Vi o lence against pa tients and staff on psy chi at ric wards is

Transcription

1 Crowding and Violence on Psychiatric Wards: Explanatory Models Shailesh Kumar, MRCPsych, MPhil 1, Bradley Ng, MBChB 2 Ob jec tive: Vi o lence is widely prevalent on acute-care psy chi at ric wards, and crowd ing has been ident i fied as a ma jor risk factor. This pa per ex plores why pa tients may re spond to crowding with vi o lence. Method: We car ried out a lit er a ture re view on Medline, us ing the key words vi o lence and crowdi ng. We con ducted an ad di - tional hand search of the ref er ences col lected from the re viewed papers. Re sults: Fac tors spe cific to the re la tion be tween crowd ing on acute-care in pa tient psy chi at ric wards and vi o lence can be di vided un der the fol low ing headings: 1) pa tient density, pri vacy, and con trol; 2) ward ar chi tec ture; 3) the so cial or ga ni za tion of psy chi - at ric wards; 4) in ter per sonal space; 5) phylogenic the o ries; and 6) an thro po log i cal the ories of hu man be hav iour. Con clu sions: We of fer ex plan a tory models for this re la tion and sug gest strat e gies to coun ter the ef fects of crowd ing. Rec om men - da tions are made for fu ture stud ies. (Can J Psy chi a try 2001;46: ) Key Words: violence, crowding, acute psychiatric wards, environmental stress Vi o lence against pa tients and staff on psy chi at ric wards is widely prev a lent and in creas ing. While it is dif fi cult to com pare prev a lence es ti mates among dif fer ent stud ies, due to vari a tions in def i ni tions of vi o lence, pop u la tion stud ied, and the source of data col lec tion, mul ti ple stud ies from dif fer ent coun tries have con firmed this to be the case (1,2). A study of 566 psy chi at ric and 898 nonpsychiatric beds from 16 psy chi - at ric and gen eral hos pi tals in the US re ported 2.54 as saults per bed yearly among psy chi at ric wards, com pared with 0.37 as - saults per bed yearly in nonpsychiatric units (3). Sim i lar esti - mates from other coun tries in clude the UK, with 0.65 at tacks lead ing to prop erty dam age or in jury to an other per son per oc - cu pied bed yearly (4); Bel gium, with 1 ag gres sive in ci dent per bed yearly (5); and Swe den, with 13 at tacks per oc cu pied bed yearly (6). In deed, the ex pe ri ence of vi o lence in the men - tal health sys tem is so wide spread that a mul ti na tional sur vey of psy chi at ric nursing staff across the US, Can ada, the UK, and South Af rica found that the re spon dents ex pected as - saults in their work with psy chi at ric pa tients (7). Var i ous stud ies have also re ported a sig nif i cant in crease with time in violent in ci dents. A Brit ish study that col lected data over 15 months re ported that the num ber of vi o lent in ci dents in the sec ond half of the study pe riod con sti tuted a highly Manuscript received November 2000, revised, and accepted April Consultant Psychiatrist, Department of Psychiatry, Rotorua Hospital, Roto - rua, New Zealand; Honorary Clinical Senior Lecturer, Division of Psychia - try, Auckland School of Medicine, Auckland, New Zealand. 2 Psychiatric Registrar, Department of Psychiatry, Nepean Hos pi tal, Sydney, New South Wales, Australia. Address for correspondence: Dr S Ku mar, Private bag 3023, Rotorua Hospi - tal, Rotorua, New Zealand. kumars@lhl.co.nz sig nif i cant 240% in crease (P < ) over the num ber in the first half (8). An other Brit ish study re ported a pro gres sive in - crease in vi o lence on an acute-care in pa tient ward from 1976 to 1984 (9). Sim i larly, an Amer i can study re ported a 32.9% in crease in as sault rates from 1978 to 1980 (10). Crowding ap pears to be emerg ing as one of many po ten tial risk fac tors for this trend of wide prev a lence and pos si ble in crease in vio - lence on acute-care psy chi at ric wards. As psy chi at ric care con tin ues to shift into the com mu nity, and as in pa tient beds de crease, crowd ing will re main a major is sue on acute-care wards (2,11 13). Two ma jor out comes of this trend can be iden ti fied: first, a group of pa tients with high re ad mis sion rates has emerged, re sult ing in a re volv ing door syn drome. Sec ond, as noted, psy chi at ric in pa tient units are get ting crowded, with bed-oc cu pancy rates of ten above 100% (13), which may be con trib ut ing to the trend (2,14,15). While crowd ing and vi o lence on acute-care in pa tient wards ap pear to be ris ing, studies of their relation have only be gun