Recovery From Schizophrenia in Community-Based Psychosocial Rehabilitation Settings Rates and Predictors

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/280037039
Recovery From Schizophrenia in Community-Based Psychosocial

Rehabilitation Settings Rates and Predictors
Article in Research on Social Work Practice · June 2015

DOI: 10.1177/1049731515588597
CITATIONS READS
3 262
5 authors, including:
Caroline S Lim Concepción Barrio

University of California, Los Angeles University of Southern California
15 PUBLICATIONS 103 CITATIONS 55 PUBLICATIONS 1,203 CITATIONS
SEE PROFILE SEE PROFILE
John Brekke
University of Southern California
129 PUBLICATIONS 3,594 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Assisted Outpatient Treatment in Los Angeles County View project
Bridge Peer Health Navigator Intervention View project
All content following this page was uploaded by Caroline S Lim on 25 September 2015.
The user has requested enhancement of the downloaded file.

Original Article
Research on Social Work Practice
1-14
Recovery From Schizophrenia in ª The Author(s) 2015
Reprints and permission:
sagepub.com/journalsPermissions.nav
Community-Based Psychosocial DOI: 10.1177/1049731515588597
rsw.sagepub.com
Rehabilitation Settings: Rates
and Predictors
Caroline Lim1, Concepción Barrio1, Mercedes Hernandez1,

Armando Barragán1, and John S. Brekke1
Abstract
Objective: We assessed the rate of recovery from schizophrenia in community-based psychosocial rehabilitation and
whether psychosocial attributes predicted the achievement of recovery beyond demographic and clinical characteristics.
Methods: We used data from 246 individuals with schizophrenia spectrum disorder collected at baseline and at 6 and
12 months after admission to psychosocial rehabilitation. Results: The proportion of participants who showed recovery during
either 6-month period and the full 1-year period was 19.86% and 7.53%, respectively. Although predictors of recovery for the
1-year period could not be reliably estimated due to its low rate, higher levels of intrinsic motivation and more positive family
relationships at baseline predicted recovery for either 6-month period after controlling for initial functioning capacity.
Conclusions: One in five individuals with schizophrenia who engage in intensive community-based psychosocial rehabilitation
can achieve periods of recovery during treatment. Psychosocial attributes at the start of treatment are important contributors to
subsequent recovery.
Keywords
schizophrenia, recovery, predictors, psychosocial rehabilitation
Schizophrenia, which affects approximately 24 million Literature Review

people worldwide, is considered a serious mental illness
(National Institute of Mental Health, 2009; World Health
Recovery From Schizophrenia
Organization, 2013). The onset of the disorder typically Accumulating evidence from several longitudinal studies of
occurs during the critical developmental period of late ado- schizophrenia, documenting a more heterogeneous and promis-
lescence and emerging adulthood, and significant social and ing long-term course, has challenged the conventional notion
occupational impairment are often the corollary of florid that schizophrenia is a mental disorder characterized by a uni-
psychotic symptoms (American Psychiatric Association, formly poor outcome (Hegarty, Baldessarini, Tohen, Waternaux,
2000). The longitudinal course interspersed with acute exacer- & Oepen, 1994; Jääskeläinen et al., 2012; McGlashan, 1988).
bation and the extent and persistence of psychosocial impair- Accordingly, favorable findings from these long-term outcome
ment between episodes distinguish schizophrenia from studies have prompted research on recovery from schizophre-
other mental disorders, making it one of the most disabling nia (Bellack, 2006). An outgrowth of this research activity
psychiatric disorders (Harrow, Grossman, Jobe, & Herbener, included developing an operational definition of recovery that
2005; McGlashan, 1988). Psychosocial dysfunction as a can be organized into two broad groups: outcome-oriented
concomitant of symptom onset places affected individuals
at risk of socioeconomic disadvantages (Susser, Lin, &
Conover, 1991; Susser, Moore, & Link, 1993). Research 1
School of Social Work, University of Southern California, Los Angeles, CA,
to determine factors that promote recovery and the propor- USA
tion of individuals with schizophrenia who have the poten-
tial for recovery is necessary and beneficial to facilitate the Corresponding Author:
Concepción Barrio, School of Social Work, University of Southern California,
development of interventions or enhancement of existing Montgomery Ross Fisher Building, 669 W. 34th Street, Los Angeles, CA 90089,
treatments to generate more favorable outcomes in individ- USA.
uals with schizophrenia. Email: [email protected]
Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

2 Research on Social Work Practice
research and process-oriented research (Jääskeläinen et al., environmental factors in understanding outcomes in schizo-
2012; Lieberman et al., 2008; Silverstein & Bellack, 2008). phrenia. The model posits that stress is a precipitating factor
Outcome-oriented definitions emphasize recovery as marked that triggers the first psychotic break among individuals with
by abatement of psychotic symptoms and restoration of func- preexisting vulnerability and an important prognostic factor
tioning to premorbid levels (e.g., Liberman, Kopelowicz, that can impinge on the course of schizophrenia by exacerbat-
Ventura, & Gutkind, 2002). Process-oriented definitions ing symptoms (Rosenthal, 1970). Notwithstanding the impor-
describe ways of coping and managing mental health despite tance of stress, the model suggests that favorable environmental
the challenges and limitations caused by the disorder (Anthony, factors such as strong social support and rich personal capital,
1993; Resnick, Fontana, Lehman, & Rosenheck, 2005; Resnick, in the context of reduced biological vulnerability, can provide
Rosenheck, & Lehman, 2004). Another notable difference a buffer against the deleterious effects of stress to prompt a more
between outcome-oriented and process-oriented definitions of favorable prognosis (Mueser & McGurk, 2004). Therefore, in
recovery is that the former were developed for the purpose of our efforts to achieve a deeper understanding of recovery from
facilitating comparative evaluations of treatments. Process- schizophrenia, it is necessary to investigate the contribution of
oriented definitions, however, were geared toward advocating psychosocial factors, particularly in the context of psychosocial
for changes in the mental health system from a paternalistic to rehabilitation in which social work practitioners are charged
a recovery-oriented model of care (Bellack, 2006). However, the with planning for the provision and delivery of therapies that tar-
mental health community has not been able to reach a consensus get intrapersonal resources and the environment.
on how best to define the concept of recovery as it applies to
schizophrenia (Bellack, 2006; Lieberman et al., 2008; Onken,
Craig, Ridway, Ralph, & Cook, 2007; Silverstein & Bellack,
Gaps in Research on Recovery
2008). Indeed, various definitions of recovery have been pro- There are several noteworthy gaps in past research that the cur-
posed (Færden, Nesvåg, & Marder, 2008). Regardless of how rent study sought to address. Most notably, empirical studies
recovery is operationalized, the research agenda for schizophre- have tended to examine the influence of nonmalleable factors
nia should include the identification of malleable factors related (e.g., gender) or factors that are mostly targets of pharma-
to recovery (Liberman, 2002; Silverstein & Bellack, 2008). To cotherapy (e.g., severity of psychopathology), thus causing a
this end, many outcome studies on schizophrenia have applied gap in knowledge about the effects of psychosocial factors on
an outcome-oriented definition of recovery in investigating its recovery from schizophrenia. Due to this emphasis on identify-
rate and predictors. ing demographic and clinical predictors, it is unclear whether
there are other unidentified malleable predictors of recovery
that are capable of being influenced using psychosocial inter-
Predictors of Recovery ventions and informal support systems of family and friends.
Researchers have applied variants of outcome-oriented defini- This gap in knowledge could hamper the development of a col-
tions of recovery to determine predictors of recovery from schi- laborative relationship among mental health clinicians, allied
zophrenia. Findings from studies focusing on biological or health professionals, and caregivers to facilitate better out-
clinical variables have indicated that recovery is predicted by comes for individuals with schizophrenia. Indeed, one of the
female gender, younger age (Lambert et al., 2008; Verma, Sub- contentious issues raised by the consumer movement is that
ramaniam, Abdin, Poon, & Chong, 2012), shorter duration of mental health professionals have often failed to include consu-
untreated psychosis (Petersen et al., 2008), better premorbid mers and family members as partners in the treatment process
adjustment (Bertelsen et al., 2009; Petersen et al., 2008), (Bellack, 2006).
absence of deficit syndrome (Petersen et al., 2008; Strauss, Second, although the influence of psychosocial correlates
Harrow, Grossman, & Rosen, 2010; Verma et al., 2012), has been evaluated qualitatively, few empirical studies have
absence of Schneiderian first-rank symptoms (Rosen, Grossman, quantitatively assessed the impact of these factors on recovery.
Harrow, Bonner-Jackson, & Faull, 2011), milder psychopathol- Findings from qualitative studies of individuals considered to
ogy at presentation (Bertelsen et al., 2009), better cognitive func- be in recovery from severe mental illnesses that included schi-
tion (Kopelowicz, Liberman, Ventura, Zarate, & Mintz, 2005), zophrenia have suggested the importance of personal attributes
higher level of medication adherence (Petersen et al., 2008), bet- and social factors in facilitating recovery. Hope of recovery
ter response to antipsychotic medication during the early course (Ridgway, 2001), self-determination (Mancini, 2007; Ochocka,
of treatment (Lambert et al., 2008; Verma et al., 2012), and Nelson, & Janzen, 2005), spirituality and religion (Barham &
absence of substance use (Petersen et al., 2008). Hayward, 1998), presence of meaningful activities (Dunn,
Wewiorski, & Rogers, 2008; Mancini, Hardiman, & Lawson,
2005), elimination of stigma (Jacobson & Greenley, 2001;
Conceptual Framework Ridgway, 2001), peer support (Mancini, 2007; Schön, Denhov,
The stress–diathesis model forms the basis for understanding & Topor, 2009), resilience (Torgalsbøen & Rund, 2010), strong
the course of schizophrenia and has guided the management clinician–client relationship (Green et al., 2008), and social
of schizophrenia (Rosenthal, 1970). The model considers the support (Cohen, 2005; Davidson, 2003; Jacobson & Greenley,
confluence of predisposing traits, clinical characteristics, and 2001; Mancini et al., 2005; Schön et al., 2009) are among the

