support a differential dosing strategy or dos - age based on sex. Limi ta tions Post hoc data analyses are by their nature less rigorous than are those planned a priori. Data derived from an open-label, noncomparative study are less ro - bust than data derived from a study with a blinded, controlled de - sign. The population contained only 30% women. Dose titration and tar get dose were consistent with the product la - bel at the study time. Current labelling recommends a 1 to 2 mg daily starting dos age and adjustment gradually over several days, based on clinical response, to a tar get dos age of 4 to 6 mg daily. to nin-s2 and do pa mine-d2 an tag o nist prop er ties. J Pharmacol Exp Ther 1988;244: Leysen JE, Gommeren W, Eens A, de Chaffoy de Courcelles D, Stoof JC, Janssen PAJ. The bio chem i cal pro file of risperidone, a new antipsychotic. J Pharmacol Exp Ther 1988;247; Claus A, Bollen J, De Cuyper H, Eneman M, Malfroid M, Peuskens J, and others. Risperidone versus haloperidol in the treat ment of chronic schizo phrenic in pa - tients: a multicentre dou ble-blind com par a tive study. Acta Psychiatr Scand 1992;85: Mul ler-spahn F. Risperidone in the treat ment of chronic schizo phrenic pa tients: an in ter na tional dou ble-blind, par al lel-group study versus haloperidol [ab stract]. Clin i cal NeuropsychoPharmacology 1992;15 (Suppl 1):90A 91A. 25. Peuskens J, Risperidone Study Group. Risperidone in the treatment of pa tients with chronic schizo phre nia: a multi-na tional, multi-cen tre, dou ble-blind, par al lel-group study versus haloperidol. Br J Psy chi a try 1995;166: Chouinard G, Jones B, Remington G, Bloom D, Adding ton D, MacEwan GW, and oth ers. A Ca na dian multicenter pla cebo-con trolled study of fixed doses of risperidone and haloperidol in the treat ment of chronic schizo phrenic pa tients. J Clin Psychopharmacol 1993;13: Marder SR, Meibach RC. Risperidone in the treat ment of schizo phre nia. Am J Psy - chi a try 1994;151: Hoyburg O, Fensbo C, Remvig J, Lingjaerde O, Sloth-Niel sen M, Salvesen I. Risperidone ver sus perphenazine in the treat ment of chronic schizo phrenic pa tients with acute ex ac er ba tions. Acta Psychiatr Scand 1993;88: Borison RL, Pathiraja AP, Di a mond BI, Meibach RC. Risperidone: Clin i cal safety and ef fi cacy in schizo phre nia. Psychopharmacol Bull 1992;28: Heinrich K, Klieser E, Hehmann E, Kinzler E, Hruschka H. Risperidone ver sus clozapine in the treat ment of chronic schizo phre nia pa tients with acute symp toms: a dou ble-blind ran dom ized trial. Prog Neuropsychopharmacol Biol Psy chi a try 1994; 18: Klieser E, Lehmann E, Kinzler E, Wurthmann C, Heinrich K. Ran dom ized dou - ble-blind con trolled trial of risperidone versus clozapine in pa tients with chronic schizo phre nia. J Clin Psychopharmacol 1995;15 (Suppl 1): Chouinard G, Kopala L, Labelle A, Beauclair L, John son SV, Singh KI, the RIS-CAN-3 Study Group. Phase-IV multicentre clin i cal study of risperidone in the treat ment of out pa tients with schizo phre nia. Can J Psy chi a try 1998;43: Amer i can Psy chi at ric As so ci a tion. Di ag nos tic and sta tis ti cal man ual of men tal dis - or ders. 3rd ed. Re vised. Wash ing ton (DC): Amer i can Psy chi at ric As so ci a tion; Kay SR, Opler LA. The Pos i tive And Neg a tive Syn drome Scale (PANSS) for schizo phre nia. Schizophr Bull 1987;13: Chouinard G, Ross-Chouinard A, Annable L, Jones BD. Extrapyramidal Symp tom Rat ing Scale [ab stract]. Can J Neurol Sci 1980;7:233.

56 August 2001 Risperidone Treatment of Outpatients With Schizophrenia 541 Résumé Traitement à la rispéridone des patients externes souffrant de schizophrénie : aucune preuve de différence selon le sexe dans la réponse au traitement Ob jec tif : Étant donné le re nou veau d inté rêt pour le rôle du sexe dans la schizo phré nie, nous avons en tre pris une ana - lyse post- hoc afin de dé ter mi ner si les diffé rences selon le sexe dans la réponse au traite ment étaient pré sen tes chez des pa tients ex ternes souf frant de schizo phré nie qui re ce vaient de la rispé ri done, dans le cadre d une étude clin ique ou - verte de 8 se maines et de phase IV. Méth ode : Trois cent trente pa tients adul tes (232 hommes, 98 femmes) ay ant reçu un di ag nos tic de schizo phré nie con - forme au DSM- III-R ont été évalués quant à l in nocuité, et 292 pa tients (206 hommes, 86 femmes), quant à l ef fi cacité. On a cessé l ad min is tra tion d an tipsy cho tiques et d an ti park in so ni ens au début de l étude. Le traite ment à la rispéri - done a com mencé par un dos age de 2 mg par jour et a été aug menté à la dose cible de 6 mg par jour au 3 e jour, puis main tenu à 6 mg par jour, jusqu au 14 e jour. Le dos age a en suite été main tenu à 6 mg par jour, plus ou moins 2 mg par jour chaque se maine, selon la réponse du pa tient. Le traite ment à la rispé ri done a été ad min is tré du rant 8 se maines, le dos age per mis variant de 4 mg à 10 mg par jour. Résul tats : Les par tici pants mas cu lins et féminins ont bien répondu au traite ment à la rispé ri done; au der nier jour d évalua tion, ils se situaient sous la ligne de départ de leurs scores à l é chelle des syn dromes po si tifs et né ga tifs (PANSS), soit une baisse de 41,0 % et 36,5 % re spec tive ment. La plupart des sujets hommes (77 %) et femmes (78 %) étaient con sidé rés répon dants au PANSS : la rispé ri done était ef fi cace à la fois con tre les symptômes po si tifs et né ga tifs de la schizo phré nie. Les deux sexes se sont amé liorés par rap port à leur ligne de départ en ce qui con cerne l in ci dence et la gravité du park in son isme, de la dystonie et de la dyski nésie. Au cune diffé rence signi fi ca tive (P > 0,05) de réponse au traite ment selon le sexe n a été ob servée dans au cun des résul tats d ef fi cacité ou dans l in ci dence ou la gravité des symptômes ex tra py ra mi daux. Con clu sions : Dans cette popu la tion de pa tients ex ternes souf frant de schizo phré nie chronique, tant les hommes que les femmes ont bien répondu à des doses vari ables de rispé ri done. Au cune diffé rence selon le sexe sig ni fi ca tive n a été mani feste, que ce soit dans la réponse au traite ment ou dans les ef fets sec on daires neu rolo g iques. L ab sence de diffé - rence selon le sexe dans la réponse au traite ment à la rispé ri done peut élimi ner le be soin d un dos age diffé ren tiel selon le sexe dans le traite ment de la schizo phré nie.

