494 by qingyunliuliu


									Changes in Services and Structure
in Community Residential Treatment
Facilities for Substance Abuse Patients
Christine Timko, Ph.D.
Michelle Lesar, B.A.
Martha Engelbrekt, B.A.
Rudolf H. Moos, Ph.D.

Objective: This study examined the extent to which community residen-                      cal diagnoses, in which program man-
tial facilities that contract with the Department of Veterans Affairs (VA)                 agers estimated the number of pa-
to treat substance abuse patients are providing more services and struc-                   tients in their caseloads with dual dis-
ture to meet the needs of a client population with increasingly severe                     orders, and in which information on
problems. Methods: A nationwide sample of 249 community residential                        patients’ psychiatric treatment histo-
facilities with VA contracts was surveyed in 1995 and again in 1998. Re-                   ries was obtained from VA databases
sults: In 1998 facilities were more likely than they were in 1995 to have                  (4,5).
psychiatrists and psychologists available to patients as well as special-                     When community residential facili-
ized counseling and psychoeducational, rehabilitation, and medical ser-                    ties have more treatment services and
vices. Facilities also provided more social and recreational activities,                   social activities available to patients,
and more structure was provided by discouraging patients’ choice of in-                    patients tend to participate more in
dividual daily living patterns. In 1998 the facilities were more likely to                 treatment. In turn, greater participa-
admit dual diagnosis patients, those with substance use disorders and                      tion results in better outcomes at dis-
psychiatric disorders. Programs that changed toward accepting dual di-                     charge, especially among more im-
agnosis patients had more services and structure than programs that                        paired substance abuse patients, such
consistently accepted only substance abuse patients. Conclusions: Com-                     as those with comorbid psychiatric di-
munity residential facilities that contract with VA are responding ap-                     agnoses (6–10). Other studies of resi-
propriately to an increasingly ill patient population by providing more                    dential substance abuse facilities (11)
services and structure. (Psychiatric Services 51:494–498, 2000)                            and psychiatric treatment facilities
                                                                                           (12) reported that patients with more
                                                                                           psychiatric impairment functioned

        he past several years have been          In the Department of Veterans Af-         better when the program provided
        a time of dramatic change for         fairs (VA), for example, clients recent-     more structure, such as when they
        substance abuse treatment na-         ly placed in community residential fa-       were taught practical skills in prepa-
tionwide. In particular, community            cilities were more than six times as         ration for release from the program.
residential facilities are assuming a         likely to misuse multiple substances         Taken together, these studies suggest
larger role in the continuum of care          and more than twice as likely to have        that more impaired substance abuse
for substance abuse patients. Many            concomitant psychiatric diagnoses            patients have better treatment out-
patients with more chronic, severe,           than patients in earlier cohorts (4,5).      comes when their treatment program
and complex disorders are now being           The finding of increased severity of         offers more services and structure.
placed in community residential facil-        problems among VA substance abuse               We examined the extent to which
ities rather than in hospital-based in-       patients was consistent in studies in        community residential facilities treat-
patient care (1–3).                           which treating clinicians made clini-        ing substance abuse patients are pro-
                                                                                           viding more services and structure to
                                                                                           meet the needs of a client population
Dr. Timko, Ms. Lesar, and Ms. Engelbrekt are research health science specialists and
                                                                                           with increasingly severe problems.