Lim et al. 3
important facilitative psychosocial factors of recovery identi- admission. Therefore, the examination of recovery during the
fied in qualitative studies. first year of intensive psychosocial rehabilitation pertained to
Third, several psychosocial factors that have been investi- the period of current treatment and does not imply a first-
gated for their association with more favorable outcomes episode sample.
marked by circumscribed deterioration in function or less The demographic and clinical variables selected were based
severe symptomatology have not been empirically tested for on findings from prior studies reporting a predictive relation-
their predictive relationships with recovery, a more demanding ship with recovery (e.g., Lambert et al., 2008; Rosenthal,
treatment outcome that requires concurrent resolution of psy- 1970). Informed by the stress–diathesis conceptual framework,
chotic symptoms and restoration of functioning. For example, we examined the predictive strength of personal attributes
previous research has demonstrated that better functional defined by intrinsic motivation and self-esteem and social fac-
outcomes in schizophrenia are predicted by higher levels tors defined by social support and family relationships on
of intrinsic motivation (Gard, Fisher, Garrett, Genevsky, & recovery. Severity of symptomatology and functioning at base-
Vinogradov, 2009; Nakagami, Hoe, & Brekke, 2010; Nakagami, line were included as covariates to minimize spurious associa-
Xie, Hoe, & Brekke, 2008; Saperstein, Fiszdon, & Bell, 2011) tions between psychosocial variables and recovery (Johnson,
and self-esteem (Brekke & Long, 2000; Roe, 2003). Despite the 2005). Several salient variables such as duration of untreated
association of intrinsic motivation and self-esteem with better psychosis and treatment characteristics could not be included
functional outcomes, previous research has not investigated in the model due to insufficient or unavailable data. We
whether these factors that are malleable to psychosocial inter- hypothesized that recovery would be predicted by more favor-
ventions are related to recovery from schizophrenia. able psychosocial attributes, namely, higher levels of intrinsic
Finally, there is a dearth of research on recovery in motivation, higher levels of self-esteem, more intact family
community-based psychosocial settings. Previous research has network relations, and higher levels of social support, after
been limited to clinical specialty programs that deliver varying adjusting for salient demographic and clinical characteristics.
intensity of psychosocial rehabilitation interventions. At pres-
ent, it is not clear how many individuals with schizophrenia
who are engaged in community-based psychosocial rehabilita- Method
tion have the potential to achieve periods of recovery. Asses-
sing recovery in such a setting is necessary because between Sample Selection
a quarter and a third of individuals with schizophrenia and Data came from two longitudinal studies conducted in Los
other severe mental illnesses participate in psychosocial Angeles, CA, designed to identify factors related to individual
rehabilitation programs (Hatfield, Gearon, & Coursey, 1996; responsiveness to psychosocial rehabilitation services in the
Lehman & Steinwachs, 1998). Assessing how readily recovery community. Participants (N ¼ 328) were recruited between
can be achieved in community-based psychosocial rehabilita- 1996 and 2005 at admission to one of six community psycho-
tion programs and understanding the specific promoters of social rehabilitation programs operating in the Los Angeles
recovery would be beneficial for social work practitioners in public mental health sector. These programs had been serving
psychosocial settings. individuals with severe mental illness for more than 50 years
through the provision of mental health services (e.g., mental
health treatment; vocational, social, and independent living ser-
Study Aims and Hypotheses vices; substance abuse treatment; and 24-hr crisis response)
The intent of this study was to fill the research gap by assessing aimed at facilitating maximum community adjustment and
recovery from schizophrenia in community-based psychosocial integration and were selected for study involvement because
settings. Specifically, the aims of this study were twofold (a) to data have demonstrated that they yielded improvement in func-
assess the percentage of individuals with schizophrenia spec- tional outcomes among participants. Both studies applied sim-
trum disorder who show periods of recovery, for either any ilar inclusion criteria: (a) age 18–55 years, (b) received a chart
6-month period or the full 1-year period, during the first year diagnosis of psychotic disorder, (c) absence of alcohol or drug
of intensive psychosocial rehabilitation for their current admis- dependence during the previous 6 months, and (d) absence
sion and (b) to investigate the effects of baseline intrapersonal of neurological illness. Participants completed a battery of
attributes (intrinsic motivation and self-esteem) and social fac- research assessments at admission and every 6 months there-
tors (family network relations and social support), hereafter after during a period of up to 2 years. For the purpose of this
referred to as psychosocial predictors, on subsequent recovery, study, analyses were conducted using data collected at baseline
after controlling for the influence of demographic and clinical and 6 and 12 months after admission. The analytic sample fea-
characteristics (gender, educational attainment, length of ill- tured 246 individuals with schizophrenia spectrum disorder:
ness, substance use, medication use, severity of symptomatol- 168 from the first longitudinal study and 78 from the second
ogy, and role and social functioning). This study included study. Participants who were not diagnosed with a schizophre-
mostly non-first-episode participants (subsequently described) nia spectrum disorder (n ¼ 62) or had missing diagnosis
who may have received treatment—pharmacotherapy, psycho- (n ¼ 20) were excluded from the analyses. Participants from
social rehabilitation, or a combination—before their current both longitudinal studies were comparable in demographic and

clinical characteristics except that participants in the latter 6 items from the intrapsychic deficit subscale of the Quality of
longitudinal study presented with more severe symptomatol- Life Scale (Heinrichs, Hanlon, & Carpenter, 1984): sense of
ogy and most had racial minority backgrounds. Both studies purpose, motivation, curiosity, and anhedonia. Sense of pur-
were reviewed and approved by the affiliated university’s insti- pose was defined as the extent to which participants expressed
tutional review board and the Los Angeles County Department realistic and integrated goals for their life, motivation measured
of Mental Health’s Research Committee. the extent to which participants were able to initiate and sustain
goal-directed activities, curiosity reflected the degree to which
participants were interested in their surroundings and ques-
Measurements tioned things they did not understand, and anhedonia measured
the extent to which participants were able to experience
Demographic and psychiatric history form. Demographic and psy-
pleasure and satisfaction. Each item was rated on a 7-point
chiatric history data used for this study were participants’ age,
Likert-type scale, with lower scores reflecting higher levels
gender, race, highest level of education completed, age at onset
of deficit.
of psychiatric illness, and length of illness. These data were
Whereas previous studies have operationalized intrinsic
collected as part of the baseline assessment. Educational attain-
motivation using only 3 items (Nakagami et al., 2008), the item
ment was determined by asking participants to indicate whether
measuring anhedonia was added to better represent the con-
they had completed high school or an equivalent credential
struct’s definition, which considers the ability to experience
(e.g., general equivalency diploma). This variable was dichot-
participatory satisfaction an important characteristic. To deter-
omized, with participants who did not have at least a high
mine whether the 4 items measured one underlying concept,
school credential forming the reference group. Age at onset
principal component factor analyses with varimax and promax
was determined by asking participants how old they were when
rotations were performed. Regardless of the rotational method
they first saw a mental health professional. Length of illness
applied, one factor was extracted, which accounted for 56.3%
was calculated by subtracting the age at onset from the age at
of the variance of the 4 items. Moreover, all items loaded
study entry and represented the period following the first onset
highly on this factor (loadings of .5 or above). The Cronbach’s
of unspecified psychiatric symptoms that were sufficiently dis-
a for these items was .73, demonstrating adequate internal
tressing to prompt participants to seek mental health treatment.
consistency among the items. Additionally, results from confir-
Diagnosis was determined using clinical records, a checklist of
matory factor analysis using maximum likelihood with miss-
symptoms according to the Diagnostic and Statistical Manual
ingness assumed to be at random revealed that the items
fourth edition criteria (American Psychiatric Association, 1994),
loaded strongly and statistically significant at p < .001 on the
and collateral reports from the admitting clinician and on-site
latent factor of intrinsic motivation, yielding a model fit of
psychiatrists.
w2(2, 244) ¼ 2.56, p ¼ .28, root mean square error of approx-
imation ¼ .034, and confirmatory factor index ¼ .998, thus
Community adjustment form. This instrument (Test et al., 1991)
indicating an excellent fit and providing preliminary evidence
was used to assess substance use, medication use, and level of
that the 4 items measured the same concept. Accordingly, rat-
intrinsic motivation.
ings assigned to the 4 items were aggregated and averaged to
Alcohol and substance use. Alcohol and substance use were generate a mean score, ranging from 1 to 7, to indicate each
determined by assessing the frequency of alcohol, marijuana, participant’s level of intrinsic motivation.
and street drug consumption during the previous 6 months
based on participants’ self-report. The variable was dichoto- Index of self-esteem. This 25-item index (Hudson, 1982) was
mized as 0 (no alcohol or substance use during the previous used to evaluate participants’ level of self-esteem. The items
6 months) and 1 (use of alcohol or substances during the pre- measure participants’ subjective evaluation of themselves and
vious 6 months). others’ perceptions of them. Responses to the items ranged
from 1 (rarely or none of the time) to 5 (most or all of the time).
Medication use. Medication use was defined as days of use of
Reverse coding was performed on 12 items as specified by
any antipsychotic medications during the previous 180 days.
Hudson (1982). Responses to all items were summed to derive
Given the bimodal distribution, with predominant response
an aggregate score for the scale, with higher scores indicating
categories of full use (180 days) and nonuse (0 days), the
greater severity of self-esteem problems. Cronbach’s a for this
variable was dichotomized as low use and regular use. In
scale was .90.
accordance with the criterion proposed by Kane, Leucht,
Carpenter, and Docherty (2003), participants who reported not
Role Functioning Scale (RFS). The RFS (Goodman, Sewell,
being on any antipsychotic medication for more than 50% of
Cooley, & Leavitt, 1993) was used to cross-sectionally evalu-
the time during the previous 6 months were coded as low use.
ate the quantity and quality of each participant’s work produc-
Intrinsic motivation. Intrinsic motivation refers to the desire to tivity, independent living and self-care skills, immediate social
engage in certain behaviors due to curiosity, interest, enjoy- network relationships, and family network relationships. Each
ment, and satisfaction derived from participation (Deci & item was rated on a 7-point Likert-type scale ranging from 1
Ryan, 1985; Ryan & Deci, 2000). It was measured by 4 of the (severely limited) to 7 (no evidence of deficit). The a coefficient

