Alternatives to Acute Hospital Psychiatric Care in East-End Montreal

Original Research

Alternatives to Acute Hospital Psychiatric Care in East-End Montreal Alain D Lesage, MD 1,2, Charles Bonsack, MD3 , Do ris Clerc, MD2, Claude Vanier, MD 2, 2 2 2 2 Maryse Charron, MD , Marc Sasseville, MD , André Luyet, MD , Dan iel Gélinas, MSW

Objective: As pres sure mounts to re duce the number of costly acute care beds, gov ernments and the lit era ture pro pose top- down ra tios. Is this rea son able and fair to the re spon sible medi cal of fi cers who, as the key care pro vid ers, will need to ad mit pa tients and de velo p dis charge plans in a reduced- beds en vi ron ment? Method: Treat ing phy si cians of all acute care in pa tients on a given day ( n = 212) and all new acute care ad mis sions over a 2- week pe riod (n = 125) com pleted an adapted ver sion of the Not ting ham Acute Beds Use Sur vey (NABUS) Ques tion naire. Re sults: On a given day, only 62 of 212 in pa tients were un suited for any al ter na tive to acute care hos pi tali za tion. A floor ra tio of 18 acute care beds per 100 000 in habi tants seems ade quate for the catch ment area in ques tion, pro vided that al ter na tives to hos pi tali zation are fully and ef fi ciently avail able. Al ter na tives es sen tially in volve an ar ray of the fol low ing: su per vised resi den tial set tings, day hos pi tals, and in ten sive home care (2 to 6 hours weekly). The ra tio of in ten sive home care work ers re quired would be 25 per 100 000 inhabitants. (Can J Psy chia try 2002;47:49–55)

Clini cal Im pli ca tions • Cli ni cians rec og nize the need for both hos pi tali za tion and al ter na tives, such as in tensive home care (IHC), day hos pi tal, and su per vised resi den tial set tings. • At ten tion should be paid not only to the number of beds, but also to the avail abil ity and ra tios of community- based serv ices such as IHC and su per vised resi den tial set tings. • The strength of cur rent as sess ment of needs pre sented here rests not on top- down evalua tion but on bottom- up as sess ment by treat ing psy chia trists. Limitations • In the end, needs as sess ment rests on clini cal judge ment, even when sys tem ati cally re corded, such as is the case here. • The as sess ment of needs for su per vised resi den tial fa cili ties could not be ac com plished with the de sign of this study, but will be on the re search agenda in the near fu ture.

Key Words: acute care beds, alternatives to hospitalization, intensive home care, day hospital, supervised residential settings

W Can J Psy chia try, Vol 47, No 1, Feb ru ary 2002

51

The Ca na dian Jour nal of Psy chia try—Orig i nal Re search

educ ing the num ber of costly long-term and acute care psy chi at ric beds in the health care sys tem is an ob jec tive that regional and national planners in Eu rope and North Amer ica con tinue to pur sue within to day’s cost-containment en vi ron ment. Na tional and in ter na tional com par i sons of variances in bed uti li za tion, as well as ex pert-based ra tios, support the top-down contention that cer tain jurisdictions manage to operate very well with fewer beds. In Canada, prov inces man age health and social service planning and bud get ing (1). Re cently, the Prov ince of Que bec (pop u la tion over 7 mil lion) an nounced a fur ther 50% re duc tion in all psychi at ric beds, bring ing the cur rent bed ra tio from 1 to 0.4 per 1000 in hab it ants and the acute care bed ra tio to 0.25 per 1000 (2). Un der this cost-cutting plan, freed re sources are to be redi rected to com mu nity ser vices for pa tients with se vere mental ill ness. These ser vices in clude proper res i den tial set tings, as ser tive com mu nity treat ment (ACT), greater avail abil ity of treat ment and re ha bil i ta tion, and sup port for rel a tives. The plan doc u ment de scribes bed uti li za tion and ra tios in other juris dic tions but makes no men tion of the ra tios for al ter na tive re sources to hos pi tal iza tion, such as res i den tial set tings. The doc u ment com pares other Ca na dian prov inces, such as Ontario (pop u la tion over 11 mil lion), where the bed ra tio is already one-half that in Que bec. It fails to add, how ever, that many Eu ro pean coun tries, in clud ing France which in part inspired Que bec’s psy chi at ric sectorization 3 de cades ago, oper ate with higher bed ra tios than those in Que bec (3).

