Sage Journals: Discover world-class research

Abstract

Objective:

Impairment in psychosocial function is common in schizophrenia. Long-acting injectable atypical antipsychotics are thought to enhance psychosocial function by boosting adherence. However, no systematic review has examined the effects of long-acting injectable atypical antipsychotics on psychosocial function in clinical trials.

Methods:

We searched major databases including Medline/PubMed, PsychINFO, EMBASE, CINAHL, Scopus, Web of Science, Cochrane Central Register of Controlled Trials and Clinical Trial Registries for randomised controlled trials that compared long-acting injectable atypical antipsychotics to placebo, oral antipsychotic medications or long-acting injectable atypical antipsychotics for all years till 2018, with no language limits. We performed a systematic review of findings on change in psychosocial function and its predictors in the included reports. Data on change in psychosocial functioning were meta-analysed using a random-effects model.

Results:

A total of 26 studies were included in systematic review, and 19 studies with 8616 adults, 68.1% males were meta-analysed. Long-acting injectable atypical antipsychotics were superior to placebo (standardised mean difference = 0.39; 95% confidence interval = [0.32, 0.47]; p < 0.001; I² = 0%; 9 studies) and oral antipsychotic medications (standardised mean difference = 0.16; 95% confidence interval = [0.01, 0.31]; p = 0.04; I² = 77%; 10 studies) for improved psychosocial function and superiority was maintained in short- and long trials. Poor psychosocial function was predicted by longer treatment duration, severe symptoms, poor cognition and poor insight. Functioning was assessed by either a single or a combination of measures, but was not the primary outcome in most studies. Other sources of bias include poor blinding and reporting of randomisation.

Conclusion:

Long-acting injectable atypical antipsychotics are beneficial for recovery of psychosocial function in comparison with placebo, but the magnitude of superiority over oral antipsychotic treatment was small. Severe psychopathology at baseline predicted poor psychosocial function. Future effectiveness trials in which post-randomisation involvement is kept to a minimum, and psychosocial function is included as primary outcome a priori, are needed to capture the real-world impact of long-acting injectable atypical antipsychotics and to address methodological biases.

Keywords

Atypical antipsychotics functional outcome long-acting injectable predictors psychosocial function quality of life schizophrenia

Introduction

Functional impairment is a major problem in schizophrenia. Globally, the disorder was estimated to account for 222.3 million disability-adjusted life years (DALYs) in 2016 (GBD 2016 DALYs and HALE Collaborators, 2016; GBD 2016 Disease and Injury Incidence and Prevalence Collaborators, 2016). Functional deficit is a core feature in the diagnosis and course of schizophrenia which can persist beyond hallmark clinical symptoms and negatively impact prognosis (Mueser and McGurk, 2004; Van Os and Kapur, 2009). While there is strong evidence for the efficacy of antipsychotic medications in treating psychotic symptoms, functional improvement is more variable and does not necessarily require complete symptom remission (Jaaskelainen et al., 2013).

Medication adherence is key to symptomatic and functional recovery (Czobor et al., 2015), and relapse rates are high following cessation of oral antipsychotics or partial adherence (Alvarez-Jimenez et al., 2012; Novick et al., 2010). Non-adherence is, however, common in schizophrenia, with discontinuation rates ranging from 44% within 1 year after first episode (Kaplan et al., 2013) and as high as 74% over longer periods (Lieberman et al., 2005; Valenstein et al., 2006).

Long-acting injectable (LAI) antipsychotic formulations are effective in reducing the rate of antipsychotic discontinuation (Tiihonen et al., 2011) and significantly lower the risk of relapse and rehospitalisation (Lafeuille et al., 2013; Leucht et al., 2011). However, extrapyramidal side-effects and poor tolerability of conventional LAIs have limited their use and led to the development of second-generation LAI to improve adherence (Tiihonen et al., 2017). Since the introduction of LAI-risperidone in 2003, several other long-acting injectable atypical antipsychotic (LAI-A) formulations (including olanzapine, paliperidone and aripiprazole) have been developed with differences in dosing flexibility, costs, pharmacodynamics–kinetics and side-effect profile (Miyamoto and Fleischhacker, 2017). Several theories have been proposed to explain the potential advantages of broader and early use of LAIs for schizophrenia. First, LAI use facilitates close monitoring of adherence which is particularly important in the early phases of psychotic illness (Tiihonen et al., 2011). Second, patients may prefer the longer dosing interval, and third, single dosing prevents the risk of self-medication and harmful use that can occur with oral preparations (Maia-de-Oliveira et al., 2013; Patel and Davis, 2005). Finally, due to higher bioavailability and lower fluctuations in serum level (Ereshefsky and Mascarenas, 2003 ), the pharmacokinetics of LAIs may promote lower effective dosing to improve tolerability. However, evidence from clinical studies suggests that for most LAIs, discontinuation rates due to side-effects are similar to oral dosing (Ostuzzi et al., 2017).

Although psychosocial function as a term is widely used in literature, there has been a lack of uniformity in its construct. The patient’s well-being/quality of life (QoL) is not measured in some studies. More recently, a multidimensional construct that includes both measures of daily function and subjective well-being/QoL is gaining support as a more reliable, valid, comprehensive and holistic indicator (Dixon et al., 2010; Harvey, 2013; Leifker et al., 2010; Menezes et al., 2006; Olagunju et al., 2018; Ro and Clark, 2009). Thus, in this study, we used a multidimensional construct of psychosocial function that includes multiple domains of basic-daily function, socio-occupational function, interpersonal relationships, adaptive function, life satisfaction and subjective well-being/QoL.

The biological basis of psychosocial function is yet to be well understood; however, antipsychotics are thought to be neuroprotective against the cytotoxic effects of active psychosis (Kim and Na, 2017; Wyatt, 1991). Psychosocial function is known to worsen with frequent relapses or longer duration of untreated psychosis. In addition, performance-based skills decline with the disruption of social relationships, unemployment and institutionalisation following relapses or chronic illness (Emsley et al., 2013). LAI-A may therefore confer functional recovery through relapse prevention via better adherence (Kaplan et al., 2013). Furthermore, LAI-A have been linked to good subjective well-being and life satisfaction, which can in turn enhance social adaptation, autonomy and psychosocial function (Kane et al., 2015; Nasrallah et al., 2004). Atypical LAIs are now recommended in practice guidelines as first-line choice for maintenance treatment in non-adherent patients (Bai et al., 2006; Galletly et al., 2016; Lehman et al., 2004).

While the efficacy of LAI-A for symptom remission and relapse prevention is well established, the functional benefit of these treatments has not been interpreted across clinical trials. The majority of comparative effectiveness trials of LAI-A versus oral medications are modelled on proxy indices of functional outcome including relapse rates and rehospitalisation. These proxy endpoints are thought to be reliable to ascertain real-world functioning (Olivares et al., 2013) and have good face validity, however, hospitalisation as an outcome translates poorly between differing healthcare contexts, does not capture patient’s personalised experience and can be a negative therapeutic goal to end-users (Burns, 2007).

The results of randomised controlled trials (RCTs) of LAIs versus orals for relapse and rehospitalisation are mixed. Compared to RCTs, naturalistic studies are consistent in reporting LAI superiority, although they are not flaw-proof (Zhornitsky and Stip, 2012). The impact of these study biases on functional outcome measures need to be considered individually for all studies (RCT and observational) to appropriately interpret the benefits of LAI over oral treatment. In addition, experts have recommended effectiveness trials, in which post-randomisation involvement (manipulation of trial-related factors including adherence such that the performance of intervention reflects ideal or well-controlled circumstances) would be kept to a minimum to better reflect routine practice (Kane et al., 2013; Kishimoto et al., 2014). Such novel trials can be informed by an analysis of study quality in existing trials using psychosocial function measures.

To summarise and interpret the impact of LAI-A on function, we conducted a systematic review and meta-analysis using a multidimensional construct for function (‘psychosocial function’) that includes multiple domains of basic-daily function, socio-occupational function, interpersonal relationships, adaptive function, life satisfaction and subjective well-being/QoL (Menezes et al., 2006; Ro and Clark, 2009). Critical appraisal of all included reports was performed to assess the quality of the implementation of standardised measures of psychosocial function. We also systematically reviewed factors predictive of improvement in function with LAI-A treatment to identify characteristics that could be used to guide the decision to commence LAIs. Our specific aims were to

Investigate the longitudinal effects of LAI-A on measures of psychosocial function among people with schizophrenia or schizoaffective disorders;

Compare the effects of LAI-A on measures of psychosocial function with placebo or oral antipsychotic medications;

Describe the predictors of change in psychosocial function included in primary studies;

Appraise the quality of measures of psychosocial function.

Materials and methods

Inclusion criteria

This systematic review was conducted in accordance with Cochrane collaboration guideline and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline (Liberati et al., 2009; Moher et al., 2009) (checklist included as Supplementary S1). All literature till March 2018, with no language limits, was searched for RCTs that compared LAI-A with placebo or active controls in people with schizophrenia or schizoaffective disorders. Eligible studies included adults aged 18 years or older of either gender that provided information on measures of psychosocial function. We included reports on placebo-controlled trials and open-label active-controlled RCTs to improve study power and to allow a comprehensive appraisal of the quality of psychosocial function measurement. Studies with flexible or fixed doses of LAI-A within the recommended therapeutic range for clinical efficacy were shortlisted (Brissos et al., 2014). Non-randomised controlled studies and single-arm prospective experimental trials were excluded (Higgins and Green, 2011). Data concerning the predictors of functional outcomes were not consistent across studies and is described as a systematic review alone.

Outcome parameters

The main outcome was longitudinal change in psychosocial function associated with LAI-A treatment as defined in primary studies in comparison with placebo or oral antipsychotic medications. We indexed psychosocial function with study-defined multidimensional measures of domains of basic-daily function, socio-occupational function, interpersonal relationship, adaptive function, life satisfaction and subjective well-being/QoL. The composite scores of the psychosocial function scales in the included studies were used. LAI-A-controlled trials, designed to compare the efficacy of different LAI-A or of different doses of the same LAI-A, were included only in systematic review. Furthermore, we collected data on factors that predicted change in psychosocial function from the included studies.

