Effects of Audits on the Behavior of Medical Professionals under the Bennett Amendment.
研究美国贝内特修正案下非财务审计对医疗专业人员行为的影响，发现审计预期和实际审计均显著缩短患者住院时长，且实际审计效果更强。
Abstract Peer review monitoring and control processes of Professional Standards Review Organizations (PSROs) under the Bennett Amendment of the Social Security statutes provides a non financial audit environment which is utilized to study effects of audits on behavior of medical professionals. The focus of this research is on the effects of audit per se (i.e., the audit process as distinguished from any resulting audit reports) in both its anticipation and occurrence by reference to data on patient length of stay. The period of audit anticipation, but no audit, exhibited significant reductions in patient length of stay relative to the period of no audit and no audit anticipation. The period of actual audit exhibited even further reductions over the period of audit anticipation only. Distinctions between audit anticipation and actual audit occurrence effects were sharpened by Exploratory Data Analysis (EDA) techniques applied to data from a Boston (U.S.) hospital. Using the universes and hypotheses defined with the help of these EDA procedures, statistical tests were then applied, using an Ottawa (Canada) hospital (not subject to such audits) as a control, and the results were found to be uniformly consistent in all respects with the hypothesized audit effects.