Dr. Moos is a research career scientist at the Center for Health Care Evaluation of the    Specifically, we surveyed a nation-
Department of Veterans Affairs Palo Alto Health Care System (152-MPD), 795 Willow          wide sample of the same community
Road, Menlo Park, California 94025 (e-mail, timko@odd.stanford.edu). Dr. Timko is also     residential facilities in 1995 and in
consulting associate professor and Dr. Moos is professor in the department of psychiatry   1998. We looked for changes over
and behavioral sciences at Stanford University Medical Center in Palo Alto, California.    time in the programs’ services and
494                                                                                  PSYCHIATRIC SERVICES   o April 2000 Vol. 51 No. 4
structure as well as changes in their         Table 1
organizational characteristics, such as       Organizational characteristics in 1995 and 1998 of 249 community residential fa-
size, patient length of stay, and             cilities serving patients with substance abuse problems
staffing. We determined whether fa-
cilities that began to admit more sub-                                                                                         Statisti-
stance abuse patients with psychiatric        Characteristics                                          1995        1998        cal test1
disorders also had more increases in
                                              Mean N of current patients                                 27.8         31.0       t=2.29∗
services and structure than facilities        Mean N of patients that program can
that consistently limited their clien-          accommodate                                              32.5         36.1       t=2.48∗
tele to patients with only substance          Length of stay
use disorders.                                  Percent of facilities with a maximum stay                58.6         63.5      χ2=1.78
                                                Maximum stay (mean N of months)                           8.4          8.1        t=.73
                                                Average stay (mean N of months)                           4.2          4.2        t=.04
Methods                                       Waiting list
Sample                                          Percent of facilities with a list                        78.3        69.1       χ2=6.82∗∗
To conduct the 1995 survey of com-              Mean N of patients on list                               18.3        16.8        t=1.22
munity residential facilities, Moos             Waiting time for admission (mean N of days)              30.4        25.1        t=2.32∗
and associates (6) obtained a list of all     Per diem charge in 1998 dollars                          $34.49      $45.18        t=4.81∗∗∗
305 community residential facilities          1   For t tests, df=248; for chi square tests, df=1
nationwide that had ongoing con-                  ∗p<.05
tracts to provide services for VA pa-
tients. A total of 299 facility managers
(99 percent) completed the survey.
To conduct the 1998 survey of com-
munity residential facilities, we ob-         Measures                                                 One PASCI subscale was used to
tained an updated list containing 321         Both the 1995 and 1998 surveys con-                   assess program structure. The 19-
such facilities nationwide that cur-          tained portions of the Policies and                   item structured policies subscale as-
rently had VA contracts. A total of 297       Services Characteristics Inventory                    sesses the degree to which the pro-
managers (93 percent) completed the           (PASCI) (13). The PASCI yields a                      gram structures patients’ patterns of
second survey.                                quantitative description of a pro-                    daily living (alpha=.71 in 1995 and
   This paper focuses on 249 commu-           gram’s policies and services so that a                .73 in 1998).
nity residential facilities from which        single program can be compared                           Most PASCI subscale items are
we had completed surveys both in              with normative samples of programs,                   scored dichotomously, with 0 indicat-
1995 and in 1998. They were all com-          and surveys of large numbers of pro-                  ing an absence or lack of the con-
munity facilities that were under con-        grams can be used to examine the im-                  struct assessed, and 1 indicating its
tract with VA to provide residential          pact of program characteristics on                    presence. The item scores are then
care for substance abuse patients who         patients’ outcomes. The PASCI is de-                  summed. To facilitate direct compar-
had received VA inpatient or outpa-           scriptive rather than prescriptive; it                isons of raw scores, percentage scores
tient services. Although these facili-        has not been used to set standards for                are obtained. For example, a score of
ties were under contract with VA,             adequate or superior policies and                     13 on the 26-item health and treat-
they served both veteran and nonvet-          services in residential treatment. In                 ment services subscale is given a per-
eran substance abuse patients.                part, the PASCI obtains information                   centage score of 50.