R

Wing (4) and Har ri son (5) sug gested the best plan ning approach for a flex i ble sys tem is to com bine top-down data from in di ca tors such as those ob tained from mod el ling uti li za tion against so cio eco nomic vari ables, with bot tom-up meth ods of needs assessment that involve ser vice pro vid ers. Only recently, how ever, have ef forts been made to de velop sim ple meth ods to as sess needs for acute care beds (6). Beck and others (7) de vel oped a pro ce dure to tap the re spon si ble med i cal of fi cer’s rec om men da tion on need for ad mis sion and on alter na tives to acute care beds. Con ducted in 3 ur ban ar eas (2 in Lon don, 1 in Nottingham), their work in di cated that further re duc tions in acute care beds were pos si ble. Fur ther, they outlined the nec es sary al ter na tives, mainly su per vised res i dential set tings and in ten sive home care. In mid 1996, Louis-H Lafontaine Hos pi tal, con cerned over forth com ing bed cuts, did a study of acute care bed uti li za tion to de ter mine whether ex ist ing beds were nec es sary and what the alternatives might be. The treating psy chiatrists of all acute care in pa tients on a given day and of all new acute care admissions over a cer tain time pe riod com pleted an adapted version of the ques tion naire that was used in the 2 British stud ies men tioned above. This ar ti cle pres ents the re sults of the inves ti ga tion. Be cause most patients at Louis-H Lafontaine Hos pi tal are res i dents of the sectorized catch ment area in east-end Montreal, the area lends it self to an 52

epidemiologically based as sess ment of needs and to in ter national com par i son (4,8).

Methods Brief De scrip tion of Ser vice The catch ment area in ques tion was pre vi ously de scribed in Lesage and oth ers (8). Briefly, psy chi at ric ser vices in Montreal were sectorized in the early 1970s. Since then, multidisciplinary teams have been re spon si ble for pro vid ing ser vices to the pop u la tion within dis tinct geo graphic sec tors of catchment ar eas. It is es ti mated that about 85% of Louis-H Lafontaine out pa tients are res i dents of the area (341 730 inhab it ants) that the hos pi tal is in tended to serve. The hos pi tal set up sat el lite multidisciplinary com mu nity psy chi at ric clinics in 7 sec tors of the catch ment area. The num ber of acute care beds avail able to each subcatchment area is not de ter mined by a quota but rather on the ba sis of need and the avail abil ity of 200 acute care beds in 1996 (ra tio of 0.58 per 1000 in hab itants). An emer gency clinic operates 24 hours daily, 7 days weekly. There are over 4000 con sul ta tions yearly, and it is visited by about 0.6% of the area’s pop u la tion an nu ally (8). It is nor mally staffed by med i cal con sul tants in at ten dance and less fre quently by reg is trars and se nior reg is trars, with the backup of con sul tants avail able 24 hours daily. The emer gency clinic op er ates a cri sis-bed unit and a buffer zone with stretch ers for acute care ad mis sions await ing a bed va cancy. In 1996, there were no day hos pi tals, and day centres were available primarily for rehabilitation pur poses. Louis-H Lafontaine psychiatric hospital, until recently the largest in Can ada, pro vides long-term in pa tient care and re ha bil i ta tion ser vices to the catch ment area, but un til the 1980s ad mit ted patients from the French-speaking com mu nity of Mon treal and the sur round ing ar eas. It is es ti mated that, at the time of the study, two-thirds of the 700 long-stay in pa tients and of the 1200 pa tients in su per vised res i den tial set tings and re ha bil i tation ser vices re sided out side the catch ment area. In the past, the res i den tial fa cil i ties com prised mainly fos ter fam i lies and hostel schemes; over the past decade, how ever, su per vised apartments, su pervised hostels, halfway houses, and crisis cen tres have also been de vel oped. Al though at the time of the study the multidisciplinary teams at the out pa tient clin ics did not pro vide home care, a group of 5 men tal health work ers super vised a group of apart ment res i dents (ra tio of 1:20). In ad dition, nurses vis ited the 400 hos tel res i dents reg u larly (in a ra tio of 1:80). In ten sive home care was un avail able then. Over the past 5 years, there has been a vir tual freeze on long-stay ad missions for the catch ment area’s pa tients. Sampling The sam pling meth ods were cho sen to ob tain a clear pic ture of both the num ber of acute care beds re quired at any time and the WCan J Psy chia try, Vol 47, No 1, Feb ru ary 2002