Database search

We searched Medline/PubMed, PsychINFO, EMBASE, CINAHL, Scopus, Web of Science, Cochrane Central Register of Controlled Trials and Clinical Trial Registries (http://clinicaltrials.gov/) for all studies (published and unpublished) to avoid publication bias. In case of PubMed, the terms used as string in combination to search in titles, abstract and as MeSH terms included (1) ‘Antipsychotic Agents’ ‘Pharmacological Action’, ‘antipsychotic’ (2) ‘schizophren’ OR ‘schizo’ (3) ‘inject’ OR ‘depot’ OR ‘long acting’ OR ‘delayed-action preparations’ (4) ‘Quality of life’ OR ‘Function’. An additional search was done with the names and standard abbreviations of scales used to measure psychosocial function and QoL. The scales included the following: ‘Global Assessment of Function’, ‘Heinrichs–Carpenter QoL scale’, ‘Short Form-36 Health Survey’, ‘Personal and Social Performance scale’, ‘Strauss–Carpenter Levels of Functioning’ and ‘Social and Occupational Functioning Assessment’. The clinical trial registration number was used to identify all published reports on each of the clinical trials for comprehensive extraction of data. References of included studies and relevant reviews were snowball searched for additional studies. Authors of studies with missing data were contacted for further information.

A.T.O. screened titles and abstracts of 1104 articles to produce a shortlist of 98 potential articles. The full texts of the selected articles were reviewed by A.T.O. for eligibility as per the a priori criteria.

Data extraction and management

Data extraction was done by A.T.O. independently using piloted forms based on the PRISMA guidelines and included the PICOS (population, intervention, comparison, outcomes and setting) data items. Data extraction forms collected the following data: first author, funding source, study design, study country or sites, participants characteristics (age range, total number, gender, diagnosis and treatment setting), trial duration, antipsychotics (LAI-A vs controls), dosage, trial arms, measures of psychosocial function, clinical scales, findings on psychosocial function (baseline, endpoint and mean change) and identified predictors with statistical significance. During data collection, discrepancies were resolved through consultation with the senior author, B.T.B., where necessary.

Assessment of quality of studies with risk-of-bias tool

We assessed the risk of bias within individual studies using the Cochrane Collaboration methods for assessing bias in controlled clinical trials (Higgins and Green, 2011). Bias assessments were conducted at both the study and the outcome levels and included items such as random sequence generation, allocation concealment, blinding of personnel, blinding of outcome assessment, incomplete data reporting, selective reporting and other sources of bias. The presence of possible publication bias was visually assessed from funnel graphs and included Egger’s test (Higgins and Green, 2011; Leucht et al., 2012).

Synthesis of results and meta-analytic calculation

We performed both qualitative and quantitative analyses of psychosocial function but could only complete qualitative analyses for predictor variables. We compared LAI-A with placebo or oral antipsychotic medications on change in psychosocial function using standardised mean difference (SMD) as the primary effect size. Analysis was conducted considering all time points, with data for the last outcome time point in each study used to calculate change from baseline. All calculations were performed using Rev. Manager 5 software (Cochrane Collaboration, 2014) and meta-regression was done with STATA (StataCorp, 2017). A p-value less than 0.05 (two-tailed) was considered statistically significant. Given the heterogeneity in the measures of psychosocial function, we used the generic inverse variance model and random effects to estimate the pooled SMDs of the included studies. If change score was not available, endpoint value was subtracted from baseline. We merged data from multi-arm trials using formula for combining groups and online calculators in Rev. Manager (Higgins and Green, 2011). Missing standard deviations were estimated from test statistics and useable data from published scales based on Cochrane methodology (Ma et al., 2008; Marshall et al., 2000). Study heterogeneity was investigated using the sensitivity test and the chi-square test of homogeneity (p < 0.05) together with the I²-statistic. We considered values of 50% heterogeneity or higher non-negligible and looked for explanations in the included studies (Higgins et al., 2003; Higgins and Green, 2011).

Sub-group analyses and meta-regression

We conducted sub-group analyses for duration of trial; however, analyses were limited due to one short-term oral-controlled trial. To check the effects of trial duration, we categorised studies into short term for trials not beyond 3 months and those greater than 3 months were grouped as long term (Wahlbeck et al., 1999). We also performed sub-group analysis to assess any difference in functional outcomes for trials that used treatment as usual (TAU) with oral antipsychotics in comparison with LAI-A. We conducted meta-regression to check effects of study settings (outpatient vs inpatient), trial design, year of publication, industry sponsorship, baseline symptom severity and inclusion of psychosocial function as primary outcome in included trials (Borenstein et al., 2009).

Results

Study selection

A total of 1631 publications were retrieved from all sources and 527 duplicates were removed. Abstracts and titles of the remaining 1104 reports were screened, and 98 full texts were shortlisted for review. Of these 98 full texts, 25 (containing 26 studies) were included, while 73 were excluded due to non-RCT studies (n = 46), duplicate data (n = 16), absence of a validated functional measure (n = 19) and wrong medications-antipsychotics (n = 21). We used main reports where many articles published findings on same RCT (details of reports used for data extraction are included in Supplementary S2). Overall, findings in 26 primary studies (one report contained two RCT studies) were included in systematic review, 19 studies from both placebo and oral-controlled trials were meta-analysed, and 7 studies with information on predictors were systematically reviewed. Figure 1 presents the PRISMA flow diagram (Liberati et al., 2009).

Figure 1.

Flow diagram of study selection.

Study characteristics

The study characteristics are shown in Tables 1 –3 and Supplementary Table S2. All 26 studies were conducted from 2003 to 2016 and included 11,097 participants with 67.1% males. Data on two trials were obtained solely from clinical trial registry (NCT00604279 and NCT00992407) because there were no publications indexed to them. Overall, 13 studies reported double-blind RCTs including all the 8 placebo-controlled studies, 3 LAI-A-controlled studies and 2 oral-controlled studies. Open-label studies were mostly oral-controlled (n = 8) and three studies were LAI-controlled. The mean trial duration was 53.52 (±9.05) weeks and ranged from 8 to 130 weeks. Of the nine placebo-controlled studies, five were short term (duration ranged from 8 to 12 weeks), while the remaining were long term (duration ranged from 14 to 52 weeks). There were 11 oral-controlled trials: 1 (Bai et al., 2006) was short trial (12 weeks) and 10 were long trials (duration ranged from 52 to 130 weeks). Concerning oral-controlled trials, LAI-A were compared with oral atypical antipsychotics in eight trials, in two of which the choice of atypical was at investigator discretion (Buckley et al., 2015; Wykes et al., 2013). Three trials compared LAI-A with TAU (oral antipsychotic medications as per investigator discretion; Alphs et al., 2016; Rosenheck et al., 2011; Schreiner et al., 2015). All six LAI-A-controlled studies (duration ranged from 24 to 52 weeks) were long term and compared LAI-A to different LAI-A (n = 4) or different LAI-A doses (n = 2). Out of 26 trials, 1 Korean, Japanese and Taiwanese trial each was single centre. Of the 23 multicentre studies, 16 took place across several international centres and 5 solely in US centres.

Table 1.

Characteristics of studies on LAI-A vs placebo.

Study	Design	Sample (N)Gender (%)	LAI-A/dosage	Functional measures	Clinical scales	Findings on psychosocial function and statistics
Berwaerts et al. (2015) ^a	RCT double-blindRatio = 1:1LAI-A vs placebo	N = 305M = 75%	Paliperidone175, 263, 350525 mg eq/3m	PSP	DSM-IV-TR, CGI-S, PANSS	PSP: there was decline in function on both formulations, although worse in placebo −4.2 (±9.70) compared to LAI-A −0.5 (±6.63). No statistical analysis was done.
Fleischhacker et al. (2014) ^a	RCT double-blindRatio = 2:1LAI-A vs placebo	N = 403M = 59.8%	Aripiprazole400 mg/m	PSP	DSM-IV-TR, CGI	PSP: significant difference in change in score between LAI-A and placebo, indicating greater functional deterioration with placebo (p < 0.001). LAI-A – mean (±SD) baseline to endpoint = 68.0 (±0.76) to 66.6 (±0.89). Placebo – mean (SD) baseline to endpoint = 69.4 (±0.97) to 63.2 (±1.37).
Fu et al. (2015) ^a	RCT double-blindRatio = 1:1LAI-A vs placebo	N = 334 M = 50.6%	Paliperidone78, 117, 156, 234 mg/m	PSP	DSM-IV, PANSS, HAM-D-21, YMRS, SCID, CGI-S-SCA, MSQ	PSP: significant difference (increase) in LAI-A vs placebo at endpoint. LS means difference = 3.3 (95% CI = [0.68, 5.95]).Participants with good function improved from 57.9% to 59% in LAI-A compared to placebo where participants with good function decreased from 50.6% to 41.1% (p = 0.002).
Hough et al. (2010)	RCT double-blindRatio = 1: 1LAI-A vs placebo	N = 410 M = 55%	Paliperidone 25, 50, or 100 mg eq/m (flexible dose)	PSP	DSM-IV, PANSS, CGI-S, PSP, EPS (SAS), AIMS, BARS	PSP: significant difference in longitudinal change in score (decrease) between LAI-A and placebo, indicating greater functional deterioration with placebo (p < 0.001). Least-squares mean difference of change in scores from baseline was placebo = −7.2 (13.03) vs LAI-A = −1.5 (11.53).
Isitt et al. (2016)	RCT double-blindRatio = 1:1LAI-A vs placebo	N = 337 M = 76.6%	Risperidone (RBP-700)90, 120 mg/m	EuroQol EQ-5D-5 L, SWN-S	DSM-IV-TR, C-SSRS, SAS, AIMS, BARS, MSQ, CGI-S, POM, PANSS	EQ-5D-5 L: significant difference in longitudinal change in score (increase) between LAI-A compared to placebo indicating improved functioning in LAI-A. Change from baseline mean score for LAI-A vs placebo was 8.184 vs 3.295, p = 0.0212.SWN-S: significantly greater improvements in LAI-A in the following scores compared to placebo: physical functioning (2.086 vs 0.750; p = 0.0093), social integration (2.868 vs 1.765; p = 0.0368) and total score (10.951 vs 6.942; p = 0.0395). Significant improvement in LAI-A relative to placebo was recorded in those on 120 mg/m dose compared to 90 mg/m.
Kane et al. (2003)	RCT double-blindRatio = 1:1LAI-A vs placebo	N = 400 M = 75.3%	Risperidone 25, 50, 75 mg/bw	SF-36	DSM-IV, PANSS, CGI, ESRS	SF-36: there was significant difference (increase) in scores (improvement from baseline) across multiple domains (BP, GH, SF, RE, MH) in favour of all the LAI-A dose group compared to placebos (p < 0.05). MC showed largest improvement after 12 weeks (2.37 vs −0.27 placebo).
Kane et al. (2014)	RCT double-blindRatio = 1:1LAI-A vs placebo	N = 340 M = 79.1%	Aripiprazole 120 mg/m	PSP	DSM-IV-TR, MINI, BARS PANSS, SAS CGI-I, CGI-S, TEAE, AIMS	PSP: LS mean (SE) change from baseline in PSP total score was significantly greater with LAI-A vs placebo at endpoint (12.3 [1.2] vs 5.2 [1.2], respectively; p < 0.0001) and at week 12 (13.0 [1.2] vs 5.5 [1.2]; p < 0.0001]
Pandina et al. (2010)	RCT double-blindRatio = 1:1LAI-A vs Placebo	N = 652 M = 67.5%	Paliperidone25,100, 150 mg/m	PSP	DSM-IV, PANSS, CGI-S	PSP: a dose-related improvement in LAI-A treatment groups including, compared to placebo, which was significant for the 100 mg eq (p = 0.007) and 150 (p < 0.001) mg eq groups.
Witte et al. (2012)	RCT double-blindRatio = 1:1:1:1LAI-A vs placebo	N = 404 M = 70.5%	Olanzapine 210, 300, 405 mg/bw	QLS, SF-36	DSM-IV or DSM-IV-TR, BPRS, PANSS	QLS: significant difference (increase) in total score in LAI-A compared to placebo (p < 0.05). Scores on IF, IPR and IR were significantly higher in LAI-A compared to placebo but no difference in COA domain (p < 0.05).SF-36: significant difference (increase) in MCS and MHF domain in LAI-A compared to placebo.

