September 27th, 2013

Higher Physician Volume Leads to Lower Mortality in HF

CardioExchange’s Editor-in-Chief, Harlan Krumholz, interviewed lead author Karen Joynt about her Circulation: Heart Failure study, which examined the relationship between physician volume, clinical outcomes, and costs among patients with heart failure (HF). The researchers found that high physician volume was associated with lower mortality, and that the association was particularly strong in lower-volume hospitals and among noncardiologist physicians.

Krumholz: You showed that physician volume is associated with lower mortality for HF, especially in low-volume institutions. Given that most patients see many doctors, how did you determine the volume of patients for each doctor?

Joynt: For this study, we only looked at the attending of record for each patient. If the attending of record was a physician that saw only two Medicare HF patients in 2009, then the patient was classified into the “low-volume” group. We didn’t evaluate whether or not there were additional physicians consulting on that patient’s care.

Krumholz: Are you confident that you were able to control for differences in the types of patients that seen by high- and low-volume doctors?

Joynt: Controlling for differences in patient factors is always tricky when using administrative data, because we just don’t have clinical variables like ejection fraction or creatinine that would let us precisely estimate each patient’s risk. We think the models, particularly for mortality, are reasonably good, but they are certainly not perfect. Intuitively, it seems likely that high-volume physicians see the sickest patients – people refer the most challenging patients to the “experts” rather than the other way around, so if anything this would have biased us against finding a relationship, but it’s hard to know for sure.

Krumholz: Why do you think that the relationship was stronger in low-volume institutions?

Joynt: I suspect that the volume-outcome relationship was strongest in low-volume institutions because the care of an HF patient depends not only on the physician, but on the whole team. The nurses, physical therapists, nutritionists, and social workers play an absolutely critical role in HF patients’ hospitalizations. A low-volume institution may have fewer systems in place, or less-uniform care, and the effect of a physician’s experience may carry more weight. At a high-volume institution, there are a lot more people who have a great deal of expertise in caring for HF patients, so even if you have a relatively low-volume physician, you still have a lot of collective experience surrounding that patient. Additionally, centers that have a lot of HF patients are more likely to have care protocols, discharge protocols, etc., that may improve outcomes even independent of the clinicians involved in the patients’ care – the “systems” effect.

Krumholz: Is there a message here for patients?

Joynt: The main take-away for patients is that both physician and hospital volume matter. If you have HF, it may be worth it to seek out a clinician who treats a lot of HF patients, or perhaps a center that treats a lot of HF patients. It’s also worth stating that it’s not realistic for all patients to see high-volume HF doctors in person if they happen to live someplace that is geographically isolated – but it doesn’t mean they shouldn’t have access to experienced specialists. We need to do a better job of developing partnerships and technologies that allow for things like telemedicine, or remote consultation, to bring the experts to the patients rather than the other way around.

One Response to “Higher Physician Volume Leads to Lower Mortality in HF”

  1. Judith Andersen, AB, MD says:

    Excellent discussion! We all know that there are brilliant specialists who can accomplish amazing outcomes in situations that they have never previously encountered, and often, with minimal help. But for most of us, experience and excellent support systems, are what matter. Building those systems should be a priority, and the paradigms that emerge from large systems should be offered to smaller systems, with contact information. Transferring expertise, whether by electronics or “hands on” strategies, to distant sites where there are good clinicians but with limited experience, who could use simulation technology to develop skills, should be a national priority. Care teams are critical: folks who know what to do, how to do it, have local knowledge and access to support, as well as knowledge of national guidelines and objectives could transform our world.