Alternatives to Acute Hospital Psychiatric Care in East-End Montreal

Table 1. Sample groups compared All pa tients ad mit ted in 1996 (n = 1926)

Part I, 1- day cen sus of Part II, 2- week in ci dence acute care ad mis sions on of new ad mis sions from 23 Oc to ber 13 to 27 No vem ber ( n = 212) (n = 92)

To tal Part I and II (n = 304)

n (%)

n (%)

n (%)

%

1 year

14 (1)

14 (7)

0 (0)

5

6–12 months

80 (4)

62 (29)

1 (1)

21

3–6 months

158 (8)

45 (21)

7 (8)

17

1–3 months

460 (24)

71 (33)

39 (42)

36

1–4 weeks

429 (22)

16 (8)

25 (27)

13

1–7 days

785 (41)

4 (2)

20 (22)

8

need for al ter na tive re sources yearly. Con se quently, we em ployed a mixed 3-part sampling method. Part 1 covered all 212 acute care in pa tients, ei ther hos pi tal ized or await ing ad mis sion on a given day (that is, Oc to ber 23, 1996). Part 2 cov ered all 125 new admissions to acute care wards or the cri sis-bed unit from No vem ber 13 to 27, 1996. Part 3 cov ered all 1926 pa tients ad mit ted to acute care wards or to the cri sis-bed unit in 1996. Questionnaire The Nottingham study group (7) provided us with its Nottingham al ter na tive to bed uti li za tion sched ule (NABUS). This instrument comprises 3 sec tions. The first cov ers the need for key el e ments of the care pack age, in clud ing res i den tial al ter na tives, home care, and day care. The sec ond in cludes rea sons for ad mis sion, while the third con tains symp toms and signs of psy chi at ric dis or ders and as sign ment of ICD-10 di agnoses. We translated the NABUS into French and then, to adapt it for lo cal psy chi a trists, we de vel oped op er a tional def ini tions for res i den tial al ter na tives. A to tal of 5 psy chi a trists did prestudy test ing on 12 cases, who were all pa tients un der their care. Re sults showed that the ques tion naire was easy to understand and there fore quick to com plete (un der 10 min utes). You can ob tain the mod i fied French ver sion of the in stru ment (the NABUS-F) from the first au thor. Procedure The com put er ized in for ma tion sys tem at Louis-H Lafontaine Hos pi tal pro vided ba sic data on the 3 sam ples, in clud ing age, sex, and date of ad mis sion. De tails on res i den tial set tings were not com put er ized and could be ob tained di rectly from pa tient files for Part 1 and 2 sam ples. Ques tion naires were sent to the treat ing con sul tants of pa tients in Part I and 2 sam ples. Re turns were checked, and staff at the De part ment of Psy chi a try W Can J Psy chia try, Vol 47, No 1, Feb ru ary 2002

and the au thors, han dled re mind ers. Of the 212 in pa tients in the Part 1 sam ple, the re turn rate was 100%, and the re turn rate for the new ad mis sions of the Part 2 sam ple was 74% (92 of 125). Analysis It seemed best to de ter mine the need for acute care beds by look ing at the 1-day cen sus (Part 1 sam ple) and to es ti mate the re quired num ber of al ter na tive re sources over a 1-year pe riod. Flannigan and oth ers pointed out, how ever, that 1-day censuses overrepresent patients with lon ger stays (6). Ta ble 1 shows that the Part 1 sam ple con tained an ex ces sive num ber of pa tients hos pi tal ized for a pe riod of over 3 months (con tin uous or frag mented), com pared with the 1926 in pa tients for all of 1996. We chose the 2-week in ci dence pe riod (Part 2) to com pen sate for this ef fect, and the com bined Part 1 and 2 sample pro vided a better sam pling of the to tal pa tients for 1996 accord ing to length of stay. We used this com bined sam ple to pro ject the num ber of re quired al ter na tive re sources based on the case-mix ap proach (4). This method as sumes that pa tients’ needs in a given length-of-stay stra tum are com pa ra ble. Using this ap proach with the com bined sam ple meant that the con fidence in ter val (CI) of es ti mates for the short-stay stra tum (that is, less than 7 days) would be larger, be cause it is based on a smaller pro por tion of all the cases in this stra tum. Then, we developed a projection scenario and chose a maximal scenario. For each stra tum of the 304 pa tients in the com bined sam ple, we de ter mined the pro por tion of pa tients ex pected to need a given ser vice (for ex am ple, in ten sive home care) out of the pa tients ex pected to re quire an al ter na tive to ad mis sion in this stra tum. Next, we pro jected these ra tios on the 1996 patients in each stra tum. Such a sce nario, in keep ing with the case-mix ap proach, as sumes that all pa tients in a stra tum share the same needs at some point dur ing the year. 53


