Focusing on risk-adjusted mortality in public reporting produces misleading figures

356
Lloyd Klein
Lloyd Klein

In Catheterization and Cardiovascular Interventions, Lloyd Klein (Rush Medical College, Chicago, USA) and others outline the 2016 revised Society for Cardiac Angiography and Interventions (SCAI) position statement on public reporting. In this interview, Klein talks to Cardiovascular News about the pros and cons of public reporting and discusses SCAI’s latest recommendations—specifically the need to reduce the importance of risk-adjusted mortality rates.

What are the advantages and disadvantages of public reporting?

The public deserve to know the outcomes of doctors who they are trusting their lives with and to the extent that they able to, they should use public reporting information to make decisions about their own healthcare—such as which doctors to see or whether they need to undergo a certain procedure. Therefore, I think both societies and physicians accept—at least in theory—that public reporting is a positive and good thing.

However, the disadvantages are that we do not really know which data should be reported. Data that include mortality statistics can be misleading because such information represents so many other factors than competence. For example, a centre may have a high mortality rate because it selects higher risk patients rather than because of the competence of its staff.

Risk adjustment is used to correct for the level of risk of the patient being treated but, in this position paper, we argue that this is not the final answer. Alone, it will not give a balanced view of the quality of a centre—it is one many of factors that we need to review if we want the public to be able to accurately assess quality.

Also, our paper makes it clear that there are no statistically valid approaches to assessing the individual operator. Applying methods derived from programmatic algorithms and applying them to operators is not a statistically valid technique.

Are you concerned that public reporting, in its current form, could lead to risk-averse behaviour?

US data indicate that in states with public reporting, the baseline acute risk of patients undergoing coronary intervention is lower than those of patients undergoing procedures in states without public reporting. As you cannot correct the data for the patients who do not undergo the procedure, interpreting this information is difficult but—as we lay out in the paper—you do get the impression that there is some kind of selection process going on. Therefore, in states with public reporting, physicians may not be choosing patients with the highest risk to undergo coronary procedures, but these patients are also the patients who are most likely to benefit from the procedure. Therefore, that is a negative consequence of public reporting for both the patient and for public policy.

 How does your paper try to address the limitations of the current public reporting approach?

We advocate a more disease-based approach so that rather just collecting the data of the patients who you are doing the procedure on, collect the mortality data of everybody with a certain disease process—such as cardiogenic shock. If these data were included in public reports, practitioners could decide on the basis of their best judgement which patients should have a high risk procedure and not be concerned that a procedure intended to save a patient’s life be withheld on the basis of how an unsuccessful outcome will cast questions on the programme or the operator’s competence.

By focusing on the disease rather than the patient, you will better able to understand that a centre has higher mortality rate because of case selection rather than lack of technical competence.

At the end the paper, we have a list of factors that we think should be made available, along with risk adjusted mortality, in public reports so that the public can see all of the factors that need to be considered when looking at a centre or a particular doctor.

We also believe that presenting all of these other factors about quality will give more recognition to work of highly competent practitioners. At present, by using mortality as the sole endpoint, interventional cardiologists who are very technically competent and who are performing very high risk procedures are not getting the credit they deserve.

How can we make public reports easier for patients to interpret?

This is one of the critical issues we raise in the report and the bottom line is that there is not an easy answer. My co-authors are some of the best experts in the world and we cannot always agree on how data should be interpreted; therefore, how can you expect patients—who may have limited medical and mathematical knowledge—to accurately interpret data?

But, again, we think reducing the importance of risk-adjusted mortality and including other factors that indicate quality will make things easier for patients.