摘要：Objectives. This study assessed the validity of a global measure of self-assessed health among Indigenous Australians. Methods. Logistic regression was used to identify factors associated with self-assessed health in a nationally representative sample. Results. Among 8782 adult respondents, poorer self-assessed health was strongly associated with several factors, including age, number of health conditions, and recent health actions. The association with health conditions was attenuated when the respondent's primary language was not English. Conclusions. Self-assessed health may be a valid measure among indigenous Australians whose primary language is English. However, although the measure draws on common experiences of health and illness, it may obscure differences in how people incorporate these experiences into social actions. There is a need to measure health status to establish differentials within and between populations, to track trends over time, and to assess changes in response to health interventions. Although health is widely recognized as a social construct, considerable emphasis in health status measurement has been placed on “objective” measures of illness such as limitations in activity, morbidity, and mortality. Within this biomedical framework, however, self-assessed health status has gained increasing attention, becoming an important component of contemporary health research 1 that is arguably as reliable as—or more reliable than—other “objective” biomedical measures for which it can be considered a proxy. 2 There are 2 general approaches to subjective health assessment. The first involves the use of multiple items, covering a number of dimensions of health, that may be designed to sum to a single index. In the second approach, a single global question is used. Global questions also appear in many validated multi-item instruments, including the General Health Questionnaire 3 and the Short-Form Health Survey, 4 and have been found to correlate with multi-item instruments such as the Nottingham Profile. 5 Because of their simplicity, ease of administration, and validated performance, global questions are now in widespread use. In the Australian National Health Survey of 1989–1990, a global measure was included for the first time. Indigenous Australians at that time composed approximately 2% of the population, and because they were not oversampled, there were insufficient data for meaningful analysis. In 1994, the Australian Bureau of Statistics conducted the first national survey of Indigenous Australians—the National Aboriginal and Torres Strait Islander Survey, or NATSIS—and included a global item on self-assessed health, making adequate data available for the first time. However, questions had long been raised about the appropriateness of using mainstream indicators in assessing health and well-being among Indigenous Australians, not least because Indigenous people are believed to have a different concept of health than other groups (National Aboriginal Health Strategy Working Party, A National Aboriginal Health Strategy [unpublished report], 1989). In this article, we distill analyses published by the Australian Bureau of Statistics 6 to examine the validity of the global measure with this population. The health status of Indigenous Australians is worse than that of non-Indigenous Australians. 7, 8 In the period 1991 through 1996, estimated life expectancies were 56.9 years for Indigenous men (vs 75.2 years for all Australian men) and 61.7 years for Indigenous women (vs 81.1 years for all Australian women). 7 As Kunitz has shown, health among this population is also relatively worse than that of Indigenous peoples in comparable colonial countries. 9 For example, life expectancy at birth among Maoris is 8 years less than that among other New Zealanders. 10 The 5 major causes of death among Indigenous Australians are, in descending order, circulatory diseases, injury and poisoning, respiratory diseases, neoplasms, and endocrine–metabolic disorders (including diabetes). 7
