Improving Health Outcomes With Less Cost? Provision of Mobile Clinic in Developing Economies
研究了政府如何通过决定移动诊所的服务频率和容量来最大化社会福利，发现移动诊所能在降低医疗成本的同时提高健康质量，尤其对快速进展的疾病效果显著。
Consider a public healthcare system consisting of a hospital, a mobile clinic (MC), and a population of potential patients. The government is concerned about the system’s healthcare spending and the population’s health outcomes. It decides the frequency and capacity of the MC service to maximize the social welfare that consists of two terms: the system’s long-run average healthcare cost and the population’s average quality-adjusted life year (QALY). We characterize the population’s natural disease progression using a Markov cohort model and derive the average healthcare cost incurred for the system and the average QALY in closed form for a given MC service. We show that the government is more likely to provide the MC service when (i) it puts more weight on the population’s average QALY, (ii) the MC treatment becomes more efficient, or (iii) the hospital treatment cost is much higher relative to the MC treatment cost. We further show that when conditions (ii) and (iii) are met or the disease progresses faster, the provision of the MC service is more likely to result in a win–win outcome, leading to both healthcare cost reduction and QALY improvement. The optimal MC delivery policy highly hinges on the disease’s progressive speed. Specifically, when the MC service is designated to serve a specific target population, once the MC service is provided, it shall be provided either every period or every other period if the disease is fast-progressive. If the disease is slow-progressive, the long-run average healthcare cost exhibits a zigzag pattern with the MC delivery cycle when the relative treatment cost-saving per person induced by the provision of the MC service is positive. Last, we conducted a real case study. Our analysis reveals that the optimal implementation of the MC program results in a 119.5% improvement in QALY and a 5% reduction in healthcare costs. The results are robust in probabilistic sensitivity analysis with a relative performance gap within 11.5% when factoring in the uncertainty of the parameters.