Table 2. Patient-care requirements over 1 year All pa tients Patients

In ten sive home care

Bed-years

Patients

Day hos pi tal Bed-years

Patients

Bed-years

Length of admissions in 1996

n

%

n

%

n

%

n

%

n

%

n

%

One year

14

1

14

6

0–7

0–0

0– 7

0–3

0–3

0–0

0–3

0–1

6–12 months

80

4

56

25

21–43

1–2

15–30

7–14

0–12

0–1

0–8

0–4

3–6 months

158

8

55

25

44–94

2–5

15–33

7–15

9–48

0–2

3–17

1–8

1–3 months

460

24

68

31

167–293

9–15

25–43

11–20

60–170

3–9

9–25

4–11

1–4 weeks

429

22

20

9

51–252

3–13

2–12

1–5

0–118

0–6

0–6

0–3

1–7 days

785

41

7

3

79–619

4–32

1–5

0–2

0–258

0–13

0–2

0–1

Total

1926

100

220

100

362–1308

19–68

58–129

26–59

69–609

4–32

12–61

5–28

Results Part 1 and 2 sam ples to gether to talled 304 pa tients, of which 45% were male and 21% were over 65 years. Or ganic brain syn drome was iden ti fied in 10%, psy cho ses in 40%, and major mood dis or ders in 38%. Comorbidity in volv ing sub stance abuse was iden ti fied in 19%, and per son al ity dis or der was iden ti fied in 28%. Most pa tients (64%) lived in their own homes at the time of admission. The others were in foster fam i lies (10%), hos tels (7%), hos tels su per vised by men tal health pro fes sion als (4%), nurs ing homes (3%), and gen eral hos pi tals (3%), or were home less (3%).

The re sults of Part 1 in di cated that, on a given day, treat ing psy chi a trists con sid ered that only 62 of the 212 pa tients in the sam ple (29%) were not suited for any al ter na tive to acute care admission. Long-stay psy chi at ric hos pi tal iza tion was ap propri ate for 15 pa tients (7%), and nurs ing homes were ap pro priate for an other 15 (7%); for the re main der, one-third (or 20% of the to tal sam ple) could be treated at home and 36% in some form of su per vised set ting. In ten sive home care (IHC) suited 31% of the patients, and day hospitals (DH) 11% (nearly one-half of these com bined with IHC). In ten sity of home care totalled at least 2 to 6 hours weekly for 69% of the com bined-sample pa tients suited for IHC. Ev i dently, fur ther analysis proved that a package of care rather than separate alternatives to hospitalization would be more appropriate. For the com bined sam ple of 304 pa tients, IHC was re quired for 59% of pa tients suited for home care and DH for 31%, compared with 50% and 13%, respectively, for patients suited for su per vised set tings. Pro por tions var ied ac cord ing to type of su per vised set ting: IHC was re quired for 50% to 81% of pa tients suited for a va ri ety of su per vised apart ments, for 53% of those suited for fos ter fam i lies, and for 29% of those suited for su per vised hos tels. Gen erally, this meant that 54