RCT: randomised controlled trials; LAI-A: long-acting injectable atypical antipsychotics; PSP: Personal and Social Performance scale; DSM-IV-TR: Diagnostic and Statistical Manual of Mental Disorders – fourth edition, text revision; CGI-S-SCA: Clinical Global Impression of Severity for Schizoaffective Disorder; PANSS: Positive and Negative Symptoms Scale; HAM-D-21: 21-item Hamilton Depression Rating Scale; YMRS: Young Mania Rating Scale; SCID: severe combined immunodeficiency; MSQ: Medication Satisfaction Questionnaire; CI: confidence interval; SAS: Simpson–Angus Scale; AIMS: Abnormal Involuntary Movement Scale; BARS: Barnes Akathisia Scale; EQ-5D-5L: five-level EuroQol five-dimensional questionnaire; SWN-S: Subjective Well-being under Neuroleptic Treatment, short version; C-SSRS: Columbia – Suicide Severity Rating Scale; POM: Preference of Medicine Questionnaire; SF-36: 36-item Short Form Health Survey; ESRS: Extrapyramidal Symptom Rating Scale; TEAE: treatment-emergent adverse events; QLS: Heinrichs–Carpenter QoL scale; COA: common objects and activities; MCS: Mental Component Summary; MHF: Mental Health Functioning; BPRS: Brief Psychiatric Rating Scale; IF: intrapsychic foundation; IR: instrumental roles; IPR: interpersonal roles; mg/bw: milligramme per 2 weeks; mg/m: milligramme per month.

Included in systematic review only.

Table 2.

Characteristics of studies on LAI-A vs oral antipsychotic medications.

Study	Design	Sample (N)Gender (%)	LAI-A/dosage	Functional measures	Clinical scales	Findings on psychosocial function and statistics
Alphs et al. (2015)	RCT open-labelRatio = 1:1LAI-A vs oral^a	N = 444M = 86.3%	Paliperidone78–234 mg/m (flexible dose)	PSP	DSM-IV, CGI-S, MINI	PSP: both formulations improved function and no significant between-group differences were observed in mean change in PSP total scores (p = 0.689).
Ascher-Svanum et al. (2014)	RCTs open-labelRatio = 1:1LAI-A vs oral^b	N = 1182 M = 66.8%	Olanzapine150–600 mg/m	QLS, SF-36; EuroQol	DSM-IV, DSM-IV-TR, MSQ, PANSS, CGI-S	QLS: significant change (increase) in total score from baseline to endpoint in LAI-A, mean (±SD) BL = 64.0 (±21.2) and mean (±SD) EP = 70.8 (±22.9). No significant difference in total scores, COA, IF, IPR and IR for LAI-A vs orals (p > 0.05).
Bai et al. (2006)	RCT single-blindRatio = 1:1LAI-A vs oral^c	N = 50 M = 48%	RisperidoneMax 50 mg/bw	GAF; SF-36	DSM-IV, SAS, PANSS, CGI, UKS, AIMS, BARS	GAF: both formulations improved general function and there was no significant difference in change of score from baseline to the endpoint between LAI-A (0.8 ± 8.1) vs oral (1.6 ± 9.4).SF-36: No significant difference between LAI-A and orals across all domains except in social function where LAI-A was superior to oral (p = 0.017).
Buckley et al. (2015)	RCT open labelRatio = 1:1LAI-A vs oral^d	N = 305 M = 71%	Risperidone12.5–50 mg/bw	SOF	DSM-IV-TR, BPRS, SANS, CGI, BARS	SOF: No significant difference in total score at endpoint in LAI-A vs oral. Estimated mean is 95% CI = 42.7 [41.6, 43.9] vs 42.1 [40.9, 43.4]; F = 0.34, df = 9, 1420; p = 0.96.SOF global rating: No difference in LAI-A vs oral. Estimated mean is 95% CI = 3.0 [2.9, 3.1] vs 3.0 [2.8, 3.1]; F = 0.90, df = 9, 1430; p = 0.53.
Fleischhacker et al. (2014)	RCT double-blindRatio = 2:2LAI-A vs oral^e	N = 662 M = 61.3%	Aripiprazole400 mg/m	PSP	DSM-IV-TR, CGI	PSP: scores remained stable in LAI-A and oral groups. LAI-A mean (±SD) baseline to endpoint = 65.3 (±0.7) to 65.9 (±0.8); oral mean (±SD) baseline to endpoint = 66.9 (±0.75) to 66.8 (±0.83).
Keks et al. (2007) ^f	RCT open labelRatio = 1:1LAI-A vs oral^b	N = 547 M = 56%	Risperidone25, 50 mg/m	W-QOLI	DSM-IV, SCI-PANSS, CGI-S, SARS	W-QOLI: there was improvement from baseline to end point on all sub-scale ratings. Clinically meaningful improvements (score changes > 0.5 points) were seen in three domains in LAI-A and orals: occupational activities, psychological well-being and symptoms/outlook.
NCT00992407 (2014)	RCT open labelRatio = 1:1LAI-A vs oral^c	N = 41 M = 60.9%	RisperidoneMax 50 mg/bw	PSP, SFS, GAF	DSM-IV, SAS PANSS, CGI-S, BDI, BARS, BAI, DAI, AIMS, Cogtest	PSP: there was improvement in both formulations from baseline to endpoint but no significant difference between LAI-A and oral (p = 0.859).GAF: there was improvement in both formulation from baseline to endpoint but no significant difference between LAI-A and oral (p = 0.957).
Rosenheck et al. (2011)	RCT double-blindRatio = 1:1LAI-A vs oral^a	N = 369 M = 90.5%	Risperidone25–50 mg/bw	QLS, PSP, QWBS	DSM-IV, CGI, DAI, PANSS, BSI, ASI, SAS, BARS	QLS, PSP and QWBS: significant change (improvement) in scores from baseline to endpoint in LAI-A. No significant superiority of LAI-A over oral was observed on all the three scales (p > 0.05). No difference between LAI-A vs oral on QLS total scores, IF, IPR and IR (p > 0.05).
Rouillon et al. (2013)	RCT open labelRatio = 1:1LAI-A vs oral^e,g	N = 666 M = 57.9%	Risperidone25–50 mg/bw	SQLS-R4, SOFAS, SF-12	AIMS, DSM-IV, ESRS, CGI-S	SOFAS: mean change in scores from baseline to endpoint was significant for LAI-A for all assessment points and endpoint (p < 0.001). Between-treatment differences in change in scores for LAI-A vs oral, respectively, were significant at treatment in 6 months (6.1 [SD = 15.2] vs 2.7 [SD = 11.0]; p = 0.02), 12 months (9.5 [SD = 11.2] vs 6.1 [SD = 10.7]; p = 0.009) and endpoint (6.6 [SD = 15.2] vs 1.1 [SD = 16.1]; p < 0.0001).
Schreiner et al. (2015)	RCT, open-labelRatio = 1:1LAI-A vs oral^a	N = 715 M = 57.9%	Paliperidone25–150 mg/m	PSP, EQ-5D, SWN-S, SF-36	DSM-IV, SAS, BARS, PANSS, AIMS CGI-S, CGI-C, TSQM	PSP: total score improved significantly from baseline to endpoint for LAI-A vs oral (mean change = 9.8 vs 8.7); both within groups p < 0.0001.SF-36, EQ-5D and SWN-S: significant improvement in scores in LAI and orals were observed. All scales showed similarity in LAI-A vs oral (p > 0.05).
Wykes et al. (2013)	RCT open labelRatio = 1:1LAI-A vs oral^d	N = 50 M = 72%	Risperidone25–50 mg/bw (flexible dose)	AQoL, PSP	DSM-IV-TR, WAI, PANSS, CGI-S, CGI-C	AQoL: significant change in LAI-A over time: QoL-illness (N = 36; z = −2.12; p = 0.034) and QoL-activities of daily living (N = 36; z = −2.82; p = 0.005).PSP: no change in score over time in LAI-A (N = 43; z = −1.08; p = 0.279). LAI-A was not significantly different from oral.

