Insured-non-insured disparity of catastrophic health expenditure in Northwest Ethiopia : a multivariate decomposition analysis

Background Financial risk protection is one indicator of universal health coverage (UHC). All people should be protected from financial risks such as catastrophic health expenditures (CHE) to ensure equitable health services. Ethiopia has launched community-based health insurance (CBHI) since 2011 to protect people from financial risk. However, out-of-pocket health expenditure is a financial barriers to achieve UHC. The insured-non-insured disparity of CHE has not been well studied in Ethiopia in general and in Debre Tabor town in particular. Therefore, this study aimed to assess the disparity of CHE between insured and non-insured households and its contributing factors in Debre Tabor town. Methods This study used the primary household survey data collected from May to June 2022 in Debre Tabor town. Data were collected from 825 household heads and analyzed using STATA version 17.0 statistical software. Logit-based multivariate decomposition analysis was conducted to determine insured-non-insured disparity of CHE. Statistical significance for all analysis was declared at a p < 0.05. Results The incidence of CHE was 17.94% and 5.58% among non-insured and insured households, respectively. About 53% and 153.20% of the insured-non-insured disparities in the magnitude of CHE were due to the difference in characteristics (endowments) and the effect of characteristics (coefficients), respectively. Age of the household head between 46 and 60 years and above 60 years, divorced and widowed marital status of household head, and chronic health conditions were the explanatory variables widening the gap in the incidence of CHE. However, do not seeking traditional medicine, family size above 4, and age of household head between 31 and 45 years were the variables contribute in reducing the gap (i.e. due to endowments) in the incidence of CHE between insured and non-insured households. Moreover, the variables that contributed to the gap in the incidence of CHE due to covariate effects were age (31-45) and marital status of household head, wealth status, family size, ownership of the household, and seeking traditional medicines. Conclusion This study revealed there is a significant disparity in the incidence of CHE between insured and non-insured households. Age, marital status and occupation of the household head, family size of household, presence of a chronically ill household member and seeking traditional medicine were significantly contributing factors for the disparity of CHE between insured and non-insured households due to endowments. The variables that contributed to the disparity in the incidence of CHE due to covariate effects were age and marital status of household