57 REVIEW PAPER Pre-, Peri-, and Postnatal Trauma in Subjects With Attention-Deficit Hyperactivity Disorder Mi chael Zappitelli, MD 1, Teresa Pinto 2, Natalie Grizenko, MD, FRCPC 3 Ob jec tive: To re view re search on pre-, peri-, and postnatal stress and their po ten tial re la tion to att en tion-def i cit hy per - ac tiv ity dis or der (ADHD). Method: We se lected and crit i cally re viewed 51 re search reports from the med i cal and psy chol ogy lit er a ture, be tween Jan u ary 1, 1976 and May 1, 2001, based on the sub jects of pre-, peri-, or postnatal stress and ADHD. Re sults: Children with ADHD show higher per cent ages of pre-, peri-, or postnatal in sult, com pared with un af fected chil dren; how ever, the rel a tive in flu ence of var i ous fac tors is still con tro ver sial. Con clu sions: The eti ol ogy of ADHD en com passes ge netic and en vi ron men tal fac tors. Pre-, peri-, and postnatal stress - ors are en vi ron men tal fac tors that may play a role in its eti ol ogy. Future re search should care fully ex am ine in ter ac tions be tween ge netic pre dis po si tion and en vi ron men tal fac tors as eti ol o gies of ADHD. (Can J Psy chi a try 2001;46: ) Key Words: prenatal, perinatal, postnatal, attention-deficit hyperactivity disorder, hypoxia, labour complications, delivery complications At ten tion-def i cit hy per ac tiv ity dis or der (ADHD) mani - fests it self in early child hood and is char ac ter ized by in - at ten tion, im pul sive ness, and hy per ac tiv ity (1). It af fects ap prox i mately 6% to 9% of chil dren from di verse so cial and cul tural back grounds, with a higher rep re sen ta tion of males to fe males (2,3). Al though the eti ol ogy of ADHD is not well un der stood, it is well es tab lished that ADHD has an im por tant ge netic compo - nent, as in di cated by fam ily (4 6), twin (7 10), and adoption (11) stud ies. Having a fam ily his tory of ADHD in creases the risk of de vel op ing the dis or der (12). As this re view pa per dis - cusses, en vi ron men tal stress ors such as preg nancy and de liv - ery com pli ca tions have been found to in crease the risk of ADHD. It is be lieved that pre- and perinatal trauma may have a di rect ef fect on the fe tal brain during a cru cial pe riod of de - vel op ment. McGrath and oth ers found that neo na tal Manuscript received April 2000, revised, and accepted June Resident, Department of Pediatrics, University of Alberta, Edmonton, Al - berta. 2 Medical Student, McGill University, Montreal, Quebec. 3 Associate Professor, Department of Psychiatry, McGill University, Mont - real, Quebec; Director, Lyall Preadolescent Day Treatment Program, Doug - las Hospital, Verdun, Quebec. Address for correspondence: Dr N Grizenko, Lyall Pavilion, Douglas Hospi - tal, 6875 Lasalle Blvd, Verdun, QC H4H 1R3 mor bid ity strongly in flu enced neu ro log i cal out come (13). With ad vanc ing med i cal tech nol o gies, it is worth while to study the dif fer ent long-term ef fects of neo na tal mor bid ity; ADHD, in creas ingly dis cussed in terms of its or i gins in neurochemical al ter ations, can be ex am ined as a neu ro log i cal out come of in sult sus tained early in life. With re gard to ADHD, it is im por tant first to ex am ine ex ist ing re la tions be - tween pre- and perinatal stress and its de vel op ment and ADHD, then, to identify the spe cific com pli ca tions that most highly cor re late with ADHD de vel op ment. In the search for ADHD s eti ol ogy, a huge num ber of fac tors that may pre dis pose to neo na tal brain in sults have been stud - ied. The lit er a ture, how ever, is very con tra dic tory on this topic. Barkley and oth ers, for ex am ple, found no greater preg - nancy and birth com pli ca tions among chil dren with ADHD than among an un af fected con trol group (12). Study meth ods vary greatly, and there is a gen eral ten dency not to con sider the role of ge net ics or fam ily his tory when ex am in ing differ - ent risk factors. This re view pa per pro vides a crit i cal over - view of the lit er a ture de scrib ing a link be tween pre-, peri-, and postnatal events and ADHD de vel op ment; syn the sizes the find ings; and in ter prets them ac cord ing to what is cur - rently known. We ob tained ref er ences for this re view from a Medline and PsycINFO search com pris ing the pe riod Jan u ary 1, 1976 to May 1, We re viewed 51 re ports cho sen for Can J Psychiatry, Vol 46, August

58 August 2001 Pre-, Peri-, and Postnatal Trauma in Subjects With ADHD 543 their re search fo cus on ob stet ri cal and de liv ery com pli ca tions and ADHD. Prenatal Stressors In Utero Toxins Re search on the ef fect of in utero tox ins on the de vel op ing fe - tus has fo cused mainly on cig a rette smok ing and al co hol ex po sure. The ac cepted neurochemical hy poth e sis be hind the pathophysiology of ADHD is a dys func tion of the do pa mine sys tems in the prefrontal cor tex (14 17). An i mal stud ies have shown that pups pre na tally ex posed to nic o tine showed de creased striatal dopaminergic re cep tor bind ing sites (18). Carboxyhemoglobin lev els are el e vated in preg nant moth ers who smoke and in their fe tuses, pos si bly leading to de creased ox y gen-bind ing ca pac ity and de creased ox y gen de liv ery to fe tal tis sues (19). Hence, 2 main mech a nisms are pro posed for the ef fects of ma ter nal cig a rette smoking on the fe tus: ma - ter nal smok ing leads to fe tal hypoxia, and nicotine causes dis - tur bances to the dopaminergic sys tem in the brain. In 1975, Denson and oth ers found that moth ers of hyperkinetic chil dren smoked sig nif i cantly more than did moth ers of a con trol group (20). Later, in a study of 2256 chil - dren, Weitzman and oth ers found that ma ter nal smoking was in de pend ently as so ci ated with many child hood be hav iour prob lems, in clud ing hy per ac tiv ity (21). All data were ob - tained from ma ter nal re ports, and ma ter nal smok ing during and af ter preg nancy was not dis tin guished. In re sponse, Fergusson and oth ers con ducted a 15-year pro spec tive study of 1265 chil dren, ex am in ing the re la tion be tween ma ter nal smok ing and several child hood be hav iours, in clud ing at ten - tion def i cit dis or der (ADD) (22). Their study was novel be - cause they eval u ated chil dren and ques tioned moth ers and teach ers se ri ally, from birth to a max i mum age of 15 years. Sev eral im por tant con found ing vari ables were in cluded in the anal y ses (for ex am ple, sex, ma ter nal age, so cio eco nomic sta - tus [SES], and pa ren tal dis cord). Smoking dur ing, but not af - ter, preg nancy was in de pend ently as so ci ated with higher teacher and ma ter nal ADD rat ings of chil dren. These findings sup port a spe cific relation be tween preg nancy smok ing and later de vel op ment of at ten tion def i cit. Un for tu nately, the ef - fect of fam ily his tory was not in cluded. In sev eral stud ies, Milberger and oth ers have spe cif i cally ex - am ined the ef fect of ma ter nal smoking on the de vel op ment of ADHD. In 1996, they com pared 140 boys with ADHD with their first-de gree bi o log i cal rel a tives and with 120 con trol sub jects (23). Twenty-two per cent of the chil dren with ADHD, com pared with 8% of the con trol sub jects, had a ma - ter nal his tory of smok ing. Re sults re mained sig nif i cant when con trolled for pa ren tal ADHD, chil dren s IQ lev els, and SES. These re sults were rep li cated in 1997 and 1998, sup port ing the hy poth e sis that nic o tine or hypoxia from ma ter nal cig a - rette smok ing causes damage to the de vel op ing brain (24,25). In a lon gi tu di nal study, Hill and oth ers studied 150 ad o les - cents who were ei ther high or low risk for de vel op ing al co hol - ism, to gether with their fam i lies (26). Var i ous psy chi at ric dis or ders were ex am ined, and the 3 main risk fac tors tested were fa mil ial risk sta tus for al co hol ism, the child s ex po sure to pre na tal al co hol, and the child s ex po sure to pre na tal cig a - rettes. Anal y ses con trolled for each of the risk fac tors, SES, and fam ily his tory of an ti so cial per son al ity dis or der (APD), when cal cu lat ing the risk of de vel op ing psy chi at ric dis or ders. The re search ers found a relation only be tween ADHD and fa - mil ial risk sta tus for al co hol ism; no re la tion was found with ex po sure to pre na tal cig a rettes or al co hol. Al though find ings per tain ing to ADHD were fairly in con clu sive, this study s strength was its de tailed fam ily his tory in for ma tion. Un for tu - nately, the sam ple size was small, and preg nancy smoking his tory was ret ro spec tive, some times span ning sev eral years. The ma ter nal use of al co hol while preg nant has been an area of in ter est, es pe cially since chil dren with fetal al co hol expo - sure (FAE) show signs of at ten tion or mem ory def i cits (27). Streissguth and oth ers ex am ined the ef fects of pre na tal alco - hol ex po sure on 462 chil dren, at birth and at age 14 years (28). Pre na tal eth a nol ex po sure was sig nif i cantly re lated to at ten - tion and mem ory def i cits in a dose-de pend ent fash ion, even when con trolled for ma ter nal smok ing. Aronson and oth ers found that 10 of the 24 chil dren whose moth ers con sumed al - co hol while preg nant had ADHD, sug gest ing at most a link be tween fetal eth a nol ex po sure and ADHD de vel op ment (29). Un for tu nately, the chil dren were not all ex am ined uni - formly; some in for ma tion was ob tained from schools, while other in for ma tion was ob tained from clin i cal ex am i na tion. By con trast, Coles and oth ers stud ied 149 Af ri can-amer i can chil dren of low SES (30). Of these, 87 had FAE, 27 had ADHD without FAE, and 35 were un af fected con trol sub - jects. Al though both the al co hol-ex posed and ADHD groups per formed more poorly on tests of in tel lec tual abilities, the sam ple with ADHD dem on strated more at ten tion prob lems and con duct dis or der (CD). This study suggests that al co hol may not play an im por tant role in the de vel op ment of ADHD per se but that at ten tion prob lems as so ci ated with FAE may be re lated to cog ni tive def i cits. The study by Hill and oth ers men tioned above also does not sup port the role of ma ter nal al - co hol in ges tion in ADHD de vel op ment, al though it does sug - gest that a fam ily his tory of al co hol ism in creases the risk of hav ing ADHD (26). This find ing, how ever, is not strongly sup ported by Bennet and oth ers study, which specifically com pared chil dren from al co holic fam i lies with chil dren from nonalcoholic fam i lies (31). The con tro versy on this sub - ject may ex ist be cause there is no clear pathophysiologic event di rectly relating ma ter nal al co hol con sump tion to spe - cific ab nor mal i ties found in ADHD.

