School of Public Policy and Administration, Xi'an Jiaotong University, 76 Yanta West Road, Xi'an 710061, China.
Health Policy. 2011 Dec;103(2-3):255-65. doi: 10.1016/j.healthpol.2011.09.005. Epub 2011 Oct 19.
Only limited empirical studies reported own-price elasticity of demand for health care in rural China. Neither research on income elasticity of demand for health care nor cross-price elasticity of demand for inpatient versus outpatient services in rural China has been reported. However, elasticity of demand is informative to evaluate current policy and to guide further policy making.
Our study contributes to the literature by estimating three elasticities (i.e., own-price elasticity, cross-price elasticity, and income elasticity of demand for health care based on nationwide-representative data. We aim to answer three empirical questions with regard to health expenditure in rural China: (1) Which service is more sensitive to price change, outpatient or inpatient service? (2) Is outpatient service a substitute or complement to inpatient service? and (3) Does demand for inpatient services grow faster than demand for outpatient services with income growth?
Based on data from a National Health Services Survey, a Probit regression model with probability of outpatient visit and probability of inpatient visit as dependent variables and a zero-truncated negative binomial regression model with outpatient visits as dependent variable were constructed to isolate the effects of price and income on demand for health care. Both pooled and separated regressions for 2003 and 2008 were conducted with tests of robustness.
Own-price elasticities of demand for first outpatient visit, outpatient visits among users and first inpatient visit are -0.519 [95% confidence interval (-0.703, -0.336)], -0.547 [95% confidence interval (-0.747, -0.347)] and -0.372 [95% confidence interval (-0.517, -0.226)], respectively. Cross-price elasticities of demand for first outpatient visit, outpatient visits among users and first inpatient visit are 0.073 [95% confidence interval (-0.176, 0.322)], 0.308 [95% confidence interval (0.087, 0.528)], and 0.059 [95% confidence interval (-0.085, 0.204)], respectively. Income elasticities of demand for first outpatient visit, outpatient visits among users and first inpatient visit are 0.098 [95% confidence interval (0.018, 0.178)], 0.136 [95% confidence interval (0.028, 0.245)] and 0.521 [95% confidence interval (0.438, 0.605)], respectively. The aforementioned results are in 2008, which hold similar pattern as results in 2003 as well as results from pooled data of two periods.
First, no significant difference is detected between sensitivity of outpatient services and sensitivity of inpatient services, responding to own-price change. Second, inpatient services are substitutes to outpatient services. Third, the growth of inpatient services is faster than the growth in outpatient services in response to income growth. The major findings from this paper suggest refining insurance policy in rural China. First, from a cost-effectiveness perspective, changing outpatient price is at least as effective as changing inpatient price to adjust demand of health care. Second, the current national guideline of healthcare reform to increase the reimbursement rate for inpatient services will crowd out outpatient services; however, we have no evidence about the change in demand for inpatient service if insurance covers outpatient services. Third, a referral system and gate-keeping system should be established to guide rural patients to utilize outpatient service.
只有有限的实证研究报告了中国农村医疗保健的自价格需求弹性。中国农村医疗保健的收入需求弹性和住院与门诊服务的交叉价格弹性都没有被报道过。然而，需求弹性对于评估当前政策和指导进一步的政策制定是有帮助的。
我们的研究通过基于全国代表性数据来估计三种弹性（即自价格弹性、交叉价格弹性和收入弹性），为文献做出了贡献。我们旨在回答有关中国农村卫生支出的三个实证问题：（1）门诊服务和住院服务哪个对价格变化更敏感？（2）门诊服务是住院服务的替代品还是互补品？（3）随着收入的增长，对住院服务的需求增长是否快于对门诊服务的需求增长？
基于国家卫生服务调查的数据，我们使用门诊就诊概率和住院就诊概率作为因变量的概率回归模型和门诊就诊数量作为因变量的零截断负二项回归模型，来隔离价格和收入对医疗保健需求的影响。我们对 2003 年和 2008 年进行了总体和分离回归，并进行了稳健性检验。
首次门诊就诊、门诊就诊者的门诊就诊次数和首次住院就诊的自价格需求弹性分别为-0.519（95%置信区间为-0.703，-0.336）、-0.547（95%置信区间为-0.747，-0.347）和-0.372（95%置信区间为-0.517，-0.226）。首次门诊就诊、门诊就诊者的门诊就诊次数和首次住院就诊的交叉价格需求弹性分别为 0.073（95%置信区间为-0.176，0.322）、0.308（95%置信区间为 0.087，0.528）和 0.059（95%置信区间为-0.085，0.204）。首次门诊就诊、门诊就诊者的门诊就诊次数和首次住院就诊的收入需求弹性分别为 0.098（95%置信区间为 0.018，0.178）、0.136（95%置信区间为 0.028，0.245）和 0.521（95%置信区间为 0.438，0.605）。这些结果均来自 2008 年的数据，与 2003 年的数据以及两个时期的总体数据结果具有相似的模式。
第一，门诊服务和住院服务对价格变化的敏感性没有显著差异。第二，住院服务是门诊服务的替代品。第三，在收入增长的情况下，住院服务的增长速度快于门诊服务的增长速度。本文的主要发现表明，需要完善中国农村的保险政策。首先，从成本效益的角度来看，改变门诊价格与改变住院价格以调整医疗保健需求一样有效。其次，目前国家医疗改革的指导方针是提高住院服务的报销率，这将排挤门诊服务；然而，如果保险涵盖门诊服务，我们没有关于住院服务需求变化的证据。第三，应建立转诊制度和守门人制度，以引导农村患者利用门诊服务。
