BETHESDA, MD – Hospitals that provide more intensive and costly care do not provide better-quality care, according to a study published in Health Affairs.

The study looks at care given to Medicare beneficiaries for three common conditions: acute myocardial infarction (or heart attack), pneumonia and congestive heart failure.

The new research by Laura Yasaitis, Elliott Fisher and Jonathan Skinner of Dartmouth and Amitabh Chandra of Harvard represents one of the first nationwide analyses of quality and spending at the level of individual hospitals.

"We found no evidence that hospitals with higher spending provided better care, whether we looked at all hospitals across the country or limited our study to academic medical centers or hospitals within a single region. In fact, in some cases hospitals that spent more provided worse care," said Yasaitis, a student at Dartmouth Medical School and a researcher at the Dartmouth Institute for Health Policy and Clinical Practice.

Previous studies found that higher spending does not lead to better care on a regional level. However, the researchers insisted that because efforts to improve the quality of care and decrease unnecessary spending are likely to start at the level of individual hospitals, it's important to ascertain the relationship between spending and quality at the hospital level.

"The fact that some hospitals in the same region are able to provide exemplary care at lower costs points to the need for better reporting of both costs and quality and for a greater understanding of what processes lead to this improvement in performance," said Chandra, a professor at Harvard University.

The authors defined a measure of spending that is not affected by the severity of patients' illnesses by examining end-of-life (EOL) spending among the 2,172 U.S. hospitals with complete data on use, spending and quality. Adjusted for the age, race, sex and disease mix of each hospital's patients, average EOL spending was $16,059 for the lowest-spending fifth of hospitals; for hospitals in the highest-spending fifth, average EOL spending was $34,742.

The study results revealed that – when looking at hospitals of all types across the country – there was a statistically significant negative relationship between hospitals' EOL spending and overall quality, meaning hospitals that spent more actually performed worse on overall quality measures than lower-spending hospitals. When the analysis was limited to academic medical centers, there was no positive association between quality and spending.

The only statistically significant finding was a negative relationship between EOL spending and the quality of treatment for heart attacks.

To assess the effect of geographic differences in care intensity, the researchers repeated their analysis while controlling for Hospital Referral Regions (HRRs), as defined by the Dartmouth Atlas of Health Care – essentially comparing each hospital only to others in the same region.

They found that controlling for HRRs eliminated the negative correlation between EOL spending and the quality of treatment for heart attacks, but strong negative correlations remained between spending and both overall quality and the quality of treatment for pneumonia.

"This shows that the failure of higher-spending hospitals to produce better care reflects more than the fact that quality is not higher in higher-spending regions," said Fisher, director of the Center for Healthcare Research and Reform at Dartmouth. “Within each region, some hospitals are able to do more with less, achieving better quality while spending fewer dollars.”