Health Care utilization with China's New Rural Cooperative Medical Scheme: Update Evidence from the 2011-2012 China Health and Retirement Study
Launched in 2003, China’s New Rural Cooperative Medical Scheme (NRCMS) is a health plan that covers the largest number of enrollees in the world. Most of its enrollees had no experience with health insurance prior to their NRCMS enrollment, hence the question of whether NRCMS could change their health care utilization behavior. Using a 2008 pilot data from Zhejiang and Gansu, we showed in an earlier study that NRCMS enrollees were indeed more likely to underutilize outpatient services than those who remained uninsured, for reasons unknown. In this study we use a national sample to reexamine the association between NRCMS enrollment and health care utilization.
Method:
We use the 2011-2012 wave of China Health and Retirement Study to explore the association between NRCMS enrollment and health care utilization. We use four outcome variables to operationalize the construct of health care utilization: 1. whether one underused outpatient care (defined as not seeing a health care professional when sick during the previous month); 2.whether one delayed inpatient care (defined as not getting hospitalized during the last 12 months even when a health care professional suggested so); 3. number of outpatient visits during the previous month; 4. number of inpatient visits during the last 12 months. We use survey-weighted logistic regressions to analyze the association between the first two outcome variables (underuse outpatient care and delay inpatient care) and NRCMS enrollment, and we use survey-weighted zero-inflated negative binomial models to analyze the association between the last two outcome variables (number of outpatient visits and number of inpatient visits) and NRCMS enrollment. We also include the following covariates in the four models: age, gender, education, income level, household registration status (“hukou,” urban vs. rural), community facility environment, self-reported health problems and geographic region (East China vs. Central China vs. West China). We interacted the region dummies with NRCMS given the different implementation paces across different regions.
Results
Compared with those uninsured, those NCRMS enrollees are significantly less likely to delay inpatient care (Odds ratio=0.464, Standard Error=0.170), have significantly more outpatient visits (Incident Rate Ratio=1.441, Standard Error=0.220) and significantly more inpatient visits (Incident rate ratio=2.066, Standard Error=0.498), after we controlled for demographic covariates and self-reported health problems in the four models. NCRMS enrollees are also less likely to underuse outpatient care than those uninsured, though the difference is not statistically significant (Incident rate ratio=0.520, Standard error=0.226). The only statistically significant pattern from the interactions between the region dummies and NCRMS enrollment is the negative association between an NCRMS enrollment in Central China and the number of inpatient visits (0.646, 0.141).
Discussion
As we tried to interpret the statistically significant negative association between NCRMS enrollment and outpatient care utilization in our analysis of the 2008 CHARLS pilot, we proposed that the low reimbursement rate of NCRMS outpatient care (while the rate for inpatient care was higher) might have discouraged new enrollees to seek outpatient care. The positive association between NCRMS enrollment and outpatient care utilization from the 2011-2012 CHARLS sample could result from the improved reimbursement for outpatient care.