There are few studies on the impact of out-of-pocket mental health care expenditures and sociodemographic factors on the probability of Mexican households to incur catastrophic healthcare expenditures (CHE).

The goal of the present study was to estimate the incidence of CHE and its main determinants among the households of persons with mental disorders (MD) in Mexico.

A cross-sectional survey was conducted, including 387 households of persons with MD. The estimation of the CHE was obtained by the health expenditure distribution method. A Logistic Regression (LR) was used to identify the determinants of probability variation of CHE occurrence. Since we expected a proportion of CHE between 20% and 80%, we assume linearity in the probability function, therefore we additionally used an Ordinary Least Squares (OLS) model.

In our sample, the incidence of CHE was 34.8%. The two mental illnesses most frequently associated with CHE were schizophrenia and hyperactive disorder (35.5% and 32.6% of CHE cases, respectively). The regression coefficients showed that for each unit (US$53.77) increase in income, the probability of CHE was reduced by 8.6%, while for each unit increase in hospitalization or medication expenditures, the probability of CHE increased by 12.9% or 19%, respectively. For each additional household member, the probability of CHE increased by 3%, and households with a male patient had a 7% greater probability of CHE.

Household income, household size, hospitalization and medication expenses, and sex of the patient were significant predictors of CHE for households caring for a person with MD.