ex tra staff would be re quired to pro vide IHC to these pa tients in ex ist ing su per vised set tings. Pro jecting the num ber of ex tra re quired res i den tial re sources proved per il ous due to the fol low ing 2 sources of es ti mate errors: 1) the es ti mates for res i dence at time of ad mis sion (from Part 1 and 2 sam ple case notes) and 2) those for rec om mended res i dence for the Part 1 and 2 sam ples. Pro jecting the need for IHC and DH involved only 1 source of estimate errors; namely, pa tients rec om mended for IHC and DH, be cause neither of these ser vices ex isted at the time of ad mis sion (that is, equal to 0). Hence, only trends can be de scribed. For the combined sample, when an alternative was suit able, su per vised apart ments were rec om mended for 27 pa tients (at ad mis sion, 4 pa tients were al ready liv ing in such fa cil i ties); fos ter fam i lies for 15 (at ad mis sion 29); hos tels for 12 (at ad mis sion 20); professionally staffed hostels for 28, including 7 in halfway houses (at ad mis sion 11); long-stay hos pi tal wards for 14 (initially 1); nurs ing home for 20 (at ad mis sion 9); and cri sis centres for 5 (at ad mis sion 0). These fig ures in di cate a need for more supervised apartments, professionally staffed hostels, and more places in nurs ing homes and long-stay wards, or the equivalent. Ta ble 2 gives the max i mal sce nario es ti mates for IHC and DH. It shows first, as a re minder, the dis tri bu tion of the case mix of pa tients ac cord ing to length of stay and num ber of bed-years oc cu pied by the group in 1996. To gether, the strata of pa tients ad mit ted for more than 3 months com prised 13% of the 1996 co hort of acute care ad mis sions but oc cu pied 56% of the bed ca pac ity. The oc cu pancy ra tio was high—220 patient-years oc cu pied the of fi cial 200 acute care beds, with over 10 pa tients on stretch ers await ing a bed on any given day. The num ber of pa tients pro jected for IHC and DH is fol lowed by the num ber of beds these pa tients oc cu pied over 1 year, in di cat ing the poten tial re duc tion in beds that could be ef fected, pro vided that, WCan J Psy chia try, Vol 47, No 1, Feb ru ary 2002


to a cer tain ex tent, these al ter na tives to ad mis sion also pre vented ad mis sion. On av er age, IHC was pro jected un der the max i mal sce nario for 835 pa tients, with CIs in di cated in Ta ble 2, and DH for 288 pa tients. Pa tients with a lon ger length of stay were more likely to be con sid ered for IHC. By con cen trat ing ef forts on the pa tients ad mit ted for more than 3 months and en vis aged for IHC (65 to 144 pa tients com bined, or 3% to 8% of all ad mit ted pa tients), one would have a po ten tial yield of 30 to 70 beds yearly. This com pares against a yield of only 1 to 5 beds yearly for the short-stay stra tum (less than 7 days), that rep re sented 4% to 32% of all pa tients. The CIs were larger in the lat ter case, ow ing to the sam pling de sign. When pro jected on a to tal pop u la tion ba sis, the re quired num ber of acute care beds (62) was 18 per 100 000 inhabitants. Based on a rec om mended op ti mal ra tio of 1:10 for IHC such as ACT (9), the max i mal sce nario would in volve be tween 106 and 383 pa tients per 100 000 or be tween 10 and 38 IHC men tal health work ers per 100 000 (25 on av er age). For DH, es ti mates were between 20 and 179 patients per 100 000 inhabitants.

Discussion The pro por tion of pa tients suited for al ter na tives to acute care hos pi tal iza tion rep re sented more than 71% of the sam ple in this lo cal Ca na dian area. This is much higher than the 29% re ported in the Nottingham study and closer to the 60% and 52% reported in the Hammersmith and Fulham and Southwark stud ies, re spec tively. These fig ures have lo cal rel e vance only, be cause they were de ter mined by the lo cal area needs for psy chi at ric care, the con fig u ra tion of the lo cal sys tem in terms of acute care bed avail abil ity and al ter na tives to ad mis sion, and bed management by local practitioners (5). In the 1980s, Louis-H Lafontaine Hospital decided to curtail access to long-stay beds, thereby creating a de facto pool of new long-stay in pa tients in acute care wards in the ab sence of al ter na tives for these pa tients. At the time of the study, res i den tial ser vices were nu mer ous, but the ar ray was re stricted; DH was not set up, and IHC was un avail able. Com par i sons are pos si ble but only on as sessed needs, such as the pop u la tion ra tio of acute care beds. The Nottingham study pro jected a need for 112 beds for 396 138 in hab it ants, aged 15 to 64 years (that is, ap prox i mately 0.21 beds per 1000) (7). The Ful ham and Hammersmith and Southwark stud ies, re spectively, fixed the ratio at 0.26 and 0.36 per 1000 in hab itants (the cur rent sit u a tion be ing 0.65 and 0.76 beds per 1000). Har ri son (5) ar gued that Nottingham was closer to the UK na tional av er age, given that needs are well rec og nized as be ing higher in more socio eco no mi cally de prived ar eas, which is the case for in ner-city Lon don. Al though the east-end Mon treal catch ment area com prised se verely un der priv i leged sec tors with needs for psy chi at ric acute care beds—as much as W Can J Psy chia try, Vol 47, No 1, Feb ru ary 2002