   The 1995 survey found that 81.9            about the organizational characteris-                    To compare the organizational
percent of patients in the 249 facili-        tics of the treatment facility such as                characteristics, services, and struc-
ties were men, 30.3 percent were be-          facility size, patient length of stay,                ture of the 249 community residential
tween the ages of 18 and 30, 42.5 per-        staffing, wait list features, and per                 facilities between 1995 and 1998,
cent were between the ages of 31 and          diem costs. PASCI scales are inter-                   paired t tests were used for mean
40, 20.6 percent were between the             nally consistent, independent, and                    scores and chi square tests were used
ages of 41 and 50, and 6.6 percent            stable (13).                                          for proportions.
were over age 50. Most of the patients           Two PASCI subscales were used in
were white (on average, 61.5 per-             this study to assess services. The 26-                Results
cent), with smaller proportions of            item subscale on availability of health               Organizational factors,
black patients (28.2 percent) and His-        and treatment services assesses the                   services, and policies
panic patients (6.1 percent). Most pa-        availability of these services in the                 Organizational characteristics. As
tients were divorced (46.7 percent) or        program (Cronbach’s alpha=.83 in                      Table 1 shows, community residential
had never married (37.3 percent). An          1995 and .84 in 1998). The other sub-                 facilities had somewhat more patients
average of 29.8 percent left school           scale, the ten-item subscale on avail-                and more program capacity in 1998
before graduating from high school,           ability of social-recreational services,              than in 1995. Most facilities set a
47.2 percent had completed high               assesses the availability of organized                maximum length of stay in both 1995
school, and 23.0 percent had at least         activities in the program (alpha=.78                  and 1998; the average maximum
some college.                                 in 1995 and .79 in 1998).                             length of stay allowed was about eight
PSYCHIATRIC SERVICES   o April 2000 Vol. 51 No. 4                                                                                      495
Table 2                                                                                             seling, including peer, religious, or
                                                                                                    spiritual counseling and vocational
Specific treatment and health services provided in 1995 and 1998 by 249 com-
                                                                                                    and educational counseling. Further-
munity residential facilities serving patients with substance abuse problems1
                                                                                                    more, in 1998 programs more fre-
                                              1995                1998                              quently provided psychoeducational
                                                                                                    services for both patients and their
                                              N of                N of                              family members as well as rehabilita-
Service                                       facilities %        facilities %        χ2†
                                                                                                    tion services such as training in daily
Counseling or psychotherapy                                                                         living skills, social skills, and stress
  Individual                                 232        93.2      236        94.8       .50         management.
  Group                                      230        92.4      235        94.4       .50            Finally, in 1998 programs made
Self-help groups                                                                                    medical care more available to pa-
  12-step groups                             242        97.2      238        95.6       .50
  Non-12-step groups                         180        72.3      161        64.7      1.90
                                                                                                    tients by providing regularly sched-
Mental health care                                                                                  uled physicians’ and nurses’ hours
  Regularly scheduled psychiatrist                                                                  and having a physician on call. Phar-
     or psychologist hours                    61        24.5      100        40.2    19.25∗∗∗       macotherapy was more commonly
  Psychiatrist or psychologist on call       102        41.0      125        50.2     6.29∗∗        offered in 1998, as was nutrition
Specialized counseling
  Couples and family                         158        63.5      169        67.9     1.69
                                                                                                    counseling. On average, between
  Peer                                       174        69.9      193        77.5     4.32∗         1995 and 1998 the availability of
  Religious or spiritual                     121        48.6      164        65.9    21.25∗∗∗       health and treatment services rose
  Vocational or educational                  172        69.1      214        85.9    24.72∗∗∗       from 59.7 percent to 68.5 percent,
Psychoeducation                                                                                     an increase of 14.7 percent. Seventy-
  For patients                               168        67.5      189        75.9      5.33∗
  For family members                          90        36.3      120        48.0      9.92∗∗
                                                                                                    five percent of the programs provid-
Rehabilitation                                                                                      ed more health and treatment ser-
  Daily living skills training               215        86.3      228        91.6     4.36∗         vices in 1998 than they did in 1995.
