Health insurance in China: variation in co-payments and psychiatric hospital utilization

Economic reform in China 30 years ago virtually eliminated all public health insurance. In the last 10 years, diverse government insurance programs have been implemented, now covering 95% of the population, primarily for inpatient care. While the development of health care in China is an incomplete work in progress and highly variable, it is unclear whether the depth of insurance coverage affects the accessibility, length of stay (LOS) of inpatient mental health services or not.

This study aims to examine the relationship between variation in insurance coverage, accessibility to inpatient mental health care and intensity of care as measured by length of stay (LOS).

Using administrative data from the Guangzhou Psychiatric Hospital (GPH), we used regression models to determine the relationship between the depth of insurance coverage and the likelihood of hospital utilization and LOS net of sociodemographic characteristics and diagnosis.

Between April 1, 2010 and March 31, 2013, 8,478 patients were discharged with ICD-10 psychiatric diagnoses with an average LOS of 75.1 (sd=244.3) days, among which 4,727 (55.8%) patients were first admissions. Logistic regression analysis showed that insurance plans with lower co-payments were significant predictors of multiple psychiatric admissions and longer LOS.

These data point to significant variability in the health insurance coverage in China and indicate a clear need for greater equalization in future years. Although the Chinese government has provided at least shallow coverage to virtually all of its citizens at this stage, further efforts are needed to expand and equalize coverage as economic development proceeds, especially in rural areas.

Although variation in health insurance plans in China are extensive and impact the accessibility and duration of psychiatric hospital care, their impact on outcomes and use of post-discharge outpatient care is unknown and requires further study.