59 544 The Canadian Journal of Psychiatry Vol 46, No 6 Al though these stud ies sug gest a re la tion be tween ma ter nal smok ing and ADHD, it is dif fi cult to draw any con clu sions about the spe cific cor re la tion be tween al co hol ex po sure and ADHD de vel op ment. Cer tain study strengths and lim i ta tions must be ex am ined. For ex am ple, sam ples were of ten drawn from pe di a tri cian or psy chi a trist re fer rals, mak ing it dif fi cult to gen er al ize find ings. Similiarly, sev eral stud ies acquired in - for ma tion from ret ro spec tive ma ter nal re ports; pro spec tive in for ma tion gath er ing from more than 1 domain (such as par - ent, teacher, or med i cal re cords) is more re li able. It is also im - por tant to com pare chil dren whose moth ers smoked during preg nancy with chil dren whose moth ers only smoked be fore or af ter preg nancy. Con founding vari ables must also be ad - dressed. As pointed out by Barkley, the re la tion of ma ter nal smok ing and al co hol use to ADHD de vel op ment cannot be stud ied without con trol ling for vari ables such as SES, family his tory, learning dis abil i ties, and so cial or fam ily en vi ron - ment (32). More over, ma ter nal al co hol con sump tion and cig - a rette smok ing in them selves are col lin ear and must be con trolled for (26). For these rea sons, large sam ple sizes, ex - am ined pro spec tively in a se rial fash ion, us ing stan dard ized ques tion naires from sev eral do mains, and dil i gently con trol - ling for con found ing vari ables, are of ut most im por tance. Ma ter nal Char ac ter is tics Dur ing Preg nancy Studies have sug gested nu mer ous ma ter nal fac tors during preg nancy that cor re late with ADHD. Hartsough and Lam - bert stud ied 301 chil dren with hy per ac tiv ity and 191 un af - fected chil dren (33). They found that young ma ter nal age, poor ma ter nal preg nancy health, eclampsia, and preg nancy par ity were fac tors pre dict ing sub se quent ADHD di ag no sis. Unique to their study, sub jects were ob tained through par - ents, schools, and phy si cians. All ob stet ri cal in for ma tion, how ever, was ob tained from ma ter nal in ter views. Chandola and oth ers com pared the birth re cords of 129 re - ferred chil dren with hy per ac tiv ity with the re main ing mem bers of a geo graph ical birth co hort (34). Fac tors as so ci - ated with in creased re fer ral were moth ers youn ger than 25 years at the time of their child s birth and moth ers not in a first mar riage. Use of birth re cords uniquely supported the re sults re li abil ity. In the Milberger and oth ers study, fam ily prob lems, ma ter nal bleed ing, and com pli ca tions of ma ter nal ac ci dents during preg nancy were pos i tively as so ci ated with the de vel op ment of ADHD (24). These stud ies sug gest that a myr iad of ma ter nal events during preg nancy may con trib ute to fu ture childhood ADHD symp - toms. More in-depth stud ies are needed, how ever, to classify the type of in sult that would most harm the de vel op ing brain. Im por tant study strengths would be pro spec tive in for ma tion gath er ing (for ex am ple, fol low ing women throughout their preg nan cies), to gether with use of med i cal re cords and data gath er ers and an a lysts with suf fi cient med i cal knowl edge to re cord and in ter pret data ac cu rately. In ret ro spec tive analy - ses, data col lec tors should be blind to the ADHD sta tus of subjects. Perinatal Hypoxia Studies have sug gested dys func tion in do pa mine trans mis - sion in the right prefrontal cor tex and striatum in pa tients with ADHD (2). It is pos si ble, then, that perinatal events re sult ing in hypoxia may ad versely af fect the de vel op ing brain acutely, lead ing to brain ischemia and ab nor mal i ties in the dopaminergic sys tem. Vaillancourt and Boksa found that rats born by ce sar ean sec - tion un der gen eral an es the sia showed in creases in dopa - mine-me di ated be hav iours (35). Brake and oth ers, dem on strated that rats born by ce sar ean section and ex posed to intra uter ine an oxia were hy per ac tive and had mesocortical do pa mine ac ti va tion im pair ments in the left prefrontal cor tex (un pub lished). These studies simply sup port the hy poth e sis of a de fec tive dopaminergic sys tem, re sult ing from neo na tal hypoxic events, leading to symp toms of ADHD. Hypoxic-ischemic brain dam age and intraventricular hemor - rhage, sec ond ary to perinatal as phyxia, can re sult in neuro - logic and in tel lec tual dys func tion and, pos si bly, psy chi at ric dis or ders (36,37). Nu mer ous fac tors may lead to perinatal as - phyxia: pre na tal risk fac tors like eclampsia and di a be tes mellitus; ob stet ri cal ab nor mal i ties like prolapsed um bil i cal cord, placenta previa, or mul ti ple preg nan cies; intrapartum risk factors, in clud ing ab nor mal fe tal po si tion or pro longed la bour; and neo na tal fac tors, such as prematurity, respiratory dis tress, or cardiopulmonary ab nor mal i ties (36). Ex am ining the re la tion be tween perinatal hypoxia and ADHD, Hartsough and Lam bert dem on strated that fe tal dis - tress during birth or de liv ery (non spe cific dis tress, head, or other birth injuries) was re ported by 17% of moth ers of chil - dren with hy per ac tiv ity, com pared with 8% of con trol-group moth ers (33). Perinatal hypoxia may af fect the brain, with re sult ing be hav - ioural dis or ders; how ever, no con clu sions can be made yet. Again, a lon gi tu di nal fol low-up, be gin ning at birth, com par - ing high- and low-risk ne o nates with re spect to hypoxic events, and taking into ac count con found ing vari ables (for ex am ple, other preg nancy com pli ca tions, ma ter nal factors, SES, and fam ily his tory) would be ideal. Low Birth Weight Low birth weight (LBW) has been ex am ined in re la tion to cog ni tive and be hav ioural de vel op ment (13,38,39). In the study by Milberger and oth ers, LBW was as so ci ated only with lower IQ scores but not with ADHD (24). Hartsough and Lam bert found no sig nif i cant dif fer ence in their stud ied

60 August 2001 Pre-, Peri-, and Postnatal Trauma in Subjects With ADHD 545 groups in terms of prematurity or LBW (33). O Callaghan and oth ers ex am ined the learn ing abil i ties and at ten tion prob - lems of 87 chil dren who were born as ex tremely LBW (ELBW) in fants (40). They found a sig nif i cant dif fer ence be - tween case and control sub jects on all learn ing lev els but not on teacher ADHD as sess ments. No other con found ing vari - ables were ac counted for, how ever. These find ings, which sug gest no link be tween birth weight and later ADHD symp - toms, have been con tra dicted by sev eral studies. Szatmari and oth ers com pared 82 five-year olds weigh ing be - tween 500 and 1000 g at birth with an age-equiv a lent con trol group (41). Psy chi at ric di ag no ses were based on in for ma tion from moth ers and teach ers. Par ent-ques tion naire re sults showed that 7.3% of ELBW chil dren dis played ADHD symptomatology, com pared with 1.4% of con trol sub jects. This as so ci a tion was spe cific to ADHD and not to other diag - no ses. When teacher and parent rat ings were com bined, 15.9% of the ELBW chil dren, com pared with 6.9% of the con trol group, were rated as hy per ac tive. Using 1 ur ban and 1 sub ur ban sam ple, Breslau and oth ers per formed cog ni tive testing on 473 chil dren, aged 6 years, who had been LBW ba bies (42). Birth weights and perinatal his to ries were ob tained from med i cal re cords. Their re sults showed that the rate of ADHD was higher in LBW than in nor mal birth weight (NBW) chil dren and that this as so ci a tion was stron ger in the ur ban than in the sub ur ban pop u