to ap pear in the lit er a ture and of ten con tra dict each other. A few stud ies have re ported no as so ci a tion be tween crowd ing and ag gres sion (16 18), but oth ers have re ported that in creased in pa tient numbers lead to more ag gres sion by pa tients against both staff and other pa tients (6,19,20). In deed, in a 12-month ret ro spec tive study of our own unit, in creased ward occu - pancy was found to be as so ci ated with both phys i cal and ver - bal ag gres sive in ci dents (21). There has been lit tle writ ten on why psy chi at ric pa tients should re spond to crowding with vi o lence and what, if any, are the pos si ble so lu tions. This pa per ex am ines the relation with a view to de vel op ing ex plan a tory models for it. To this ef fect, we re view the lit er a ture from 6 dif fer ent ar eas that ex - plain a rise in stress levels in crowded sit u a tions. We pro pose that high stress level on a crowded ward acts as a pre cip i tat ing fac tor for vi o lence through at least 2 mech a nisms: First, on a crowded ward, pa tients ap praisal of their en vi ron ment may Can J Psychiatry, Vol 46, June

2 434 The Canadian Journal of Psychiatry Vol 46, No 5 eas ily be come dis torted, leading to mis in ter pre ta tion of ward ac tiv i ties and cul mi nat ing in as saults (22). Sec ond, it is a well-known fact that when stress level crosses a certain thresh old our tol er ance level de creases, leading to a re duced abil ity to cope that may, in turn, lead peo ple to lose con trol and as sault oth ers. The con tri bu tions from the 6 re viewed ar - eas are united in the con text of acute-care psy chi at ric wards. Finally, we sug gest some strat e gies to coun ter the ef fects of crowd ing and make rec om men da tions for future stud ies. For this re view, we de fined vi o lence as any un wel come phys i cal or sex ual con tact or threat to kill or harm with a weapon. We did not in clude sim ple verbal threats or damage to prop erty. We ac knowl edge that such in ci dents may be dis tress ing to the vic tims or to peo ple wit ness ing the in ci dent, but we be lieve that their im pact is less se vere. Area 1: Density, Privacy, and Control Den sity, pri vacy, and con trol are 3 in ter re lated vari ables de - scribed in the psy cho log i cal lit er a ture that may help in un der - stand ing the re la tion be tween crowd ing and vi o lence. These vari ables are gen er ally used when in ves ti gat ing the im pact of ar chi tec ture on hu man mood and be hav iour (23). Den sity has been de fined as the num ber of in di vid u als per unit of space (24). One re view has iden ti fied 2 types of den sity in re - la tion to crowd ing and vi o lence on psy chi at ric wards: so cial den sity, or num ber of peo ple in a given area, and spa tial den sity, or size of an area used by a given num ber of in di vid - u als (14). With an in crease in so cial den sity, stress increases, and an in di vid ual s pri vacy and control, ac tual and per ceived, de crease (23). In such sit u a tions, so cial in ter ac tions are of ten im posed and can oc cur at in con ve nient times, which may have an ad verse ef fect on in di vid ual frus tra tion tol er ance. This can lead to fewer or even im proper so cial in ter ac tions (25). Pri vacy is de fined as an in di vid ual s abil ity to pro tect in ter per sonal space of ten a very sub jec tive con cept. In ter - per sonal space is the space in which 2 or more peo ple come into any type of in ter ac tion (26). When equally ac ces si ble to all par ties, in ter per sonal space is thought to pro tect against stress and vi o lence (27). In di vid ual strat e gies for main tain ing per sonal space de pend on cul ture, room size, level of ac - quain tance, and other so cial and psy cho log i cal vari ables (27). For ex am ple, the in ter per sonal space re quired by an Ameri - can would be larger than that re quired by a Med i ter ra nean per son. Sim i larly, it tends to be smaller when we are with peo - ple we know well as op posed to when we are with strangers. As so cial den sity in creases, an in di vid ual s control over en - vi ron ment de creases, which is in turn as so ci ated with self-re - ported, be hav ioural, and bio chem i cal in di ces of stress (23). As more peo ple are forced to in ter act and share a com mu nal space as is the case on a crowded psy chi at ric ward their pri vacy, the so cial den sity of the ward, and the control they have on their en vi ron ment are eas ily dis rupted. Con