What do 100 top hospitals lists tell us, our patients?

The annual publication of the 100 Top Hospitals — Benchmarks for Success ratings is now in its eighth year.

The ranking is published by HCIA-Sachs in Baltimore. With that track record in place, a group of researchers recently conducted the first study of how top 100 hospitals compare with their nonranked peers.

A team led by Jersey Chen, MD, MPH, a researcher at Yale University in New Haven, CT, examined performance in acute myocardial infarction (AMI).1 "Whether or not the top 100 hospitals have different outcomes, processes of care, or resource use compared with other hospitals has not been studied before," wrote the investigators.

The findings show:

• Peer organizations and top 100 hospitals had similar results on 30-day mortality, use of aspirin, beta-blockers, and reperfusion therapy (thrombolytic medications or angioplasty).

• Top 100 hospitals had lower length of stay and in-hospital costs, as well as similar or lower re-admission rates.

Reason for studying the 100 top hospitals

Among the plethora of report cards and similar quality measures, this ranking is one of the less common variety which incorporates financial and clinical performance. The majority of studies assess clinical outcomes such as mortality rates, or process-based measures such as the proportion of patients who receive particular therapies or procedures.

Highlights of 100 top hospitals methodology

The 100 top hospitals rating is compiled by the health care consulting firms of HCIA-Sachs Inc. in Baltimore and William M. Mercer Inc. in New York City and other cities throughout the United States.

Data used in calculations

1. MedPAR (Medicare Provider Analysis and Review) for clinical indicators.

2. Medicare cost report for cost indicators.

Comparison groups

1. Teaching hospitals with cardiovascular residency programs.

2. Teaching hospitals without cardiovascular residency programs.

3. Nonteaching hospitals.

Performance measures

1. Risk-adjusted medical patient mortality index.

2. Risk-adjusted surgical patient mortality index.

3. Risk-adjusted postoperative infection index.

4. Risk-adjusted postoperative hemorrhage index.

5. Percentage of coronary artery bypass graft (CABG).

6. Percentage of percutaneous transluminal coronary angioplasty patients with CABG surgeries during the same admission.

7. Severity-adjusted average length of stay.

8. Wage- and severity-adjusted average cost.

Types of "peer" hospitals compared with top 100 hospitals

1. Small rural (fewer than 200 beds).

2. Small urban (fewer than 200 beds).

3. Nonteaching (at least 200 beds).

4. Teaching (at least 200 beds).

For study purposes, a top 100 hospital was defined as an institution that appeared in the ranking at least once between 1994 and 1996. Top 100 hospitals were compared with peers on:

• Outcomes (30-day mortality).

• AMI treatment (in-hospital use of aspirin, beta-blockers, and reperfusion therapy — defined as thrombolytic medications or angioplasty).

• Resource consumption (length of stay, costs, re-admission rates, and use of cardiac procedures).

Key findings from a sample of 149,177 patients in 4,672 peer hospitals

• A significantly higher proportion of top 100 hospitals included private, for-profit organizations, especially within the small rural, small urban, and nonteaching hospital categories.

• In the nonteaching and teaching hospital categories, the top 100 were more likely than others to have on-site cardiac catheterization, coronary angioplasty, and open-heart surgery facilities.

• Among ideal candidates for aspirin and beta-blocker therapies, the proportion of patients receiving therapy in the top 100 hospitals was similar to that of patients in peer hospitals.

• No significant difference appeared in reperfusion therapy between the top 100 hospitals and peer hospitals in nonteaching and teaching institutions.

• The mean observed and risk-adjusted 30-day mortality rates for the top 100 hospitals did not differ significantly from peers across all hospital groups.

• On average, the top 100 hospitals had lower in-hospital costs per AMI admission (from $1,014 lower in nonteaching hospitals to $1,855 lower in small rural hospitals).

• Length of stay correlated highly with in-hospital cost, explaining 61% of the variation.

• The percentage of patients who survived the initial hospitalization and were re-admitted for all causes, or for reinfarction within 180 days of discharge, was either the same for peer groups and top 100 hospitals or lower in top 100 hospitals.

• Total days in the hospital 180 days after admission, or after discharge, was significantly lower for top 100 hospitals compared with peers.

• Considerable variation appeared within respective hospital categories for 30-day mortality, use of aspirin and beta-blockers, length of stay, and in-hospital costs. Taken as a group, the average length of stay in the top 100 hospitals was lower than for peer hospitals, but there were many facilities whose individual lengths of stay were lower than the overall average for the top 100. (See tables, p. 89.)

"It is important to emphasize that the top 100 hospitals were not homogeneous in their performance, and this substantial overlap should temper decisions solely made on the 100 top hospitals designation," wrote Chen and colleagues. They emphasized that, "many peer hospitals had comparable or better performance than the average top 100 hospitals in a number of areas."

The goal of ranking is to select hospitals with superior performance in financial management, operations, and clinical practices.

However, the researchers cautioned, "The results of this study suggest that the report may be more appropriately suited to identifying hospitals with lower lengths of stay and costs rather than hospitals with superior outcomes or higher quality of care."


1. Chen J, Radford MJ, Wang Y, et al. Performance of the 100 Top Hospitals’: What does the report card report? Health Aff 1999; 18:53-68.