Academic hospitals in general and nonacademic hospitals located in the Northeast and Midwest appear to offer greater quality than nonteaching facilities or those in other geographic regions for certain conditions, according to a recent study by Ashish K. Jha, M.D., and colleagues at the Harvard School of Public Health.
The authors found that academic hospitals performed slightly better than nonacademic facilities in treating heart attack and heart failure, but underperformed in the case of pneumonia. Not-for-profit hospitals significantly outscored for-profits in all measures in those conditions, they said.
However, the data do not support the idea that “'good' hospitals are easy to identify or are consistent in their performance across conditions,” the authors pointed out.
The study was the first to use national hospital data submitted as part of the Hospital Quality Alliance (N. Engl. J. Med. 2005;353:265–74).
The HQA is a joint effort of the Centers for Medicare and Medicaid Services, the Joint Commission on Accreditation of Healthcare Organizations, the American Hospital Association, and consumer groups, including the AARP. Under the Medicare Modernization Act, hospitals have financial incentives to report quarterly to CMS on a specific set of quality indicators. Initial data became publicly available in November 2004, which led Dr. Jha and colleagues to conduct their analysis.
In all, 3,558 hospitals reported on their performance in the first half of 2004, citing quality measures for acute myocardial infarction, heart failure, and pneumonia. To be included, each hospital had to report on at least 25 discharged patients. Of the participating hospitals, 16% were for-profit, 8% were members of the Council of Teaching Hospitals, and 62% were in urban settings.
The authors measured the mean performance for the 3,558 hospitals and variability of performance across the country.
They also sought to determine whether a high level of performance in treating one disease translated into equally good care in the other conditions. They investigated whether profit status, number of beds, geographic area, and academic involvement affected performance.
The heart attack measures included whether aspirin and ?-blockers were given within 24 hours of admission and at discharge and if ACE inhibitors were given to patients with left ventricular systolic dysfunction. For heart failure, the measures included assessment of left ventricular function and whether an ACE inhibitor was given. For pneumonia, hospitals were measured on the timing of initial antibiotics, vaccination, and assessment of oxygenation.
The hospitals' scores reflected the proportion of patients who satisfied the criterion for the performance measure.
Overall, hospitals did best at conducting an oxygenation assessment in pneumonia patients; the mean performance score was 98%, plus or minus 5%. Hospitals had their lowest scores in rates of vaccinations for pneumonia: 43%, plus or minus 27%.
There were huge variations among regions. For the pneumonia composite score, there was a 23% difference between top-ranked Oklahoma City and bottom-ranked San Bernardino, Calif. There was a 12% difference between high and low performers on heart attack, and a 21% difference between the top and bottom for heart failure.
The authors found that high performance scores for acute myocardial infarction predicted equally good performance in heart failure, but not for pneumonia. For instance, 73% of the hospitals that were in the top decile of scores for acute MI were in the top quartile for heart failure, and less than 1% were in the bottom quartile. But only 33% of the hospitals in that top decile for acute MI were in the top quartile for pneumonia.
Limitations of the study include the fact that the investigators limited their evaluation to only 10 measures of the quality of care for 3 clinical conditions, and the results focused on process measures rather than patient outcomes, the authors noted.