External classification for ethnicity in health survey to monitor Roma vs non-Roma differences

• Published in: European journal of public health 2020-09, Vol.30 (Supplement_5)
• Main Authors: Sándor, J, Janka, E, Vincze, F, Ádány, R
• Format: Article
• Language: English
• Subjects: Adults; Appeals; Classification; Data collection; Employment; Ethnicity; Health behavior; Health surveys; Indicators; Minority & ethnic groups; Morbidity; Polls & surveys; Population; Public health; Regression analysis; Regression models; Reliability; Romani people; Social behavior; Social capital; Statistical analysis; Statistical inference
• ISSN: 1101-1262; 1464-360X
• DOI: 10.1093/eurpub/ckaa166.561

Abstract Background The Roma population of European countries is characterized by a poor and poorly described health status. The lack of detailed data on health loss among Roma can be partly attributed to legal limitations based lack of reliable methodology to effectively identify them. On the other hand, it is among factors responsible for restricted effectiveness of policy to narrow gap between Roma and non-Roma health status. Objectives We aimed to compare directly results from population based health survey on Roma vs non-Roma differences based on self-declared and observer-reported Roma classification. Methods In a population based health interview survey of random sample of 1849 subjects aged 18-64 years representing 965680 adults, both self-reporting and observer-reporting on Roma was applied. Data collection covered 52 indicators of morbidity, functional loss, health behavior, social capital, and use of health services. Health status differences between the self-declared Roma (N = 124) and the non-Roma (N = 1725) as well as the observer-reported Roma (N = 179) and non-Roma populations (N = 1670) were described by logistic regression models controlled for age, sex, education and employment. Results There was no significant Roma vs non-Roma difference in the respect of 33 indicators, and 14 indicators showed significant difference by both Roma definitions. Differences between the observer- and self-reported Roma ethnicity in statistical inference were observed for 7 indicators. In ascertaining significant Roma vs non-Roma difference, self-reporting was more effective for 2, and observer-reporting was for 5 indicators. Conclusions The application of observer-reported ethnicity classification cannot increase the practical value of health survey on Roma to non-Roma differences, compared to the self-reporting based approach. Therefore, the use of observer-reported Roma ethnicity in health surveys to improve the reliability of Roma specific indicators seems to be not justified. Key messages Application of external classification for ethnicity cannot increase the effectiveness of population based health interview survey in monitoring Roma vs non-Roma differences. The use of observer-reported Roma ethnicity in health surveys to improve the reliability of Roma specific indicators seems to be not justified.