While early intervention in psychosis (EIP) programs have been increasingly implemented across the globe, many initiatives from Africa, Asia and Latin America are not widely known. The aims of the current review are (a) to describe population-based and small-scale, single-site EIP programs in Africa, Asia and Latin America, (b) to examine the variability between programs located in low-and-middle income (LMIC) and high-income countries in similar regions and (c) to outline some of the challenges and provide recommendations to overcome existing obstacles.

EIP programs in Africa, Asia and Latin America were identified through experts from the different target regions. We performed a systematic search in Medline, Embase, APA PsycInfo, Web of Science and Scopus up to February 6, 2024.

Most EIP programs in these continents are small-scale, single-site programs that serve a limited section of the population. Population-based programs with widespread coverage and programs integrated into primary health care are rare. In Africa, EIP programs are virtually absent. Mainland China is one of the only LMICs that has begun to take steps toward developing a population-based EIP program. High-income Asian countries (e.g. Hong Kong and Singapore) have well-developed, comprehensive programs for individuals with early psychosis, while others with similar economies (e.g. South Korea and Japan) do not. In Latin America, Chile is the only country in the process of providing population-based EIP care.

Financial resources and integration in mental health care, as well as the availability of epidemiological data on psychosis, impact the implementation of EIP programs. Given the major treatment gap of early psychosis in Africa, Latin America and large parts of Asia, publicly funded, locally-led and accessible community-based EIP care provision is urgently needed.

Although relapse in psychosis is common, a small proportion of patients will not relapse in the long term. We examined the proportion and predictors of patients who never relapsed in the 10 years following complete resolution of positive symptoms from their first psychotic episode.

Patients who previously enrolled in a 12-month randomized controlled trial on medication discontinuation and relapse following first-episode psychosis (FEP) were followed up after 10 years. Relapse of positive symptoms was operationalized as a change from a Clinical Global Impression scale positive score of <3 for at least 3 consecutive months to a score of ⩾3 (mild or more severe). Baseline predictors included basic demographics, premorbid functioning, symptoms, functioning, and neurocognitive functioning.

Out of 178 first-episode patients, 37 (21%) never relapsed during the 10-year period. Univariate predictors (p ⩽ 0.1) of patients who never relapsed included a duration of untreated psychosis (DUP) ⩽30 days, diagnosed with non-schizophrenia spectrum disorders, having less severe negative symptoms, and performing better in logical memory immediate recall and verbal fluency tests. A multivariate logistic regression analysis further suggested that the absence of any relapsing episodes was significantly related to better short-term verbal memory, shorter DUP, and non-schizophrenia spectrum disorders.

Treatment delay and neurocognitive function are potentially modifiable predictors of good long-term prognosis in FEP. These predictors are informative as they can be incorporated into an optimum risk prediction model in the future, which would help with clinical decision making regarding maintenance treatment in FEP.