Lim et al. 5
for the 4 items was .50, indicating that the items were not measuring social contacts, and (d) receive no psychiatric hospi-
closely related and hence should be analyzed separately to talization defined by a score of 4 on the S-CS item measuring
determine their predictive relationship with recovery. For the duration of hospitalization for psychiatric treatment. Two lev-
purpose of this study, work productivity, social network rela- els of recovery were examined: meeting the aforementioned
tionships, and family network relationships were examined. criteria for either 6-month period (less demanding criteria) or
Work productivity is a measure of role functioning. In essence, the full 1-year period (more stringent criteria) during the first
the variable measures the ability to function in school, indepen- year of follow-up.
dent employment, or at home. Immediate social network is an The Remission in Schizophrenia Working Group proposed
evaluation of social functioning, essentially the ability to that symptomatic remission is achieved if the symptomatology
develop and maintain close relationships with friends. Family unique to schizophrenia, namely, delusions, hallucinations, dis-
network relationships indicate the extent to which individuals organized speech, grossly disorganized or catatonic behavior,
are able to maintain positive relationships with family members and negative symptoms, are of minimal intensity for at least
as characterized by consistent contacts and reciprocity. 6 months (Andreasen et al., 2005). To this end, selected BPRS
items that correspond to key features of schizophrenia were
Social support. An adapted version of the self-rated Medical used to determine whether participants achieved symptomatic
Outcome Study Social Support Survey (Sherbourne & Stewart, remission: grandiosity, suspiciousness, unusual thought con-
1991) was used to assess perceived social support. The 18-item tent, hallucinatory behavior, conceptual disorganization, man-
survey assesses level of social support in the dimensions of nerisms or posturing, blunted affect, emotional withdrawal,
emotional and informational support, tangible support, affec- and motor retardation. Participants who received BPRS ratings
tionate support, and positive social interaction. Items in each of mild or less ( 3) for these items and were not hospitalized
domain were rated from 1 (none of the time) to 5 (all of the during the previous 6 months were classified as in remission.
time). The ratings were aggregated and divided by the number Given the constraints of administering a structured BPRS
of items completed to achieve an index of each participant’s assessment schedule every 2 weeks during a 6-month period,
level of social support ranging from 1 to 5. Cronbach’s a for the length of psychiatric hospitalization was used as an indirect
the scale was .95. indicator of the stability of symptom severity to be consistent
with extant studies’ methodology (Eberhard, Levander, &
Brief Psychiatric Rating Scale (BPRS). The 24-item BPRS (Lukoff, Lindström, 2009; Novick, Haron, Suarez, Vieta, & Naber, 2009;
Nuechterlein, & Ventura, 1986) was administered once every 6 Strauss et al., 2010).
months to assess the presence and severity of psychiatric symp-
toms during the prior 2 weeks. Every item was rated on a
Likert-type scale ranging from 1 (not present) to 7 (extremely
Statistical Analysis Plan
severe) on the basis of each participant’s self-report and beha- Descriptive statistics and rate of recovery for the sample were
vior during interview. derived using univariate analyses. Due to the study’s modest
sample size, which precluded the inclusion of all proposed
Strauss and Carpenter Outcome Scale (S-CS). The 3-item S-CS baseline explanatory variables in the multivariate model for
(Strauss & Carpenter, 1972) was used to evaluate outcomes prediction of recovery, the selection of variables was guided
of schizophrenia during the previous 6 months in terms of dura- by the purposeful method proposed by Hosmer, Lemeshow,
tion of hospitalization for psychiatric treatment, frequency of and Sturdivant (2013b). First, Pearson chi-square tests for cate-
social contacts, and engagement in useful employment, includ- gorical explanatory variables or two-sample t-tests for contin-
ing work as a homemaker or student. Each item was rated uous explanatory variables were performed to investigate the
on a 5-point Likert-type scale. Higher scores indicated better predictive relationship between the investigated baseline vari-
outcomes. ables and recovery during the first year of follow-up (see
Figure 1). For categorical factors with few observations, Fish-
er’s exact test was performed instead of the regular chi-square
Operational Definition of Recovery test. Explanatory variables that were statistically significantly
We followed the precedent of Harrow, Grossman, Jobe, and associated with recovery at p < .10 in the bivariate analyses
Herbener (2005), Strauss, Harrow, Grossman, and Rosen were candidates for entry in the multivariate model. Second,
(2010), and Ventura et al. (2011) in operationalizing recovery a direct binary logistic regression analysis was performed to
to enable comparison. Recovery was defined as simultaneously identify statistically significant psychosocial predictors of
fulfilling the following set of criteria for a specified duration: recovery after adjusting for baseline demographic and clinical
(a) be in symptomatic remission per the operational definition characteristics. Explanatory variables that were not statistically
proposed by the Remission in Schizophrenia Working Group significant in the multivariable model were removed. Third,
(Andreasen et al., 2005), (b) demonstrate adequate work func- explanatory variables that were not statistically significant at
tioning defined by a score of 2 or higher on the S-CS item mea- the bivariate level were added to the multivariate model sepa-
suring useful employment, (c) demonstrate adequate social rately to identify variables that were statistically significantly
functioning defined by a score of 2 or higher on the S-CS item related to recovery in the presence of other variables. Variables

Table 1. Demographic and Clinical Descriptors of Sample at Baseline

Symptomatic
Remission
(N ¼ 246).
(n = 52)
Variable M (SD) n (%)
Demographic characteristics
21
Age 37.93 (9.35)
Male 161 (66.80)
14 6
Ethnicity
11 African American 101 (42.80)
Adequate Social Adequate Work European American 84 (35.59)
Functioning 20 11 Functioning Latino 30 (12.71)
6
(n = 51) (n = 34) Asian 10 (4.24)
Other ethnicities 11 (4.66)
Marital statusa
Married or living together 4 (1.65)
Figure 1. Rate of recovery for the 1-year period (n ¼ 146). Single 176 (72.43)
Divorced or separated 63 (25.93)
that were not statistically significant were removed, resulting in Completed high school 153 (65.38)
Clinical characteristics
the final model. The purposeful selection method has been Length of illness (years) 14.43 (10.14)
shown to be superior to stepwise selection in retaining sta- Medication use
tistically significant explanatory variables and confounders Low use 57 (23.65)
(Bursac, Gauss, Williams, & Hosmer, 2008). Regular use 184 (76.35)
Diagnostic statistics were calculated to determine whether Alcohol and substance use 43 (17.70)
the final multivariate model satisfied the assumptions of logis- Severity of symptomatologyb 40.31 (11.34)
tic regression. Area under the receiver–operating characteristic Role functioningc 2.00 (1.60)
Social functioningc 3.06 (1.66)
curve was estimated for the final model as a measure of the
a
short listed predictors’ ability to discriminate between partici- Total does not sum to full sample or 100% due to missing data.
b
pants who did or did not show periods of recovery. Lowess Measured with the Brief Psychiatric Rating Scale. Higher scores indicate more
severe symptomatology.
smooth and fractional polynomials methods were used to check c
Measured with the Role Functioning Scale. Higher scores indicate higher levels
the assumption that the logit of the outcome variable (recovery of impairment.
outcome) was a linear combination of the explanatory variables
(Hosmer, Lemeshow, & Sturdivant, 2013a). To test for model
quarters (76.35%) of participants were on antipsychotic medi-
specification errors, a link test was performed. To check for
cation more than 50% of the time during the previous 6 months
multicollinearity, tolerance levels and variance inflation factors
before enrollment in psychosocial rehabilitation. Moreover,
were generated. Tolerance values less than .10 and variance
a majority (82.30%) did not use any substances during the
inflation factor values greater than 10 suggest the presence of
6 months preceding admission. Notwithstanding the high rate
redundant predictors. Preselected two-tailed directional tests
of medication use, the lower baseline RFS scores indicate that
and statistical significance were determined at a ¼ .05. Analy-
the sample featured participants who presented with marked
ses were performed with STATA 13.0 using data from partici-
functional deficits (M ¼ 2.00, SD ¼ 1.60) and impaired social
pants who completed all three assessments during the 1-year
function (M ¼ 3.06, SD ¼ 1.66).
study period and had complete information for all variables
included in the analytic model.
Attrition Rate
Results Outcome data with respect to recovery status during the 1-year
study period were obtained for 59.35% (n ¼ 146) of the 246
Sample Characteristics participants. The rate of attrition in the sample was 22.36%
Table 1 displays the demographic and clinical profile of the (n ¼ 55) at 6 months and 32.93% (n ¼ 81) by 12 months.
sample (N ¼ 246). Of note, the study had a racially diverse Approximately one fifth (n ¼ 52, 21.13%) of the sample missed
sample with majorities of participants between the age of 30 at least one follow-up assessment and 17.07% (n ¼ 42) missed
and 50 (66.67%), men (66.80%), singles (72.43%), and high both 6- and 12-month assessments. Six participants (2.44%)
school graduates or an equivalent credential (65.38%). Regard- had incomplete outcome data, which precluded the assessment
ing the sample’s clinical characteristics, the average and med- of their recovery status.
ian length of illness was 14.43 years (SD ¼ 10.14) and 12.50 Table 2 reveals that participants who missed at least one
years, respectively, suggesting that most participants were not follow-up assessment did not differ from participants who
experiencing their first episode. On average, the participants completed all follow-up assessments in terms of most baseline
presented with very mild psychiatric symptoms as assessed demographic and clinical characteristics. However, partici-
by the BPRS (M ¼ 40.31, SD ¼ 11.02). More than three pants who missed at least one follow-up assessment had

