Introduction
Relapse is highly prevalent in people with severe mental illness (SMI). Identifying early warning signs (EWS) indicative of worsening is important to prevent full-blown relapses. While SMI is traditionally classified using pre-determined criteria, recent transdiagnostic approaches acknowledge that SMIs share underlying processes and symptoms. Identifying transdiagnostic relapse predictors may help build effective relapse prevention tools across SMIs.


Purpose
In this scoping review, we aim to identify time-varying factors that may serve as EWS of relapse in major depressive disorder (MDD), bipolar I or II disorder (BD), and non-affective psychotic disorders (NAPD).


Methods
The search strategy included terms related to the disorders of interest, prediction, relapse, and longitudinal study designs. Predictors and relapse needed to be measured through self-report. Eligible studies were required to assess potential predictors and relapse at least twice. No interval could exceed 14 days, and the total assessment period had to span at least 14 days. Intensive repeated assessments (e.g., experience sampling) had to be administered at least once a day for a minimum of 5 consecutive days to capture fluctuations in relapse-related symptoms over time.


Findings
Transdiagnostic predictors of relapse across MDD, BD, and NAPD appear limited. Fifteen articles were included. Sleep problems were the only factor identified as a predictor across all three disorders. Negative affect was shared between BD and NAPD only. Other predictors were disorder-specific. The limited findings may partly be explained by methodological choices and our strict inclusion criteria. That is, insufficient measurement intervals was the leading reason for exclusion (55 of 93 articles).


Discussion
Additional transdiagnostic predictors may exist but remain unidentified because they have not been measured frequently enough or over a sufficiently long period. To improve early detection of relapse, future studies should implement frequent and consistent symptom monitoring to enhance the identification of transdiagnostic EWS across diagnoses.