Hospital volume may influence condition-specific mortality rates

Admission to high-volume hospitals was associated with reductions in 30-day mortality in patients with acute myocardial infarction, heart failure and pneumonia — but only up to a certain threshold of patient volume, study results indicated.

“[U]nderstanding this relationship may help determine whether operator experience accounts for better outcomes, as reported in studies examining percutaneous coronary intervention and coronary artery bypass grafting,” the researchers wrote.

They conducted cross-sectional analyses of Medicare administrative claims for fee-for-service beneficiaries hospitalized between 2004 and 2006. The researchers included data on 734,972 myocardial infarction (MI) hospitalizations from 4,128 hospitals; 1,324,287 heart failure (HF) hospitalizations from 4,679 hospitals; and 1,418,252 pneumonia hospitalizations in 4,673 hospitals. Using hierarchical logistic regression for each condition, the researchers estimated “the change in odds of death within 30 days associated with an increase of 100 patients in the annual hospital volume.”

Increased condition-specific hospital volume was associated with a reduced 30-day death rate for patients with acute MI (adjusted OR=0.89; 95% CI, 0.88-0.90), HF (adjusted OR=0.91; 95% CI, 0.90-0.92) and pneumonia (adjusted OR=0.95; 95% CI, 0.94-0.96). However, these benefits were attenuated as hospital volumes increased beyond certain thresholds (per 100 additional annual patients). Reductions in 30-day mortality were no longer significant when hospitals attained a volume of 610 (95% CI, 539-679) patients with acute MI, 500 (95% CI, 433-536) patients with HF and 210 (95% CI, 142-284) patients with pneumonia, according to the researchers.

“Because more than 60% of patients were receiving care at large-volume hospitals, whereas approximately 15% were receiving care at small-volume hospitals, a policy aimed at universally increasing hospital volume in order to reduce death rates would not have uniform benefits,” they wrote. “Nevertheless, large proportions of patients received care at ‘below-threshold’ hospitals, particularly for acute MI.” – by Eric Raible

Ross JS. N Engl J Med. 2010;362:1110-1118.

This was an interesting study, but I am not sure if the results will have much of an effect on practice. Part of the problem with these types of studies is that although there may be a difference between small institutions or hospitals and the larger facilities, what we should be looking at is: What is it about the large hospitals that help them obtain better outcomes that can be used in the smaller hospitals to improve theirs? That is unfortunately something that you do not get from this kind of study. The results provide a big picture but do not tell you what you need to know to have an effect on care.

The other thing I thought was interesting about the study was the difference in magnitude between what the researchers defined as small hospitals and large hospitals. They seemed to lump medium-sized and larger hospitals together. A very small hospital is often going to be in a small community, and what kinds of resources it will have and what it will be able to bring to bear is unclear. It is difficult to generalize and define what a small hospital is in this study. There has to be some intermediate-sized facility included. Another thing to consider, and the researchers point this out in the article, is that there were some smaller hospitals that did a very good job and some larger hospitals that did not do that well, so there is something to be learned on both sides. For me, it comes back to processes of care in both large and small hospitals — and not so much the gross outcome measures.

– Larry Dean, MD
Professor of Medicine and Surgery
University of Washington School of Medicine, Seattle

Follow InfectiousDiseaseNews.com on Twitter.