Analysis of Inpatient Services Utilization by Wealth Classes under the Linkage between NCMS and MA --- the Case of Three Counties in Rural China

Presenter: Xuefei Gu

Abstract

Introduction: In late 2002, Chinese government resolved to introduce and financially contribute to two major programs of rural health financial support to prevent "illness induced poverty". They are the New Rural Cooperative Medical Scheme (NCMS), targeting all rural communities and Medical Assistance (MA) targeting the poor in rural and urban areas. MA finances are principally directed at funding poor farmers’ contributions to NCMS. In terms of inpatient services, the poor (MA targeting groups) pay a lower proportion of medical care expenses out-of-pocket than better off because they can get reimbursement from NCMS and MA. Some previous research showed that farmers’ use of health services, particularly inpatient services, had risen dramatically under NCMS, but the poor used less than non-poor. Except for the high co-payment, complicated reimbursement procedure was another important reason (NCMS is administered by health authorities and MA is charged by the Ministry of Civil Affairs). So the linkages between two programs including scheme design and management level became a big problem, and a few counties had done some experiments on it. The study attempts to answer if the inpatient services utilization by the poor can be improved under the linkage between NCMS and MA.

Methodology: The study is based on data collected from 2007 and 2008 household surveys conducted in three counties located in China' Hubei, Anhui and Qinghai Provinces. We assess self-reported inpatient services utilization by asking respondents the number of hospital admissions if s/he had major diseases and doctor suggested hospital admission in previous year. One or more admissions last year was coded as 1, and no admissions as 0. The study population were divided into 5 groups by wealth status. Wealth index was used as a proxy for poverty variable and other sociodemographic characteristics were considered. Logistic regression was used to analyse the association between inpatient service utilization and socio-economic status.

Results: The odds of the 2nd poorest quintile getting inpatient services for major illnesses is 11% higher than the poorest quintile, and the odds of middle quintile is 42% higher than that of the poorest quintile, but are not statistically significant. Compared to 2006, the odds of hospital admission for all farmers is increased by 30% in 2007 (not significant). Separate analysis shows that the poorest quintile in county A used inpatient services more than the 2nd and third quintiles.

Discussion: The inpatient service utiliazation by the poor in county A is much higher compared to that in other two counties, the most important reason being the reimbursement procedure that is more simple and convenient for the poor. We could increase the poor's inpatient service utilization by improving the linkages between NCMS and MA.