twice as high as in other subcatchment ar eas (8)—over all, the sec tor’s in dex of de pri va tion com pared with that for the whole of the Mon treal area (10). This study’s needs as sess ment of 18 acute care beds per 100 000 in hab it ants re mains in the low end, com pared with the UK stud ies, and is lower than the Quebec gov ern ment’s pro posed 25 per 100 000 in 5 years (2). Further, the study pro vided es ti mates for al ter na tives to admission. The ma jor pat tern was for a pack age of su per vised residential set tings com bined with IHC. There might have been some con fu sion among cli ni cians in rat ing su per vised apart ments plus IHC. Even af ter con sid er ing this, the pat tern in Montreal resembles Nottingham’s in terms of the large need for res i den tial a l ternatives and the Fulham and Hammersmith pat tern in terms of the need for IHC. Ra tios for res i den tial set tings could not be pro duced be cause of the limited in for ma tion avail able, wherein the yielded CIs were too large to be use ful. More over, if an as sess ment of res i den tial needs for pa tients cur rently on long-stay wards or in the var ious res i den tial set tings is ab sent, pre dict ing the num ber of locally re quired fa cil i ties is im pos si ble. Other stud ies in the UK (11) and in Mon treal (12) sug gest that most long-stay in patients could be re lo cated to mod er ate- to high-supervised settings, but rarely in supervised apart ments. The trend in the needs of our com bined sam ple of pa tients from the area who were ad mit ted at least once to acute care wards in the past year sug gests that the num ber of places in fos ter fam i lies and hos tel schemes should be re duced. Fur ther, the pres ent lim ited number of places in supervised apartments and professionally staffed su per vised hos tels should be in creased. The de sign made it pos si ble to rea son ably es ti mate the pop ula tion-based ra tios of IHC and DH. Com par a tive data are rare, and data orig i nate mostly from ACT pro gram lit er a ture. In the province of On tario, guide lines for de vel op ing ACT teams esti mate that 15% of the pop u la tion with se vere men tal ill ness (2% of the gen eral pop u la tion or a ra tio of 300 per 100 000 inhabitants) require ACT (13), even though these estimates were de creased as Pro grams for As ser tive Com mu nity Treatment (PACT) were deployed. In Quebec, the Coun cil for Health Technology Evaluation suggested lower ratios for ACT teams, be tween 70 to 200 pa tients who re quire ACT per 100 000 (14). Our es ti mates for IHC prob a bly re coup 1) patients re quir ing PACT-type pro grams with their ca pac ity to de liver home care daily for years to pa tients with the most severe men tal ill ness; and 2) those pa tients (more nu mer ous) who may require for some months, intensive, but not daily acute home care, be fore be ing re ferred back to usual com munity psy chi at ric team. Finally, our es ti mates did not in clude the needs of long-stay in pa tients or su per vised-setting res idents who could move to more au ton o mous res i den tial settings with ACT. 55


The study’s lim i ta tions must be con sid ered. Those most of ten cited in such bot tom-up stud ies fo cus on the value of the clini cian’s judge ment, which is per ceived as in vari ably id io syncratic. Har ri son (5) has ar gued that this is coun ter bal anced by the na ture of the bot tom-up ap proach, rec og niz ing that needs also rest on the key de ci sion mak ers of ad mis sion and discharge, who will have to op er ate in the real en vi ron ment of reduced beds. In addition, their judgement is systematically elicited and recorded via meth ods and in stru ments that are sim ple and fea si ble. Sec ond, a max i mal sce nario was adopted be cause a min i mal sce nario would have in volved mak ing the un rea son able as sump tion that all in pa tients for whom no alter na tive to acute care was suit able on a given day would later have no need for home care, DH, or any of the dif fer ent res iden tial set tings. For pro vin cial, re gional, and lo cal plan ners in Que bec, the results of the pres ent study will come as a sur prise. Pro vin cial plan ners and non profit or ga ni za tions gen er ally con sider psychi a trists as hos pi tal-centred and resistant to bed cuts (15). The re sults came as a sur prise to Louis-H Lafontaine Hos pital, where it was ex pected that lo cal psy chi a trists, who were not fa mil iar with DH and IHC, would be more likely to recom mend al ter na tives more akin to hos pi tal care, such as DH. In stead, they fa voured care pack ages that de liver treat ment in the homes or res i den tial set tings of pa tients in their com munity. This con forms with the cur rent trend to ward ACT and sup ported hous ing that is gain ing mo men tum in North America (9,16). On one hand, Louis-H Lafontaine Hos pi tal now uses these results to sup port its re source de vel op ment, and on the other, it uses them to sup port down siz ing. The beds ra tio should not be seen as ab so lute, and the op ti mal num ber of beds for the area should be based on a more reasonable oc cupancy ratio to avoid changing thresholds for required hospitalization at peak pe ri ods dur ing the year (for ex am ple, 85% in stead of the cur rent 120%). Fur ther, it should take into ac count the timely de vel op ment of res i den tial al ter na tives. Spe cifically, the pi oneer ing work of Test and Stein (17) sug gested that IHC could be de vel oped most ef fi ciently first for the small num ber of patients who have used acute bed fa cil i ties most. We hope that, through this re search, Mon treal’s clin i cal and ad min is tra tive planners will proceed with caution, set up alternative resources be fore clos ing beds, and as sess needs on a reg u lar basis. Large ur ban ar eas may prove to have greater needs than ex pected, and clos ing beds too rap idly could place un bear able pres sure on re main ing beds, as was the case in Lon don (18). The cur rent lo cal area–based study and its use ful ness should en cour age other re gions to do sim i lar sur veys. With time, as local-area, provider-based, bottom-up ratios are published