la tion, even when con found ing vari ables (ma ter nal smok ing, alco - hol use, SES, and ma ter nal anx i ety) were ac counted for. They sug gested that ex am in ing ur ban and sub ur ban pop u la tions sep a rately con trolled for many con found ing vari ables. Bot - ting and oth ers com pared 136 LBW chil dren, aged 12 years, with 148 full-term NBW chil dren matched for age and sex (43). The re sults showed that LBW chil dren were more likely to have ADHD, but not oppositional de fi ant dis or der (ODD) or CD. These re sults sug gest a the ory of an in de pend ent re la tion be - tween LBW and ADHD. Find ings are clearly con tra dic tory, how ever, and study meth od ol o gies vary in con trol ling for con found ing vari ables, sam ple sizes, and data gath er ing. More over, re port ers were not al ways blind to sub jects birth weights. The fact that find ings are so con tra dic tory is not sur - pris ing be cause, al though LBW has a clear relation to fu ture in tel li gence level and learn ing abil i ties (13,39), it does not pro vide a clear eti ol ogy for some spe cific hypoxic brain-dam ag ing event. LBW is also as so ci ated with many other dif fer ent risk fac tors (13), making it a dif fi cult caus ative fac tor to ex am ine. Postnatal Events Hypoxia Rat ne o nates ex posed to an oxia within the first 24 hours of life dem on strated in creased mo tor ac tiv ity in am bu la tion, rear ing, and sniff ing at 10, 15, and 20 days of age, com pared with con trols (44). By day 25, the only dif fer ence remaining was in creased sniff ing in the an oxia-ex posed rats. Though a cor re la tion be tween postnatal an oxia and hy per ac tiv ity may be drawn from this ex per i ment, it is some what dif fi cult to make a di rect cor re la tion with hu mans, be cause ro dents are less sen si tive to an oxia (45). In a ret ro spec tive anal y sis, Chandola and oth ers found a re la - tion be tween re fer ral for ADHD and in fants who re ceived Car diff Bag Re sus ci ta tion and had a 1-min ute Apgar score of be low 7 or a 5-min ute Apgar score of be low 9 (34). These fac - tors, though, were strongly cor re lated with each other and not sig nif i cant when con trolled for con found ing vari ables. This find ing sug gests either an in ter ac tion be tween these postnatal fac tors or a cu mu la tive ef fect from all factors. In a study of 11 chil dren (aged 6 to 15 years) with ADD, Lou and oth ers showed through use of xe non 133 in ha la tion and CT, that ce re bral hypoperfusion of the fron tal lobes was pres - ent in all (46). This, they be lieve, is linked to early hypoxic-ischemic dam age. The sam ple cho sen, how ever, lacked uni for mity and dem on strated such con found ing vari - ables as dif fer ences in ges ta tional con di tions, birth weights, and use of med i ca tion. Fur ther studies on the ef fect of hypoxemia and hypotension on the de vel op ing brain were done by Low and oth ers, who pro spec tively stud ied 130 preterm ne o nates (47). The in fants were eval u ated for echosonographically de mon stra ble ce re - bral le sions (EDCL) dur ing the neo na tal pe riod, at ei ther 3 or 6 months. Among the preterm ne o nates who ex pe ri enced hypotension or hypoxemia in the early hours of life, 34% showed EDCL (namely, intraventricular hem or rhage, 21%; ventriculomegaly, 18%; and hyperechoic parenchymal le - sions, 8%), com pared with preterm ne o nates who had not ex - pe ri enced hypotension or hypoxemia, where the rate of EDCL was 13%. When hypotension and hypoxemia were both present, the chance of having EDCL sur passed 50%. In 1997, Whitaker and oth ers ex am ined the re la tion be tween neo na tal cra nial ul tra sound ab nor mal i ties (re flect ing perinatal brain in jury) and psy chi at ric dis or ders in 685 LBW chil dren aged 6 years (37). Compared with nor mal cra nial ultrasounds, cra nial ul tra sound ab nor mal i ties near birth inde - pend ently in creased the risk of hav ing a future di ag no sis of ADHD even when neo na tal com pli ca tions, pre na tal factors, and so cial dis ad van tage were ac counted for. These studies are based on the hy poth e sis that brain injury may lead to ADHD. All pre vi ously men tioned risk fac tors in this ar ti cle re fer to events and phys i cal or men tal states af fect - ing the preg nant mother, fe tus, or new born that might place a child at in creased risk of de vel op ing a brain ab nor mal ity that could lead to ADHD. Im aging, how ever, po ten tially provides a way to quan tify the dam age done and, the o ret i cally, pre dict

61 546 The Canadian Journal of Psychiatry Vol 46, No 6 the con tri bu tion of the phys i cal as pect of ADHD etiology in a given in di vid ual. Conclusion Re search gen er ally sup ports the hy poth e sis that pre-, peri-, and postnatal stress ors may play an im por tant role in the de - vel op ment of ADHD. A wide spec trum of pre na tal events as - so ci ated with even tual symp toms of ADHD has emerged. These range from chronic in sults through out preg nancy, such as ma ter nal smok ing, high blood pres sure, age, and the mother s emo tional state to more acute events like ma ter nal bleed ing, eclampsia, or abruption. It is un clear which events those lead ing to chronic fetal hypoxia dur ing ges ta tion or the acute events have the major role in ab nor mal de vel op - ment of neu ral dopaminergic sys tems; this is a re search area worth ex plor ing. The lit er a ture shows that the role of ma ter - nal eth a nol in ges tion is still un clear; it is a difficult risk factor to study due to the presence of sev eral con found ing vari ables (for ex am ple, SES, ma ter nal smok ing, pres ence of ma ter nal psy chi at ric ill ness, or symp toms of FAE). Among perinatal events, acute hypoxic damage re sult ing from long la bour, ce - sar ean de liv er ies with hypoxia, and hypoxic-ischemic dam - age to the brain may also con trib ute to a future pre sen ta tion with ADHD. All these fac tors need more de tailed study to achieve more def i nite con clu sions. The role of LBW, which may simply be an in di ca tor of prematurity or intrauterine prob lems, also re mains un clear. Finally, it has re cently been dis cov ered that early brain imag - ing ab nor mal i ties may in di cate in creased risk for ADHD de - vel op ment. This find ing is par tic u larly in ter est ing be cause pre sum ably these ab nor mal i ties re sult from all previously men tioned risk fac tors. This leads to the question whether fur - ther re search should fo cus on this method of as sess ing risk, rather than looking at spe cific events that lead to the same end. The pur pose of all this re search is to con trib ute to our un der - stand ing of what causes ADHD. There is still much un cer - tainty, which partly ex plains many of the lim i ta tions existing in the stud ies re viewed lim i ta tions that must be ex am ined to pro pose fu ture di rec tions. As men tioned, a ma jor limitation of many studies is the lack of ac count ing for con found ing vari ables: with a dis or der so ob vi ously multifactorial, one must ac count for vari ables that might af fect the caus ative role of spe cific fac tors. Vari ables that should al ways be in cluded are SES, fam ily his tory, and the pres ence of learn ing disor - ders. More over, stud ies should con trol for other pos si ble caus ative fac tors when as sess ing a spe cific one. For ex am ple, if one is to as sess the role of LBW in ADHD de vel op ment, it is im per a tive to con trol for ma ter nal smoking and other intrauterine risk fac tors be cause they may be col lin ear with the factor studied (LBW), which could un pre dict ably al ter re - sults. The stron gest stud ies are those with large, rep re sen ta - tive sam ple sizes (iden ti