MORE DISEASE: HOW MAJOR A FACTOR IN HIGHER UTILIZATION

更多疾病：高利用率的主要因素

• 批准号: 2871182
• 负责人: Michael Shwartz
• 金额: $ 6.44万
• 依托单位: BOSTON UNIVERSITY (CHARLES RIVER CAMPUS)
• 依托单位国家: 美国
• 财政年份: 1998
• 资助国家: 美国
• 起止时间: 1998-09-30 至 2001-03-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/2871182
• 关键词: Medicare /Medicaid; clinical research; epidemiology; geographic site; health care service evaluation; health services research tag; hospital utilization; human data; human old age (65+); outpatient care; statistics /biometry

Many studies have reported large differences in age/sex adjusted rates of hospital utilization across small areas, often leading to the conclusion that hospitalization rates in the higher rate areas can be reduced to those in lower "benchmark" areas. An assumption underlying these recommendations is that age and sex adjustment has fully accounted for differences in underlying disease in the different areas. The hypothesis that we will examine is that there is more underlying disease in higher rate areas: As a proxy for underlying disease, we will use rates of treatment of an "outpatient only" basis. We will study Medicare admission in Massachusetts in 1995 for 22 medical conditions, 17 of which have shown in earlier work to be in the top 50% in terms of small area variations. From our earlier work, we known that the drive of variations in inpatient admissions are variations in a number of individuals admitted (as opposed to re-admissions). Thus, we will focus on individuals treated in two categories: as inpatients and as outpatients only. We will use the 70 small geographic areas in Massachusetts previously created. For each are and each condition, we will calculate relative rates of individual inpatient admissions, individuals treated as outpatients only, and individuals treated as inpatients or outpatients only as follows: (observed number-expected number)/expected number. Empirical Bayes (EB) techniques will be used to estimate "true" relative rates in order to guard against the impact that extreme rates from small areas might have on results. In the analysis, we will focus on the correlation of inpatient and outpatient EB-estimated rates (a high positive correlation would support the "more disease" hypothesis; a high negative correlation the "practice style" hypothesis); R-squared, indicating the percent of variation in inpatient rates explained by variation in outpatient only rates; and a comparison of EB estimates of systematic variation in individuals admitted to systematic variation in inpatients plus outpatients. Empirical support for the "more disease" hypothesis might shift some research focus from ways to reduce inpatient utilization to study of the causes of more disease.

许多研究报告称，年龄/性别调整后的发病率存在巨大差异。 医院在小范围内的利用，通常会得出这样的结论 高发病率地区的住院率可以降低至 那些处于较低“基准”区域的人。这些背后的假设 建议充分考虑年龄和性别调整 不同地区基础疾病存在差异。假设 我们要检查的是，在较高的人群中存在更多的潜在疾病 率区域：作为潜在疾病的代表，我们将使用 治疗以“仅限门诊”为基础。 我们将研究 1995 年马萨诸塞州 22 名医疗人员的医疗保险准入情况 条件，其中 17 个条件在早期工作中已显示在前 50% 之内 小面积变化方面。从我们早期的工作中，我们知道 住院人数变化的驱动因素是许多方面的变化 被录取的个人（而不是重新入学）。因此，我们将重点关注 接受两类治疗的个人：住院患者和门诊患者 仅有的。我们将使用马萨诸塞州的 70 个小地理区域 之前创建的。对于每个是和每个条件，我们将计算 个别住院患者的相对比率，接受治疗的个人 仅限门诊患者以及作为住院患者或门诊患者接受治疗的个人 仅如下：（观察数-预期数）/预期数。 经验贝叶斯 (EB) 技术将用于估计“真实”相对值 利率，以防范小额利率极端利率的影响 领域可能会产生结果。在分析中，我们将重点关注 住院患者和门诊患者 EB 估计率的相关性（高 正相关将支持“更多疾病”的假设；高 负相关的“实践风格”假设）； R 平方， 表明住院率的变化百分比，解释为 仅门诊费率的变化；以及 EB 估计值的比较 个体的系统变异承认存在系统变异 住院病人加门诊病人。 “更多疾病”的实证支持 假设可能会将一些研究重点从减少住院病人的方法上转移 用于研究更多疾病的原因。