Lim et al. 7
Table 2. Baseline Characteristics of Participants Who Completed All or Missed at Least One Follow-Up Assessment.
Variable Completers (n ¼ 152) Noncompleters (n ¼ 94) Test Statistic p
Agea 38.24 (0.78) 37.43 (0.92) t(241) ¼ 0.14 .52
Male 100 (66.67) 61 (67.03) w2(1, 241) ¼ 0.00 .95
Ethnicityb .59
African American 63 (41.45) 38 (45.24)
European American 57 (37.50) 27 (32.14)
Latino 21 (13.82) 9 (10.71)
Asian 6 (3.95) 4 (4.76)
Other ethnicities 5 (3.29) 6 (7.14)
Marital statusb .91
Married or living together 3 (1.97) 1 (1.10)
Single 111 (73.03) 65 (71.43)
Divorced or separated 38 (25.00) 25 (27.47)
Completed high school 104 (70.27) 49 (56.98) w2(1, 234) ¼ 4.25 .04
Length of illness (years)a 14.27 (0.85) 14.73 (1.11) t(224) ¼ 0.14 .74
Medication use .63
Low use 37 (24.67) 20 (21.98) w2(1, 241) ¼ 0.23
Regular use 113 (75.33) 71 (78.02) w2(1, 243) ¼ 0.43
Alcohol and substance use 25 (16.45) 18 (19.78) t(236) ¼ 1.91 .51
Severity of symptomatologya,c 39.22 (0.91) 42.10 (1.22) w2(1, 242) ¼ 1.96 .06
Role functioninga,d 2.15 (0.14) 1.74 (0.13) w2(1, 242) ¼ 0.90 .05
Social functioninga,d 3.13 (0.13) 2.93 (0.17) .37
Note. Values represent n (%) unless otherwise noted.
a
Values represent M (SD).
b
Fisher’s exact test was used to determine if groups differed at statistically significant levels.
c
Measured with the Brief Psychiatric Rating Scale. Higher scores indicate more severe symptomatology.
d
Measured with the Role Functioning Scale. Higher scores indicate higher levels of impairment.
statistically significantly lower educational attainment (56.98% maintain symptomatic remission and adequate work functioning
vs. 70.27% completed high school, respectively), w2(1, 234) ¼ (n ¼ 17, 11.64%) or adequate work and social functioning (n ¼
4.25, p < .05, and higher levels of functional impairment at 17, 11.64%). Achieving an adequate level of work functioning
study entry (RFS work productivity score of 1.74 vs. 2.15, was an impediment to meeting the recovery criteria for the major-
respectively), w2(1, 242) ¼ 1.96, p ¼ .05. Moreover, there were ity of participants. More than half (n ¼ 77, 52.74%) did not parti-
marginally statistically significant differences in the overall cipate in any useful work during the 1-year follow-up period.
severity of symptomatology (BPRS score of 42.10 vs. 39.22, When the less demanding recovery criteria were considered,
respectively), t(236) ¼ 1.91, p ¼ .06. These findings indicate which required the fulfillment of the aforementioned outcome
that participants with possibly lower premorbid role function- criteria for either 6-month period rather than for the full 1-year
ing and who were more ill at study entry were more likely to period during the first year of follow-up, the rate of recovery
miss a follow-up assessment. nearly tripled (n ¼ 29, 19.86%). In addition to the participants
who met criteria for recovery for 1-year period, another 12.33%
Rate of Recovery (n ¼ 18) of participants met criteria for recovery during either
6-month period. Eight (5.48%) participants showed recovery
Figure 1 shows the proportion of participants who achieved the during the first half of the 1-year follow-up period, but subse-
outcomes germane to fulfilling the recovery criteria for 1-year quently lost their recovery status because they did not maintain
period. Applying the more stringent recovery criteria, which an adequate level of work or social functioning or both.
required participants to be in symptomatic remission and con- Regarding the 10 (6.85%) participants who achieved recovery
currently demonstrate adequate work and social functioning for during the second half of the follow-up period, inadequate
1 year, yielded a rate of 7.53% (n ¼ 11). When considering the symptomatic remission or work functioning delayed their
entire 1-year follow-up period, 35.62% (n ¼ 52) of the sample achievement of recovery during the first 6 months.
sustained symptomatic remission, 34.93% (n ¼ 51) had ade-
quate social functioning, and 23.29% (n ¼ 34) had adequate
work functioning. More than one sixth (n ¼ 25, 17.12%) of par-
Group Differences in Baseline Investigated Factors
ticipants simultaneously maintained symptomatic remission The low rate of participants who met recovery criteria for the
and adequate social functioning, however, a considerable 1-year period relative to the number of participants who did not
smaller proportion of participants were able to concurrently experience recovery limited the number of predictors that could

Table 3. Results of Bivariate Analyses Relating Baseline Demographic, Clinical, and Psychosocial Variables to Recovery.a
Variable Not in Recovery (n ¼ 117) In Recoveryb (n ¼ 29) Test Statistic P
Male 77 (66.96) 19 (65.52) w2(1, 144) ¼ 0.02 .88
Completed high school 76 (67.26) 23 (79.31) w2(1, 142) ¼ 1.59 .21
Length of illness (years)c 15.05 (10.33) 12.03 (10.07) t(137) ¼ 1.40 .16
Medication used .64
Low use 30 (26.09) 6 (20.69)
Higher use 85 (73.91) 23 (79.31)
Alcohol and substance use 16 (13.68) 6 (20.69) w2(1, 146) ¼ 0.89 .35
Severity of symptomatologyc,e 40.15 (11.01) 35.41 (10.68) t(141) ¼ 2.08 .04
Role functioningc,f 1.86 (1.49) 3.45 (2.26) t(144) ¼ 4.59 <.01
Social functioningc,f 2.84 (1.51) 4.31 (1.66) t(144) ¼ 4.28 <.01
Psychosocial characteristics
Intrinsic motivationc,g 3.82 (1.23) 4.95 (0.99) t(144) ¼ 4.62 <.01
Self-esteemc,g 65.75 (19.40) 63.40 (19.50) t(144) ¼ 0.59 .56
Family network relationshipsc 3.68 (1.96) 4.75 (1.58) t(144) ¼ 2.70 <.01
Social supportc,h 3.24 (1.06) 3.38 (0.97) t(144) ¼ 0.65 .52
Note. Values represent n (%) unless otherwise noted.
a
n ¼ 146.
b
Recovery for either 6-month period during the 1st year of follow-up.
c
Values represent M (SD).
d
Fisher’s exact test was used to determine if groups differed at statistically significant levels.
e
Measured with the Brief Psychiatric Rating Scale. Higher scores indicate more severe symptomatology.
f
Measured with the Role Functioning Scale. Higher scores indicate higher levels of impairment.
g
Higher scores indicate higher levels of intrinsic motivation.
h
Higher scores indicate higher levels of social support.
be estimated precisely in a multivariable logistic regression Predictors of Recovery

model (Argresti, 2007). Therefore, analyses of bivariate and
The bivariate analyses identified the following baseline variables
multivariate relationships were performed with recovery for
as candidates for the multivariable logistic regression model:
either 6-month period as the outcome variable, which was more
severity of symptomatology, role functioning, social functioning,
prevalent in this sample. Additional analyses revealed that par-
intrinsic motivation, and family network relationships. Severity
ticipants who showed recovery for the 1-year period and parti-
of symptomatology did not retain its statistical significance in the
cipants who showed recovery for either 6-month period did not
fitted multivariable model at p < .05 (results not shown) and was
differ at statistically significant levels with regard to all inves-
removed. When the explanatory variables that were not statisti-
tigated baseline characteristics.
cally significant at the bivariate level were entered individually
Table 3 displays the results of the bivariate analyses com-
into the multivariate model, the coefficient for substance use was
paring participants who showed recovery for 6 months and par-
on the border of statistical significance (p ¼ .08) and the remain-
ticipants who were not in recovery during the 1-year follow-up
ing variables remained statistically nonsignificant.
period, on the investigated baseline characteristics. Compared
The results of the final fitted model of recovery on intrinsic
to participants who were not in recovery during the 1-year
motivation and family network relationships after adjusting for
follow-up period, participants who showed 6-month periods
role and social functioning are shown in Table 4. For every unit
of recovery presented with statistically significantly less severe
increase in role functioning, the odds of achieving recovery
symptomatology (40.15 vs. 35.41, respectively) t(141) ¼ 2.08,
increased by 1.34, and for every unit increase in social func-
p < .05, less functional impairment (1.86 vs. 3.45, respec-
tioning, the odds of achieving recovery increased by 1.40. After
tively), t(144) ¼ 4.59, p < .01, better social functioning
controlling for baseline functioning, intrinsic motivation and
(2.84 vs. 4.31, respectively), t(144) ¼ 4.28, p < .01, higher
family network relationships at baseline reliably predicted
levels of intrinsic motivation (3.82 vs. 4.95, respectively),
the achievement of recovery for either 6-month period during
t(144) ¼ 4.62, p < .01, and more intact family network rela-
the follow-up period. For every unit increase in intrinsic moti-
tionships (3.68 vs. 4.75, respectively), t(144) ¼ 2.70, p < .01.
vation, the odds of recovery was estimated to increase by 1.68
Moreover, participants who showed periods of recovery
(95% confidence interval [CI] ¼ [1.01, 2.79]) for a given level
appeared more likely to have completed high school (67.26%
of role and social functioning. Likewise, for every unit increase
vs. 79.31%, respectively), w2(1, 142) ¼ 1.59, p ¼ .21, and
in family network relationships, the odds of recovery were esti-
tended to have shorter lengths of illness (15.05 vs. 12.03,
mated to increase by 1.33 (95% CI ¼ [1.01, 1.74]) after control-
respectively), t(137) ¼ 1.40, p ¼ .16, although the differences
ling for role and social functioning.
were not statistically significant.

