Ayanian JZ, Weissman JS, Chasan-Taber S, Epstein AM; Association for Health Services Research. Meeting.
Abstr Book Assoc Health Serv Res Meet. 1997; 14: 168.
Department of Health Care Policy, Harvard Medical School, Boston, MA 02115, USA.
RESEARCH OBJECTIVE: To assess the quality of care provided for patients with congestive heart failure and pneumonia in teaching and non-teaching hospitals. STUDY DESIGN: We studied a stratified random sample of 1767 elderly Medicare beneficiaries hospitalized in 1991 and 1992 for congestive heart failure and pneumonia in 71 major teaching hospitals (>0.25 interns and residents per bed), 172 other teaching hospitals, and 328 non-teaching hospitals in Illinois, Massachusetts, New York, and Pennsylvania. Quality of care was assessed with methods that have been shown to predict 30-day mortality for these conditions, including structured implicit review by physicians and detailed explicit process measures collected by nurses. Quality ratings were standardized so that one unit represented one standard deviation in quality across the full cohort. Thus, differences in ratings can be interpreted as effect sizes. Linear regression was used to adjust for patient age, sex, race, median household income by Zip code, state, RAND Sickness-at-Admission Score, and hospital teaching status, for-profit status, and location in a Metropolitan Statistical Area. PRINCIPAL FINDINGS: Physician reviewers rated overall quality as very good to excellent significantly more often in major and other teaching hospitals than in non-teaching hospitals for patients with congestive heart failure (65.0%, 45.4%, 33.2%) and pneumonia (62.3%, 45.4%, 38.1%) (both P<0.001 by Pearson chi-square test). Adjusting for demographic, clinical, and hospital factors, implicit ratings of care were significantly better in major and other teaching hospitals than non-teaching hospitals for both congestive heart failure (effect sizes of 0.65 and 0.17, respectively) and pneumonia (0.46 and 0.30). Adjusted explicit ratings were also significantly better in both types of teaching hospitals for congestive heart failure (0.35 and 0.26) and pneumonia (0.26 and 0.32) (p<0.01 by Student's t test for each two-way comparison with non-teaching hospitals). On explicit subscales for each condition, physicians' cognitive and technical diagnostic care were rated better in major and other teaching hospitals than in non-teaching hospitals, but nursing care was rated better in non-teaching hospitals than in major teaching hospitals (all p<0.05). CONCLUSIONS: Teaching hospitals performed significantly better than non-teaching hospitals on implicit and explicit measures of the overall quality of care which have been shown to predict mortality within 30 days of admission, and the adjusted differences in quality reflect effect sizes of moderate magnitude. RELEVANCE TO CLINICAL PRACTICE AND POLICY: These differences in quality of care are particularly important because congestive heart failure and pneumonia are the two most common causes of hospitalization in the Medicare program, together accounting for over one million admissions in 1994. At a time when teaching hospitals face increasing financial constraints from public and private payers, these findings demonstrate that the added clinical value of teaching hospitals extends beyond tertiary care to common medical conditions.
Publication Types:
Keywords:
- Aged
- Continental Population Groups
- Heart Failure
- Hospitals
- Hospitals, Private
- Hospitals, Teaching
- Humans
- Illinois
- Massachusetts
- Medicare
- New York
- Pennsylvania
- Pneumonia
- economics
- education
- mortality
- hsrmtgs
Other ID:
UI: 102233652
From Meeting Abstracts