Comparing interventions for early psychosis: a systematic review and component network meta-analysis

Abstract

Background: 'Early Intervention in Psychosis' (EIP) services have been associated with improved outcomes for early psychosis. However, these services are heterogeneous and many provide different components of treatment. The impact of this variation on the sustained treatment effects is unknown.

Methods: We performed a systematic review and component network meta-analysis (cNMA) of randomised controlled trials (RCTs) that compared specialised intervention services for early psychosis. We searched CENTRAL (published and unpublished), EMBASE, MEDLINE, CINAHL, PsycINFO and Web of Science from inception to February 2023. Primary outcomes were negative and positive psychotic symptoms at 3-month and 1-year follow-up and treatment dropouts. Secondary outcomes were depressive symptoms and social functioning at 1-year follow-up. We registered a protocol for our study in PROSPERO (CRD42017057420).

Findings: We identified 37 RCTs including 4599 participants. Participants' mean age was 25.8 years (SD 6.0) and 64.0% were men. We found evidence that psychological interventions (this component grouped all psychological treatment intended to treat, or ameliorate the consequences of, psychotic symptoms) are beneficial for reducing negative symptoms (iSMD -0.24, 95% CI -0.44 to -0.05, p = 0.014) at 3-month follow-up and may be associated with clinically relevant benefits in improving social functioning scores at 1-year follow-up (iSMD -0.52, 95% CI -1.05 to 0.01, p = 0.052). The addition of case management has a beneficial effect on reducing negative symptoms (iSMD -1.17, 95% CI -2.24 to -0.11, p = 0.030) and positive symptoms (iSMD -1.05, 95% CI -2.02 to -0.08, p = 0.033) at 1-year follow-up. Pharmacotherapy was present in all trial arms, meaning it was not possible to examine the specific effects of this component.

Interpretation: Our findings suggest psychological interventions and case management in addition to pharmacotherapy as the core components of services for early psychosis to achieve sustained clinical benefits. Our conclusions are limited by the small number of studies and sparsely connected networks.

Funding: National Institute for Health and Care Research.

Keywords: Component network meta-analysis; Early Intervention in Psychosis; First episode psychosis; Mental health services; Schizophrenia.

Fig. 1

Prisma flow diagram.

Fig. 2

Network structure for primary outcomes. Network structure for the 5 primary outcomes examined in this article. Nodes represent combinations of components, and lines denote trials performing the corresponding comparison. The width of the lines is proportional to the number of trials comparing each pair of treatments. The size of the nodes is proportional to the number of randomised participants.

Fig. 3

Forest plot of component network meta-analysis for primary and secondary outcomes. This plot shows the estimated component-specific incremental standardised mean differences of adding the row-defining component to an EIP package (including pharmacotherapy as standard), for each outcome. For dropouts, incremental risk ratios were converted to incremental standardised mean differences for plotting purposes. It was not possible to examine the effect of family interventions or social interventions for the outcome ‘depressive symptoms’ due to insufficient studies to isolate these components.