The tension between expertise and convenience in complex medical care

This Study of the Week made me think. I hope it does the same for you.

A truth the medical profession does not like to discuss is the gradient of skill amongst doctors. It’s weird, because gradients of skill exist in every part of life. Run a local 5K and it will be obvious.

In the place of skill gradients, medical leaders like to discuss access to care. In a perfect world, people who live in rural settings should have equal access to high-tech medical care.

Well.

Given these two ideas, let’s look at a nice observational study published in JAMA Cardiology regarding procedural volume and outcomes for two of the more complex procedures in cardiology—TAVR and M-TEER. (TAVR = transcatheter aortic valve implantation and M-TEER = mitral valve transcatheter edge-to-edge repair, AKA, mitral clipping.)

Our government actually designed a superb healthcare policy when they required that all patients who have these procedures have their data entered into a mandatory registry wherein outcomes (at least short-term) are recorded. The mandatory nature of this registry makes it stand out against voluntary registries (or survey’s) where doctors can exclude patients that have bad outcomes.

I have heard the Society of Thoracic Surgeons (STS)/American College of Cardiology (ACC) Transcatheter Valve Therapies (TVT) Registry, called a national gem. Surgeon Joseph Bavaria likes to use the following quote: (I am not sure of the origin)

Science tells us what we can do, trials tell us what we should do, and registries tell us what we are doing.

The study is fairly simple. It makes correlations between procedural volume and outcomes. They have big numbers (N = 358,000 patients who had TAVR and 51,400 who had M-TEER).

Here are the two main results:

• In adjusted analyses, a higher risk of 30-day mortality (2.4% vs 2.0%; odds ratio [OR], 1.13; 95% CI, 1.02-1.26; P = .02) and in-hospital complications (OR, 1.09; 95% CI, 1.03-1.16; P = .005) was observed for low-volume TAVR operators (<15/y) compared with high-volume operators (>37/y).

• For MTEER, in-hospital complications (OR, 1.31; 95% CI, 1.11-1.56; P = .002) were higher for low-volume operators (<8/y) compared with high-volume operators (>16/y), while 30-day mortality was not different (OR, 1.16; 95% CI, 0.96-1.41; P = .12).

The authors conclude:

We demonstrate a persistent inverse association between operator volume and patient outcomes for both TAVR and MTEER. These findings may help inform future policies aimed at ensuring optimal outcomes.

Conclusion

The first thing to write is to congratulate the authors and the structural cardiology community for publishing these provocative data. It’s laudable—and we should commend academics when they do commendable things.

Now I suppose before we wrestle with the implications of these results, we should assess the validity of the findings.

Here I would be confident that the worse outcomes with lower volume operators is signal not noise. Mainly because of conditional probability. Bayesian thinking has it that anything that requires high levels of skill improves with repetition. Also reassuring is the mandatory nature of the registry; there is no cherry picking.

Now we have two tensions.

The first tension is that every high-volume operator was once a low-volume operator. When a doctor gets out of training, he or she has a procedural log documenting training cases. But that’s like hitting in a batting cage. The real test doesn’t start until regular independent practice begins. The question is when does an operator cross the line between average and highly skilled. I don’t know the answer. And surely no patient can know either.

Second tension: how do you incorporate this data into the calls for greater access to the procedure in underserved populations?

I have had two experiences in different geographies.

In my home of Louisville KY there are many centers doing these procedures. We are not a big city so some of these places are surely low in volume. But patients have tons of choices—and the drive is short.

When I visited Calgary a few years ago, I learned that complex procedures are done in only two centers in the entire province of Alberta—Calgary and Edmonton. Thus, there are more procedural centers in one corner of KY than the entire province of Alberta.

The advantage of the Canadian approach is that your doctor is one of the busiest and experienced in the world—a true expert. The disadvantage is the wait and drive.

I will put my cards down: I believe the Canadian system wherein expertise is concentrated in a few centers is preferable. It may not have been 20 years ago when procedures were a fraction of the complexity they are now.

If I were a patient, I’d want to know my doctor was not in the second inning of his/her career, or that they (and their hospital) did only a few cases per year.

Of course, one wish I have is that it was easier for patients to know this information.

One caveat: the practice of medicine is always variable. You can seek out what you believe is the best doctor in the best center in the world and have a terrible outcome. And…you can blindly pick the closest hospital with the best parking (and a piano in the lobby) and have a great outcome.

Yet, on average, we prefer high-volume doctors and hospitals because the odds favor a better outcome. How to accomplish that for all is a tough darn question.