RCT: randomised controlled trials; LAI-A: long-acting injectable atypical antipsychotics; PSP: Personal and Social Performance scale; DSM-IV-TR: Diagnostic and Statistical Manual of Mental Disorders – fourth edition, text revision; CGI-S-SCA: Clinical Global Impression of Severity for Schizoaffective Disorder; QLS: Heinrichs–Carpenter QoL scale; MSQ: Medication Satisfaction Questionnaire; PANSS: Positive and Negative Symptoms Scale; SAS: Simpson–Angus Scale; AIMS: Abnormal Involuntary Movement Scale; BARS: Barnes Akathisia Scale; GAF: Global Assessment of Function; SF-36: 36-item Short Form Health Survey; SOF: Scale of Functioning; BPRS: Brief Psychiatric Rating Scale; SANS: Scale for Assessment of Negative Symptoms; W-QOLI: Wisconsin Quality-of-Life Index; SARS: Simpson–Angus Rating Scale; DAI: Drug Attitude Inventory; Cogtest: cognitive battery test; BSI: Brief Symptom Index; ASI: Addiction Severity Index; QWBS: Quality of Well-Being Scale; SQLS-R4: Schizophrenia Quality-of-Life Scale, Revision 4; SOFAS: Social and Occupational Functioning Assessment Scale; ESRS: Extrapyramidal Symptom Rating Scale; TSQM: Treatment Satisfaction Questionnaire for Medication; IF: intrapsychic foundation; IR: instrumental roles; IPR: interpersonal roles; AQoL: Assessment of Quality of Life; mg/bw: milligramme per 2 weeks; mg/m: milligramme per month; SWN-S: Subjective Well-being under Neuroleptic Treatment, short version.

Treatment as usual; ^bOral olanzapine; ^cOral risperidone; ^dOral atypical at discretion of clinician; ^eOral aripiprazole; ^fIncluded in systematic review only; ^gOral quetiapine.

Table 3.

Characteristics of studies on LAI-A vs LAI-A.

Study	Design	Sample (N)Gender (%)	LAI-A/dosage	Functional measures	Clinical scales	Findings on psychosocial function and statistics
Koshikawa et al. (2016) ^a	RCT open-labelRatio = 1:1LAI-A vs LAI-A	N = 30 M = 52.3%	RisperidoneMax 50 mg/bwPaliperidone^b Max 150 mg/m	SFS, UPSA-B	DSM-IV-TR, PANSS, JART, SECT, Cogtest, DIEPSS	SFS: the two LAI-A groups showed improvement in scores from baseline, with LAI-P showing greater improvement in total score and two sub-scales: independence-competence and independence-performance, compared to the LAI-R group (total score: F = 5.03, df = 1, 18, p = 0.038; competence: F = 14.04, df = 1, 18, p = 0.001; and performance: F = 9.14, df = 1, 18, p = 0.007), respectively.UPSA-B: significant difference (increase) in total scores (improvement) for both LAI-A groups with greater improvement in LAI-P. Mean change (SD) LAI-R = 1.36 (13.85) vs LAI-p = 5.78 (8.79), and no significant difference in the two groups F = 1.70; p = 0.211.
Meltzer et al. (2014) ^a	RCT double-blindRatio = 1:1LAI-A vs LAI-A	N = 160 M = 72.5%	Risperidone 50 mg/bwRisperidone^b 100 mg/bw	GAF, PSP	DSM-IV-TR, PANSS, SAS, CGI-S, AIMS, BARS, Cogtest	GAF: significant change (increase) in scores from baseline (improvement) was reported in the two doses of LAI-A (p < 0.05). Effect size for improved GAF scores was 1.67 and 1.94 for the 100- and 50-mg doses, respectively.PSP: significant change (increase) in scores from baseline (improvement) was reported in the two doses of LAI-A (p < 0.05). The ES for improvement between baseline and 24 weeks was 1.29 and 1.10 for the 50- and100-mg dose groups, respectively. Improvement in function began at 6 weeks and was progressive but had plateaued by 24 weeks.
Naber et al. (2015) ^a	RCT single-blindRatio = 1:1LAI-A vs LAI-A	N = 295 M = 56.9%	Aripiprazole 400 mg/mPaliperidone^b 50–150 mg/m (flexible dose)	QLS; SWN-S TooL	DSM-IV-TR, CGI-S, IAG, SWN-S, TooL, WoRQ	QLS: significant change (increase) in scores as measured with LSM treatment differences at week 28 favouring AOM 400 vs PP in patients ⩽35 years on QLS (10.7; 95% CI: [0.70, 20.7]; p = 0.037). No difference between AOM vs PP in COA, IPR and IR (p > 0.05), but AOM was superior to PP in IF (p = 0.039).
NCT00604279 (2013)^a	RCT open-labelRatio = 1:1LAI-A vs LAI-A	N = 452M = 40%	Paliperidone50, 100, 150 mg/mRisperidone^b 25, 37.5, 50 mg/bw	PSP	DSM-IV, SVAS, PANSS, CGI-S	PSP: there was improvement in function in both LAI-A groups, mean change 16.8 (±14.76) vs 18.6 (±13.92). No between-group statistics.
Pandina et al. (2011) ^a	RCT double-blindRatio = 1:1LAI-A vs LAI-A	N = 1221 M = 55%	Paliperidone50–150 mg/bwRisperidone^b 25, 37.5, 50 mg/bw (flexible dose)	PSP	DSM-IV, PANSS, CGI-S, SDS	PSP: there was improvement in longitudinal change in score in both paliperidone, 8.5 (±11.82) and risperidone 8.8 (±11.65). No significant difference between the two drugs (p > 0.05).
Simpson et al. (2006) ^a	RCT double-blindRatio = 1:1LAI-A vs LAI-A	N = 323 M = 62.3%	Risperidone25 mg/bwRisperidone^b 50 mg/bw	LOF, PSP, SQLS	DSM-IV, AIMS, PANSS, CGI-S, ESRS, Cogtest	PSP: longitudinal change in scores (increase) was significantly different for both LAI-A doses. LAI-A-25 mg mean (SD) baseline to endpoint = 61.8 (±14.8) to 63.8 (±15.0). LAI-A-50 mg mean (SD) baseline to endpoint = 62.5 (±13.5) to 64.4 (±13.1). No difference in scores between both LAI-A treatment doses group.

RCT: randomised controlled trials; LAI-A: long-acting injectable atypical antipsychotics; UPSA-B: University of California San Diego (USCD) Performance-based Skills Assessment, Brief; DSM-IV-TR: Diagnostic and Statistical Manual of Mental Disorders – fourth edition, text revision; PANSS: Positive and Negative Symptoms Scale; SAS: Simpson–Angus Scale; JART: Japanese Adult Reading Test; SECT: Social Emotional Cognition Task; Cogtest: cognitive battery test; DIEPSS: Drug-Induced Extrapyramidal Symptoms Scale; GAF: Global Assessment of Function; AIMS: Abnormal Involuntary Movement Scale; BARS: Barnes Akathisia Scale; QLS: Heinrichs–Carpenter QoL scale; SWN-S: Subjective Well-being under Neuroleptic treatment, short version; CGI-S-SCA: Clinical Global Impression of Severity for Schizoaffective Disorder; LSM: least-squares mean; AOM: Aripiprazole once monthly; PP: paliperidone palmitate; IF: intrapsychic foundation; IR: instrumental roles; IPR: interpersonal roles; COA: common objects and activities; SVAS: Sleep Visual Analogue Scale; SDS: Schedule for Deficit Syndrome; LOF: Strauss–Carpenter Levels of Functioning; SQLS: Schizophrenia Quality-of-Life Scale; ESRS: Extrapyramidal Symptom Rating Scale; mg/bw: milligramme per 2 weeks; mg/m: milligramme per month.

Included in systematic review only; ^bComparator LAI-A.

Diagnoses of schizophrenia and schizoaffective disorder were based on the Diagnostic and Statistical Manual of Mental Disorders – fourth edition (DSM-IV) criteria in all 26 studies. Most studies recruited participants with diagnosis of schizophrenia (n = 16) compared to either schizophrenia or schizoaffective disorder (n = 9) and only schizoaffective disorder in one study (Fu et al., 2015). Most of the studies enrolled both inpatients and outpatients (n = 13) compared to studies with either outpatients (n = 11) or inpatients (n = 2) alone. In addition to DSM-IV diagnostic scales, 37 other clinical scales were used to measure predictors of psychosocial function, including assessment of severity of illness using Clinical Global Impression scale (CGI) in 24 studies, the Positive and Negative Symptom Scale (PANSS) in 20 studies and the Brief Psychiatric Rating Scale (BPRS) in 2 studies (Buckley et al., 2015; Witte et al., 2012). Extrapyramidal side-effects were assessed using both Abnormal Involuntary Movement Scale (AIMS) and Barnes Akathisia Scale (BARS) in nine studies, Simpson–Angus Rating Scale (SARS) in seven studies and Drug-Induced Extrapyramidal Symptoms Scale (DIEPSS) in one study. A cognitive test battery (Cogtest) was assessed in four studies.

In 12 studies, LAI-risperidone was compared with placebo (n = 2), oral comparators (n = 7) and other LAI-A (n = 3). LAI-paliperidone was compared with placebo in four studies, and two each with orals and other LAI-A. Two studies compared LAI-aripiprazole with placebo, while one study each compared it with oral aripiprazole and with LAI-paliperidone. One study each compared LAI-olanzapine with placebo and oral olanzapine.