se - quently, pa tients ex pe ri ence in creased stress lev els, which then pre cip i tates vi o lence. Con versely, in en vi ron ments where den sity, pri vacy, and con trol are in op ti mal bal ance, vi - o lence is less likely to oc cur. It has been pro posed that stress can be re duced and pri vacy to pa tients can be en sured through such simple en vi ron men tal strat e gies as al lo cat ing in di vid ual rooms to pa tients, pre vent ing crowd ing, and matching the ward ac tiv i ties to the needs and ca pac i ties of pa tients (22). The role played by the ar chi tec ture of acute-care psy chi at ric wards in the oc cur rence of vi o lence needs to be ex am ined be - cause pri vacy, den sity, and control are fac tors that can be ma - nip u lated by pay ing at ten tion to ar chi tec ture. Area 2: Violence and the Architecture of Acute-Care Psychiatric Wards There is often a lack of com mu ni ca tion be tween hos pi tal ar - chi tects, pa tients, and staff work ing on acute-care psy chi at ric wards. Con se quently, the ar chi tects of psy chi at ric wards of - ten over look or un der es ti mate the im pact of den sity and the loss of pa tients pri vacy and control over their en vi ron ment (28,29). As a re sult, their de sign may par tially con trib ute to the oc cur rence of vi o lence, given the sig nif i cant dis crep ancy be tween liv ing con di tions in the com mu nity and those of a psy chi at ric ward. Liv ing con di tions in most West ern coun - tries al low peo ple to have a great de gree of pri vacy, con trol, and per sonal liv ing space to con duct ac tiv i ties. The pa tient on a psy chi at ric ward, on the other hand, is faced with sharing com mu nal rooms, such as the kitchen and lounge, largely with strang ers. These ar eas are of ten cramped al ready, and a high ward pop u la tion may worsen the sit u a tion. Areas per - ceived to be pri vate, such as bed rooms, are smaller, and there is less ob jec tive area that the pa tient can define as his or her own. Not only are pa tients ob jec tively crowded, but they may also strongly ex pe ri ence sub jec tive crowd ing in such high den sity sit u a tions, be cause they see more people around them and ex pe ri ence in creased in tru sion. A sense of frus tra tion and an ger are ex pected out comes, es pe cially when they feel that they have lost con trol over their en vi ron ment. This situation can be eas ily re versed by pay ing at ten tion to ar chi tec tural prin ci ples that en hance the ther a peu tic en vi ron ment (29). Such strat e gies in clude ded i cat ing space for so cial in ter ac - tion, clearly in di cat ing a room s in tended use, mak ing ar eas vi su ally dis tinct so that the in tended use of dif fer ent parts can be de lin eated from their ap pear ance, using colours to en hance ac tiv i ties and spaces, us ing var i ous ma te ri als to pro vide dif - fer ent tac tile and vi sual ex pe ri ences, using lighting to help de - fine space, and fi nally, making the spaces that have spe cial mean ing to pa tients stand out (28). En hanced di a logue be - tween ar chi tects and mental health pro fes sion als can en list ar - chi tec ture as a ther a peu tic tool and re duce en vi ron men tal stress, which in turn should lead to re duced vi o lence on psy - chi at ric wards. Area 3: Violence and Social Organization of Acute-Care Psychiatric Wards The im pact of a dif fer ent liv ing en vi ron ment be comes clearer when the ward s so cial or ga ni za tion is con sid ered. The em - pha sis on psy chi at ric wards is to so cial ize and to par tic i pate in ac tiv i ties, which in creases in tru sion upon the pa tient s per - sonal liv ing area and ac tiv i ties. Sim i larly, staff main tain ing rou tine ob ser va tions and dis pens ing med i ca tions frequently en ter the pa tients bed rooms or personal space, and pa tients might per ceive staff as forc ing them to so cial ize. Other pa - tients may also in trude, and this may be more ap par ent in com mu nal rooms. Fur ther, con cen tra tion of more se verely ill

