Latent variable analysis of adherence to antipsychotics among South Carolina medicaid beneficiaries with schizophrenia or schizoaffective disorder

Schizophrenia and schizoaffective disorder are chronic mental illnesses that may lead to positive symptoms, negative symptoms, or cognitive symptoms. This study examined latent classes within medication adherence profiles for long-acting injectable antipsychotics (LAIs) and oral antipsychotics (OAPs). Latent profile analysis (LPA) and multinomial logistic regression models were used to explore the association between patient characteristics and latent class membership. Four latent classes were identified using LPA with observed adherence measures. The classes were labeled as "best adherent", "intermittent adherent", "early drop-off", and "worst adherent", with their respective estimated prevalences of 58 %, 17 %, 9 %, and 16 %. Multinomial logistic regression showed that patients on LAIs were more likely to belong to the "best adherent" group than to the non-adherent groups ("intermittent adherent", "early drop-off", and "worst adherent") when compared to those on OAPs. These findings may contribute to the development of strategies for medication prescription and disease management for people with schizophrenia or schizoaffective disorder.