A total of 17 different measures of psychosocial function were used across studies. And 13 studies each used single and multiple scales. Personal and Social performance scale (PSP) was used in 17 studies, Short Form Health Survey in 6 studies, Heinrichs–Carpenter QoL scale (QLS) in 5 studies, EuroQol in 2 studies, Global Assessment of Function (GAF) in 3 studies, University of California San Diego (USCD) Performance-based Skills Assessment, Brief (UPSA-B) and Social and Occupational Functioning Assessment Scale (SOFAS) in 1 study each. Only three studies assessed psychosocial function as primary study objective a priori. All the included studies were industry sponsored except two.

Assessment of quality of studies with risk-of-bias tool

The findings on study quality using the risk-of-bias tool are presented in Figures 2 and Supplementary S3. We included 13 double-blind RCTs studies, 11 open-labelled and 2 rater-blinder studies. All 26 studies were randomised but only 11 studies described the process of sequence generation for randomisation including 7 that used interactive web software, 3 that allocated by randomly permuted blocks and 1 study by stratified lot. All participants were aware of treatment allocation in two studies. There was no confirmation of effectiveness of blinding in any of the 26 studies. While selective reporting was low in 24 studies, it was high and unclear in 1 study each. Incomplete outcome data were high in 6 studies due to early withdrawal, drop out from relapse and lack of efficacy, and low in the remaining 20 trials. Other bias was high in three studies, low in one study and unclear in the remaining trials.

Figure 2.

Quality assessment of studies with risk-of-bias tool.

Findings on outcome

Qualitative description of effects of LAI-A across studies on psychosocial function

Qualitative findings on the effects of LAI-A on psychosocial function are presented in Tables 1 –3. All the 26 studies reported positive effects of LAI-A on longitudinal change in psychosocial function except three (Berwaerts et al., 2015; Fleischhacker et al., 2014; Hough et al., 2010). LAI-A superiority was reported in three oral-controlled and eight placebo-controlled trials. Head-to-head comparisons of LAI-A were reported in six trials and showed mixed results (Koshikawa et al., 2016; Meltzer et al., 2014; Naber et al., 2015; Pandina et al., 2011; Simpson et al., 2006). Data showing the clinical significance of LAI-A benefits on psychosocial function were reported in few placebo and oral-controlled trials. Findings varied across studies and between LAI-A and comparators. For example, in one study, patients treated with LAI-paliperidone with good functioning (PSP score > 70) increased from 57.9% at baseline to 59.0% at endpoint, whereas there was reduction in the proportion of placebo-treated subjects with good functioning from 50.6% at baseline to 41.1% at trial endpoint. The difference between LAI-A versus placebo was significant (Fu et al., 2015). In one oral-controlled trial, the proportion of patients treated with LAI-paliperidone with good functioning (PSP score > 70) increased from 6.4% at baseline to 36.7% at endpoint, while those on oral antipsychotics with good functioning increased from 6.5% to 34.6% at endpoint (Schreiner et al., 2015). Similarly, 19.2% of LAI-olanzapine-treated patients had a good level of functioning (QLS score ⩾ 84.5) at baseline, which increased to 27.5% (p < 0.05), while the figures for oral olanzapine were 14.2% and 24.5%, respectively (Ascher-Svanum et al., 2014). In terms of clinically meaningful data, Rouillon et al. (2013) reported a five-point increase in mean SOFAS score in patients treated with LAI-risperidone. At trial endpoint, while SOFAS scores continued to represent serious impairment for patients treated with oral quetiapine, the scores improved from serious impairment at baseline to moderate functional difficulty at endpoint for patients on LAI-risperidone.

Effects of LAI-A on psychosocial function compared to placebo

Figures 3 –5 show the overall effects of LAI-A compared to placebo on psychosocial function. When the nine included studies were pooled, LAI-A treatment showed significant superiority in functional measures over placebo (SMD = 0.39; 95% confidence interval [CI] = [0.32, 0.47]; p < 0.001; nine studies, n = 2862; I² = 0%). Both short-term (SMD = 0.39; 95% CI = [0.28, 0.50]; p < 0.001; five studies, n = 1430; I² = 0%) and long-term (SMD = 0.40; 95% CI = [0.29, 0.50]; p < 0.001; four studies, n = 1421; I² = 0%) studies showed consistent superiority of LAI-A over placebo with respect to improving psychosocial function.

Figure 3.

All LAI-A vs placebo studies.

Figure 4.

Short-term trials of LAI-A vs placebo.

Figure 5.

Long-term trials of LAI-A vs placebo.

Effects of LAI-A on psychosocial function compared to oral antipsychotic medications

LAI-A superiority in total score on functional outcomes was reported in three oral-controlled trials (Alphs et al., 2016; Buckley et al., 2015; Rouillon et al., 2013). Seven trials, Ascher-Svanum et al. (2014), Bai et al. (2016), Fleischhacker et al. (2014), NCT00992407, Rosenheck et al. (2011), Schreiner et al. (2015) and Wykes et al. (2013), reported no statistically significant difference in total functional improvement between LAI-A and orals. However, pooled effects of all the 10 oral-controlled studies showed that LAI-A-treated patients displayed a small but significantly greater improvement in functional outcomes compared to oral antipsychotic medications (SMD = 0.16; 95% CI = [0.01, 0.31]; p = 0.04; 10 studies, n = 3540; I² = 77%), and this small advantage was maintained in all long-term trials when the short-term trial (Bai et al., 2006) was excluded (SMD = 0.17; 95% CI = [0.01, 0.33]; p = 0.03; 9 studies, n = 3490; I² = 79%) (see Figure 6).

Figure 6.

All LAI-A vs orals studies.

Effects of LAI-A across sub-domains of the psychosocial scales

The majority of the studies did not report on sub-domains of the functional scales used. Four studies reported positive change in QLS total scores among LAI-A and also published sub-domain data (Ascher-Svanum et al., 2014; Naber et al., 2015; Rosenheck et al., 2011; Witte et al., 2012). Compared to placebo, LAI-A showed significant improvement in three domains of QLS including intrapsychic functioning (adaptive function and life satisfaction), instrumental roles (socio-occupational function) and interpersonal roles (interpersonal relationship) but no difference in common objects and activities (basic-daily function) (Witte et al., 2012). LAI-risperidone and oral olanzapine treatment showed similar improvements in the sub-domains of occupational activities, psychological well-being and symptoms/outlook based on the Wisconsin Quality-of-Life Index (W-QOLI; Keks et al., 2007). Individual studies suggested improvements for LAI-A over oral antipsychotics in the social function scale of the 36-item Short Form Health Survey (SF-36; Bai et al., 2006) and illness and activities of daily living scales of the Assessment of Quality of Life (AQoL; Wykes et al., 2013). Details are included in Tables 1 –3.

Investigation of heterogeneity and meta-regression

Heterogeneity was present in the oral-controlled trials (χ² = 39.08; df = 9; p = 0.0001; I² = 77%). All the 10 oral-controlled studies were long term except one, and open-labelled except two that were double-blind (Fleischhacker et al., 2014; Rosenheck et al., 2011). Visual inspection showed that heterogeneity effects were due to Alphs et al. (2016) that showed superiority of LAI-paliperidone over oral antipsychotic medications (SMD = 0.56; 95% CI = [0.36, 0.77]; p < 0.00001; n = 388) and two studies including Buckley et al. (2015) and Rouillon et al. (2013) that demonstrated significant superiority of LAI-risperidone over oral antipsychotic medications (SMD = 0.43; 95% CI = [0.20, 0.66]; p = 0.0002; n = 305) and (SMD = 0.35; 95% CI = [0.19, 0.51]; p < 0.0001; n = 593), respectively. In addition to open-blind design, these studies allowed flexibility in the choice and dosage of oral antipsychotic medications and included only intermittent telephone monitoring of compliance, resulting in high variability in treatment. When all four oral TAU comparator trials were excluded (Alphs et al., 2016; Rosenheck et al., 2011; Schreiner et al., 2015), heterogeneity was reduced but functional outcome still favoured LAI-A over oral atypical antipsychotic medications (SMD = 0.16; 95% CI = [0.07, 0.24]; p = 0.0003; seven studies; I² = 69%].

Meta-regression did not show trial effects of study settings (outpatient vs inpatient), trial design, year of publication, industry sponsorship, baseline symptom severity and inclusion of psychosocial function as primary outcome. Egger’s test was not significant for publication bias considering all studies (intercept = 0. 22; 95% CI = [−0.122, 0.563]; p = 0.790) and when placebo-controlled trials (intercept = 0.785; 95% CI = [0.176, 1.39]; p = 0.168) and oral-controlled trials (intercept = 0.193; 95% CI = [−0.247, 0.635]) were considered alone (details included in Supplementary S4–S7).

Predictors of change in psychosocial function

Predictors of psychosocial function are presented in Table 4. Considering the seven studies that addressed predictors, poor psychosocial function outcome was predicted by baseline factors including high general psychopathology in one study (Witte et al., 2012), high negative symptoms in four studies, (Gharabawi et al., 2007; Koshikawa et al., 2016; Pandina et al., 2013; Witte et al., 2012) and severe illness based on poor CGI-S (Peuskens et al., 2012). Furthermore, positive change in insight, longer duration of treatment, lesser disorganised thoughts and higher cognitive performance predicted better psychosocial function (Gharabawi et al., 2007; Pandina et al., 2013). Witte et al. (2012) reported that high baseline general psychopathology predicted poorer instrumental role, intrapsychic foundations, common objects and activities, interpersonal relations and overall function. More severe negative symptoms predicted poorer recreational and prosocial functioning with medium effect size (Koshikawa et al., 2016). In the study of Gharabawi et al. (2007), a combination of predictive factors explained larger variability (27%) in psychosocial function compared to the 2.1% explained by insight, 10.3% by negative symptoms and 1.6% explained by duration of treatment alone.

Table 4.

Predictors of change in functioning in LAI-A trials.

