Original articlesThe cross-cultural validity of self-reported use of health care: A comparison of survey and registration data
Introduction
Native and immigrant groups often differ as to health status 1, 2, 3, 4, 5, and utilization of health care 2, 4, 6. For adequate health planning, information is needed on the use of health care in these groups 2, 4, 6, 7, 8. Theoretically, the recording of ethnicity in various health care registers is the easiest route to obtaining this information. However, practical problems such as high costs and unwillingness of care providers to record ethnicity may hamper this route 7, 8.
Alternatively, surveys comprising various ethnic groups may be used to obtain information of ethnic differences in the use of health care. However, language problems and cultural differences (e.g., regarding the perception of time and sensitivity for socially accepted behavior) might equally cause a culturally determined information bias regarding health care utilization 9, 10, 11.
No evidence is available on the cross-cultural validity of responses to survey questions regarding use of various types of care. Results regarding health-related behavior, however, show that cross-cultural validity is affected by methodological factors like questionnaire design [11], interviewer selection, and interviewer training 10, 11. Furthermore, it is affected by cultural differences in responses, like the willingness to use extreme response categories [11]. This study therefore examines the validity of data on health care utilization, on the basis of a data set containing both registered and self-reported use of native and immigrant (foreign-born) people.
Section snippets
Methods
The study was based on a linkage of data from a survey and from a health insurance register, in Amsterdam, the Netherlands.
Results
Self-reported utilization for all three types of care is higher than registered utilization, but the differences are not statistically significant, neither overall nor for Dutch- and foreign-born respondents separately (Table 1.)
The overall concordance between reported and registered use of care is fair regarding all three types of care: kappa varies from 0.58 to 0.77 (Table 1). Regarding all three types of care, kappa is lower for foreign-born than for Dutch-born respondents, though without
Discussion
This analysis shows that self-reports yield somewhat higher utilization rates than registration, without systematic ethnic differences regarding this aspect. Furthermore, the concordance between self-report and registration is generally fair, albeit somewhat lower for immigrants, especially those from Turkey and Morocco. Only regarding (ambulatory) physiotherapy, ethnicity measured in detail modifies the association between reported and registered utilization.
The results of this study show that
Acknowledgements
The Amsterdam Municipal Health Service (GGEGD) provided data for this study.
References (28)
- et al.
Improving question wording in surveys of culturally diverse populations
Ann Epidemiol
(1997) - et al.
Bias, prevalence and kappa
J Clin Epidemiol
(1993) - et al.
Comparison of self-reported and medical record health care utilization measures
J Clin Epidemiol
(1996) - et al.
A reappraisal of the kappa coefficient
J Clin Epidemiol
(1988) Cancer in Mediterranean migrants, based on studies in France and Australia
Cancer Causes Control
(1995)- et al.
Health of migrants and migrant health policy, the Netherlands as an example
Soc Sci Med
(1995) - et al.
Impaired mobility and impaired working capacity among foreign born people and native born Swedes
J Epidemiol Community Health
(1998) Reported health, lifestyles and use of health care of first-generation immigrants in the Netherlandsdo socioeconomic factors explain their adverse position?
J Epidemiol Community Health
(1998)- et al.
Race and self assessed health statusthe role of socioeconomic factors in the USA
J Epidemiol Community Health
(1996) - et al.
Ethnic differences in general practitioner consultations
Br Med J
(1989)