  Social skills training                     210        84.3      231        92.8    12.12∗∗∗          Social and recreational ser-
  Stress management training                 118        47.4      201        80.7    62.84∗∗∗       vices. As shown in Table 3, compared
Medical care
  Regularly scheduled physicians’ hours       94        37.8      145        58.2    24.75∗∗∗
                                                                                                    with programs in 1995, programs in
  Regularly scheduled nurses’ hours           53        21.3       87        34.0    17.56∗∗∗       1998 were more likely to offer a num-
  Physician on call                           94        37.8      115        46.2     4.94∗         ber of recreational activities. The ac-
  Pharmacotherapy                             75        30.1      108        43.4     9.94∗∗        tivities included those with an infor-
  Nutrition counseling                       147        59.0      177        71.1    10.01∗∗        mational focus, such as movies, class-
1 Mean±SD scores on the availability of health and treatment services subscale were 59.73±17.70
                                                                                                    es, lectures, and discussion groups,
  in 1995 and 68.51±18.58 in 1998 (t=8.29, df=248, p<.001). Scores ranged between 10 and 93 in      and those with a more purely social
  1995 and between 0 and 100 in 1998.                                                               aim, such as social or coffee hours and
† df=1
                                                                                                    clubs or social groups. In 1998 pro-
 ∗∗p<.01                                                                                            grams were more likely to offer reli-
∗∗∗p<.001                                                                                           gious services on a frequent basis. On
                                                                                                    average, between the two surveys, the
                                                                                                    availability of social and recreational
                                                                                                    services increased from 44.9 percent
months. Patients’ actual average                   tered nurses, aides, and social work-            to 53.7 percent, a rise of 19.6 percent.
length of stay was about four months               ers, with one exception. In 1998 pro-            Fifty-nine percent of the programs
in both 1995 and 1998. In 1998 facili-             grams had more full-time-equivalent              provided more social and recreational
ties were less likely to have a waiting            alcohol or drug addiction counselors             services in 1998 than they did in
list, and patients did not wait as long            than they did in 1995 (4.25 versus               1995.
to be admitted to the program. The                 3.59; t=2.57, df=248, p<.01).                       Structured policies. As Table 4
reduced waiting time may be partly                    Health and treatment services.                indicates, in 1998 programs had be-
due to the larger size of the facilities           As shown in Table 2, in both 1995                come more structured than they
in 1998. Even corrected for inflation              and 1998 most community residen-                 were in 1995, in that a number of pa-
(by converting 1995 dollars to 1998                tial facilities offered individual and           tients’ behaviors were more likely to
dollars using the consumer price in-               group counseling or psychotherapy                be discouraged or considered intol-
dex), the amount patients were                     as well as 12-step and non-12-step               erable. Policies covered the type and
charged for a day of treatment was                 self-help groups. In 1998 programs               placement of furniture in patients’
higher in 1998 than in 1995.                       were more likely to have a psychia-              rooms; televisions, radios, and stere-
   The community residential facili-               trist or a psychologist, or both, avail-         os in patients’ rooms; pets; sleeping
ties did not differ between 1995 and               able to patients, either on an on-call           late; smoking; and leaving the pro-
1998 in the number of full-time-                   basis or by having regularly sched-              gram on evenings and weekends.
