Public health expenditure and health system responsiveness for low-income individuals: results from 63 countries.

Improvement in overall responsiveness to people's expectations is an important goal for any health system; socioeconomic equity in responsiveness is equally important. However, it is not known if socioeconomic disparities in responsiveness can be reduced through greater public health expenditures. This article assesses the relationship of the proportion of public health expenditure over total health expenditure (PPHE) with responsiveness for poorest individuals and the difference in responsiveness between the richest and poorest individuals. We used data from six responsiveness dimensions (prompt attention, dignity, choice, clarity of information, confidentiality and quality of basic amenities) of outpatient services from World Health Survey data from 63 countries. Hierarchical Ordered Probit (HOPIT) models assessed the probability of 'very good' responsiveness in each domain among the poorest and richest individuals for each country, correcting for reporting heterogeneity through vignettes. Linear regression models were then used to assess the association between predicted probabilities from HOPIT models and PPHE, adjusting for (log) Gross Domestic Product per capita. The study findings showed that higher PPHE was associated with a higher probability of 'very good' responsiveness for each domain among the poorest individuals, and with smaller pro-rich disparities in responsiveness between the richest and poorest individuals. In conclusion, increasing PPHE may improve the responsiveness of health services for the poorest individuals and reduce disparities in responsiveness between the richest and poorest individuals.