Lim et al. 9
Table 4. Results of Fitting the Final Multivariable Model.a extant findings. Longitudinal studies that have applied the
same set of criteria to examine the rate of recovery in individ-
Baseline Characteristics Coeff. SE OR 95% CI p
uals with schizophrenia during the early stages of treatment
Intrinsic motivation 0.52 0.26 1.68 [1.01, 2.79] .047 have noted rates between 1% and 18% (Harrow et al., 2005;
Family network relationships 0.28 0.14 1.32 [1.01, 1.74] .044 Strauss et al., 2010; Ventura et al., 2011). Comparing the rate
Role functioning 0.29 0.13 1.34 [1.03, 1.74] .030 of recovery that lasted for 1 year across studies suggests that
Social functioning 0.33 0.15 1.40 [1.05, 1.86] .023 individuals with less severe forms of schizophrenia (i.e., schi-
Note. CI ¼ confidence interval; Coeff. ¼ estimated unstandardized coefficient; zophreniform disorder and absence of negative symptoms)
OR ¼ odds ratio; SE ¼ standard error. tended to have better prognoses. Nonetheless, the lower rates
a
n ¼ 146. of recovery that lasted for 1 year in the present study and the
aforementioned studies suggest that achievement of uninter-
As a set, the predictors reliably distinguished between parti- rupted remission for 1 year is demanding during the early
cipants’ recovery outcomes during the first year of follow-up, stages of treatment. As noted previously, achieving an adequate
w2(1, 146) ¼ 37.94, p < 0.01. The set of predictors had a sensi- level of work functioning was an impediment to meeting the
tivity value of 34.48%, a specificity value of 94.87%, a positive more stringent recovery criteria. Harrow et al. (2005) and
predictive value of 62.50%, and a negative predictive value of Strauss et al. (2010) found that rate of recovery tended to
85.38%, for an overall classification rate of 82.88%. The model increase over time, suggesting that recovery might be a more
yielded a receiver–operating characteristic curve of .86, indi- distal outcome for the majority of individuals with schizophre-
cating excellent discrimination of outcomes (Hosmer et al., nia, given that functional restoration is a more laborious task.
2013a). These results indicate that the model fit the data by Nonetheless, the finding that more than 1 in 20 participants
adequately reflecting the observed outcome in the sample. in this sample achieved recovery that lasted for 1 year chal-
Lowess smooth and fractional polynomial tests found no statis- lenges the long held view that schizophrenia is characterized
tical evidence of nonlinearity in the logit for the three continu- by a uniformly poor outcome. In this study, most participants
ous explanatory variables in the model. The link test was not had experienced a more persistent course of schizophrenia and
significant (p ¼ .13), confirming that the multivariate model the rate of recovery for 1-year period in this study is not lower
was properly specified, thereby indicating that the model than the rates found in first-episode samples, suggesting that
included meaningful predictors and excluded any irrelevant schizophrenia is not necessarily characterized by a deteriorat-
variables. Both measures of tolerance and variance inflation ing course.
factors confirmed the absence of multicollinearity in the esti- Our finding regarding the higher rate of recovery for either
mated model. Taken together, these findings demonstrate that 6-month period compared with recovery for the 1-year period is
the assumptions of multivariate logistic regression were not consistent with the findings by Ventura et al. (2011) who docu-
violated. mented a rate of 10% when they applied the less stringent
recovery criteria to their sample of individuals with first-
episode schizophrenia. The marginal discrepancy in the rate
Discussion of recovery for the 6 month-period between the present study
In the present study, the proportion of participants who showed and the study by Ventura et al. (2011) may be attributed to
periods of recovery for either 6-month period or the full 1-year differences in research and therapeutic methodology. First,
period was 19.86% and 7.53%, respectively. The hypotheses Ventura et al. (2011) conducted symptom assessments biweekly,
that higher levels of positive psychosocial attributes at baseline whereas we completed the assessments once every 6 months
would increase the likelihood of recovery during the follow-up and confined the assessment period to the previous 2 weeks.
period were partially supported. Although predictors of recov- Due to our less rigorous assessment schedule, the rate of recov-
ery that lasted for 1 year could not be reliably estimated due to ery may be inflated in the present study. Second, participants in
its low rate in this sample, recovery that lasted for either 6- Ventura et al.’s study were not enrolled in specific psychoso-
month period during the follow-up period was predicted by cial interventions, which may have hindered the restoration
higher levels of intrinsic motivation and more positive family of adequate functioning in the sample. Indeed, Ventura et al.
relationships at baseline after controlling for initial role and (2011) noted that achieving adequate role functioning was an
social functioning. Collectively, these findings indicate that impediment to meeting the recovery criteria in their first-
in this sample of individuals with schizophrenia spectrum dis- episode sample.
order who were engaged in community-based psychosocial There has been little research relating psychosocial attributes
rehabilitation, about 20% achieved periods of recovery that to recovery in schizophrenia, with the exception of several stud-
lasted for 6 months. Furthermore, our findings suggest that psy- ies (Albert et al., 2011; Bertelsen et al., 2009; Harrow, Hans-
chosocial attributes at the start of treatment are important con- ford, & Astrachan-Fletcher, 2009). A central finding from
tributors to subsequent recovery. this study indicates that for a given level of functioning at base-
Our finding regarding the lower occurrence of recovery that line, recovery for either 6-month period during the first year of
lasted for 1 year during the first year of follow-up after enroll- follow-up was predicted by higher levels of intrinsic motivation
ment in psychosocial rehabilitation (7.53%) is consistent with and more positive family network relationships at enrollment in

psychosocial rehabilitation. Because individuals who are intrin- for our statistically nonsignificant findings. First, the data had
sically motivated are driven by curiosity, interest, derived limited variation in values of medication use and substance
enjoyment, and satisfaction, it is possible that higher levels of use because the majority of participants reported high medi-
intrinsic motivation prompt recovery in individuals with schizo- cation use and no substance use at baseline. The relative
phrenia through better engagement in both pharmacotherapy homogeneity in these explanatory variables could have lim-
and psychosocial rehabilitation interventions (Medalia & ited the detection of significant effects on recovery. Second,
Brekke, 2010), thereby increasing the likelihood of sympto- length of illness is conceptually distinct from the duration of
matic remission and functional restoration, respectively. Con- untreated psychosis. The former considers periods of active
cerning the influence of family relationships on recovery, the psychosis and remission. In contrast, the latter is a measure
evidence that emerged from this study suggests that having bet- of the length of active illness. It is possible that longer peri-
ter family network relationships at enrollment in psychosocial ods of active psychosis, rather than the aggregated length of
rehabilitation is statistically significantly predictive of both illness, impinge on the likelihood of recovery. Third, these
symptomatic remission and adequate functioning sufficient to discrepancies may suggest that beyond the first psychotic
generate periods of recovery, thus expanding previous research episode, psychosocial attributes may play a more significant
on the association between familial relations and clinical or role in recovery relative to clinical factors, however, this is a
functional outcomes in schizophrenia (Cechnicki, Bielańska, hypothesis that has yet to be empirically tested.
Hanuszkiewicz, & Daren, 2013; Guada, Hoe, Floyd, Barnour,
& Brekke, 2012). More positive familial relationships may be
indicative of low levels of expressed emotion in the family. Limitations
Maintaining more positive ties with family members may also Several limitations to this research must be acknowledged.
afford participants more opportunities to establish meaningful First, intrinsic motivation and family network relationships
connections and increase the likelihood of finding work and predicted shorter periods of recovery that lasted at least
age-appropriate roles through referrals from individuals in 6 months but less than 1 year. It is unclear whether these
affiliated social networks. psychosocial attributes would effect longer periods of recov-
In addition to intrinsic motivation and family relationships, ery. Second, intrinsic motivation and negative symptoms share
we found that higher levels of social and role functioning at conceptual similarities. We sought to determine their empirical
baseline were statistically significant predictors of recovery independence by conducting principal component factor anal-
at follow-up. Although this finding suggests that participants ysis with varimax rotation on 4 items from the Quality of Life
who had less functional impairment at admission to psychoso- Scale that measured intrinsic motivation and 3 items from the
cial rehabilitation were more likely to achieve periods of recov- BPRS that measured negative symptoms. The 7 items were
ery, it also underscores the importance of intervening more empirically summarized by two factors that accounted for
intensively during the early phase of psychosocial rehabilita- 62.5% of the variance. The first factor included the 4 items
tion with individuals who present with marked psychosocial measuring intrinsic motivation, with each item loading highly
impairments to enhance their chances of recovery. on the factor (loadings above .4). The second factor included
Contrary to expectations, social support and self-esteem did the 3 BPRS items measuring negative symptoms, which loaded
not predict recovery, although the data revealed a trend toward highly on the factor (loadings above .7). Similar results were
recovery with respect to these factors. It is possible that social derived when we applied the oblique rotational method. These
support influences certain domains of functional outcomes but findings suggest that intrinsic motivation and negative symp-
does not directly influence the remission of psychotic symptoms are empirically independent of each other, consistent with
toms. When outcomes germane to the recovery criteria mea- the findings from previous studies (Nakagami et al., 2008;
sured at the 6-month follow-up were individually regressed Saperstein et al., 2011). If the items used to measure intrinsic
on level of baseline social support, social support was only sig- motivation were in fact measuring negative symptoms, our
nificantly associated with social functioning, suggesting that it study findings suggest that negative symptoms typified by
may not be a predictor of symptomatic remission or improved motivation are crucial targets of pharmacotherapy to increase
role functioning. Concerning self-esteem, it is possible that the odds of recovery. Third, the study was not sufficiently pow-
recovery engenders higher levels of self-esteem but not vice ered to detect the effects of other investigated explanatory vari-
versa. Indeed, previous studies have found support for the ables on recovery in the multivariate model. Peduzzi, Concato,
cross-sectional association between better functional outcomes Kemper, Holford, and Feinstein (1996) proposed that there
and improvement in self-esteem (Davis, Kurzban, & Brekke, should be 10 outcomes of each type for every predictor in the
2012). multivariate model. Because 29 participants in this sample
It is noteworthy that most of the investigated clinical vari- showed periods of recovery, the multivariate logistic regression
ables—length of illness, medication use, and substance use— model should not include more than three explanatory vari-
did not predict recovery in this sample, although converse ables. Because an insufficiently powered study is more prone
effects were documented in studies of first-episode samples to committing a Type II error, findings regarding the statisti-
(Albert et al., 2011; Bertelsen et al., 2009; Petersen et al., 2008; cally nonsignificant effects of the investigated demographic
Verma et al., 2012). There are several possible explanations and clinical characteristics should be interpreted conservatively.

