Background

People with schizophrenia often experience long-term psychosocial disabilities and frequent relapse. Family plays a key role in caring for ill relatives, which in turn probably contributes to high levels of distress and burdens for the family. Family-based interventions have been developed and applied to family members and their relatives with schizophrenia to improve their outcomes. This is an update of a Cochrane review that was last updated in 2011, which has been split into this review, one on group- versus individual-based family interventions and one on family-based cognitive versus behavioural management interventions.

Objectives

To assess the effects of family-based interventions for people with schizophrenia or schizophrenia-like disorders and their families compared with standard care.

Search methods

We searched the following electronic databases from inception until April 2023: CENTRAL, Medline, Embase, PsycInfo, CINAHL, WHO International Clinical Trials Registry Platform (ICTRP), Clinicaltrials.gov, SinoMed, China Network Knowledge Infrastructure (CNKI), Wanfang, and Chinese Scientific Journals Database (VIP). We also searched the reference lists of included studies and accessible reviews for additional references.

Selection criteria

We included randomised controlled trials (RCTs) that compared the effects of family-based interventions for people with schizophrenia or schizophrenia-like disorders and their families and reported at least one patient’s and one family member’s outcomes. In this update, we only investigated standard care as the comparator.

Data collection and analysis

We used standard Cochrane methods. The review authors independently screened studies, extracted data, and assessed risk of bias for each study using the Cochrane risk of bias tool for RCTs. We pooled data and estimated effects with the mean difference (MD), standardised mean difference (SMD), or risk ratio (RR) with 95% confidence interval (CI). We judged the certainty of evidence using GRADEpro GDT. We divided the outcomes into short-term (≤1 month postintervention), medium-term (>1–6 months postintervention), and long-term follow-up (>6 months postintervention), if available.

Main results

We identified 26 RCTs in this review, with 1985 people with schizophrenia or schizophrenia-like disorders, and 2056 family members. Most family-based interventions were conducted on a weekly or biweekly basis, with duration ranging from five weeks to two years. We had substantial concerns regarding the methodological quality of the included studies given that we judged all studies at high risk of performance bias and several studies at high risk of detection, attrition or reporting bias.

Low-certainty evidence indicated that family-based interventions may reduce patients’ relapse at one month or less postintervention (RR 0.66, 95% CI 0.49–0.89; 4 RCTs, 229 participants). We downgraded the evidence by two levels due to imprecision (small number of participants) and high risk of performance, detection and attrition bias.

Compared to standard care, family-based interventions probably reduce caregiver burden at one month or less postintervention (MD −5.84, 95% CI −6.77 to −4.92; 8 RCTs, 563 participants; moderate-certainty evidence) and may result in more family members shifting from high to low expressed emotion (RR 3.90, 95% CI 1.11 to 13.71; 2 RCTs, 72 participants; low-certainty evidence).

Family interventions may result in little to no difference in patients’ death (RR 0.48, 95% CI 0.18 to 1.32; 6 RCTs, 304 participants; low-certainty evidence) and hospital admission (≤1 month postintervention; RR 0.81, 95% CI 0.51 to 1.29; 2 RCTs, 153 participants; low-certainty evidence) in comparison with standard care.

Due to the heterogeneous measures and various follow-up periods, we were unable to provide pooled effect estimates for patients’ compliance with medication and quality of life. We were very uncertain whether family interventions resulted in enhanced compliance with medication and improved quality of life for patients. We downgraded the evidence to very low certainty due to high risk of bias across studies, inconsistency (different directions of effects across studies), and imprecision (small number of participants or CIs of most studies including the possibility of no effect).