Using Health Status to Predict Out-of-Pocket Payment for Health Care under Taiwan’s NHI system

Presenter: Chunhuei Chi, Oregon State University

Abstract

I. Motivation
The motivation for this study was to estimate the association between an individual’s health status measured by SF-36 and the size of out-of-pocket (OOP) payment for health care under Taiwan’s Health Insurance (NHI) Program. Further, the authors will also estimate how age and education can predict out-of-pocket payment for health care in Taiwan.

II Objectives
The objective of this study is to estimate the association between health status as measured by SF-36 and out-of-pocket payment for health care under Taiwan’s NHI, and to compare that with the association between age, education and out-of-pocket payment.

III. Research Methods
This project used two methods for studying out-of-pocket payment. The first method is the sample households survey and prospective household diary records. A random sample of 500 households central Taiwan, including both metropolitan and rural areas. Using individual as the unit of analysis, all members of the sample households were included as study sample, with a total effective sample size of 1,062 subjects. Besides a face-to-face interview, we also rely on household diary to record the use of health care, and OOP health care expenditure over 2 months period. The second source of data is the use of claim record. In order to ascertain the quality and accuracy of our survey and diary data, we supplemented it with the claim records from the NHI’s Central Regional Bureau, by matching the sample data with claim records. Health status of respondents was measured by SF-36 Taiwan version, by using a composite health score in our analysis.

IV. Results
Using two-parts model to estimate the association between SF-36 score and out-of-pocket payment, this study estimated that for an increase of 1 in the SF-36 score, we expected the respondents to reduce their OOP by USD$100 per year. Compared with health status, we estimated that by increasing 1 year of age, the expected OOP will increase USD$43 per year, while education was not a statistically significant predictor of OOP.

V. Conclusions
This study provides support that SF-36 is a significant predictor for an individual’s OOP under Taiwan’s NHI system. Further, when compared with age, SF-36 composite score has stronger predicting power in OOP. This suggests that although age has often being suggested as an important predictor of health care expenditure, it is health status per se that is more significantly associated with health care expenditure. Being older in age alone does not have to cause health expenditure to rise. Ultimately, it is health status that influences utilization of health care and expenditure. While it is a common sense that a young and unhealthy person can consume more health care than an old and healthy person, this empirical study provides a strong evidence to support this argument. Instead of worrying about how aging society can cause health care expenditure to rise, result of this study suggests that focusing more on keeping people at all age healthy, besides being an important goal of all societies, it will also be far more effective in cost control,