Study	Predictors/correlates of change in functional outcome in LAI-A	Statistics (mean difference, regression coefficient and effect size)
Gharabawi et al. (2007)	Predictors of change in PSP scores Change in insight (PANSS item G12)Change in negative factors (PANSS negative factor)Duration of treatment	Contribution of predictors to variability in PSP scores Partial r² for insight = 0.0224Partial r² for negative symptoms = 0.1689Partial r² for duration of treatment (months) = 0.0619R² for model = 0.2742
	Predictors of change in LOF scores Change in insight (PANSS item G12)Change in negative factors (PANSS negative factor)Duration of treatment	Contribution of predictors to Variability in LOF scores Partial r² for insight = 0.0021Partial r² for negative symptoms = 0.1034Partial r² for duration of treatment (months) = 0.0161R² for model = 0.1502
	Correlates of change in SQLS scores Change in insight correlated significantly with change in SQLS psychosocial sub-scale	Correlation with SQLS Insight with SQLS psychosocial sub-scale r = 0.149
Isitt et al. (2016)	Dose of LAI-A was linked with improvement as those on higher dose reported significant improvement	nr
Koshikawa et al. (2016)	Correlates of change in SFS scores PANSS negative symptomsSignificant negative correlations were found between negative symptoms assessed with PANSS and two sub-scales of the SFS (i.e. recreation p = 0.036 and prosocial p = 0.028)	Correlation coefficient (r) of PANSS negative score vs recreation: r = −0.46prosocial: r = −0.48
Pandina et al. (2013)	Correlates of change in PSP scores Higher endpoint LSM in NCS T-scores was associated with better functioning (p = 0.006)Lower mean endpoint scores (i.e. improved) on PANSS total score, disorganised thoughts, and positive and negative factors were associated with better function (all, p < 0.0001)	nr
Peuskens et al. (2012)	Predictors of change (improved) in QLS at 6 months Higher PANSS positive scores at baselineLow CGI-S scores at baseline	Regression coefficient, OR at 6 months PANSS positive (OR = 2.840; 95% CI = [1.834, 4.399]; p < 0.0001)CGI-S (OR = 0.619; 95% CI = [0.463, 0.828]; p = 0.0012)
	Predictors of change (improved) in QLS at 12 months High PANSS positive scores at baseline	Regression coefficient, OR at 12 months PANSS positive (OR = 1.537; 95% CI = [1.129, 2.093]; p = 0.0063)
Potkin et al. (2017)	Predictors of change (improvement) in QLS at 24 months Type of LAI-A performed better in people aged below 35 years	The odds of improved function was greater in LAI-aripiprazole compared to LAI-paliperidone in participants <35 years (OR = 10.7; 95% CI: [0.70, 20.7]; p = 0.037) but there was no difference in patients >35 years
Witte et al. (2012)	Correlates of change in QLS scores Changes in PANSS total, PANSS negative and PANSS general psychopathology scores showed significant negative correlations with QLS scores (p < 0.0001)	Correlation coefficient (r) of PANSS total score with QLS total = −0·44QLS instrumental role = −0.23QLS intrapsychic foundations = −0.43QLS common objects and activities = −0.28QLS interpersonal relations = −0.37

LAI-A: long-acting injectable atypical antipsychotics; PSP: Personal and Social Performance scale; PANSS: Positive and Negative Symptoms Scale; LOF: Strauss–Carpenter Levels of Functioning; SQLS: Schizophrenia Quality-of-Life Scale; nr: not reported; LSM: least-squares mean; NCS: cognitive composite scores; CGI-S: Clinical Global Impression – Severity; OR: odds ratio; CI: confidence interval; QLS: Heinrichs–Carpenter QoL scale; r: correlation coefficient.

Discussion

To our knowledge, this systematic review and meta-analyses is the first to analyse the effects and predictors of LAI-A on psychosocial function compared to placebo or oral antipsychotics. We were able to review evidence from 26 studies, including 11,097 participants treated with second-generation LAIs where psychosocial function was indexed by a wide range of measures including QoL (Ro and Clark, 2009).

We found that LAI-A were superior to placebo in improving psychosocial function with a medium effect size of SMD = 0.39, consistent over short- and long-term trials. LAI-A also demonstrated a small but significant benefit (SMD = 0.16) compared to oral antipsychotic medications for functional outcome in long-term clinical trials. While there was clear evidence of clinically meaningful benefit for LAI-A treatment over placebo, where reported absolute differences between oral and LAI-A treatment in terms of percentage of patients achieving good function were small, this benefit is smaller than that reported in mirror-image and some large cohort studies, potentially due to higher levels of oral adherence and under representation of non-compliant patients in selective and closely monitored clinical trials (Kane et al., 2013; Tiihonen et al., 2011; Zhornitsky and Stip, 2012). Thus, there is a need for true effectiveness trials, in which post-randomisation involvement would be kept to a minimum to better reflect routine practice (Kane et al., 2013; Kishimoto et al., 2014). Non-adherence with oral antipsychotic medication is associated with high rates of relapse and poor functional outcomes (Leucht et al., 2011). Adherence is affected by many factors including insight, side-effects, patient treatment goals, attitude to treatment and the effectiveness of the treatment itself (Kikkert et al., 2006). Depot treatments can reduce day-to-day fluctuations in compliance that occur with oral dosing and allow clear monitoring of adherence, leading to early psychoeducation and intervention.

Systematic review of predictors in seven studies showed that patients with more severe symptoms (Ascher-Svanum et al., 2014; Koshikawa et al., 2016; Simpson et al., 2006; Witte et al., 2012), cognitive impairment (Pandina et al., 2013) and poor insight (Simpson et al., 2006) at decision to trial LAI-A were less likely to improve in psychosocial function. Our findings are consistent with the existing evidence of a clear association between cognitive impairment, severity of psychopathology and poor insight with poor functional recovery (Bowie and Harvey, 2006). This association is complex but modifiable, such that adherence, less severe psychopathology and functional recovery are associated with better insight, while insight is cognition-dependent, particularly on executive performance and working memory (Aleman et al., 2006; Donohoe et al., 2005; Rajji et al., 2014; Rossell et al., 2003). Evidence suggests that insight (Pijnenborg et al., 2013) and cognition can be improved by a wide range of interventions including cognitive behavioural therapy, assertive community treatment and cognitive remediation. These interventions are recommended for the promotion of functional recovery and in combination with pharmacotherapy lead to the best psychosocial outcomes (Galletly et al., 2016; Mueser et al., 2013; Pfammatter et al., 2006). Further research is required to predict poor functional outcomes and fully personalise the selection of medications and interventions (Jaaskelainen et al., 2013). More fundamentally, there is a need to develop better understanding of the pathophysiology of functional decline and studies of the biological basis of function are ongoing (Clark et al., 2018; Wojtalik et al., 2017).

On a different note, our findings underscore the impact of antipsychotics on overall outcome, and by extension support the benefits of maintenance antipsychotics in patients with schizophrenia. Several studies have reported decline in functional recovery and increased relapse rates with cessation of maintenance antipsychotic treatment (Thompson et al., 2018). Notwithstanding the small advantage of LAI-A over oral antipsychotic medications on psychosocial function, careful interpretation of this finding is required. The choice of antipsychotic formulation for individual patient should entail a balanced consideration of clinical factors, patient attributes and evidence from clinical guidelines through a shared decision process (Correll et al., 2016).

There are several limitations to this study. First, in all but two included studies, psychosocial function was not the primary outcome measure. While some trials reported statistical corrections to address potential effects of this limitation on findings, the quality of evidence would be much improved if cofounding effects and study power were considered in relation to psychosocial function a priori (Freemantle, 2001). Given the increasing burden of mental illness-related disability and the emerging understanding that symptomatic improvement is not prerequisite nor necessarily sufficient for improved function, there is a need to consider functional measures as primary outcomes in studies of antipsychotic efficacy (GBD 2016 DALYs and HALE Collaborators, 2016; Whiteford, 2017). In keeping with a broad composite concept of psychosocial function, 50% of the studies reviewed considered both QoL and general functional measures. Standardisation of these approaches will make future trials of depot efficacy more comparable.

Similar to previous systematic reviews of the literature, we found poor reporting of the randomisation process, concealment and sequence generation was common (Leucht et al., 2011). We used a random-effects model to address heterogeneity in psychosocial function scales and grouped trials into clinical, meaningful placebo-controlled versus oral-controlled groups during analyses and sensitivity analyses. The oral-controlled data showed heterogeneity of 77%, accounted for by three studies that showed superiority of LAI-paliperidone and LAI-risperidone over orals (Alphs et al., 2016; Buckley et al., 2015; Rouillon et al., 2013). The oral arms of these trials were not well standardised or monitored. For example, LAI-A were not necessarily compared to the same atypical oral; however, recommended dosing was generally implemented. The outcome of meta-analysis was not changed when studies using a TAU oral comparator were excluded.

Only two non-industry trials could be included suggesting risk of industry-sponsored bias (Buckley et al., 2015; Koshikawa et al., 2016). Studies varied in active comparators and drug dosage; however, all trials used doses within the recommended ranges. Trial length did not appear to influence functional outcome in depot over placebo treatment, suggesting a ceiling effect as optimal function was met within the first 3 months (Carter, 2006; Swartz et al., 2007). Most of the trials did not look at predictors of psychosocial function and useable data were poorly reported.

Conclusion

We conclude that LAI-A improve psychosocial function, but in clinical trials, the magnitude of effect is small when compared to oral antipsychotic medications. Patients with more severe symptoms and cognitive impairment at decision to trial were less likely to improve in psychosocial function and may benefit from more intensive psychosocial therapies in adjunct to pharmacological treatment. Consideration of baseline attributes at diagnosis could help to stratify patients who may benefit from early adjunct interventions. Further analysis of predictors of long-term function is needed, potentially leading to trials of indicated psychosocial intervention. Future clinical trials should improve on the quality of reporting and design, and include composite psychosocial function as a measure of efficacy a priori.