56

and com pared against inter na tional, na tional, and ex pert-based, top-down ra tios, the for mer will in flu ence the latter and be come part of the new top-down ra tios, thus bring ing the plan ning and eval u a tion cy cle full cir cle (19).

References 1. Free man SJJ. An over view of Can ada’s men tal health sys tem. New Dir Ment Health Serv 1994;61:11–20. 2. MSSS. Ministère de la Santé et des Ser vices Sociaux du Que bec. Ori en ta tions pour la trans for ma tion des ser vices de santé mentale. Que bec: Gouvernement du Que bec; 1997. 3. Goldberg D. Com mu nity psy chi a try in Eu rope: as sess ment and eval u a tion. In: Tansella M, ed i tor. Making ra tio nal men tal health ser vices, Epidemiologia e Psichiatria Sociale Mono graph Sup ple ment 1. Rome: Il Pensiero Scientifico Editore; 1997. 4. Wing JK. Epidemiologically-based men tal health needs as sess ments. Re view of re search on psy chi at ric dis or ders (ICD-10, F2-F6). Lon don: Royal Col lege of Psy chi a trists; 1992. 5. Har ri son G. How many psy chi at ric beds: to wards a needs based port fo lio of res iden tial care. In: Tansella M, ed i tor. Making ra tio nal men tal health ser vices. Epidemiologia e Psichiatria Sociale Mono graph. Sup ple ment 1. Rome: Il Pensiero Scientifico Editore; 1997. 6. Flannigan CB, Glover GR, Feeney ST, Wing JK, Bebbington PE, Lewis SW. Inner Lon don col lab o ra tive au dit of ad mis sions in two health dis tricts. I: in tro duction, meth ods and pre lim i nary find ings. Br J Psy chi a try 1994;165:734–42. 7. Beck A, Croud ace TJ, Singh S, Har ri son G. The Nottingham acute bed study: alter na tives to acute psy chi at ric care. Br J Psy chi a try 1997;170:247–52. 8. Lesage AD, Clerc D, Uribé I, Cournoyer J, Fa bian J, Tourjman V, Van Haaster, Chang L-H. Es ti mating lo cal-area needs for psy chi at ric care: a case study. Br J Psy chi a try 1996;169: 49–57. 9. Allness DJ, Knoedler WH. The Pact model of com mu nity based treat ment for per sons with se vere and per sis tent men tal ill nesses: a man ual for Pact start-up. Arlington (VI): Na tional Al li ance for the Men tally Ill, 1998. 10. RRSSSM. Régie Régionale de la santé et des ser vices sociaux de Montréal-centre. Responsabilité sectorielle des cen tres hospitaliers de la région de Montréal-centre quant à la dis pen sa tion des ser vices psychiatriques aux adultes, révision de 1996. Montréal, 1996. 11. Beecham J, Hal lam A, Knapp M, and Baines B. Costing care in the hos pi tal and in the com mu nity. In: Leff J, ed i tor. Com mu nity care: il lu sion or re al ity? Chichester: John Wiley and Sons; 1997. 12. Lesage AD, Morissette R. Res i den tial and pal lia tive needs of per sons with se vere men tal ill ness who are sub ject to long-term hos pi tal iza tion. Chronic Dis eases in Can ada. Pop u la tion and Pub lic Health Branch: Health Can ada; 1993;Spring:13–8. 13. On tario Min is try of Health. As ser tive Com mu nity Treat ment Team Guide line; 1997. 14. CÉTS. Suivi intensif en équipe dans la communauté pour personnes atteintes de trou bles mentaux graves. Rap port 99-1. Montréal: CÉTS, 1999. xvi -88 p. (ISBN 2-550-34445-6). 15 Mercier C, White D. Men tal health pol icy in Que bec: chal lenges for an in te grated sys tem. New Dir Ment Health Serv 1994;61:41–52. 16. Deci PA, Santos AB, Hiott WB, Schoenwald S, Dias JK. Dis sem i na tion of as sertive com mu nity treat ment pro grams. Psychiatr Serv 1995;46:676–8. 17. Marx AJ, Test MA, Stein LI. Extrahospital man age ment of se vere men tal ill ness. Arch Gen Psy chi a try 1973;29:505–11. 18. King’s Fund. Lon don’s men tal health. In: John son S, Ramsay R, Thornicroft G, Brooks L, Elliott P, Peck E, ed i tors. The re port to the King’s Fund Lon don Commis sion. Lon don: King’s Fund; 1997. 19. Wing JK. The cy cle of plan ning and eval u a tion. In: Wilkinson G, Free man H, edi tors. The pro vi sion of men tal health ser vices in Brit ain. The way ahead. Lon don: Gaskell; 1986.