Lim et al. 11
Fourth, as previously noted, the use of a less rigorous symptom by the social environment. Therefore, the delivery of mental
assessment schedule may have resulted in the misclassification health interventions that emphasize consumer choice over deci-
of outcomes such that participants were more likely to be mis- sions about their treatment could encourage recovery in indi-
classified as meeting the recovery criteria, and hence the rate of viduals with schizophrenia. Given the empirical support for
recovery may be inflated in the present study. This methodolo- the effectiveness of family psychoeducation in reducing the
gical limitation is not unique to the present study and has been odds of relapse (Kreyenbuhl, Buchanan, Dickerson, & Dixon,
widely reported in other studies (Bobes, Ciudad, Álvarex, & 2010), social work clinicians are encouraged to offer psychoe-
San, 2009; Novick et al., 2009; Petersen et al., 2008; Robinson, ducation intervention to family members as a means to foster
Woerner, & McMeniman, 2004; Schennach et al., 2012). We or enhance positive relationships. Additionally, incorporating
attempted to minimize the rate of misclassification by using supportive family members as partners in the treatment pro-
the length of psychiatric hospitalization during the previous cess could facilitate better outcomes for individuals with
6 months as an indicator of the stability of symptomatic remis- schizophrenia.
sion, which represents a methodological improvement over Regarding future research implications, we offer several
previous studies that tended to assume stability of symptoms recommendations. First, longitudinal studies with more fre-
in the absence of a rigorous assessment schedule. It is neverthe- quent follow-up intervals are needed to reduce the misclassifi-
less an inadequate indicator, given that contextual factors may cation of recovery outcomes. Research has generally used two-
influence participants’ odds of receiving psychiatric hospitali- wave designs to identify predictors of recovery. This limitation
zation. Selective attrition of participants with possibly lower is noteworthy because such designs preclude the identification
premorbid role functioning and who were more ill at study of changes in recovery status over time. Therefore, more multi-
entry could have also inflated the rate of recovery. Fifth, the wave studies are needed to classify the varied trajectories of
presence of selection bias may have influenced the magnitude schizophrenia and identify the predictors of these outcomes.
of the observed association between the investigated explana- Considering that the investigated clinical factors found to exert
tory variables and recovery. Because the sample featured par- significant influence on recovery in first-episode samples did
ticipants with a more persistent course of schizophrenia and not predict recovery in our sample, studies should include both
who experienced higher levels of functional disability at study first-episode and multiple-episode participants to determine
entry, predictors found to be associated with recovery in the whether recovery after a first episode of psychosis is predicted
present study may not be generalized to all individuals with by the same factors that predict recovery after multiple epi-
schizophrenia, particularly individuals with first-episode psy- sodes. Because schizophrenia is a complex disorder whose pro-
chosis and individuals not engaged in psychosocial rehabilita- gression is influenced by varied factors that have not been
tion. Sixth, the multivariate model yielded a pseudo R2 of satisfactorily investigated, future studies need to be sufficiently
26.06%, which attests to the presence of uninvestigated predic- powered to determine the contributory effects of a larger set of
tors, for example, treatment factors, service characteristics, and explanatory variables. Related to the measure of intrinsic moti-
neurocognition. Notwithstanding these limitations, this study vation, further validation of the expanded definition is needed.
offers evidence suggesting the potential of psychosocial attri- We also need studies with racially and ethnically diverse sam-
butes in effecting recovery from schizophrenia and lays the ples to examine the contribution of sociocultural and contextual
foundation for further research on the contribution of nonclini- factors to recovery from schizophrenia (Brekke & Barrio,
cal factors. 1997; Lopez, Barrio, Kopelowicz, & Vega, 2012).
Declaration of Conflicting Interests

Conclusion
The authors declared no potential conflicts of interest with respect to
In the context of psychosocial rehabilitation, individuals with the research, authorship, and/or publication of this article.
schizophrenia have highly heterogeneous trajectories. Although
the majority of participants in psychosocial rehabilitation may Funding
not show recovery during the early course of treatment, a sub- The authors disclosed receipt of the following financial support for
group of individuals with more positive family relationships, the research, authorship, and/or publication of this article: Data for
higher levels of intrinsic motivation, and less social and func- this study came from research supported by grants MH53282 and
tional impairment have an increased probability of achieving MH01628 awarded to John S. Brekke by the National Institute of
early recovery. Findings from this study have important clini- Mental Health.
cal implications, the most central being that intrinsic motiva-
tion and family network relationships are important targets of References
psychosocial interventions to promote recovery from schizo- Albert, N., Bertelsen, M., Thorup, A., Petersen, L., Jeppesen, P., Le
phrenia. Choi and Medalia (2010) found that intrinsic motiva- Quack, P., . . . Nordentoft, M. (2011). Predictors of recovery from
tion in individuals with schizophrenia was enhanced with psychosis: Analyses of clinical and social factors associated with
psychosocial interventions that were contextualized and perso- recovery among patients with first-episode psychosis after 5 years.
nalized and that promoted independent decision making, sug- Schizophrenia Research, 125, 257–266. doi:10.1016/j.schres.
gesting that intrinsic motivation is both malleable and shaped 2010.10.013

