Session: What Influences Hospital Quality?
Room: Hollister 306
Time: Mon 16:45-18:15

Presenter: Richard Smith (University of South Florida, St. Petersburg. College of Business)

Discussant: Richard Lindrooth (University of Colorado, Denver)

Abstract

Background: In 1999, the Institute of Medicine’s (IOM) report, “To Err is Human: Building a Safer Health System,” documented the alarmingly high incidence of preventable medical errors occurring in U.S. hospitals. In the wake of these reports, the federal government sponsored the development of discharge-level indicators of patient safety. These indicators allow for the direct comparison of potentially preventable medical errors across hospitals.

Objective: This study analyzes the risk-adjusted rate of preventable errors in hospitals relative to annual discharges, controlling for other relevant factors, such as case mix, nurse staffing, technology, teaching and ownership status. Results are used to estimate the hospital scale that minimizes potentially preventable medical errors, comparing these results to the established literature on scale economies for hospitals, as well as to the more recent literature on service volume and patient outcomes.

Data: Primary data source for this analysis is the National Inpatient Sample (NIS)from the Healthcare Cost and Utilization Project (HCUP) at the Agency for Healthcare Research and Quality (AHRQ), which provides discharge-level information for a nationally representative sample of community hospitals. These data will be supplemented with hospital-level information from the American Hospital Association’s Annual Survey of Hospitals Database. The outcome measure, hospital rate of patient safety, is developed from a select
group of indicators using the latest release of the patient safety indicators from AHRQ.

Methodology: The data are used to construct a panel of hospitals between 2005 and 2007, which allows for controlling unobserved differences across hospitals, including other dimensions of quality. A simple linear, fixed-effects model is used to estimate the log of the hospital annual rate of patient-safety events against the log of annual discharges, the key explanatory variable, and other control variables. Fitted values of the patient safety rates are then used
to estimate the number of annual discharges (as well as bed size) that minimize the rate of potentially preventable errors.

Key Terms

Hospitals, market structure, quality

Authors:

Richard Smith (University of South Florida St. Petersburg. College of Business)