equivalent staff members in various                uled hours. Programs also were more              Programs did not change significant-
treatment positions, such as regis-                likely in 1998 to offer specialized coun-        ly between 1995 and 1998 on poli-
496                                                                                           PSYCHIATRIC SERVICES   o April 2000 Vol. 51 No. 4
cies about times for waking up,               Table 3
bathing, and curfews. In 1998 pa-             Specific social and recreational services provided in 1995 and 1998 by 249 com-
tients were required to go to bed             munity residential facilities serving patients with substance abuse problems1
earlier than they were in 1995 (on
average, about 10 p.m. versus 11                                                               1995                  1998
p.m.; t=2.38, df=248, p<.05). On av-          Services offered at least
erage, during the three-year period,          once or twice a week                             N         %           N           %           t†
programs showed an increase of 5.1            Films and movies                                149        59.8        170         68.1        2.03∗
percent in structured policies. Fifty-        Classes and lectures (not part of
eight percent of the programs                   therapy)                                        66       26.4        100         40.2        5.06∗∗∗
demonstrated increased structure.             Religious services                                63       25.5         88         35.4        2.60∗∗
   Correlations. Between 1995 and             Social or coffee hour                             83       33.2        131         52.6        5.53∗∗∗
                                              Club or social group                              37       14.7         54         21.6        3.24∗∗∗
1998 changes in health and treatment          Discussion groups (not part of
services showed a moderate positive             therapy)                                      102        40.9        118         47.5        2.30∗
correlation with changes in social and
                                              1 Mean±SD scores on the availability of social-recreational services subscale were 44.94±24.47 in
recreational services (r=.38, p<.001),
                                                1995 and 53.74±24.82 in 1998 (t=5.06, df=248, p<.001). Scores ranged between 0 and 100 in both
but they were not associated with
changes in program structure.                 † Because items were rated on a 4-point scale (1, very rarely or never; 2, once or twice a month; 3,

Changes in social and recreational              once or twice a week; and 4, three times a week or more), t tests were conducted. To emphasize
services had a modest positive corre-           clinically relevant findings, the percentages of facilities offering services at least once or twice a
                                                week, instead of item means, are presented.
lation with changes in program struc-           ∗ p<.05
ture (r=.16, p<.01).                           ∗∗ p<.01
                                              ∗∗∗ p<.001

Types of patients treated
In 1995 a total of 67 percent of pro-
grams (N=167) limited their admis-            fered more social and recreational                      (57.03 versus 48.04; t=2.19, df=165,
sions to patients with substance use          services (mean subscale score of                        p<.04) but also had more health and
disorders only, whereas 33 percent            46.68 versus 38.63; t=1.98, df=165,                     treatment services (70.49 versus
(N=82) also admitted patients with            p<.05). In 1998, compared with pro-                     61.69; t=2.61, df=165, p<.01) and
substance use and concomitant psy-            grams that accepted only substance                      structured policies (67.11 versus
chiatric disorders. In 1998 a total of        abuse patients, programs accepting                      62.33; t=2.04, df=165, p<.05).
24 percent of programs (N=60) limit-          both substance abuse patients and                         Analysis of change scores—1995
ed admissions to patients who had             dual diagnosis patients not only had                    subscale scores subtracted from the
only substance use disorders, whereas         more social and recreational services                   corresponding scores for 1998—
76 percent (N=189) also treated dual
diagnosis patients. This change repre-
sents a significant shift toward treat-       Table 4
ing patients with both substance              Structured policies in 1995 and 1998 of 249 community residential facilities serv-
abuse and psychiatric problems (χ2=           ing patients with substance abuse problems1
12.40, df=1, p<.001).
   Types of patients, services, and                                                            1995                  1998
structure. Between 1995 and 1998 a
total of 116 community residential fa-        Behavior is discouraged                          N of                  N of
                                              or not tolerated                                 facilities %          facilities %            t†
cilities changed from serving only pa-
tients with substance abuse problems          Having own furniture in room                    187        75.2        206         82.8        2.89∗∗
to also serving patients with sub-            Moving furniture around in room                 113        45.4        130         52.2        2.58∗∗
stance abuse and psychiatric disor-           Keeping a small pet                             211        84.9        225         90.4        4.09∗∗∗
ders. We examined whether these fa-           Skipping breakfast to sleep late                225        90.3        230         92.4        2.18∗
                                              Having a TV in room                             178        71.6        191         76.6        2.62∗∗
cilities offered more services and had        Having a radio or stereo in room                 44        17.7         67         26.8        2.53∗∗
more structure in 1998, compared              Smoking in the program facilities               108        43.3        156         62.5        6.61∗∗∗
with the 51 facilities that served only       Going out in the evenings                        33        13.1         51         20.6        3.13∗∗
substance abuse patients to the exclu-        Spending the weekend away from
sion of dual diagnosis patients in both         the program                                     35       14.1         67         22.7        3.62∗∗∗
1995 and 1998.                                1 Mean±SD scores on the structured policies subscale were 58.50±16.79 in 1995 and 63.56±17.09
   Results of t tests showed that in            in 1998 (t=5.84, df=248, p<.001). Scores ranged between 0 and 95 in 1995 and 0 and 100 in 1998.
