March 26th, 2010

Hospital Volume and Outcomes: Size Matters, but Only to a Point

This week we welcome CardioExchange Contributor Joseph Ross, MD, to answer questions about his study in NEJM, Hospital Volume and 30 Day Mortality for Three Common Medical Conditions (co-authored with CardioExchange Editor, Harlan Krumholz, MD).

CardioExchange Editors: What does this research add to our current knowledge about volume and outcomes?

The relationship between greater volume and lower mortality outcomes had been fairly well demonstrated for surgeries and procedures, but less was known about the relationship between volume and mortality for medical conditions. We studied hospitalizations for AMI, heart failure, and pneumonia, three of the most common reasons for hospitalization. We found that, on average, higher volume hospitals achieve lower mortality rates. However, our study takes the next step in understanding the volume-outcome relationship by demonstrating that while volume matters, it only matters to a point. The association between greater volume and lower mortality progressively attenuates as volume increases, such that we identified a volume threshold after which we would predict there is no longer a significant association between volume and mortality. 

CardioExchange Editors: Does this mean that patients should generally choose high volume hospitals over lower volume hospitals?

On the contrary, our research suggests that patients should attempt to receive care at hospitals that are achieving better outcomes! In addition, there are other reasons why we believe volume is not a good surrogate for quality and should not be used to choose a hospital. First, we studied care for three acute medical conditions, and if a patient urgently needs care, it’s best to go to whichever hospital is closest — where they can receive care more promptly. Second, in our health care delivery system, patients rarely have a choice between a small hospital and a large hospital. The small hospitals in our study managed between 10-18% of patients for the three conditions. These small hospitals serve as important resources for their communities, providing acute care nearer to people in remote areas. Finally, among the more interesting findings from our paper was the substantial amount of variation in outcomes, among hospitals of all sizes.This speaks to the capacity of some small hospitals to achieve excellent outcomes. Our research suggests that we need to identify the strategies and services being offered by hospitals achieving excellent outcomes. Perhaps they do a better job of coordinating discharge care or following-up with patients after discharge. Large hospitals may be more likely to have the economies of scale to efficiently offer these services, but we should work toward ensuring that all hospitals have the capacity and resources to provide them.

CardioExchange Editors: Has anyone looked at the cost associated with the treatment of these common conditions in high volume versus low volume hospitals? If so, how might these data influence hospital reimbursement based on whether an institution is considered a Center of Excellence?

To my knowledge, no one has examined that. This question certainly raises an important point — but I would not suggest we use hospital volume as a marker to make reimbursement decisions such as these, unless it is to be sure that the smallest hospitals have sufficient financial capacity to offer key strategies and services to improve outcomes, such as a nurse dedicated to following patients after discharge to ensure that they keep a scheduled appointment with their community physician or had an opportunity to order their prescription medications.

CardioExchange Editors: How do you think physicians should act on these data?

Our study is the first to demonstrate the dynamic relationship between volume and outcomes, showing that the magnitude of the association between the two narrows at greater and greater volumes. This raises all sorts of interesting questions for clinical practice that deserve inquiry. Is this dynamic true for surgical care and procedures as well? Is there some volume-threshold that exists for care of all medical conditions? Does it extend to physician volume as well? As I said above, I do not think that physicians should take our study and begin to refer exclusively to higher and higher volume hospitals — the relationship is more complicated than that. Physicians should be working to identify the strategies that lead to better outcomes of care and ensuring that these strategies are in place at their hospital.

Comments are closed.