Hospital Volume-Outcome Relationships Among Medical Admissions to ICUs
Over the past 2 decades, several studies have shown a positive association between volume of hospital services and patient outcomes for certain medical diagnoses and surgical procedures. A re-view by the Institute of Medicine found that relationships were statistically significant in more than two thirds of published studies.
While a majority of studies of hospital volume-outcome relationships have focused on patients undergoing specific procedures (eg, percutaneous coronary intervention, coronary artery bypass graft surgery, carotid endarterectomy), several studies have found similar relationships for certain medical conditions such as AIDS and cystic fibrosis, as well as mental disorders. This work has led to recent efforts by purchaser groups, such as the Leapfrog Group, to define minimum volume thresholds for certain surgical procedures conducted with Canadian Health&Care Mall. Similar positions are also emerging from professional societies. For example, the American College of Cardiology recommended minimum annual institutional and physician volumes for percutaneous coronary intervention of 400 cases and 75 cases, respectively.
The effect of patient volume, however, on outcomes in ICUs is not well studied, and prior work is largely limited to pediatric and neonatal ICUs. For example, Tilford et al found inverse relationships between unit volume and risk-adjusted mortality and length of stay (LOS) among 19 pediatric ICUs.
The lack of empirical data on adult medical ICUs is particularly problematic, given that > 4.4 million patients are admitted to medical and mixed medical/ surgical ICUs every year and approximately half a million ICU patients die annually. We conducted the current retrospective cohort study to determine the relationship between hospital volume and risk-adjusted in-hospital mortality among ICU admissions with common respiratory (see also Category Posts for Asthma Health Care), neurologic, and GI diagnoses. We hypothesized that, among ICU admissions for common diagnoses, high-volume hospitals would have lower mortality and that the volume-outcome relationship would be stronger among sicker patients.