Bivariate and multivariate analysis.

To estimate the survival function and identify risk factors for loss to follow-up (LTFU) among elderly patients with schizophrenia treated at the National Institute of Mental Health (INSM) in Peru. A retrospective cohort study was conducted using clinical records from 138 randomly selected patients diagnosed with schizophrenia. Eligible patients had attended at least one appointment between April 15, 2018, and April 15, 2022. Clinical and attendance data were recorded for each psychiatric appointment until the end of the follow-up period or LTFU. The survival function for LTFU was estimated, and associations with type of appointment (telemedicine vs. face-to-face) and period of first appointment (pre- vs. post-telemedicine implementation) were analyzed using Cox regression. The overall survival function at the end of the follow-up period was 47%. Patients whose treatment began at INSM after the implementation of telemedicine (TM) had a survival function of 49%, compared with 75% for those who initiated their treatment before TM. LTFU was significantly associated with initiating treatment in the Emergency Department (ED) rather than through outpatient care and with starting treatment post-COVID-19. Our findings indicate a higher rate of LTFU among patients who began treatment at INSM during the post COVID-19 period or through the Emergency Department. These results highlight the need for targeted interventions to address these risk factors and improve continuity of care for this vulnerable population.