3 June 2001 Crowding and Violence on Psychiatric Wards: Explanatory Mod els 435 pa tients in shrink ing in pa tient fa cil i ties and fre quent rehospitalization of se verely dis turbed pa tients who spend shorter pe ri ods out side hos pi tals col lec tively lead to ward tur - moil a fac tor as so ci ated with vi o lence on acute-care inpa - tient wards (1,30). In crowded con di tions, so cial with drawal may be used as a cop ing strat egy that is per haps sim i lar to the ur ban dweller s wish to be iso lated and anon y mous, given the ur ban char ac - ter is tics of den sity and het er o ge ne ity (31). We see sim i lar i ties be tween pa tients so cial in ter ac tions on crowded acute-care psy chi at ric wards and those of ur ban dwell ers both hav ing a ten dency to be su per fi cial and tran si tory. The natural re - sponse to with draw in crowded sit u a tions may be met with dis ap proval by staff on ther a peu tic grounds. Yet, if pa tients are forced to go against this nat u ral re sponse due to the expec - ta tions of the ward cul ture, stress in creases and vi o lence may en sue. The in creas ing di ag nos tic het er o ge ne ity of psy chi at ric inpa - tients may also be an im por tant fac tor in in creas ing stress, due to the lack of a con sis tent ther a peu tic ward en vi ron ment. One study, for ex am ple, re ported that pa tients with psy cho sis need an en vi ron ment scoring low on the an ger and ag gres sion and high on the sup port subscales of WAS (Ward At mo sphere Scale), as op posed to nonpsychotic pa tients, who need low lev els of staff control and in ter me di ate an ger and ag gres sion lev els (32). Al though it is im prac ti cal to pro vide dif fer ent ward en vi ron ments to pa tients with dif fer ing di ag no ses, the con di tions for high stress levels will ex ist on crowded wards with such a mix of patients. Area 4: Violence and Body Buffer Zone It would ap pear that sub jec tive crowd ing, when peo ple per - ceive an en vi ron ment as crowded, is more likely to pre cip i tate vi o lence than ob jec tive crowd ing. The for mer, but not the lat - ter, has been as so ci ated with ad verse mental health out comes (33 35). In this re gard, we dis tin guish be tween the vari able of den sity as de scribed above and body buffer zone. The for mer is an ob jec tive mea sure of crowding and the lat ter a sub jec tive sense that shapes our per cep tion of crowd ing. In one study, Nijman spec u lated that sub jec tive crowding was im por tant in pre cip i tat ing vi o lence (20). Body buffer zone in - flu ences our per cep tion of what is our space and when we feel that somebody has in truded upon it, and it could be an im por - tant de ter mi nant for sub jec tive crowding. A body buffer zone is de fined as the area that de mar cates what is per ceived as in ner vs outer self (36). Anx i ety oc curs if an other per son en ters this area. Studies have shown that vio - lent pris on ers re quire a greater buffer zone than do non vi o lent pris on ers (37,38), with Kinzel re port ing that vi o lent pris on ers fre quently mis in ter preted oth ers as loom ing or rush ing to ward them (37). One can spec u late whether pa tients with a his tory of vi o lence or poor im pulse control might also have large body buffer zones and whether, when in truded upon in a crowded sit u a tion, they re spond with ag gres sion. The con cept of body buffer zones can also help us un der stand the interindividual dif fer ences in vi o lent re sponse to crowd ing. Per sonal and cultural dif fer ences ex ist among peo - ple, which may af fect body buffer zone size and, conse - quently, the sense of sub jec tive crowd ing (37,38). It is also pos si ble that, with in creas ing ur ban iza tion across the world, more peo ple are living in dense and cir cum scribed ar eas (33,39) and, there fore, ex pe ri enc ing a pro gres sive restriction in the amount of space avail able to them. They may per ceive their body buffer zones to have di min ished, mak ing them less tol er ant of crowd ing. With the added ef fects of high stress lev els and a com pro mised abil ity to con trol im pul sive out - bursts, it is pos si ble that some pa tients may be primed to re - spond vi o lently to in tru sions and tur moil on crowded psy chi at ric wards. Area 5: Phylogenic Models of Violence and Crowding From a phylogenic point of view, it has been pro posed that vi - o lence serves an in stinc tive and adap tive func tion in situa - tions where food, ter ri tory, and mates are con tested be tween 2 or more an i mals (40). Such com pe ti tion for re sources may oc - cur in times of over pop u la tion or crowding (41). Ag gres sion is seen as a nat u ral re ac tion to crowd ing that dis perses ani - mals over a wide area, thereby con serv ing scant food re - sources (42,43). It has been pos tu lated that most fights are no more than tri als of strength fol lowed by dis en gage ment and rapid with drawal by the