Predictors of carer-burden in schizophrenia: a five-year follow-up study in Butajira, Ethiopia.

OBJECTIVE: We assessed carer-burden and its predictors in a traditional rural Ethiopian community in order to establish the longitudinal course of carer-burden and factors predicting changes.
METHODS: Using a 5-year follow-up data from the ongoing Butajira outcome study on SMI, carer-burden was assessed annually with the Family Interview Schedule (FIS). Multilevel modeling was used to identify clinical predictors of severity and rate of change of burden.
RESULTS: Scores in all domains of carer burden decreased over time, although the greatest reduction was seen in the first year. In a univariate analyses, longitudinal reduction in burden score was predicted by longer period in remission during follow-up, while negative and positive symptom severity scores predicted higher burden score. In the fully adjusted model, poor social support predicted higher burden score (beta=0.38, 95%CI 0.04, 0.72), and longer period in remission predicted lower level of carer-burden (beta = -0.49, 95%CI = -0.89, - 0.10). Reduction in positive symptoms was associated with the instantaneous rate of reduction of burden score (beta = -0.03, 95%CI - 0.05, -0.01).
CONCLUSION: There is a significant reduction in carer-burden over the years in all burden domains. Providing accessible mental health care has the potential to alleviate carer-burden, as positive symptoms are believed to be more amenable to intervention. The study also indicates that remission is associated with reduction in carer-burden.