A decomposition analysis for socioeconomic inequalities in health status associated with the COVID-19 diagnosis and related symptoms during Brazil's first wave of infections

Recent studies have shown that COVID-19 affects different population groups asymmetrically. This work uses data from the National Survey of Households—PNAD COVID-19/IBGE—to quantify the socioeconomic inequality in health during the first wave of COVID-19 infections in Brazil. We use the concentration curve, the concentration index, and a decomposition analysis to verify the factors that most influence the inequalities in the specified health variables. We find a positive concentration index for the incidence rate, indicating a greater concentration of diagnoses (number of tests) among groups with higher income levels. When considering symptoms similar to a COVID-19 infection, inequality practically disappears. Among people with higher income, a pre-existing disease has a more significant contribution to the concentration of COVID-19 in the presence of correlated symptoms than in its diagnosis. Tests of dominance support the findings. Moreover, the decomposition results show that if the inequalities were explained only by race (non-white) and place of living (North and Northeast), there would be a concentration of COVID-19 among the poorest.