IJE Advance Access published online on August 2, 2008
International Journal of Epidemiology, doi:10.1093/ije/dyn146
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Differential health reporting by education level and its impact on the measurement of health inequalities among older Europeans
1 Department of Applied Economics, Erasmus School of Economics, Erasmus University, Rotterdam, The Netherlands.
2 Tinbergen Institute, Amsterdam, The Netherlands.
3 Department of Balkan, Slavic & Oriental Studies, University of Macedonia, Thessaloniki, Greece.
4 Department of Health Policy and Management, Erasmus University Medical Centre, Rotterdam, The Netherlands.
* Corresponding author. Department of Applied Economics (Room H13-09), Erasmus School of Economics, P.O. Box 1738, 3000 DR Rotterdam, The Netherlands. E-mail: bagoduva{at}few.eur.nl
 |
Abstract |
|---|
Background This study aims to establish whether health reporting differs by education level and, if so, to determine the extent to which this biases the measurement of health inequalities among older Europeans.
Methods Data are from the Survey of Health, Ageing and Retirement in Europe (SHARE) covering eight countries. Differential reporting of health by education is identified from ratings of anchoring vignettes that describe fixed health states. Ratings of own health in six domains (mobility, pain, sleep, breathing, emotional health and cognition) are corrected for differences in reporting using an extended ordered probit model. For each country and health domain, we compare the corrected with the uncorrected age–sex standardized high-to-low education rate ratio for the absence of a health problem.
Results Before correction for reporting differences across the 48 combinations of country by health domain, there was no inequality in health by education (P > 0.05) in 32 of 48 cases. However, there were reporting differences by education (P < 0.05) in 29 out of 48 cases. In general, higher educated older Europeans are more likely to rate a given health state negatively (except for Spain and Sweden). Correcting for these differences generally increases health inequalities (except for Spain and Sweden) and results in the emergence of inequalities in 18 cases (P < 0.05), which may be considered ‘statistically significant’. The greatest impact is in Belgium, Germany and The Netherlands, where inequalities (P < 0.05) appear only after adjustment in four of the six health domains.
Conclusions These results emphasize the need to account for differences in the reporting of health. Measured health inequalities by education are often underestimated, and even go undetected, if no account is taken of these reporting differences.
Keywords Health measurement, self-rated health, socioeconomic status, reporting differences, vignettes
Accepted 18 June 2008