A propensity score matching analysis using statistical methods for estimating the impact of intervention: The cost of malaria and its impact on the health system

Abstract

Colombia contributes 17% of all malaria cases in Latin America and the Caribbean and continues to have major health and economic costs. A propensity score-matched cost-of-illness study was conducted using an incidence-based approach focused on direct healthcare costs in 2019. The statistical analysis was designed to determine the economic burden attributed to malaria on the Colombian health system, a middle-income country with universal coverage. Propensity score analysis effectively adjusted for confounding, thus facilitating comparability between groups with and without malaria. Direct costs attributable to malaria were estimated using a bottom-up strategy using fractions attributable to the population. We used administrative databases containing patient-level information on the consumption of healthcare services and reports on healthcare spending published by the Colombian Ministry of Health. All expenses were reported in Colombian pesos and converted to US dollars. We show malaria imposes a significant economic burden on the Colombian health system. The most affected population is part of the subsidized regime, which assumes the greatest financial burden. The cost of complicated malaria far exceeds that incurred for uncomplicated malaria. Propensity score matching adjusted for significant differences in comorbidities between groups, allowing for the closest estimate of malaria-attributed costs in the country.

Keywords

• Malaria
• Costs and cost analysis
• Cost of illness
• Health expenditures
• Health care costs