International Comparisons of Health Care Responsiveness using Anchoring Vignettes from SHARE

Presenter: Nicolas Sirven, IRDES (Institute for Research and Development in Health Economics)

Abstract

Health care responsiveness, or the extent to which the process of care delivery matches patients’ expectations, stands for an important criterion for evaluating health care systems. However, expectations and satisfaction with access may be influenced by a range of socio-demographic or health variables, which distribution is not equivalent across countries and consequently necessitate some adjustment. Responses to subjective questions are also likely to reflect a different ‘response style’ in regions or countries, reflecting historical, cultural or environmental circumstances. The vignettes method, designed for correcting subjective responses and thus avoiding the effects of a response style bias, has recently been applied in the Survey of Health, Ageing, and Retirement in Europe (SHARE). The aim of this exploratory study is to compare health care responsiveness in ten countries participating in SHARE (2006), adjusting for a range of socio-demographic and health factors, to describe the effect of correction for differences in response styles using a set of vignettes, and to investigate such differences at the aggregated level. More than 6,000 respondents aged 50 years old and over were asked to assess the quality of health care responsiveness in three domains: waiting time for medical treatment, quality of the conditions in visited health facilities, and communication and involvement in decisions about the treatment. Chopit models estimates suggest that reporting heterogeneity is influenced by both individual (socio-economic, health) and national characteristics. Our results suggest that the use of vignettes improve the quality of the relationship between satisfaction in health care in a country and the national level of health expenditures. In particular, it seems that each euro spent increases satisfaction in “conditions of the health facilities” and “communication with doctors”, more than we would have though if correction for response heterogeneity would not have been taken into account.