1995 these two sets of programs did           † Because items were rated on a 4-point scale (1, encouraged; 2, allowed; 3, discouraged; and 4, in-

not differ in health and treatment ser-         tolerable), t tests were conducted. To emphasize clinically relevant findings, the percentages of fa-
                                                cilities rating behaviors as discouraged or intolerable, instead of item means, are presented.
vices or structured policies. However,          ∗ p<.05
the programs that changed toward ac-           ∗∗ p<.01
                                              ∗∗∗ p<.001
cepting dual diagnosis patients of-
PSYCHIATRIC SERVICES   o April 2000 Vol. 51 No. 4                                                                                                 497
showed that, compared with pro-            toms than those in day treatment.                Severity of substance use disorders among
                                                                                            psychiatric inpatients. Journal of Nervous
grams that accepted only substance         The better outcomes of residential               and Mental Disease 182:164–167, 1994
abuse patients, the programs accept-       clients were attributed in part to the
                                                                                          3. Lehman AF, Myers CP, Thompson J, et al:
ing substance abuse and dual diagno-       program’s having more structure,                  Implications of mental and substance use
sis patients had larger increases in       such as strictly enforcing rules.                 disorders: a comparison of single and dual
health and treatment services (17.6           Results of the 1995 survey support             diagnosis patients. Journal of Nervous and
                                                                                             Mental Disease 181:365–370, 1993
percent versus 13.4 percent; t=1.9,        the hypothesis of Guydish and col-
df=165, p<.05).                            leagues that more program structure            4. Humphreys K, Hamilton EG, Moos RH, et
                                           is related to better patient retention.           al: Policy-relevant program evaluation in a
                                                                                             national substance abuse treatment system.
Discussion and conclusions                 Specifically, in our study more pro-              Journal of Mental Health Administration
The two surveys of a nationwide sam-       gram structure was positively corre-              24:373–385, 1997
ple of 249 community residential fa-       lated with the percentage of patients          5. Humphreys K, Moos RH, Hamilton EG:
cilities showed that programs were         who successfully completed treat-                 Psychiatric services in VA substance abuse
more likely to serve dual diagnosis pa-    ment in 1995 (r=.24, p<.001). (This               treatment programs. Psychiatric Services
                                                                                             47:1203, 1996
tients in 1998 than in 1995 and that in    outcome was not assessed in 1998.)
1998 programs provided more ser-           Thus program structure may be asso-            6. Moos RH, King MJ, Burnett EB, et al:
                                                                                             Community residential program policies,
vices and more structure. Facilities       ciated with better substance abuse                services, and treatment orientations influ-
that changed their admission policies      treatment outcomes and psychosocial               ence patients’ participation in treatment.
to admit dual diagnosis patients were      functioning because it facilitates pa-            Journal of Substance Abuse 9:171–187,
most likely to have a service-intensive    tients’ completion of recommended
treatment program in 1998.                 treatment regimens.                            7. Hitchcock H, Stainbach R, Roque G: Ef-
   Currently, community residential           Our findings suggest that commu-               fects of halfway house placement on reten-
                                                                                             tion of patients in substance abuse after-
facilities are more likely than they       nity facilities that contract with VA             care. American Journal of Drug and Alco-
were previously to have psychiatrists      are adapting appropriately to an in-              hol Abuse 21:379–390, 1995
and psychologists available to patients    creasingly ill resident population by          8. Moos RH, King MJ: Participation in com-
as well as specialized counseling and      providing more services and struc-                munity residential treatment and substance
psychoeducational, rehabilitation, and     ture. These facilities now are more               abuse patients’ outcomes at discharge.
