DETERMINANTS OF CESAREAN SECTION: EXPLAINING RECENT CHAN

在20世纪60年代、70年代和80年代的大部分时间里，美国的总体剖腹产率迅速上升。这一比率在20世纪80年代末趋于平稳，并开始缓慢下降，根据最新的数据，这一趋势仍在继续。1995年，这一比例为20.8%，许多权威机构仍认为这一比例过高。20世纪80年代的高剖腹产率引起了医学和经济方面的广泛关注。是什么导致了这么高的利率？经济因素在多大程度上起了作用？剖宫产率的稳定和下降也提出了重要的问题。这种变化的本质究竟是什么？它的解释是什么？拟议的研究试图回答这些问题，在一个精心设计的决定因素研究的背景下，使用一个异常丰富的数据集。我们的具体目标是构建和开发纽约州1986年和最近可用年份（可能是1996年）的适当数据集，并每年对剖宫产概率的决定因素进行两阶段回归研究；然后将结果用于模拟练习，以确定1986年至1996年间剖宫产手术变化的原因，区分行为变化（系统参数）和变量变化（系统独立变量）。这两个阶段包括使用胎儿窘迫和难产标签作为剖宫产指征的决定因素；以及（使用回归指标中这些适应症的预测值）剖宫产概率的决定因素。有证据表明，使用这些标签是内生的。在模拟中，1996年的变量与1986年的参数一起使用，以确定如果1986年的行为保持不变，1996年将发生多少次剖宫产，以及剖宫产率将是多少。结果可以与1996年的实际值进行比较。假设剖宫产手术数与实际剖宫产手术数之间的差异被解释为1986-1996年行为改变导致的额外（正或负）1996次剖宫产手术。剖腹产率也是如此。该研究有助于我们对剖宫产决策的理解；以及我们对非临床因素对临床决策的影响的理解，包括对诊断和治疗的选择。除其他事项外，这使人们更容易理解政策在这一重要领域已经做出和能够做出的贡献。

The overall c-section rate rose rapidly in the United States in the 1960s, 1970s, and most of the 1980s. The rate leveled off in the late 1980s, and began a slow decline, which continues according to the most recent data available. It stood at 20.8 percent in 1995, a rate still regarded by many authorities as too high. The high c-section rate of the 1980s attracted widespread concern, both on medical and on economic grounds. What was responsible for the high rate? To what extent did economic factors play a role? The leveling off and decline of the c- section rate has also raised important questions. Exactly what is the nature of the change? What is its explanation? The proposed research seeks to answer such questions, within the context of a carefully designed determinants study using an unusually rich data set. Our specific aims are to construct and develop an appropriate data set for New York State for 1986 and the most recent available year, probably 1996, and for each year to perform a two-stage regression study of the determinants of the probability of cesarean delivery; and then to use the results for a simulation exercise, to determine the causes of changes in cesarean sections between 1986 and 1996, distinguishing between changes in behavior (system parameters) and changes in variables (system independent variables). The two stages consist of determinants of use of the labels fetal distress and dystocia as indications for cesarean section; and (using predicted values of these indications among the regressors), determinants of the probability of cesarean delivery. There is evidence suggesting use of these labels is endogenous. In the simulation, the 1996 variables are used with the 1986 parameters, to determine how many cesarean sections would have occurred in 1996 if 1986 behavior had been maintained, and what the cesarean section rate would have been. The result can be compared with the actual 1996 values. The difference between the hypothetical and the actual number of cesarean sections is interpreted as the additional (positive or negative) 1996 cesarean sections attributable to 1986-1996 behavior change. The same is true of the c-section rate. The research contributes to our understanding of cesarean delivery decisions; and to our understanding of the influence of non-clinical factors on clinical decisions, including choice of diagnosis as well as of treatment. Among other things, this makes it easier to understand the contributions that policy has made and can make in this important area.