The impact of Taiwan's national health insurance on physician and patient behavior: Vaginal delivery versus cesarean section

On March 1, 1995, the government of Taiwan implemented a National Health Insurance (NHI) that had increased insured population from 60% to 97% by July 1998. NHI applies several policies to control the rapid growth of health care expenditure, such as case payment and cost-sharing. This study analyzes the influence of NHI on patients and health care providers in deciding between vaginal delivery and cesarean section (C-section). The data consist of 25,946 patients who were delivered by 26 physicians between March 1, 1994 and February 29, 1996 in three branches of Chang Gung Memorial Hospital (CGMH) in Taiwan.;The study develops a theoretical and two empirical models. The theoretical model is used to describe patient and physician behavior and to derive hypotheses regarding factors that influence C-section rates. The empirical analysis consists of two parts: (1) analysis of physician practice variation. Linear regression is used to analyze physicians' monthly C-section rates with respect to professional level, financial incentive, the implementation of NHI, and other physician characteristics. (2) analysis of C-section probability. Logit regression is used to analyze effects of a variety of factors on the probability of C-section, including insurance status, a cultural factor, physician propensity to perform C-section, and convenience. Panel data analysis and other econometric techniques are used to improve model estimation.;Results show that physicians' monthly C-section rates are significantly associated with professional level and financial incentive. In addition, their monthly C-section rates had decreased by 2% (p < 0.05) after the introduction of NHI. A patient's probability of C-section was studied in three types of patients: (1) patients covered by NHI after its implementation; (2) patients insured before NHI by three major social insurance plans; (3) patients with no insurance coverage before NHI. The probability of C-section is significantly associated with insurance status, physician's propensity to perform C-section, and convenience. Patients insured under NHI were more likely to select C-section than the previously uninsured, but their probabilities of C-section were lower than in those previously insured.;Findings of the study have important implications for policy formulation. Financial motive influences both patients and physicians. Failure to consider the policy comprehensively will vitiate its effectiveness.