American Psychiatric Association. (1994). DSM-IV: Diagnostic and Davis, L., Kurzban, S., & Brekke, J. (2012). Self-esteem as a mediator
statistical manual of mental disorders. Washington, DC: American of the relationships between role functioning and symptoms of
Psychiatric Publishing. individuals with severe mental illness: A prospective analysis of
American Psychiatric Association. (2000). Diagnostic and statistical modified labeling theory. Schizophrenia Research, 137, 185–189.
manual of mental disorders (4th ed.). Washington, DC: American doi:10.1016/j.schres.2012.02.003
Psychiatric Publishing. Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-
Andreasen, N. C., Carpenter, W. T., Kane, J. M., Lasser, R. A., Marder, determination in human behavior. New York, NY: Plenum.
S. R., & Weinberger, D. R. (2005). Remission in Schizophrenia: Dunn, E. C., Wewiorski, N., & Rogers, E. S. (2008). The meaning and
Proposed criteria and rationale for consensus. American Journal of importance of employment to people in recovery from serious
Psychiatry, 162, 441–449. doi:10.1176/appi.ajp.162.3.441 mental illness: Results of a qualitative study. Psychiatric Rehabi-
Anthony, W. A. (1993). Recovery from mental illness: The guiding litation Journal, 32, 59–62. doi:10.2975/32.1.2008.59.62
vision of the mental health service system in the 1990s. Psychoso- Eberhard, J., Levander, S., & Lindström, E. (2009). Remission in schi-
cial Rehabilitation Journal, 16, 11–23. doi:10.1037/h0095655 zophrenia: Analysis in a naturalistic setting. Comprehensive Psy-
Argresti, A. (2007). Building and applying logistic regression models. chiatry, 50, 200–208. doi:10.1016/j.comppsych.2008.08.010
In D. J. Balding, N. A. C. Cressie, G. M. Fitzmaurice, H. Golstein, Færden, A., Nesvåg, R., & Marder, S. R. (2008). Definitions of the
I. M. Johnstone, G. Molenberghs, . . . S. Weisberg (Eds.), An intro- term ‘recovered’ in schizophrenia and other disorders. Psycho-
duction to categorical data analysis (p. 138). Hoboken, NJ: pathology, 41, 271–278. doi:10.1159/000141921
John Wiley. Gard, D. E., Fisher, M., Garrett, C., Genevsky, A., & Vinogradov, S.
Barham, P., & Hayward, R. (1998). In sickness and in health: Dilem- (2009). Motivation and its relationship to neurocognition, social
mas of the person with severe mental illness. Psychiatry, 61, cognition, and functional outcome in schizophrenia. Schizophrenia
163–170. doi:10.1521/00332747.1998.11024826 Research, 115, 74–81. doi:10.1016/j.schres.2009.08.015
Bellack, A. S. (2006). Scientific and consumer models of recovery in Goodman, S. H., Sewell, D. R., Cooley, E. L., & Leavitt, N.
schizophrenia. Schizophrenia Bulletin, 32, 432–442. doi:10.1093/ (1993). Assessing levels of adaptive functioning: The Role
schbul/sbj044 Functioning Scale. Community Mental Health Journal, 29,
Bertelsen, M., Jeppesen, P., Petersen, L., Thorup, A., Øhlenschlæger, 119–131. doi:10.1007/BF00756338
J., Quach, P. L., . . . Nordentoft, M. (2009). Course of illness in a Green, C. A., Polen, M. R., Janoff, S. L., Castleton, D. K., Wisdom, J.
sample of 265 patients with first-episode psychosis—Five-year P., Vuckovic, N., . . . Oken, S. L. (2008). Understanding how
follow-up of the Danish OPUS trial. Schizophrenia Research, clinician-patient relationships and relational continuity of care
107, 173–178. doi:10.1016/j.schres.2008.09.018 affect recovery from serious mental illness: STARS study results.
Bobes, J., Ciudad, A., Álvarex, E., & San, L. (2009). Recovery from Psychiatric Rehabilitation Journal, 32, 9–22. doi:10.2975/
schizophrenia: Results from a 1-year follow-up observation study 32.1.2008.9.22
of patients in symptomatic remission. Schizophrenia Research, Guada, J., Hoe, M., Floyd, R., Barbour, J., & Brekke, J. S. (2012).
115, 58–66. doi:10.1016/j.schres.2009.07.003 How family factors impact psychosocial functioning for African
Brekke, J. S., & Barrio, C. (1997). Cross-ethnic symptom differences American consumers with schizophrenia. Community Mental
in schizophrenia: The influence of culture and minority status. Health Journal, 48, 45–55. doi:10.1007/s10597-010-9365-6
Schizophrenia Bulletin, 23, 305–316. doi:10.1093/schbul/23.2.305 Harrow, M., Grossman, L. S., Jobe, T. H., & Herbener, E. S. (2005).
Brekke, J. S., & Long, J. D. (2000). Community-based psychosocial Do patients with schizophrenia ever show periods of recovery?
rehabilitation and prospective change in functional, clinical, and A 15-year multi-follow-up study. Schizophrenia Bulletin, 31,
subjective experience variables in schizophrenia. Schizophrenia 723–734. doi:10.1093/schbul/sbi026
Bulletin, 26, 667–680. doi:10.1093/oxfordjournals.schbul.a033485 Harrow, M., Hansford, B. G., & Astrachan-Fletcher, E. B. (2009).
Bursac, Z., Gauss, H. C., Williams, D. K., & Hosmer, D. W. (2008). Pur- Locus of control: Relation to schizophrenia, to recovery, and to
poseful selection of variables in logistic regression. Source Code for depression and psychosis: A 15-year longitudinal study. Psy-
Biological and Medicine, 16, 3–17. doi:10.1186/1751-0473-3-17 chiatry Research, 168, 186–192. doi:10.1016/j.psychres.2008.
Cechnicki, A., Bielańska, A., Hanuszkiewicz, I., & Daren, A. (2013). 06.002
The predictive validity of expressed emotions (EE) in schizophre- Hatfield, A. B., Gearon, J. S., & Coursey, R. D. (1996). Family mem-
nia: A 20-year prospective study. Journal of Psychiatric Research, bers’ ratings of the use and value of mental health services: Results
47, 208–214. doi:10.1016/j.jpsychires.2012.10.004 of a national NAMI survey. Psychiatric Services, 47, 825–831.
Choi, J., & Medalia, A. (2010). Intrinsic motivation and learning in doi:10.1176/ps.47.8.825
a schizophrenia spectrum sample. Schizophrenia Research, 118, Hegarty, J. D., Baldessarini, R. J., Tohen, M., Waternaux, C., &
12–19. doi:10.1016/j.schres.2009.08.001 Oepen, G. (1994). One hundred years of schizophrenia: A meta-
Cohen, O. (2005). How do we recover? An analysis of psychiatric analysis of the outcome literature. American Journal of Psychiatry,
survivor oral histories. Journal of Humanistic Psychology, 45, 151, 1409–1416. doi:10.1176/ajp.151.10.1409
333–354. doi:10.1177/0022167805277107 Heinrichs, D. W., Hanlon, T. E., & Carpenter, W. T., Jr. (1984). The
Davidson, L. (2003). Living outside mental illness: Qualitative studies quality of life scale: An instrument for rating the schizophrenic def-
of recovery in schizophrenia. New York: New York University icit syndrome. Schizophrenia Bulletin, 10, 388–398. doi:10.1093/
Press. schbul/10.3.388

Lim et al. 13
Hosmer, D. W., Lemeshow, S., & Sturdivant, R. X. (2013a). Assessing Lopez, S. R., Barrio, C., Kopelowicz, A., & Vega, W. A. (2012). From
the fit of the model. In D. J. Balding, N. A. C. Cressie, & G. M. documenting to eliminating disparities in mental health care
Fitzmaurice (Eds.), Applied logistic regression (3rd ed., for Latinos. American Psychologist, 67, 511–523. doi:10.1036/
pp. 153–227). Hoboken, NJ: John Wiley. a0029737
Hosmer, D. W., Lemeshow, S., & Sturdivant, R. X. (2013b). Model Lukoff, D., Nuechterlein, K. H., & Ventura, J. (1986). Manual for the
building strategies and methods for logistic regression. In D. J. expanded Brief Psychiatric Rating Scale. Schizophrenia Bulletin,
Balding, N. A. C. Cressie, & G. M. Fitzmaurice (Eds.), Applied 12, 594–602.
logistic regression (3rd ed., pp. 89–153). Hoboken, NJ: John Mancini, M. A. (2007). The role of self-efficacy in recovery from
Wiley. serious psychiatric disabilities: A qualitative study with fifteen
Hudson, W. (1982). The clinical measurement package: A field psychiatric survivors. Qualitative Social Work, 6, 49–74.
manual. Chicago, IL: Dorsey Press. doi:10.1177/1473325007074166
Jääskeläinen, E., Juola, P., Hirvonen, N., McGrath, J. J., Saha, S., Mancini, M. A., Hardiman, E. R., & Lawson, H. A. (2005). Making
Isohanni, M., . . . Miettunen, J. (2012). A systematic review and sense of it all: Consumer providers’ theories about factors facilitat-
meta-analysis of recovery in schizophrenia. Schizophrenia Bulle- ing and impeding recovery from psychiatric disabilities. Psychia-
tin, 39, 1296–1306. doi:10.1093/schbul/sbs130 tric Rehabilitation Journal, 29, 48–55. doi:10.2975/29.2005.48.55
Jacobson, N., & Greenley, D. (2001). What is recovery? A conceptual McGlashan, T. H. (1988). A selective review of recent North Ameri-
model and explication. Psychiatric Services, 52, 482–485. can long-term follow-up studies of schizophrenia. Schizophrenia
doi:10.1176/appi.ps.52.4.482 Bulletin, 14, 515–542. doi:10.1093/schbul/14.4.515
Johnson, D. (2005). Two-wave panel analysis: Comparing statistical Medalia, A., & Brekke, J. (2010). In search of a theoretical structure
methods for studying the effects of transitions. Journal of for understanding motivation in schizophrenia. Schizophrenia
Marriage and Family, 67, 1061–1075. doi:10.1111/j.1741-3737. Bulletin, 36, 912–918. doi:10.1093/schbul/sbq073
2005.00194.x Mueser, K. T., & McGurk, S. R. (2004). Schizophrenia. Lancet, 363,
Kane, J. M., Leucht, S., Carpenter, D., & Docherty, J. P. (2003). 2063–2072. doi:10.1016/S0140-6736(04)16458-1
Expert consensus guideline series. Optimizing pharmacologic Nakagami, E., Hoe, M., & Brekke, J. S. (2010). The prospective rela-
treatment of psychotic disorders. Introduction: Methods, commen- tionships among intrinsic motivation, neurocognition, and psycho-
tary, and summary. Journal of Clinical Psychiatry, 64, 5–19. social functioning in schizophrenia. Schizophrenia Bulletin, 36,
Kopelowicz, A., Liberman, R. P., Ventura, J., Zarate, R., & Mintz, J. 935–948. doi:10.1093/schbul/sbq043
(2005). Neurocognitive correlates of recovery from schizophre- Nakagami, E., Xie, B., Hoe, M., & Brekke, J. S. (2008). Intrinsic moti-
nia. Psychological Medicine, 35, 1165–1173. doi:10.1017/ vation, neurocognition and psychosocial functioning in schizo-
S0033291705004575 phrenia: Testing mediator and moderator effects. Schizophrenia
Kreyenbuhl, J., Buchanan, R. W., Dickerson, F. B., & Dixon, L. B. Research, 105, 95–104. doi:10.1016/j.schres.2008.06.015
(2010). The schizophrenia patient outcomes research team National Institute of Mental Health. (2009). Schizophrenia (NIH
(PORT): Updated treatment recommendations 2009. Schizophre- Publication No. 09-3517). Retrieved from http://www.nimh.
nia Bulletin, 36, 94–103. doi:10.1093/schbul/sbp130 nih.gov/health/publications/schizophrenia/nimh-schizophrenia-
Lambert, M., Naber, D., Schacht, A., Wagner, T., Hundermer, H.-P., booklet.pdf
Karow, A., . . . Schimmelmann, B. G. (2008). Rates and predictors Novick, D., Haron, J. M., Suarez, D., Vieta, E., & Naber, D. (2009).
of remission and recovery during 3 years in 392 never-treated Recovery in the outpatient setting: 36-month results from the schi-
patients with schizophrenia. Acta Psychiatrica Scandinavica, zophrenia outpatient health outcomes (SOHO) study. Schizophre-
118, 220–229. doi:10.1111/j.1600-0447.2008.01213.x nia Research, 108, 223–230. doi:10.1016/j.schres.2008.11.007
Lehman, A. F., & Steinwachs, D. M. (1998). Patterns of usual Ochocka, J., Nelson, G., & Janzen, R. (2005). Moving forward:
care for schizophrenia: Initial results from the schizophrenia Negotiating self and external circumstances in recovery. Psychia-
patient outcomes research team (PORT) client survey. Schizo- tric Rehabilitation Journal, 28, 315–322. doi:10.2975/28.2005.
phrenia Bulletin, 24, 11–20. doi:10.1093/oxfordjournals.schbul. 315.322
a033303 Onken, S. J., Craig, C. M., Ridway, P., Ralph, R. O., & Cook, J. A.
Lieberman, J. A., Drake, R. E., Sederer, L. I., Belger, A., Keefe, R., (2007). An analysis of the definitions and elements of recovery:
Perkins, D., & Stroup, S. (2008). Science and recovery in schizo- A review of the literature. Psychiatric Rehabilitation Journal,
phrenia. Psychiatric Services, 59, 487–496. doi:10.1176/appi. 31, 9–22. doi:10.2975/31.1.2007.9.22
ps.59.5.487 Peduzzi, P., Concato, J., Kemper, E., Holford, T. R., & Feinstein, A. R.
Liberman, R. P. (2002). Future directions for research studies and (1996). A simulation study of the number of events per variable in
clinical work on recovery from schizophrenia: Questions with logistic regression analysis. Journal of Clinical Epidemiology, 49,
some answers. International Review of Psychiatry, 14, 337–342. 1373–1379. doi:10.1016/S0895-4356(96)00236-3
doi:10.1080/0954026021000016987 Petersen, L., Thorup, A., Øqhlenschlæfer, J., Christensen, T. Ø.,
Liberman, R. P., Kopelowicz, A., Ventura, J., & Gutkind, D. (2002). Jeppesen, P., Krarup, G., . . . Nordentoft, M. (2008). Predictors of
Operational criteria and factors related to recovery from schizophre- remission and recovery in a first-episode schizophrenia spectrum
nia. International Review of Psychiatry, 14, 256–272. doi:10.1080/ disorder sample: 2-year follow-up of the OPUS trial. Canadian
0954026021000016905 Journal of Psychiatry, 53, 660–670.