fied from school, phy si cian or psy chol o gist, and par ent re fer rals) and with sound di ag nos tic meth ods. Child be hav iour in for ma tion should in clude par ent and teacher re ports, and any med i cal data should ide ally in - clude data from pa tient charts. Pro spec tive stud ies are ideal, elim i nat ing re call bias and al low ing for se rial as sess ment of child biopsychosocial sta tus. More spe cif i cally, stud ies that ex am ine ma ter nal smok ing should com pare the ef fects of smok ing dur ing preg nancy and smok ing only af ter preg nancy to truly test the the ory that nic o tine causes intrauterine hypoxia or dam age to the brain s dopaminergic sys tem. An av e nue that has not been ex plored very much is the role of pre- and perinatal risk fac tors in af fect ing the vari able ex pres - sion of ADHD. Children with ADHD may be in at ten tive, hy - per ac tive and im pul sive, or both, and many have comorbid dis or ders such as ODD and CD (48 50). De spite this vari able phenotypic ex pres sion, most chil dren ben e fit from methylphenidate ther apy. Ap prox i mately one-third of chil - dren do not, how ever (51). Hence, the the ory ex ists that ADHD may not be a sin gle syn drome but, rather, sev eral dif - fer ent sub types that may dic tate treat ment re sponse. More - over, it is not known whether these dif fer ent ex pres sions of ADHD are ac tual sub types or sim ply dif fer ences in se ver ity. Pre- or perinatal com pli ca tions may be in volved in mod u lat - ing the phenotypic ex pres sion of ADHD. It is also not known whether methylphenidate re spond ers and nonresponders dif - fer in their pre- and perinatal his to ries. ADHD con sists of a spec trum, so that al though a child may not ful fill cri te ria for di ag no sis, cer tain symp toms may be pres ent, and chil dren di ag nosed with ADHD can have differ - ent levels of symp tom se ver ity. It is pos si ble that the severity of ADHD pre sen ta tion may be linked to in ter ac tions be tween ge netic sus cep ti bil ity and en vi ron men tal stressors. Are these fac tors ad di tive in pro duc ing phenotypic ex pres sion of ADHD, or are they in de pend ent, pro duc ing dif fer ent types of the dis or der? These are other ques tions that well-de signed stud ies on pre- and perinatal risk fac tors may help to elu ci date. Fu ture re search should ex plore the im por tance of ge netic and en vi ron men tal fac tors, such as perinatal events, in dif fer ent sub groups of ADHD (for ex am ple, in at ten tive vs hy per ac - tive). The in volve ment of these pos si ble etiologic fac tors should also be stud ied both in chil dren with pure, non comorbid ADHD and in those with comorbid ADHD. In ad - di tion, it is im por tant to re search the rel e vance of ge netic and en vi ron men tal fac tors in the ex pres sion of ADHD se ver ity by ex plor ing whether these fac tors cor re late with in creas ing symptomatology. An ideal way of doing this would be to com pare chil dren with ADHD with their sib lings, who pre - sum ably have sim i lar ge netic back grounds and home en vi - ron ments but may have dif fer ent perinatal histories. Clar ifying the ge netic and en vi ron men tal in ter ac tions in ADHD will lead to a better un der stand ing of the dis or der s

62 August 2001 Pre-, Peri-, and Postnatal Trauma in Subjects With ADHD 547 bi o log i cal as pect and may lead to im proved ther a peu tic ap - proaches. Un der stand ing what early events play a role in ADHD de vel op ment may eventually help cli ni cians to de sign pre ven ta tive mea sures for at-risk families. Ac knowl edge ment The authors thank Ma rina Ter Stepanian for her as sis tance. References 1. Amer i can Psy chi at ric As so ci a tion. Di ag nos tic and sta tis ti cal man ual of mental dis - or ders, 4th. ed. Wash ing ton (DC): Amer i can Psy chi at ric As so ci a tion; Lou, HC. Eti ol ogy and pathogenesis of at ten tion-def i cit hy per ac tiv ity dis or der (ADHD): sig nif i cance of prematurity and perinatal hypoxic-haemodynamic encephalopathy. Acta Paediatrica 1996;85: Tannock, R. At ten tion deficit hy per ac tiv ity dis or der: ad vances in cognitive, neurobiological, and ge netic re search. J Child Psychol Psy chi a try 1998;39: Hechtman L. Fam ilies of children with at ten tion def i cit hy per ac tiv ity dis or der: a re - view. Can J Psy chi a try 1996;41(6): Sam uel VJ, George P, Thornell A, Curtis S, Tay lor A, Brome D, and oth ers. A pi lot con trolled family study of DSM-IIR and DSM-IV ADHD in Af ri can Amer i can chil dren. J Am Acad Child Adolesc Psy chi a try 1999;38(1): Faraone SV, Biederman J, Milberger S. An ex plor atory study of ADHD among sec - ond de gree rel a tives of ADHD chil dren. Biol Psy chi a try 1994;35: Nadder TS, Silberg JL, Eaves LJ, Maes HH, Meyer JM. Ge net ics ef fects on ADHD symptomatology in 7 to 13 year-old twins: re sults from a tele phone sur vey. Behav Genet 1998;28: Eaves LJ, Silberg JL, Meyer JM, Maes HH, Simonoff E, Pickles A, and oth ers. Ge - net ics and de vel op men tal psychopathology: the main ef fects of genes and en vi ron - ment on be hav ioural prob lems in the Virginia twin study of ad o les cent be hav ioural de vel op ment. J Child Psychol Psy chi a try 1997;154: Levy F, Hay DA, McStephen M, Wood C, Wladman I. At ten tion deficit hy per ac tiv - ity dis or der di men sions: a twin study of in at ten tion and impulsivity- hy per ac tiv ity. J Am Acad Child Adolesc Psy chi a try 1997;36: Good man R, Stevenson J. A study of hy per ac tiv ity II. The eti o log i cal role of genes, fam ily re la tion ships and perinatal ad ver sity. J Child Psychol Psy chi a try 1989;30: Cadoret RJ, Stew art MA. An adop tion study of at ten tion def i cit/hy per ac tiv ity/ag - gres sion and their re la tion to adult an ti so cial per son al ity. Compr Psy chi a try 1991;32: Barkley RA, DuPaul GJ, McMurray MB. Com pre hen sive eval u a tion of at ten tion def i cit dis or der with and with out hy per ac tiv ity as de fined by re search cri te ria. J Con sult Clin Psychol 1990;58: McGrath MM, Sullivan MC, Lester BM, Oh W. Lon gi tu di nal neu ro logic fol low-up in neo na tal in ten sive care unit sur vi vors with var i ous neo na tal mor bid i ties. Pe di at - rics 2000;106: Savgolden T, Ser geant JA. At ten tion def i cit/hy per ac tiv ity dis or der from brain dysfunctions to be hav iour. Behav Brain Res 1998;94: Castellanos FS, Elia J, Kruesi MJ, Gulotta CS, Mefford IN, Pot ter WZ, and others. Cerebrospinal fluid monoamine me tab o lites in boys with at ten tion def i cit hy per ac - tiv ity dis or der. Psy chi a try Res 1994;52: Ernst M, Zametkin AJ, Matochik JA, Jons PH, Co hen RM. DOPA de car box yl ase ac tiv ity in at ten tion def i cit hy per ac tiv ity dis or der in adults. A (flu o rine-18) fluorodopa pos i tron emis sion tomographic study. J Neurosci 1998;18: Fumagalli F, Jones S, Bosse R, Jaber M, Giros B, Missale C, and oth ers. In ac ti va - tion of the do pa mine trans porter re veals es sen tial roles of do pa mine in the con trol of lo co mo tion, psychostimulant re sponse, and pi tu itary func tion. Adv Pharmacol 1998;42: Fung YK, Lau YS. Ef fects of pre na tal nic o tine ex po sure on rat striatal dopaminergic and nic o tinic sys tems. Pharmacol Biochem Behav 1989;33: Longo L. The bi o log i cal ef fects of car bon mon ox ide on the preg nant woman, fe tus and new born in fant. Am J Obstet Gynecol 1977;129: Denson R, Nanson J, McWatters J. Hyperkinesis and ma ter nal smok ing. Ca na dian Psy chi a try As so ci a tion Jour nal 1975;20: Weitzman M, Gortmaker S, Sobol A. Maternal smoking and be hav iour prob lems of chil dren. Pe di at rics 1992;90: Fergusson DM, Horwood J, Lynskey MT. Ma ter nal smoking be fore and af ter preg - nancy: ef fects