Supplemental Material

Revised_supplmentary_material_for_ANZJPF31012019 – Supplemental material for Long-acting atypical antipsychotics in schizophrenia: A systematic review and meta-analyses of effects on functional outcome

Supplemental material, Revised_supplmentary_material_for_ANZJPF31012019 for Long-acting atypical antipsychotics in schizophrenia: A systematic review and meta-analyses of effects on functional outcome by Andrew T Olagunju, Scott R Clark and Bernhard T Baune in Australian & New Zealand Journal of Psychiatry

Footnotes

Acknowledgements

The authors would like to thank Suzanne Edwards for providing statistical advice and support. This research was supported by an Australian Government Research Training Programme (RTP) Scholarship. The abstract was presented in the New Investigator Award Programme at the 2018 Annual Meeting of the American Society of Clinical Psychopharmacology, Miami, Florida.

Author contribution

All authors were involved in the study design and interpretation of results. Article selections were independently performed by A.T.O. Data extraction was completed by A.T.O. independently using a piloted form. Clarification was resolved by consultation with the senior author (B.T.B.). A.T.O. and S.R.C. drafted the manuscript with supervision from B.T.B. All authors participated in critical revision of manuscript drafts and approved the final version.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

ORCID iD

Andrew T Olagunju

Supplemental material

Supplemental material for this article is available online.

References

Aleman

Agrawal

Morgan

et al . (2006) Insight in psychosis and neuropsychological function: Meta-analysis. British Journal of Psychiatry 189: 204–212.

Alphs

Mao

Lynn

et al . (2016) A pragmatic analysis comparing once-monthly paliperidone palmitate versus daily oral antipsychotic treatment in patients with schizophrenia. Schizophrenia Research 170: 259–264.

Alvarez-Jimenez

Priede

Hetrick

et al . (2012) Risk factors for relapse following treatment for first episode psychosis: A systematic review and meta-analysis of longitudinal studies. Schizophrenia Research 139: 116–128.

Ascher-Svanum

Novick

Haro

et al . (2014) Long-term functional improvements in the 2-year treatment of schizophrenia outpatients with olanzapine long-acting injection. Neuropsychiatric Disease and Treatment 10: 1125–1131.

Bai

Chen

et al . (2006) A comparative efficacy and safety study of long-acting risperidone injection and risperidone oral tablets among hospitalized patients: 12-week randomized, single-blind study. Pharmacopsychiatry 39: 135–141.

Berwaerts

Liu

Gopal

et al . (2015) Efficacy and safety of the 3-month formulation of paliperidone palmitate vs placebo for relapse prevention of schizophrenia: A randomized clinical trial. JAMA Psychiatry 72: 830–839.

Borenstein

Hedges

Higgins

JPT

et al . (2009) Meta-Regression in Introduction to Meta-Analysis. Chichester: John Wiley & Sons.

Bowie

Harvey

(2006) Cognitive deficits and functional outcome in schizophrenia. Neuropsychiatric Disease and Treatment 2: 531–536.

Brissos

Veguilla

Taylor

et al . (2014) The role of long-acting injectable antipsychotics in schizophrenia: A critical appraisal. Therapeutic Advances in Psychopharmacology 4: 198–219.

10.

Buckley

Schooler

Goff

et al . (2015) Comparison of SGA oral medications and a long-acting injectable SGA: The PROACTIVE study. Schizophrenia Bulletin 41: 449–459.

11.

Burns

(2007) Hospitalisation as an outcome measure in schizophrenia. British Journal of Psychiatry 191: S37–S41.

12.

Carter

(2006) Understanding the glass ceiling for functional outcome in schizophrenia. The American Journal of Psychiatry 163: 356–358.

13.

Clark

Schubert

Olagunju

et al . (2018) Trajectories in response to Aripiprazole maintenance study: Six-months follow-up. Australian and New Zealand Journal of Psychiatry 52: 140–141.

14.

Cochrane Collaboration (2014) Review Manager (RevMan) [Computer program], version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration.

15.

Correll

Citrome

Haddad

et al . (2016) The use of long-acting injectable antipsychotics in schizophrenia: Evaluating the evidence. Journal of Clinical Psychiatry 77(Suppl. 3): 1–24.

16.

Czobor

Van Dorn

Citrome

et al . (2015) Treatment adherence in schizophrenia: A patient-level meta-analysis of combined CATIE and EUFEST studies. European Neuropsychopharmacology 25: 1158–1166.

17.

Dixon

Dickerson

Bellack

et al . (2010) The 2009 schizophrenia PORT psychosocial treatment recommendations and summary statements. Schizophrenia Bulletin 36: 48–70.

18.

Donohoe

Corvin

Robertson

(2005) Are the cognitive deficits associated with impaired insight in schizophrenia specific to executive task performance? Journal of Nervous and Mental Disease 193: 803–808.

19.

Emsley

Chiliza

Asmal

(2013) The evidence for illness progression after relapse in schizophrenia. Schizophrenia Research 148: 117–121.

20.

Ereshefsky

Mascarenas

(2003) Comparison of the effects of different routes of antipsychotic administration on pharmacokinetics and pharmacodynamics. Journal of Clinical Psychiatry 64(Suppl. 16): 18–23.

21.

Fleischhacker

Baker

Eramo

et al . (2014) Effects of aripiprazole once-monthly on domains of personal and social performance: Results from 2 multicenter, randomized, double-blind studies. Schizophrenia Research 159: 415–420.

22.

Freemantle

(2001) Interpreting the results of secondary end points and subgroup analyses in clinical trials: Should we lock the crazy aunt in the attic? BMJ 322: 989–991.

23.

Turkoz

Simonson

et al . (2015) Paliperidone palmitate once-monthly reduces risk of relapse of psychotic, depressive, and manic symptoms and maintains functioning in a double-blind, randomized study of schizoaffective disorder. Journal of Clinical Psychiatry 76: 253–262.

24.

Galletly

Castle

Dark

et al . (2016) Royal Australian and New Zealand College of Psychiatrists clinical practice guidelines for the management of schizophrenia and related disorders. Australian and New Zealand Journal of Psychiatry 50: 410–472.

25.

GBD 2016 DALYs and HALE Collaborators (2016) Global, regional, and national disability-adjusted life-years (DALYs) for 333 diseases and injuries and healthy life expectancy (HALE) for 195 countries and territories, 1990-2016: A systematic analysis for the Global Burden of Disease Study 2016. Lancet 390: 1260–1344.

26.

GBD 2016 Disease and Injury Incidence and Prevalence Collaborators (2016) Global, regional, and national incidence, prevalence, and years lived with disability for 328 diseases and injuries for 195 countries, 1990-2016: A systematic analysis for the Global Burden of Disease Study 2016. Lancet 390: 1211–1259.

27.

Gharabawi

Bossie

Turkoz

et al . (2007) The impact of insight on functioning in patients with schizophrenia or schizoaffective disorder receiving risperidone long-acting injectable. Journal of Nervous and Mental Disease 195: 976–982.

28.

Harvey

(2013) Assessment of everyday functioning in schizophrenia: Implications for treatments aimed at negative symptoms. Schizophrenia Research 150: 353–355.

29.

Higgins

Thompson

Deeks

et al . (2003) Measuring inconsistency in meta-analyses. BMJ 327: 557–560.

30.

Higgins

JPT

Green

(ed.) (2011) Cochrane Handbook for Systematic Reviews of Interventions, version 5.1.0. London: Cochrane.

31.

Hough

Gopal

Vijapurkar

et al . (2010) Paliperidone palmitate maintenance treatment in delaying the time-to-relapse in patients with schizophrenia: A randomized, double-blind, placebo-controlled study. Schizophrenia Research 116: 107–117.

32.

Isitt

Nadipelli

Kouassi

et al . (2016) Health-related quality of life in acute schizophrenia patients treated with RBP-7000 once monthly risperidone: An 8-week, randomized, double-blind, placebo-controlled, multicenter phase 3 study. Schizophrenia Research 174: 126–131.

33.

Jaaskelainen

Juola

Hirvonen

et al . (2013) A systematic review and meta-analysis of recovery in schizophrenia. Schizophrenia Bulletin 39: 1296–1306.

34.

Kane

Eerdekens

Lindenmayer

et al . (2003) Long-acting injectable risperidone: Efficacy and safety of the first long-acting atypical antipsychotic. American Journal of Psychiatry 160: 1125–1132.

35.

Kane

Kishimoto

Correll

(2013) The comparative effectiveness of long-acting injectable vs. oral antipsychotic medications in the prevention of relapse: A case study in CER in psychiatry. Journal of Clinical Epidemiology 66: S37–S41.

36.

Kane

Peters-Strickland

Baker

et al . (2014) Aripiprazole once-monthly in the acute treatment of schizophrenia: Findings from a 12-week, randomized, double-blind, placebo-controlled study. Journal of Clinical Psychiatry 75: 1254–1260.

37.

Kane

Sanchez

Baker

et al . (2015) Patient-centered outcomes with aripiprazole once-monthly for maintenance treatment in patients with schizophrenia: Results from two multicenter, randomized, double-blind studies. Clinical Schizophrenia & Related Psychoses 9: 79–87.

38.

Kaplan

Casoy

Zummo

(2013) Impact of long-acting injectable antipsychotics on medication adherence and clinical, functional, and economic outcomes of schizophrenia. Patient Preference and Adherence 7: 1171–1180.

39.

Keks

Ingham

Khan

et al . (2007) Long-acting injectable risperidone v. olanzapine tablets for schizophrenia or schizoaffective disorder: Randomised, controlled, open-label study. British Journal of Psychiatry 191: 131–139.

40.

Kikkert

Schene

Koeter

MWJ

et al . (2006) Medication adherence in schizophrenia: Exploring patients’, carers’ and professionals’ views. Schizophrenia Bulletin 32: 786–794.

41.

Kim

(2017) Neuroprotection in schizophrenia and its therapeutic implications. Psychiatry Investigation 14: 383–391.

42.

Kishimoto

Robenzadeh

Leucht

(2014) Long-acting injectable vs oral antipsychotics for relapse prevention in schizophrenia: A meta-analysis of randomized trials. Schizophrenia Bulletin 40: 192–213.

43.

Koshikawa

Takekita

Kato

et al . (2016) The comparative effects of risperidone long-acting injection and paliperidone palmitate on social functioning in schizophrenia: A 6-month, open-label, randomized controlled pilot trial. Neuropsychobiology 73: 35–42.

44.