weaker (44). Death and se ri ous phys i cal in jury are avoided through rit u al ized threat and sub - mis sion ges tures, and con se quently, a so cial hi er ar chy is formed, lead ing to unity and sta bil ity (45). All so cial animals, in clud ing hu man be ings, have ag gres sive be hav iour and dom i nance rit u als as eco log i cal ad ap ta tions (42). As hu mans have moved up the phylogenic ech e lon, disap - proval of ag gres sion in in ter per sonal in ter ac tions has in - creased (46); there is in creas ing re li ance on in tel lec tual strengths, and less on phys i cal con fron ta tion, to establish dom i nance within a par tic u lar so cial hi er ar chy. Where this ac quired in hi bi tion is ab sent or di min ished, how ever, one may expect to see re-emer gence of ag gres sion as a coping strat egy to es tab lish dom i nance. This phe nom e non can typi - cally be ob served in acute-care psy chi at ric wards be cause, in acute states of most mental ill ness, peo ple are less pre oc cu - pied with so ci etal ex pec ta tions and norms. In deed, more ver - bal con fron ta tions than phys i cal as saults were reported in a study of pa tients with head in jury (47) and on an acute-care psy chi at ric ward (21). Area 6: Anthropology and Violence An thro po log i cal stud ies have sug gested that prim i tive hu man be ings did not de sire serious or con tin u ous com bat with neigh bours (48). Among prim i tive hunt ing tribes, there would be an equi lib rium with neigh bour ing tribes, and war was a rare event (43). Ste reo typed ag gres sion that avoids se ri - ous in jury or death is seen in tribal peo ples, such as Aus tra lian Ab orig ines (49). In mod ern so ci ety, on the other hand, the frus tra tion of over crowd ing and over pop u la tion are of ten as - so ci ated with in creas ing vi o lence and civil dis tur bance (50). It is thought that ag gres sion which is meant to avoid se ri ous in jury is cul tur ally trans mit ted (50), and it is pos si ble that we con tinue to use it as a cop ing strat egy in sit u a tions where our

4 436 The Canadian Journal of Psychiatry Vol 46, No 5 in di vid u al ity co mes un der threat. Prim i tive pat terns of ag - gres sion in hu man be ings for the pur poses of self-pres er va - tion or trials of strength, as op posed to more se vere phys i cal con fron ta tion, may ex plain the use of ver bal, rather than phys i cal, ag gres sion on crowded psy chi at ric wards, as re - ported by Ng and oth ers (21). Conclusions, Possible Solutions, and Future Research None of the above models on their own can ex plain the re la - tion be tween vi o lence and crowd ing. A pos si ble ex pla na tion may only come from com bin ing them. From a phylogenic and an thro po log i cal per spec tive, hu mans are pre dis posed to ex - hibit violent be hav iour in the in ter ests of self-pres er va tion in stress ful or crowded sit u a tions. Such vi o lence is less likely to cause serious in jury or death and could well man i fest as ver - bal ag gres sion or tri als of strength. In ad di tion, the stress aris - ing from the ad mis sion pro cess and the disinhibiting ef fects of some mental dis or ders must also be con sid ered. Fur ther, it ap pears that the so cial or ga ni za tion and ar chi tec ture of the acute-care psy chi at ric ward may have an ad di tive ef fect on such pre dis po si tions, es pe cially when the loss of pri vacy and con trol over one s en vi ron ment is con sid ered. The re sult ing low ered frus tra tion tol er ance may pre cip i tate vi o lent in ci dents. It is dif fi cult to offer any so lu tions to the prob lem of crowding and vi o lence. Larger in pa tient units or more beds may only tem po rarily ad dress the prob lem, be fore they, too, be come over crowded. While crowd ing may be un avoid able, some sug ges tions may be made to mit i gate its ad verse ef fects on peo ple s men tal health. Hope fully, this may re duce the oc cur - rence of vi o lence on acute psy chi at ric wards. None the less, it must be noted that a re cent study found that the ad di tion of ex - tra space on an acute-care psy chi at ric ward did re duce ag gres - sive in ci dents (20). This study, how ever, did not re port on the ex tent to which pa tients used the added space an el e ment that may have sig nif i cant im pact on fac tors such as dis per sion of pa tients and sub jec tive per cep tion of crowd ing (52). In a study of Thai house holds, Fuller and oth ers (33) of fered sev eral so lu tions