on be hav ioural out comes in mid dle child hood. Pe di at rics 1993;92: Milberger S, Biederman J, Faraone SV, Chen L, Jones J. Is ma ter nal smok ing dur - ing preg nancy a risk fac tor for at ten tion def i cit hy per ac tiv ity dis or der in chil dren? Am J Psy chi a try 1996;153: Milberger S, Biederman J, Faraone S, Guite J, Tsuang MT. Preg nancy, de liv ery and in fancy com pli ca tions and at ten tion def i cit hy per ac tiv ity dis or der: is sues of gene-en vi ron ment in ter ac tion. Biol Psy chi a try 1997;41: Clinical Implications Research suggests that certain common pre-, peri-, and postnatal stresses may have a significant role in the etiology of attentiondeficit hyperactivity disorder (ADHD). Identifying specific early-life events that predispose children to ADHD may lead to its earlier detection or prevent certain behaviours in at-risk fami lies. The role of maternal alcohol ingestion during pregnancy in the etiol - ogy of ADHD is still unclear; however, many effects of fetal al co hol exposure include attention and behaviour problems. Limitations This review only examines pregnancy and delivery complications with relation to ADHD and not to other childhood behaviour disor - ders. Reviewed studies are of mixed qual ity without standardized formats, and many lack objective data collection for pregnancy, delivery, and neonatal complications. There is no origi nal research material presented here, and more re - search is needed to validate many findings. 25. Milberger S, and oth ers. Further ev i dence of an as so ci a tion be tween ma ter nal smok ing dur ing preg nancy and at ten tion def i cit hy per ac tiv ity dis or der. J Clin Child Psychol 1998;27: Hill SY, Low ers L, Locke-Wellman J, Shen S. Ma ter nal smok ing and drinking dur - ing preg nancy and the risk for child and ad o les cent psy chi at ric dis or ders. J Stud Al - co hol 2000;61: Nanson JL, Hiscock M. At ten tion def i cits in children ex posed to al co hol pre na tally. Al co hol Clin Exp Res 1990;14: Streissguth AP, Sampson PD, Olson HC, Bookstein FL, Barr HM, Scott M, and oth ers. Ma ter nal drink ing dur ing preg nancy: at ten tion and short-term mem ory in 14-year old off spring: a lon gi tu di nal pro spec tive study. Al co hol Clin Exp Res 1994;18: Aronson M, Hagberg B, Gillberg C. At ten tion def i cits and au tis tic spec trum prob - lems in chil dren ex posed to al co hol dur ing ges ta tion: a fol low-up study. Dev Med Child Neurol 1997;39: Coles CD, Platzman KA, Raskind-Hood CL, Brown RT, Falek A, Smith IE. A com - par i son of children af fected by pre na tal al co hol ex po sure and at ten tion def i cit, hy - per ac tiv ity dis or der. Al co hol Clin Exp Res 1997;21: Bennet LA, Wolin SJ, Reiss D. Cog ni tive, be hav ioural, and emo tional problems among school-age chil dren of al co holic par ents. Am J Psy chi a try 1988; 145: Barkley RA. At ten tion deficit hy per ac tiv ity dis or der: a hand book for di ag no sis and ther apy. New York: The Guilford Press; Hartsough CS, Lambert NM. Med i cal fac tors in hy per ac tive and nor mal chil dren: pre na tal, de vel op men tal, and health his tory find ings. Am J Orthopsychiatry 1985;55: Chandola CA, Robling MR, Pe ters TJ, Mel ville-thomas G, McGuffin P. Pre- and perinatal factors and the risk of sub se quent re fer ral for hy per ac tiv ity. J Child Psychol Psy chi a try 1992;33: Vaillancourt C, Boksa P. Ce sar ean sec tion birth with gen eral an es the sia in creases do pa mine-me di ated be hav iour in the adult rat. Neuroreport 1998;9: Lou HC. Perinatal hypoxic-ischemic brain dam age and intraventricular hem or - rhage. Arch Neurol 1980;37: Whitaker AH, Van Rossem R, Feldman JF, Schonfeld IS, Pinto-Mar tin JA, Torre C, and others. Psy chi at ric out comes in low-birth-weight children at age 6 years: re - la tion to neo na tal cra nial ul tra sound ab nor mal i ties. Arch Gen Psy chi a try 1997;54: The Scot tish low birthweight study group. The Scot tish low birthweight study: II. Lan guage at tain ment, cognitive sta tus, and be hav ioural problems. Arch Dis Child 1992;67: Hack M, Breslau N. Very low birth weight in fants: ef fects of brain growth dur ing in fancy on in tel li gence quo tient at 3 years of age. Pe di at rics 1986;77: O Callaghan MJ, Burns YR, Gray PH, Harvey JM, Mohay H, Rog ers YM, and oth - ers. School per for mance of ELBW chil dren: a con trolled study. Dev Med Child Neurol 1996;38: Szatmari P, Saigal S, Rosenbaum P, Campbell D, King, S. Psy chi at ric dis or ders at five years among chil dren with birthweights 1000g: a regional per spec tive. Dev Med Child Neurol 1990;32:

63 548 The Canadian Journal of Psychiatry Vol 46, No Breslau N, Brown GG, DelDotto JE, Kumar S, Ezhuthschan S, and oth ers. Psy chi - at ric sequelae of low birth weight at 6 years of age. J Abnorm Child Psychol 1996;24: Bot ting N, Powls A, Cooke RWI. At ten tion deficit hy per ac tiv ity dis or ders and other psy chi at ric out comes in very low birthweight chil dren at 12 years. J Child Psychol Psy chi a try 1997;38: Speiser Z, Korczyn AD, Teplitzky I, Gitter S. Hy per ac tiv ity in rats fol low ing postnatal an oxia. Behav Brain Res 1983;7: Da vis JN. The use of small an i mals to study the ef fects of hypoxia. Adv Neurol 1979;26: Lou HC, Henriksen L, Bruhn P. Fo cal ce re bral hypoperfusion in chil dren with dysphagia and/or at ten tion def i cit dis or der. Arch Neurol 1984;41: Low JA, Froese AB, Smith JT, Galbraith RS, Sauerbrei EE, Karchmar EJ. Hypotension and hypoxemia in the preterm new born dur ing the four days fol low - ing de liv ery iden tify in fants at risk of echosonographically demonstratable ce re bral le sions. Clin In vest Med 1991;15: Hazell P. The over lap of at ten tion def i cit hy per ac tiv ity dis or der with other com mon men tal dis or ders. Jour nal of Pe di at rics and Child Health 1997;33: Lalonde J, Turgay A, Hud son JL. At ten tion def i cit hy per ac tiv ity dis or der sub types and comorbid dis rup tive be hav iour dis or ders in a child and ad o les cent mental health clinic. Ca na dian Jour nal of Child Psy chi a try 1998;43: Pliszka SR. Comorbidity of at ten tion def i cit hy per ac tiv ity dis or der with psy chi at r ic dis or der: an over view. J Clin Psy chi a try 1998;59 (Suppl 7): Gelenberg AJ, Bassuk EL, Schoonover SC. The prac ti tio ner s guide to psy cho ac - tive drugs, 3rd ed. New York: Plenum Pub lishing; Résumé Traumatisme prénatal, périnatal et postnatal chez les sujets souffrant du trouble d hyperactivité avec déficit de l attention Ob jec tif : Pas ser en re vue la re cher che sur le stress pré na tal, péri na tal et post na tal et sur sa re la tion pos si ble avec le trou ble d hy per ac tiv ité avec déficit de l at ten tion (THADA). Méth ode : Nous avons choisi 51 rap ports de re cher che dans la docu men ta tion médi cale et psy cholo gique, en tre le 1 er jan vier 1976 et le 1 er mai 2001, portant sur le stress pré na tal, péri na tal ou post na tal et le THADA, puis nous en avons fait l analyse cri tique. Résul tats : Les en fants souf frant du THADA af fichent des pour cent ages plus élevés de trau ma tismes pré na taux, péri - na taux ou post na taux que les en fants qui n en souf frent pas. Ce pend ant, l in flu ence relat ive de di vers fac teurs est en - core con tro versée. Con clu sions : L é ti olo gie du THADA englobe des fac teurs gé né tiques et en vi ron ne men taux. Les facteurs de stress pré - na taux, péri na taux et post na taux sont des fac teurs en vi ron ne men taux qui peu vent jouer un rôle dans l é ti olo gie du trou ble d hy per ac tiv ité. Les fu tures études doivent ex am iner at ten tive ment les in ter ac tions en tre la prédis po si tion gé - né tique et les fac teurs en vi ron ne men taux comme éti olo gies du THADA.

