Who is at greatest risk of catastrophic health expenditures? A Heckman probit model of careseeking and financial protection in Peru

Many families in Peru lack financial protection from catastrophic health expenditures, and are both limited in their access to health care and exposed to the risk of impoverishment and suffering. Identifying who is most at risk for burdensome health expenditures could help policymakers target vulnerable groups for improved financial protection mechanisms. Unfortunately, some sick individuals may not seek or spend money on health care because they cannot afford it. They thus fail to be “captured” by common catastrophic expenditure (CE) indicators, such as spending more than 20% of total household consumption on health care. The purpose of this analysis was to identify risk factors for CEs, using a Heckman probit model to simultaneously control for factors affecting the probability of seeking care and factors influencing total out-of-pocket costs among those who do seek care. I analyzed the Peruvian National Survey of Households for 1998-2002. Standard probit regressions indicated no significant difference in risk of CEs between the poorest and richest asset quintiles, but those in the lower quintiles were much less likely to seek care. Using the Heckman probit, the poor were at significantly (p<0.0001) greater risk of CEs once their lower rates of health care utilization were adjusted for. Increases in the education of the household head, formal sector employment, urban residence, and coverage by national health insurance programs were other protective factors. Results indicate that poor and rural groups should be targeted for better financial protection mechanisms, which would increase their access to health care as well.