for crowd ing, in clud ing spend ing time away from the house alone, so cial iz ing with oth ers out side the home, with draw ing or ig nor ing the de mands of oth ers, or turn ing some living ar eas into sep a rate rooms to heighten in - di vid ual pri vacy and con trol. There were also in di rect so lu - tions, such as long com mutes to and from work that re duced ob jec tive crowding in fam ily house holds. Ex trap o lating from this study, we pro pose that dis per sion of pa tients on a crowded ward may help to re duce vi o lent in ci dents. Ac tiv - ities can ei ther be ar ranged out side the unit (for ex am ple, group out ings or walks) or in side the unit (us ing ac tiv ity and oc cu pa tional ther apy rooms). Day or over night leave may be pos si ble for se lected pa tients, al though in this re spect, care ful con sid er ation must be given to is sues like safety and ap pro - pri ate tim ing of leave. Cli ni cians should be aware and screen for a sense of sub jec tive crowd ing, par tic u larly in those pa - tients who may have poor impulse con trol. These rec om men - da tions will need to be sub stan ti ated by well-de signed studies. Clinical Implications With the trend to reduce acute-care inpatient beds, crowd ing is likely to become a significant issue. The association between crowd ing and violence on acute-care psy - chiatric wards needs further investigation. Explanatory models for this relation can be developed in an interdis - ciplinary context. Limi ta tions The explanatory models are largely hypothetical and will need fur - ther corroboration. The applicability of models extrapolated from the unaffected popu - lation to the psychiatric population may be limited. Cross-cultural applicability of the proposed models may be limited. Three ar eas of fur ther re search are needed in this im por tant as pect of psy chi at ric care. First, fur ther nat u ral is tic stud ies are needed to con firm the re la tion be tween crowding and ag - gres sion. To derive mean ing ful con clu sions, these stud ies will need to be able to min i mize the ef fects of ex ter nal ob ser - va tion on the in ci dence of vi o lence on acute-care psy chi at ric wards. Sec ond, cross-cul tural stud ies are needed to clarify the sig nif i cance of ob jec tive and sub jec tive crowd ing. Finally, in ter ven tion ist stud ies are re quired to de ter mine the suc cess of any dis persal method in re duc ing pa tient vi o lence. Any ex - per i men tal stud ies looking at re duc ing vi o lence by in creas ing the available ob jec tive space for pa tients should also ex am ine pa tients sub jec tive sense of space and crowd ing. Acknowledgements The authors thank Dr R Liang and Dr L Blackstock for comments on ear lier drafts. References 1. Haller RM, Deluty RH. As saults on staff by psy chi at ric in-pa tients. A crit i cal re - view. Br J Psy chi a try.1988;152: Da vis S. Vi o lence by psy chi at ric in pa tients: a re view. Hospital and Com mu nity Psy chi a try 1991;42: Reid WH, Bol linger MF, Ed wards G. As saults in hos pi tals. Bull Am Acad Psy chi a - try Law.1985;13: Pearson M, Wil mot E, Padi M. A study of violent be hav iour among in-pa tients in a psy chi at ric hos pi tal. Br J Psy chi a try.1986;149: Nijman HLI, Allertz WWF, Merckelbach HLGJ, Campo JLMG, Ravelli DP. Ag - gres sive be hav iour on an acute psy chi at ric ad mis sions ward. Eu ro pean Jour nal of Psy chi a try 1997;11: Palmstierna T, Huitfeldt B, Wistedt B. The re la tion ship of crowding and ag gres sive be hav iour on a psy chi at ric in ten sive care unit. Hos pi tal and Com mu nity Psychiatryiatry 1991;42: Poster EC. A mul ti na tional study of psy chi at ric nursing staffs be liefs and con cerns about work safety and patient assault. Arch Psychiatr Nurs 1996;10: James DV, Fineberg NA, Shah AK, Priest RG. An in crease in vi o lence on an acute psy chi at ric ward: a study of as so ci ated fac tors. Br J Psy chi a try 1990;156: No ble P, Rodger S. Vi o lence in psy chi at ric in-pa tients. Br J Psy chi a try 1989;155: Lehmann LS, Padilla M, Clark S, Loucks S. Training per son nel in the pre ven tion and man age ment of vi o lent be hav iour. Hosp Com mu nity Psy chi a try 1983;34: Storch DD. Crowding and the fis cal cri sis. Hos pi tal and Com mu nity Psy chi a try 1992;43: MILMIS Pro ject Group. Mon i toring inner Lon don mental ill ness ser vices. Psy chi - at ric Bul le tin 1995;19: Deahl M, Turner T. Gen eral psy chi a try in no man s land. Br J Psy chi a try 1997;171:6 8.