64 REVIEW PAPER Body Mass Index in Persons With Schizophrenia Sha lom Coodin, MD, FRCPC 1 Back ground: Schizo phre nia has been as so ci ated with sev eral health con cerns and risks. Over all mortal ity among per - sons with schizo phre nia has been shown to be about twice that of the gen eral pop u la tion. There is growing con cern that per sons with schizo phre nia may also be at risk for being over weight or obese, com pared with the gen eral pop u la tion. To ex am ine this pos si bil ity, the au thor com pared the dis tri bu tion of body mass in dex val ues (BMI = kg / m 2 ) in people with schizo phre nia with that of the Ca na dian pop u la tion as a whole. Method: Weights and heights were ob tained for 183 pa tients re ceiv ing treat ment in a hos pi tal-based pro gram for per - sons with schizo phre nia. These BMI val ues were com pared with the results of Sta tis tics Cana da s National Pop u la tion Health Sur vey (NPHS), which pro vided av er age BMI val ues for the gen eral pop u l a tion. Re sults: The av er age BMI in the study sam ple was 29.02, with the av er age for men be ing (range to 49.22, SD 6.25) and the av er age for women, (range to 45.71, SD 6.45). This is com pared with the NPHS average BMI of 26.3 for men and 24.3 for women. The prev a lence of obe sity (BMI >30) in the sam ple was 42.08%, 3.5 times that of the Ca na dian av er age of 12% and 2.8 times that of the 15% prev a lence in Man i toba. In this sam ple, 26.78% had a BMI in the ac cept able range, in con trast to the 48% of those in the NPHS who had a weight ap pro pri ate to their height. Con clu sions: This anal y sis pro vides ev i dence that the BMI dis tri bu tion of the sam ple pop u la tion is dif fer ent from that of the na tional pop u la tion as rep re sented in the NPHS data. The data in di cate that pa tients with schizo phre nia are sig nif - i cantly heavier than the gen eral pop u la tion. (Can J Psy chi a try 2001;46: ) Key Words: body mass index, BMI, obesity, schizophrenia Schizo phre nia has been as so ci ated with sev eral health con - cerns and risks. Per sons with schizo phre nia are more likely to smoke (1), to abuse al co hol and street drugs (2), and to suf fer with un treated med i cal prob lems (3). More over, they have a much higher risk of suicide than does the gen eral pop u la tion (4). Per sons with schizo phre nia have been found to have an in creased risk of dying from car dio vas cu lar ill ness, with some re search ers finding a near dou bling in ex cess mor - tal ity from this cause (5). Overall mor tal ity in per sons with schizo phre nia has been shown to be about twice that of the gen eral pop u la tion (5). In an ex ten sive re view of the mor tal ity lit er a ture, Har ris and Barraclough re port an in creased risk of nat u ral and un nat u ral (sui cide and other vi o lent causes) Manuscript received Janu ary 2001, revised, and accepted June Assistant Professor, Department of Psychiatry, Faculty of Medicine, Uni - versity of Manitoba, Winnipeg, Manitoba; Staff Psychiatrist, Schizophrenia Treatment and Education Program, PsycHealth Centre, Winnipeg, Mani - toba. Address for correspondence: Dr S Coodin, PZ 331, 771 Bannatyne Avenue, Winnipeg, MB R3E 3N4 scoodin@hsc.mb.ca deaths (6). Deaths from in fec tious dis ease, en do crine, men tal, cir cu la tory, re spi ra tory, di ges tive, and gen i to uri nary sys tem dis or ders all were sig nif i cantly in creased in per sons with schizo phre nia, com pared with the gen eral pop u la tion (6). In ad di tion, per sons with schizo phre nia are of ten caught in the low est so cio eco nomic level. Winkleby and oth ers com ment that One of the stron gest and most con sis tent pre dic tors of a per son s mor bid ity and mor tal ity ex pe ri ence is that per son s so cio eco nomic sta tus. This find ing per sists across all dis - eases with few ex cep tions, con tin ues throughout the life span, and ex tends across nu mer ous risk fac tors for dis ease (7). There is grow ing rec og ni tion that per sons with schizo phre nia may also face an in creased risk of car ry ing ex cess weight, com pared with the gen eral pop u la tion. The cost of obe sity, as with the cost of schizo phre nia, is con sid er able. In Can ada, the to tal di rect costs of obe sity in 1997 were es ti mated to be over $1.8 bil lion (8), pri mar ily due to hy per ten sion, type II diabe - tes, and cor o nary artery dis ease (9). Con ven tional and sec - ond-gen er a tion antipsychotic med i ca tions have been as so ci ated with weight gain (10). Life style and ill ness-re lated fac tors, such as lack of mo ti va tion and pov erty, may also con - trib ute to this prob lem. Can J Psychiatry, Vol 46, August

65 550 The Canadian Journal of Psychiatry Vol 46, No 6 Table 1. Comparison of Body Mass Index (BMI) in the Schizophrenia Treatment and Education Program (STEP) patients vs National Population Health Survey (NPHS) (shaded cells in table) results Age in years n Un der weight BMI < 18.5 n / % Ac cept able BMI n / % Over weight BMI n / % Obese BMI >30 n / % years Total STEP / / 26.78% 51 / 27.87% 77 / 42.08% NPHS Canada (%) NPHS Manitoba (%) STEP Men / 5 36 / / / NPHS Men (%) STEP Women 63 0 / 0 13 / / / NPHS Women (%) years Total STEP 19 2 / / / / NPHS (%) STEP Men 18 2 / / / / NPHS Men (%) STEP Women / 100% NPHS Women (%) years Total STEP / / / NPHS (%) STEP Men / / / NPHS Men (%) STEP Women / / NPHS Women (%) years Total STEP 55 1 / / / / NPHS (%) STEP Men 37 1 / / / / NPHS Men (%) STEP Women / / / NPHS Women (%) years Total STEP 35 3 / / / / NPHS (%) STEP Men 18 3 / / / / NPHS Men (%) STEP Women 17 0 / 0 3 / / / NPHS Women (%) years Total STEP 25 0 / 0 7 / / / NPHS (%) STEP Men / / / NPHS Men (%) STEP Women / / / NPHS Women (%)

66 August 2001 Body Mass Index in Persons With Schizophrenia 551 Grouping Cate gory (Sam ple vs NPHS data) Table 2: Exact χ 2 and P-values, STEP vs NPHS data χ 2 Ex act P-value Age years STEP vs NPHS < STEP vs Manitoba < STEP men < STEP women < Age years STEP men and women < STEP men < STEP women (ns) Age years STEP men and women STEP men (ns) STEP women Age years STEP men and women < STEP men < STEP women < Age years STEP men and women STEP men < STEP women Age years STEP men and women STEP men STEP women (ns) Table 3. BMI and education Education n Un der weight BMI < 18.5 Ac cept able BMI Over weight BMI Obese BMI > 30 n / % n / % n / % n / % Grade 8 or less 33 0 / 0 11 / / / Grade / / / / Diploma or certificate or university degree 15 1 / / / / Data unavailable 16 0 / 0 6 / / / Table 4. BMI and primary income source Primary income source n Un der weight BMI < 18.5 Ac cept able BMI Over weight BMI Obese BMI > 30 n / % n / % n / % n / % Social assistance or disability income / / / / Employment insurance or disability income 17 0 / 0 1 / / / Family income 12 0 / 0 5 / / / Other 4 0 / 0 1 / / / Data unavailable 5 0 / 0 2 / / / 0 Table 5. BMI, education, and income BMI distribution χ 2 Ex act P-value Education Income The body mass in dex (BMI), also known as the Quetelet in - dex, has been widely used for clin i cal and epidemiologic pur - poses. BMI is cal cu lated by di vid ing weight in ki lo grams by height in metres squared. The use of the BMI has lim i ta tions be cause it does not dis tin guish be tween fat and lean mass and is gen er ally not ad justed for age or sex. While hydrodensitometry and dual-en ergy X-ray ab sorp tion may