Manuscript received February 2000, revised October 2001, and accepted November 2001. 1 Re search Pro fes sor, Cen tre de re cher che Fernand-Seguin, Hôpital Louis-H Lafontaine, Unité 218, 7401 Hochelaga, Montréal, Que bec. 2 Na tional Health Scholar, Hôpital Louis-H Lafontaine, af fil i ated with the Uni ver sity of Mon treal, Montréal, Que bec. 3 Guest Scholar, Uni ver sity De part ment of Psy chi a try, Lausanne, Swit zer land. Ad dress for cor re spon dence: Dr AD Lesage, Cen tre de re cher che Fernand-Seguin, Hôpital Louis-H Lafontaine, Unité 218, 7401 Hochelaga, Montréal, PQ H1N 3M5 e-mail: [email protected]

WCan J Psy chia try, Vol 47, No 1, Feb ru ary 2002


Rés umé : Solutions de rechange pour un hôpital psychiatrique de soins actifs de l’Est de Montréal Ob jec tif : Comme la pres sion s’in ten si fie en vue de réduire l e nom bre de lits de soins ac tifs coûteux, les gou ver ne ments et la docu men ta tion pro posent des ra tios à la baisse. Est- ce rai son na ble et juste pour les médecins re sponsa bles qui, à ti tre de prin ci paux fournis seurs de soins, dev ront hos pi taliser des pa tients et éla borer une plani fi ca tion des con gés dans un mi lieu où les lits sont réduits?

Méth ode : Les médecins trai tants de 1) tous les pa tients hos pi tal isés aux soins ac tifs un jour en par ticu lier ( n = 212) et de 2) toutes les nou velles hos pi tali sa tions aux soins ac tifs sur une pé ri ode de 2 se maines (n = 125) ont rem pli une

W Can J Psy chia try, Vol 47, No 1, Feb ru ary 2002

ver sion adaptée du ques tion naire de Not ting ham sur l’u tili sa tion des lits de soins ac tifs (NABUS). Résul tats :Con cer nant la journée en par ticu lier, seule ment 62 des 212 pa tients hos pi tal isés étaient in ap tes à l’hos pi tali sa tion aux soins de re change ou ac tifs. Un ratio plancher de 18 lits de soins ac tifs par 100 000 ha bi tants sem ble adéquat pour l’aire de re cru te ment en ques tion, pourvu que les so lu tions de re change à l’hos pi tali sa tion soi ent pleine ment et ef fi cace ment ac ces si bles. Les so lu tions de re change com por tent es sen tiel le ment la gamme des mi lieux rési den tiels pro té gés suivants : les hôpi taux de jour et les soins in ten sifs à domi cile (de 2 à 6 heures par se maine). Le ra tio des tra vail leurs de soins in ten sifs à

57