5 June 2001 Crowding and Violence on Psychiatric Wards: Explanatory Mod els Drinkwater J, Gudjonsson GH. The nature of vi o lence in psy chi at ric hos pi tals. In: Howells K, Hollin CR, ed i tors. Clin i cal ap proaches to vi o lence. Chichester: Wiley; Lanza ML. Nurses as patient as sault vic tims: an up date, syn the sis, and rec om men - da tions. Arch Psych Nurs 1992;6: Kalogerakis MG. The assaultive psy chi at ric pa tient. Psychiatr Q 1971;45: Lanza ML, Kayne HL, Hicks C, Milner J. En vi ron men tal char ac ter is tics re lated to pa tient as sault. Is sues in Mental Health Nursing 1994;15: Hardie TJ. Crowding and vi o lent be hav iour: the in flu ence of patient den sity on vi o - lent and self harm ing be hav iour at a me dium se cure unit. Med Sci Law 1999;39: Owen C, Tarantello C, Jones M, Ten nant C. Vi o lence and ag gres sion in psy chi at ric units. Psychiatr Serv 1998;49: Nijman HLI, Rec tor G. Crowding and ag gres sion on in pa tient psy chi at ric wards. Psychiatr Serv 1999;50: Ng B, Kumar S, Ranclaud M, Rob in son E. Ward vi o lence: a pos i tive re la tion ship with ward crowd ing. A twelve month ret ro spec tive study. Psychiatr Serv. Forth - com ing. 22. Nijman HL, a-campo JM, Ravelli DP, Merckelbach HL. A ten ta tive model of ag - gres sion on in pa tient psy chi at ric wards. Psychiatr Serv 1999;50: Spencer A, Baum A. Ar chi tec ture and health. In: Baum A, Newman S, Weinman J, West R, McManus C, ed i tors. The Cam bridge handbook of psy chol ogy, health and med i cine. Cam bridge: Cam bridge Uni ver sity Press; p Baum A, Koman S. Dif fer en tial re sponse to an tic i pated crowd ing: psy cho log i cal ef fects of so cial and spatial den sity. J Pers Soc Psychol 1976;34: Baum A, Valins S. Ar chi tec ture and so cial be hav iour: psy cho log i cal stud ies of so - cial den sity. Hillsdale (NJ): Erlbaum; Domachowski W. Func tions of in ter per sonal space. Przeglad-Psychologiczny 1980;23; Codol JP. Espace per son nel, dis tance interindividuelle et densite sociale (Per sonal space, in ter per sonal dis tance and so cial den sity). Re - vue-de-psychologie-appliquee 1978;28: Gulak MB. Ar chi tec tural guide lines for state psy chi at ric hos pi tals. Hos pi tal and Com mu nity Psy chi a try 1991;42: Shrivastava R, Kumar S, Ja cob son RR. Psy chi at ric units in dis trict hos pi tals: de sign is sues. In ter na tional Med i cal Jour nal 1999;6; Krakowski M, Czobor P. Vi o lence in psy chi at ric pa tients: The role of psychosis, fron tal lobe im pair ment, and ward tur moil. Compr Psy chi a try 1997;38: Wirth I. Ur bani sa tion as a way of life. Am J Sociol 1938;44: Friis S. Char ac ter is tics of a good ward at mo sphere. Acta Psychiatr Scand 1986; 74: Fuller TD, Ed wards JN, Vorakitphokatorn S, Sermsri S. Chronic stress and psy cho - log i cal well be ing: ev i dence from Thai land on house hold crowd ing. Soc Sci Med 1996;42: Booth A. Ur ban crowd ing and its con se quences. New York: Praeger; Gove WR, Hughes M. Over crowding in the house hold. New York: Ac a demic Press; Horowitz MJ, Duff D, Stratton L. The body buffer zone: an ex plo ra tion of per sonal space. Arch Gen Psych 1964;11: Kinzel AF. Body buffer zone in vi o lent pris on ers. Am J Psy chi a try 1970;127: Hildreth AM, Derogatis LR, McCusker K. Body buffer zone and vi o lence: a re as - sess ment and con fir ma tion. Am J Psych 1971;127: Marsella AJ, Escudero M, Gordon P. The ef fects of dwelling den sity on men tal dis - or ders in Filipino men. J Health Soc Behav 1970;11: Free man D. Ag gres sion: in stinct or symp tom? Aust N Z J Psy chi a try 1971; 5: Rus sell C, Rus sell WMS. Vi o lence, mon keys and man. London: MacMillan; Wynne- Ed wards VC. An i mal dis per sion in re la tion to so cial be hav iour. Lon don: Ol i ver and Bod; Zegans LS. To wards a unified theory of hu man ag gres sion. Br J Med Psychol 1971;44: Lorenz K. Rit u al ized fight ing. In: Carthy JD, Ebling FJ, ed i tors. The nat u ral his tory of ag gres sion. New York and Lon don: Ac a demic Press; p Roth M. Hu man vi o lence as a viewed from the psy chi at ric clinic. Am J Psy chi a try 1972;128: Lion. JR. Ag gres sion. In: Kaplan HI, Saddock BJ, ed i tors. Com pre hen sive text - book of psy chi a try. 6th ed. Bal ti more: Wil liams and Wilkins; p Grafman J, Schwab K, War den D, Pridgen A, Brown HR, Salzar AM. Frontal lobe in ju ries, vi o lence and ag gres sion: a re port of the Viet nam head injury study. Neu - rol ogy 1996;46: Malinowski B. An an thro po log i cal anal y sis of war. Amer i can Jour nal of So ci ol ogy 1941;46: Jones IH. Ste reo typed ag gres sion in a group of Aus tra lian Western Desert Ab orig - ines. Br J Med Psychol 1971;44: Rus sell C, Rus sell WM. The natural his tory of vi o lence. J Med Eth ics 1979; 5: Kumar S, Ng B, Rob in son E. The crowded ward! Psychiatr Serv. 1999;50: Résumé La violence dans les uni tés psychiatriques et le surpeuplement : des modèles explicatifs Ob jec tif : La vio lence dans les uni tés de soins psy chia tri ques aigus est très répan due et l on a re connu que le sur peu ple ment est un im por tant fac teur de risque. Le pré sent ar ti cle cher che à sa voir pour quoi les pa tients peu vent répon dre au sur peu ple ment par la vio lence. Méth ode : Une re vue de la docu men ta tion a été menée dans Med line à l aide des mots clés vio lence et sur peu ple ment. Une re - cher che manu elle ad di tion nelle des réfé rences re cueil lies dans les articles ex ami nés a aussi été menée. Résul tats : Les fac teurs spé ci fiques de la re la tion en tre le sur peu ple ment des unités de soins psyc hia tri ques ai gus et la vio lence peu vent se di viser selon les caté go ries suivantes : 1) la den sité, l in ti mité et le con trôle des pa tients, 2) l ar chi tec ture de l u nité, 3) l or gani sa tion so ci ale des unités psy chia tri ques, 4) l espace in ter per son nel, 5) les théo ries phy logé niques et 6) les théo ries an - thro polo giques du com por te ment hu main. Con clu sions : Les modèles ex pli ca tifs de cette re la tion sont of ferts et des straté gies sont suggé rées pour con trer les ef fets du sur - peu ple ment. Des re com man da tions sont faites pour de fu tures études.