67 552 The Canadian Journal of Psychiatry Vol 46, No 6 pro vide greater ac cu racy in es ti mat ing body fat, these are gen er ally re search, rather than clin i cal, mea sures. Sim i larly, mea sures of skinfold thick ness and bioimpedance do not pro vide sig nif i cantly greater ac cu racy than does the BMI. BMI has gen er ally been ac - cepted as a rea son able guide for clin i cal and epidemiologic purposes. In their as sess ment of BMI and mor tal ity, Calle and oth ers found that the op ti mal BMI for lon gev ity falls be tween 20.5 and 24.9 for men and women of all ages (9). They com ment that, These data of fer sup port for the use of a sin gle rec om mended range of body weight through out life. Manson and oth ers, in fol - low ing a large group of women in the Nurses Health Study, found body weight and mortal - ity from all causes to be di rectly re lated. Fur - ther, they found that an in crease in weight of 10 kg or more after age 18 years was as so ci - ated with in creased mor tal ity in midlife (11). Figure STEP distributions (n = 183) vs NPHS and Manitoba. Pre vi ous reports ex am in ing BMI val ues in per sons with schizo phre nia have been lim ited by such fac tors as sin gle-sex sam ples, self-re - ported di ag no ses, or self-reported weight and height data, rather than cli ni cian-mea sured in - for ma tion and sam ples with pa tients in volved in clin i cal trials of med i ca tions or other in ter - ven tions. While some in ves ti ga tors con sider self-re ported height and weight to be ac cu rate, oth ers have ob served that self-re port data un - der es ti mate obe sity by about 10% (12). There are few re ports of un re stricted sam ples of pa - tients with schizo phre nia. Allison and others, us ing Na tional Health In ter view Sur vey data from the US, con cluded that men with schizo - phre nia had mean BMIs sim i lar to those of men with out schizo phre nia, while women with schizo phre nia were found to have sig nif i cantly higher BMIs on av er age (13). Method Figure 2. STEP distributions vs NPHS, all men (n = 120). To com pare the weights of this pop u la tion with those of the gen eral pop u la tion, weights and heights were ob tained for 183 pa tients re ceiv - ing treat ment in the Schizo phre nia Treat ment and Ed u ca tion Pro gram (STEP), a clin i cal pro - gram op er at ing out of a uni ver sity teaching hos pi tal. During the 3-month pe riod of May 1 July 31, 2000, height and weight data were ob tained for all pa tients at tend ing STEP. All pa tients in the dataset had di ag no ses of schizophreniform dis or der, schizo phre nia,

68 August 2001 Body Mass Index in Persons With Schizophrenia 553 Figure 3. STEP distributions vs NPHS, all women (n = 63) schizoaffective dis or der, de lu sional dis or der, or psy chotic ill - ness not oth er wise spec i fied (NOS). All mea sure ments were taken by a nurse or physician. BMI val ues for all pa tients were de ter mined and com pared with the re sults of the National Pop u la tion Health Sur vey (NPHS) of Ca na dian re spon dents be tween age 20 and 64 years (14) as well the val ues for Manitoba re ported therein. In for ma tion on high est level of ed u ca tion achieved and pri - mary in come source are col lected rou tinely for all pa tients in STEP, and this anal y sis used the most re cent en tries. To com pare the find ings in the study pop u la tion with those re - ported in the NPHS, the in ter na tional stan dard en dorsed by the World Health Or ga ni za tion (WHO) and the US Na tional In sti tutes of Health was used. The BMI cat e go ries are as fol - lows: 18.5 or less (un der weight), 18.6 to 24.9 (ac cept able weight), 25.0 to 29.9 (over weight), and 30.0 and more (obese). These val ues dif fer some what from the Ca na dian stan dard: less than 20.0 (un der weight), 20.0 to 24.9 (ac cept - able weight), 25.0 to 27.0 (some ex cess weight), and more than 27.0 (over weight). BMI was cal cu lated for people be - tween ages 20 and 64 years and ex cludes preg nant women. Com par i sons were made by sex and age, with group ings fol - low ing those used in the NPHS. To de ter mine whether the BMI dis tri bu tions of STEP pa tients were con sis tent with those of the gen eral pop u la tion, a good ness-of-fit anal y sis was per formed using StatXact4 for Win dows sta tis ti cal soft - ware (15). While the con ven tional Pearson s chi-square test might be used for this pur pose, it loses va lid ity when the sam - ple size is small and when more than 20% of ex pected counts are less than 5, as found in this sam ple. Mehta and Patel have de rived a col lec tion of ex act chi-square tests for such cat e gor - i cal data (16). The re sult ing sta tis tic mea sures the dis crep - ancy be tween the BMI dis tri bu tion of STEP pa tients and the BMI dis tri bu tion of the gen eral pop u la tion: the larger the chi-square value, the greater the dis crep ancy. To de ter mine the thresh old level for the chi-square val ues, the P-value of the ob served test sta tis tic was cal cu lated. When the P-value is less than 0.05, the null hy poth e sis (no dif fer ence be tween the sam ple and the ref er ent NPHS pop u la tion) is re jected. Results Weights and heights were gath ered for 120 men and 63 women. Av er age age overall was 39.6 years (range 20.2 to 64.1, SD 11.5), with an av er age for men of 37.5 years (range 20.2 to 59.9, SD 11.0) and 43.7 years (range 22.6 to 64.1, SD 11.8) for women. The av er age BMI in the study sam ple was 29.0, with the av er age for men in this sam ple being 28.5 (range 15.5 to 49.2, SD 6.25), and for women, 30.0 (range 19.3 to 45.7, SD 6.45). The NPHS found an av er age BMI of 26.3 for men and 24.3 for women (Fig ures 1, 2, and 3). With 3 ex cep tions, analysis of the BMI dis tri bu tion strat i fied by age and sex (Ta bles 1 and 2) and using ex act chi-square