Lafeuille

Laliberte-Auger

Lefebvre

et al . (2013) Impact of atypical long-acting injectable versus oral antipsychotics on rehospitalization rates and emergency room visits among relapsed schizophrenia patients: A retrospective database analysis. BMC Psychiatry 13: 221.

45.

Lehman

Lieberman

Dixon

(2004) Practice guideline for the treatment of patients with schizophrenia, second edition. American Journal of Psychiatry 161: 1–56.

46.

Leifker

Patterson

Bowie

et al . (2010) Psychometric properties of performance-based measurements of functional capacity: Test-retest reliability, practice effects, and potential sensitivity to change. Schizophrenia Research 119: 246–252.

47.

Leucht

Heres

Kane

et al . (2011) Oral versus depot antipsychotic drugs schizophrenia – A critical systematic review and meta-analysis of randomised long-term trials. Schizophrenia Research 127: 83–92.

48.

Leucht

Tardy

Komossa

et al . (2012) Antipsychotic drugs versus placebo for relapse prevention in schizophrenia: A systematic review and meta-analysis. Lancet 379: 2063–2071.

49.

Liberati

Altman

Tetzlaff

et al . (2009) The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: Explanation and elaboration. Annals of Internal Medicine 151: W65–W94.

50.

Lieberman

Stroup

McEvoy

et al . (2005) Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. New England Journal of Medicine 353: 1209–1223.

51.

Liu

Hunter

et al . (2008) Performing meta-analysis with incomplete statistical information in clinical trials. BMC Medical Research Methodology 8: 56.

52.

Maia-de-Oliveira

Bressan

Elkis

(2013) Why we should use long-acting injectable antipsychotics more frequently. Revista Brasileira de Psiquiatria 35: 217–218.

53.

Mannaert

Vermeulen

Remmerie

et al . (2005) Pharmacokinetic profile of long-acting injectable risperidone at steady-state: Comparison with oral administration. Encephale 31: 609–615.

54.

Marshall

Lockwood

Bradley

et al . (2000) Unpublished rating scales: A major source of bias in randomised controlled trials of treatments for schizophrenia. British Journal of Psychiatry 176: 249–252.

55.

Meltzer

Lindenmayer

Kwentus

et al . (2014) A six month randomized controlled trial of long acting injectable risperidone 50 and 100mg in treatment resistant schizophrenia. Schizophrenia Research 154: 14–22.

56.

Menezes

Arenovich

Zipursky

(2006) A systematic review of longitudinal outcome studies of first-episode psychosis. Psychological Medicine 36: 1349–1362.

57.

Miyamoto

Fleischhacker

(2017) The use of long-acting injectable antipsychotics in schizophrenia. Current Treatment Options in Psychiatry 4: 117–126.

58.

Moher

Liberati

Tetzlaff

et al . (2009) Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Annals of Internal Medicine 151: 264–269W264.

59.

Mueser

McGurk

(2004) Schizophrenia. Lancet 363: 2063–2072.

60.

Mueser

Deavers

Penn

et al . (2013) Psychosocial treatments for schizophrenia. Annual Review of Clinical Psychology 9: 465–497.

61.

Naber

Hansen

Forray

et al . (2015) Qualify: A randomized head-to-head study of aripiprazole once-monthly and paliperidone palmitate in the treatment of schizophrenia. Schizophrenia Research 168: 498–504.

62.

Nasrallah

Duchesne

Mehnert

et al . (2004) Health-related quality of life in patients with schizophrenia during treatment with long-acting, injectable risperidone. Journal of Clinical Psychiatry 65: 531–536.

63.

Novick

Haro

Suarez

et al . (2010) Predictors and clinical consequences of non-adherence with antipsychotic medication in the outpatient treatment of schizophrenia. Psychiatry Research 176: 109–113.

64.

Olagunju

Clark

Baune

(2018) Clozapine and psychosocial function in schizophrenia: A systematic review and meta-analysis. CNS Drugs 32: 111011–111023.

65.

Olivares

Sermon

Hemels

et al . (2013) Definitions and drivers of relapse in patients with schizophrenia: A systematic literature review. Annals of General Psychiatry 12: 32.

66.

Ostuzzi

Bighelli

et al . (2017) Does formulation matter? A systematic review and meta-analysis of oral versus long-acting antipsychotic studies. Schizophrenia Research 183: 10–21.

67.

Pandina

Bilder

Turkoz

et al . (2013) Identification of clinically meaningful relationships among cognition, functionality, and symptoms in subjects with schizophrenia or schizoaffective disorder. Schizophrenia Research 143: 312–318.

68.

Pandina

Lane

Gopal

et al . (2011) A double-blind study of paliperidone palmitate and risperidone long-acting injectable in adults with schizophrenia. Progress in Neuro-Psychopharmacology & Biological Psychiatry 35: 218–226.

69.

Pandina

Lindenmayer

Lull

et al . (2010) A randomized, placebo-controlled study to assess the efficacy and safety of 3 doses of paliperidone palmitate in adults with acutely exacerbated schizophrenia. Journal of Clinical Psychopharmacology 30: 235–244.

70.

Patel

David

(2005) Why aren’t depot antipsychotics prescribed more often and what can be done about it? Advances Psychiatric Treatment 11: 203–213.

71.

Peuskens

Porsdal

Pecenak

et al . (2012) Schizophrenia symptoms and functioning in patients receiving long-term treatment with olanzapine long-acting injection formulation: A pooled analysis. BMC Psychiatry 12: 1301–1311.

72.

Pfammatter

Junghan

Brenner

(2006) Efficacy of psychological therapy in schizophrenia: Conclusions from meta-analyses. Schizophrenia Bulletin 32: S64–S80.

73.

Pijnenborg

van Donkersgoed

David

et al . (2013) Changes in insight during treatment for psychotic disorders: A meta-analysis. Schizophrenia Research 144: 109–117.

74.

Potkin

Loze

Forray

et al . (2017) Multidimensional assessment of functional outcomes in schizophrenia: Results from qualify, a head-to-head trial of aripiprazole once-monthly and paliperidone palmitate. International Journal of Neuropsychopharmacology 20.

75.

Rajji

Miranda

Mulsant

(2014) Cognition, function, and disability in patients with schizophrenia: A review of longitudinal studies. Canadian Journal of Psychiatry 59: 13–17.

76.

Clark

(2009) Psychosocial functioning in the context of diagnosis: Assessment and theoretical issues. Psychological Assessment 21: 313–324.

77.

Rosenheck

Krystal

Lew

et al . (2011) Long-acting risperidone and oral antipsychotics in unstable schizophrenia. New England Journal of Medicine 364: 842–851.

78.

Rossell

Coakes

Shapleske

et al . (2003) Insight: its relationship with cognitive function, brain volume and symptoms in schizophrenia. Psychological Medicine 33: 111–119.

79.

Rouillon

Eriksson

Burba

et al . (2013) Functional recovery results from the risperidone long-acting injectable versus quetiapine relapse prevention trial (ConstaTRE). Acta Neuropsychiatrica 25: 297–306.

80.

Schreiner

Aadamsoo

Altamura

et al . (2015) Paliperidone palmitate versus oral antipsychotics in recently diagnosed schizophrenia. Schizophrenia Research 169: 393–399.

81.

Simpson

Mahmoud

Lasser

et al . (2006) A 1-year double-blind study of 2 doses of long-acting risperidone in stable patients with schizophrenia or schizoaffective disorder. Journal of Clinical Psychiatry 67: 1194–1203.

82.

StataCorp (2017) Stata Statistical Software: Release 15. College Station, TX: StataCorp LLC.

83.

Swartz

Perkins

Stroup

et al . (2007) Effects of antipsychotic medications on psychosocial functioning in patients with chronic schizophrenia: Findings from the NIMH CATIE study. American Journal of Psychiatry 164: 428–436.

84.

Thompson

Winsper

Marwaha

et al . (2018) Maintenance antipsychotic treatment versus discontinuation strategies following remission from first episode psychosis: Systematic review. BJPsych Open 4: 215–225.

85.

Tiihonen

Haukka

Taylor

et al . (2011) A nationwide cohort study of oral and depot antipsychotics after first hospitalization for schizophrenia. American Journal of Psychiatry 168: 603–609.

86.

Tiihonen

Mittendorfer-Rutz

Majak

et al . (2017) Real-world effectiveness of antipsychotic treatments in a nationwide cohort of 29823 patients with schizophrenia. JAMA Psychiatry 74: 686–693.

87.

Valenstein

Ganoczy

McCarthy

(2006) Antipsychotic adherence over time among patients receiving treatment for schizophrenia: A retrospective review. Journal of Clinical Psychiatry 67: 1542–1550.

88.

Van Os

Kapur

(2009) Schizophrenia. Lancet 374: 635–645.

89.

Wahlbeck

Cheine

Essali

et al . (1999) Evidence of clozapine’s effectiveness in schizophrenia: A systematic review and meta-analysis of randomized trials. American Journal of Psychiatry 156: 990–999.

90.

Whiteford

(2017) The disparity between changes in the prevalence of mental illness and disability support rates in Australia. Medical Journal of Australia 206: 486–487.

91.

Witte

Case

Schuh

et al . (2012) Effects of olanzapine long-acting injection on levels of functioning among acutely ill patients with schizophrenia. Current Medical Research and Opinion 283: 315–323.

92.

Wojtalik

Smith

Keshavan

et al . (2017) A systematic and meta-analytic review of neural correlates of functional outcome in schizophrenia. Schizophrenia Bulletin 43: 1329–1347.

93.

Wyatt

(1991) Neuroleptics and the natural course of schizophrenia. Schizophrenia Bulletin 17: 325–351.

94.

Wykes

Rose

Williams

et al . (2013) Working alliance and its relationship to outcomes in a randomized controlled trial (RCT) of antipsychotic medication. BMC Psychiatry 13: 28.

95.

Zhornitsky

Stip

(2012) Oral versus long-acting injectable antipsychotics in the treatment of schizophrenia and special populations at risk for treatment nonadherence: A systematic review. Schizophrenia Research and Treatment 2012: 407171.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.21 MB