Improving recovery and remission rates in early-stage psychosis requires understanding its trajectory, encompassing remission, recovery, and relapse, particularly after the first episode of psychosis (FEP). Early-stage psychosis is defined in this study as a clinical diagnosis within ≤2 years of antipsychotic treatment and a duration of adequate antipsychotic treatment of more than 4 weeks but less than or equal to 2 years. Previous research on FEP reveals varying remission rates (36% to 58%) and recovery rates (17% to 38%) across studies, highlighting the need for a better understanding of the predictors of remission and recovery. Predictors identified in the literature include better premorbid functioning, milder baseline symptoms, early treatment response, shorter DUP, better premorbid adjustment, higher education levels, better neurocognition, premorbid IQ, superior occupational status, being married, medication adherence, favorable personality, and fewer negative baseline symptoms. Relapse is a significant concern, with rates ranging from 28% to 80% at 1 to 5 years post-treatment. Factors increasing relapse risk include medication non-adherence, persistent substance use disorders, critical caregiver comments, and poorer premorbid adjustment. Long-acting injectable antipsychotics (LAIA) offer a potential strategy to improve medication adherence and prevent relapse, although their use remains low. The Korea Early Psychosis Study (KEPS) provides a valuable opportunity to investigate these issues in a long-term prospective cohort study within a Korean context. This study aimed to identify various three-year outcomes in early-stage psychosis (remission, recovery, relapse, medication adherence) and examine predictors of full recovery.

Existing literature shows inconsistent findings regarding remission and recovery rates in first-episode psychosis (FEP). Some systematic reviews report mean symptomatic remission rates ranging from 36% to 58% with follow-up durations varying from 6 months to 7 years. Recovery rates are similarly variable, ranging from 17% to 38% over a mean follow-up of 7.2 years. Studies investigating predictors of remission and recovery have identified a number of factors, including premorbid functioning, symptom severity at baseline, early treatment response, duration of untreated psychosis (DUP), gender, education level, neurocognition, premorbid IQ, occupational status, marital status, medication adherence, personality traits, and the presence of negative symptoms. Relapse rates following treatment for FEP also vary substantially across studies, with a wide range of reported percentages. Medication non-adherence, substance use disorders, critical caregiver comments, and poorer premorbid adjustment have all been identified as factors increasing relapse risk. The use of long-acting injectable antipsychotics (LAIA) is suggested as a strategy to improve outcomes and reduce relapse rates, but their usage is currently low.

This prospective, naturalistic observational cohort study utilized data from the Korea Early Psychosis Study (KEPS), focusing on patients with early-stage psychosis diagnosed with schizophrenia spectrum disorders (SSD) or psychotic disorder not otherwise specified (PNOS). Data were collected from January 2015 to July 2020 from 11 participating hospitals. The early-stage psychosis criterion was defined as a duration of adequate antipsychotic treatment of more than 4 weeks but less than or equal to 2 years. A total of 534 patients were included in the analysis, with varying durations of follow-up. Diagnoses were established using DSM-IV criteria and the Korean version of the Mini-International Neuropsychiatric Interview, with consensus reached by two experienced psychiatrists. Sociodemographic data (age, sex, education, insurance type, job type), family history of psychotic disorders, DUP, duration of illness (DI), Positive and Negative Syndrome Scale (PANSS), Clinical Global Impression (CGI) scales, Social and Occupational Functioning Assessment Scale (SOFAS), Calgary Depression Scale for Schizophrenia (CDSS), Columbia-Suicide Severity Rating Scale (C-SSRS), and comorbid mental disorders were assessed at baseline and at various follow-up points. Self-rated variables including the Big Five Inventory (BFI-10), Brief Core Schema Scales (BCSS), Brief Resilience Scale (BRS), Brooding Scale (BS), Diet History Questionnaire (DHQ), Early Trauma Inventory Self Report-Short Form (ETISR-SF), Family Adaptability and Cohesion Evaluation Scales III (FACES-III), Family Intimacy (FI), Physical Activity Rating (PAR), and the Korean version of the Subjective Well-being Under Neuroleptics-Short Form (K-SWN) were collected at baseline. Symptomatic remission was defined as a PANSS score ≤3 on eight specific items. Full recovery required both symptomatic recovery (PANSS score ≤2 on the same eight items) and functional recovery (adequate social interaction and occupational functioning). Drop-out and relapse were also defined and tracked. Logistic regression was used to identify predictors of full recovery at year 3.

Over the three-year follow-up period, the rates of symptomatic remission in total subjects were 76.10%, 69.20%, 79.50%, and 79.10%, while the rates of full recovery were 22.80%, 26.40%, 28.60%, and 39.60% at 6, 12, 24, and 36 months, respectively. Drop-out rates increased over time, reaching 51.1% at 36 months, while relapse rates also increased, reaching 38.9% at 36 months. Medication adherence was high (87.8% to 93.9% across the follow-up periods), and the prescription rate of LAIA increased from 9.7% at baseline to 18–26% at follow-up. Stepwise logistic regression revealed that significant predictors for full recovery were shorter duration of untreated psychosis (DUP) (odds ratio [OR] 0.534, p=0.001), higher family intimacy (FI) (OR 2.262, p=0.003), and higher physical activity intensity (PAR) (OR 1.231, p=0.040). Subgroup analysis showed higher remission and recovery rates in schizophreniform disorder and PNOS compared to schizophrenia in the early stages of follow-up, though this trend reversed at later time points. The utilization of mental health welfare facilities remained low across the entire period despite the high eligibility of patients.

The findings of this study demonstrate relatively high remission and full recovery rates compared to some previous studies, but also reveal a considerable gap between remission and recovery, highlighting the need for intensive psychosocial interventions to bridge this gap and improve functional outcomes. The significant predictors of full recovery (shorter DUP, higher family intimacy, and higher physical activity intensity) provide valuable insights for targeted interventions. The study's findings support the importance of early intervention to reduce DUP, strengthening family support to foster intimacy, and promoting physical activity to enhance recovery. The relatively low rate of LAIA prescription suggests that increased efforts are needed to promote the use of this medication for relapse prevention. The low utilization of mental health welfare facilities warrants further investigation into potential barriers to access and utilization of these resources.

This three-year longitudinal study of early-stage psychosis patients revealed high remission and recovery rates, but also highlighted the significant gap between the two. Key predictors of full recovery were identified as shorter DUP, strong family intimacy, and higher intensity of physical activity. The study underscores the necessity of comprehensive psychosocial interventions targeting these factors to enhance recovery outcomes. Future research should focus on refining early intervention strategies, fostering family support, and promoting engagement in exercise programs. Further investigation is warranted regarding the barriers affecting mental health welfare facility utilization and the optimal strategies for promoting LAIA use.

The study has several limitations. First, drop-out patients were not included in the calculation of remission, recovery, relapse, and adherence rates, potentially underestimating these rates. Second, the study did not include known predictors of outcome such as premorbid functioning and cognition. Third, the definition of early-stage psychosis might have recruited heterogeneous subjects. Fourth, the study relied solely on self-reported medication adherence, potentially affecting the accuracy of these measurements. Fifth, childhood trauma assessment relied on self-report rather than interviews, which might limit its reliability. These limitations should be considered when interpreting the study's results.