The challenge of measuring need for health care in household surveys – a preliminary analysis for South Africa F. Meheus, O. Okorafor, D. McIntyre Inaugural Conference of the African Health Economics and Policy Association Conference (AFHEA) Accra, 10 th -12 th March 2009 Abstract n° PS06/7 SHIELD project: Strategies for Health Insurance for Equity in Less Developed Countries
22
Embed
SHIELD project: Strategies for Health Insurance for Equity in Less Developed Countries
Abstract n° PS06/7. The challenge of measuring need for health care in household surveys – a preliminary analysis for South Africa. SHIELD project: Strategies for Health Insurance for Equity in Less Developed Countries. F. Meheus, O. Okorafor, D. McIntyre - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
The challenge of measuring need for health care in household surveys – a preliminary analysis for South Africa
F. Meheus, O. Okorafor, D. McIntyre
Inaugural Conference of the African Health Economics and
Policy Association Conference (AFHEA)
Accra, 10th-12th March 2009
Abstract n° PS06/7
SHIELD project: Strategies for Health Insurance for Equity in Less Developed Countries
Mortality and morbidity
• Indicators of health outcome (mortality and/or morbidity) are frequently used in health equity analysis.
• Mortality indicators are for example life expectancy, or overall or disease specific mortality rates.
• Morbidity indicators can be either objective or subjective:
– Objective or observed morbidity indicators are those measures that can be directly observed (e.g. physical examination or laboratory investigations);
– Subjective or self-perceived indicators are assessments of health status/illness/symptoms/functional disability by the individual.
• Morbidity indicators are usually obtained from household surveys.
• A common application of mortality/morbidity measures in health equity analysis is looking at the (socio-) economic gradient of these measures.
Need for health care
• But health equity analysis can also be concerned with the utilization of and payment for health services.
• A common application is to examine the degree of horizontal inequity in utilization of health care services relative to socio-economic status.
– This is based on the principle of "equal treatment for equal need", irrespective of other characteristics such as income, race, place of residence, etc.
– Deviations in the degree to which health care is distributed according to need are measured by an index of horizontal inequity.
• Need for health care is usually estimated using a self-perceived indicator, mainly self-assessed health or self-reported illness.
Self-perceived indicators
• There are different types of self-perceived indicators (SPI) :
– Medical, functional and subjective indicators (Blaxter 1993)
• Medical indicators measure health as a deviation from physiological norms (e.g. presence of certain diseases).
• Functional indicators: measures ill-health as a lack of ability to perform "normal" tasks or roles (e.g. days of restricted activity).
• Subjective indicators: health is defined in terms of the individual's perception of his/her health (e.g. self-assessed health).
• The most frequent indicators are self-assessed health (subjective) and self-reported illness (subjective/medical).
Are self-perceived indicators reliable?Evidence from developed countries
• They have shown to be correlated with actual health; e.g Cutler &
Richardson, 1997; Kaplan & Baron-Epel, 2003.
• They appear to be a good predictor of:
– Mortality and morbidity; e.g. Idler & Benyamini 1997;
– Subsequent use of medical care; e.g. van Doorslaer, Jones,
& Koolman, 2004;
• As such, they have often been used to describe socio-economic inequalities in health
– Various studies have observed an economic gradient in self-assessed health status; e.g. Smith 1999; Case et al 2002; Currie et al
2004; Kunst et al 1995; van Doorslaer and Koolman 2004. :
• However few evidence from developing countries:
– Greater likelihood of subsequent mortality observed in individuals who perceive their health as poor (Frankenberg & Jones (2004) in Indonesia and Yu et al. (1998) in China);
– A similar relationship was observed between self-assessed health and physical functioning (Zimmer et al 2000, Wu & Rudkin 2000; Rahman & Barsky 2003);
• Use of self-perceived indicators (SPI) remain underutilized in developing countries although included in many household surveys.
Are self-perceived indicators reliable?Evidence from developing countries
Scepticism towards SPI
• Despite being reliable, valid and a cost-effective means of health assessment, SPI are viewed with much scepticism in developing countries
– Especially when assessed in relation to socio-economic status (SES) or other covariates of health and wealth (e.g. age, sex).
• While indicators such as life expectancy or mortality have shown consistent inequalities to the disadvantage of the poor, self-perceived indicators based on self-reported illness have shown counterintuitive findings in developing countries.
• If those with lower SES report less illness given true health, then socio-economic inequality will be underestimated by using self-reported illness.
Previous findings
• Several studies have shown lower "recognition" of illness by
low-income groups than higher income groups (Baker & Van
der Gaag 1993, McIntyre & Gilson 2000, McIntyre et al.
1998, Makinen et al. 2000, Sen 2002)
• Several explanations:
– Poor cannot afford to be ill (Sauerborn et al. 1996);
– Adaptation and scale of reference bias (e.g. Groot 2000)
– Perception of own illness is contingent on social experience
(Sen 1993, 2002)
• Self-assessed health does not seem to be influenced by this.
Objective & approach
• Examine the economic gradient in self-assessed health (SAH) with self-reported illness (SRI).
Provides a useful test of the validity of self-perceived indicators in relation to socio-economic status (SES).
• Approach:
– Use own data from household survey in South Africa where both SAH and SRI are measured.
• Compare degree of inequality of both measures by looking at proportions.
• Examine degree of inequality by calculating concentration index.
Data sources
• Data from South Africa
• Carried out own household survey in 2008 (CASE-survey)
• Comprised all nine South African provinces and is nationally
representative
• Sample size = 4,800 households
Data was collected as part of a larger project (SHIELD –
Strategies for Health Insurance and Equity in less
Developed countries) examining the financing and benefit
incidence in three African countries, with specific focus
on health insurance.
Indicators included in the study
• Self-assessed health (SAH)– Question: "How would you describe your current health status?"