Resnick, S. G., Fontana, A., Lehman, A. F., & Rosenheck, R. A. International Journal of Social Psychiatry, 55, 336–347.
(2005). An empirical conceptualization of the recovery orientation. doi:10.1177/0020764008093686
Schizophrenia Research, 75, 119–128. doi:10.1016/j.schres.2004. Sherbourne, C. D., & Stewart, A. L. (1991). The MOS social support
05.009 survey. Social Science and Medicine, 32, 705–714. doi:10.1016/
Resnick, S. G., Rosenheck, R. A., & Lehman, A. F. (2004). An 0277-9536(91)90150-B
exploratory analysis of correlates of recovery. Psychiatric Services, Silverstein, S. M., & Bellack, A. S. (2008). A scientific agenda for the
55, 540–547. doi:10.1176/appi.ps.55.5.540 concept of recovery as it applies to schizophrenia. Clinical
Ridgway, P. (2001). Restorying psychiatric disability: Learning from Psychology Review, 28, 1108–1124. doi:10.1016/j.cpr.2008.03.004
first person narrative accounts of recovery. Psychiatric Rehabilita- Strauss, G. P., Harrow, M., Grossman, L. S., & Rosen, C. (2010). Peri-
tion Journal, 24, 335–343. doi:10.1037/h0095071 ods of recovery in deficit syndrome schizophrenia: A 20-year
Robinson, D. C., Woerner, M. G., & McMeniman, M. (2004). Symp- multi-follow-up longitudinal study. Schizophrenia Bulletin, 36,
tomatic and functional recovery from a first episode of schizophre- 788–799. doi:10.1093/schbull/sbn167
nia. American Journal of Psychiatry, 161, 473–479. doi:10.1176/ Strauss, J. S., & Carpenter, W. T. (1972). The prediction of out-
appi.ajp.161.3.473 come in schizophrenia: I. Characteristics of outcome. Archive
Roe, D. (2003). A prospective study on the relationship between self- of General Psychiatry, 27, 739–746. doi:10.1001/archpsyc.1972.
esteem and functioning during the first year after being hospita- 01750300011002
lized for psychosis. Journal of Nervous and Mental Disease, Susser, E. S., Lin, S. P., & Conover, S. A. (1991). Risk factors for
191, 45–49. doi:10.1097/00005053-200301000-00008 homelessness among patients admitted to a state mental hospital.
Rosen, C., Grossman, L. S., Harrow, M., Bonner-Jackson, A., & Faull, American Journal of Psychiatry, 148, 1659–1664. doi:10.1176/
R. (2011). Diagnostic and prognostic significance of schneiderian ajp.148.12.1659
first-rank symptoms: A 20-year longitudinal study of schizophrenia Susser, E. S., Moore, R., & Link, B. (1993). Risk factors for homeless-
and bipolar disorder. Comprehensive Psychiatry, 52, 126–131. ness. Epidemiologic Reviews, 15, 546–556.
doi:10.1016/j.comppsych.2010.06.005 Test, M. A., Knoedler, W. H., Allness, D. J., Burke, S. S., Brown, R.
Rosenthal, D. (1970). Genetic studies of schizophrenia. In N. L., & Wallisch, L. S. (1991). Long-term community care through
Garmezy, R. L. Solomon, L. V. Jones, & H. W. Stevenson (Eds.), an assertive continuous treatment team. In C. A. Tamming &
Genetic theory and abnormal behavior (pp. 92–200). New York, S. C. Schulz (Eds.), Advances in neuropsychiatry and psychophar-
NY: McGraw-Hill. macology: Vol. 1, Schizophrenia research (pp. 239–246). New
Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the York, NY: Raven Press.
facilitation of intrinsic motivation, social development, and well- Torgalsbøen, A.-K., & Rund, B. R. (2010). Maintenance of recovery
being. American Psychologist, 55, 68–78. doi:10.1037/0003- from schizophrenia at 20-year follow-up: What happened? Psychiatry,
066X.55.1.68 72, 70–82. doi:10.1521/psyc.2010.73.1.70
Saperstein, A. M., Fiszdon, J. M., & Bell, M. D. (2011). Intrinsic moti- Ventura, J., Subotnik, K. L., Guzik, L. H., Hellemann, G. S., Gitlin, M.
vation as a predictor of work outcome after vocational rehabilita- J., Wood, R. C., & Nuechteriein, K. H. (2011). Remission and
tion in schizophrenia. Journal of Nervous and Mental Disease, recovery during the first outpatient year of the early course of schi-
199, 672–677. doi:10.1097/NMD.0b013e318229d0eb zophrenia. Schizophrenia Research, 132, 18–23. doi:10.1016/
Schennach, R., Riedel, M., Obermeier, M., Jäger, M., Schmauss, M., j.schres.2011.06.025
Laux, G., . . . Möller, H.-J. (2012). Remission and recovery and Verma, S., Subramaniam, M., Abdin, E., Poon, L. Y., & Chong, S. A.
their predictors in schizophrenia spectrum disorder: Results from (2012). Symptomatic and functional remission in patients with
a 1-year follow-up naturalistic trial. Psychiatric Quarterly, 83, first-episode psychosis. Acta Psychiatrica Scandinavica, 126,
187–207. doi:10.1007/s11126-011-9193-z 282–289. doi:10.1111/j.1600-0447.2012.01883.x
Schön, U.-K., Denhov, A., & Topor, A. (2009). Social relationships as World Health Organization. (2013). Schizophrenia. Retrieved from
a decisive factor in recovering from severe mental illness. http://www.who.int/mental_health/management/schizophrenia/en/

View publication stats

Recovery From Schizophrenia in Community-Based Psychosocial Rehabilitation Settings Rates and Predictors

Uploaded by

Copyright:

Available Formats

Recovery From Schizophrenia in Community-Based Psychosocial Rehabilitation Settings Rates and Predictors

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Recovery From Schizophrenia in Community-Based Psychosocial Rehabilitation Settings Rates and Predictors

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Recovery From Schizophrenia in Community-Based Psychosocial

Article in Research on Social Work Practice · June 2015

Caroline S Lim Concepción Barrio

SEE PROFILE SEE PROFILE

Assisted Outpatient Treatment in Los Angeles County View project

Bridge Peer Health Navigator Intervention View project

The user has requested enhancement of the downloaded file.

Caroline Lim1, Concepción Barrio1, Mercedes Hernandez1,

Schizophrenia, which affects approximately 24 million Literature Review

Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

Table 1. Demographic and Clinical Descriptors of Sample at Baseline

Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

Variable Completers (n ¼ 152) Noncompleters (n ¼ 94) Test Statistic p

Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

Variable Not in Recovery (n ¼ 117) In Recoveryb (n ¼ 29) Test Statistic P

be estimated precisely in a multivariable logistic regression Predictors of Recovery

Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

Declaration of Conflicting Interests

Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

Downloaded from rsw.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on September 25, 2015

You might also like