                                                                                             Journal of Substance Abuse Treatment
medical services. The facilities now       likely to offer services of highly                14:71–80, 1997
offer more social and recreational ac-     trained mental health professionals,
                                                                                          9. Rosenheck R, Gallup P: Involvement in
tivities and are more likely to discour-   such as psychiatrists and psycholo-               outreach and residential treatment pro-
age patients from choosing individual      gists, and to offer regularly scheduled           gram for homeless mentally ill veterans.
patterns of daily living. These treat-     physicians’ and nurses’ hours and                 Journal of Nervous and Mental Disease
                                                                                             179:750–754, 1991
ment programs are now somewhat             pharmacotherapy. Probably because
larger, and patients do not have to        of these more expensive services, per        10. Lehman AF, Herron JD, Schwartz RP, et al:
wait as long to be admitted.               diem costs at community residential              Rehabilitation for adults with severe mental
                                                                                            illness and substance use disorders: a clini-
   Although recent studies have not        facilities have increased by more than           cal trial. Journal of Nervous and Mental
focused on increases in services and       30 percent.                                      Disease 181:86–90, 1993
structure in residential substance            Research is needed to examine             11. Conrad KJ, Hultman CI, Pope AR, et al:
abuse care, they have emphasized           cost-effective methods of providing              Case managed residential care for home-
that higher levels of services and         the increased services and structure             less addicted veterans: results of a true ex-
                                                                                            periment. Medical Care 36:40–53, 1998
structure are beneficial for patients.     that impaired substance abuse pa-
For example, Nuttbrock and col-            tients need. Although service- and           12. Timko C, Moos RH: Determinants of the
                                                                                            treatment climate in psychiatric and sub-
leagues (14) found that patients had       structure-intensive programs may be              stance abuse programs: implications for im-
better outcomes in a therapeutic           more expensive in the short term, the            proving patient outcomes. Journal of Ner-
community that provided psychiatric        long-term costs to patients and soci-            vous and Mental Disease 186:96–103, 1998
and psychological care on demand           ety of failing to provide adequate care      13. Timko C: Policies and services in residen-
and had highly structured rules of         may be far greater. o                            tial substance abuse programs: compar-
conduct than they did in community                                                          isons with psychiatric programs. Journal of
                                           Acknowledgments                                  Substance Abuse Treatment 7:43–59, 1995
residences that did not dispense in-
house services and aimed to be less        This research was supported by the           14. Nuttbrock LA, Rahav M, Rivera JJ, et al:
                                           Health Services Research and Develop-            Outcomes of homeless mentally ill chemi-
restrictive. Patients in the therapeutic                                                    cal abusers in community residences and a
                                           ment Service (grant IIR-95-011) and the
community had reduced levels of            Mental Health Strategic Health Group             therapeutic community. Psychiatric Ser-
substance use and psychopathology                                                           vices 49:68–76, 1998
                                           Department of the Veterans Health Ad-
and better functioning.                    ministration.                                15. Guydish J, Wedegar D, Sorensen JL, et al:
   Guydish and associates (15) report-                                                      Drug abuse day treatment: a randomized
                                           References                                       clinical trial comparing day and residential
ed that clients in residential sub-                                                         treatment programs. Journal of Consulting
stance abuse programs were less like-      1. Lamb HR: The new state mental hospitals       and Clinical Psychology 66:280–289, 1998
ly to drop out of treatment and               in the community. Psychiatric Services
                                              48:1307–1310, 1997
showed greater improvement in so-
cial problems and psychiatric symp-        2. Lehman AF, Myers CP, Corty E, et al:

498                                                                               PSYCHIATRIC SERVICES   o April 2000 Vol. 51 No. 4

To top