Author + information
- Spencer B. King III, MD, MACC, Editor-in-Chief, JACC: Cardiovascular Interventions∗ ()
- ↵∗Address correspondence to:
Dr. Spencer B. King III, Saint Joseph’s Heart and Vascular Institute, 5665 Peachtree Dunwoody Road NE, Atlanta, Georgia 30342.
I hate the term “risk avoidance.” Why? It is not because I don’t want to reduce risk for patients. Risk and risk reduction are critical components in all our clinical trials and observational studies. We think about it in all our clinical decisions. How do I reduce the risk for my patient? If we could avoid risk for her, wouldn’t it be even better? These are noble thoughts when applied to the patient, but now risk avoidance has become a dirty word (2 words) when applied to the doctor or hospital. Conversations in doctors’ lounges across the country are about nonmedical issues, mostly sports or politics, or medical issues. The medical ones that are most constructive are about difficult cases and informal consultation (of course, with HIPAA [the Health Insurance Portability and Accountability Act] in mind), or depressing ones about electronic medical records, outcome reporting, and reimbursement issues. In cardiology, when we are discussing patient risk and selection of therapy to reduce that risk, the conversations are upbeat. However, the risk being increasingly talked about is the physicians’ risk. For me that is depressing. Opinion pieces sent to JACC: Cardiovascular Interventions, as well as physician complaints about public reporting and the use of outcome measures tied to reimbursement, are increasingly dominating the pessimistic conversations.
Let’s think about 2 ends of the spectrum. On the one hand are outcome measures, usually death, which is of great interest to the physician and the patient. After all, we are here to save lives or to decrease the risk of dying. But risk avoidance is now used to describe a method of improving scorecards by avoiding treatment for patients who are at high risk of dying. Extensive efforts at risk adjustment have not convinced many physicians that taking on high-risk patients does not put their scorecard in jeopardy. I have learned that risk adjustment is much more favorable to those who treat high-risk patients, but it is not perfect. I find it hard to believe that there are many physicians who withhold therapy they think will be beneficial because of the chance that the outcome may not be favorable. There are, however, very high-risk situations, such as patients who present after cardiac arrest with anoxic encephalopathy or patients who are in refractory shock and perhaps other scenarios in which the physician views the intervention to have a questionable chance of altering the outcome. Among surgeons, this has always been part of the discussion. “I don’t think the patient will make it through surgery” is the powerful balancing concept. For those performing coronary interventions, the risk of the procedure is less, the time to decide is less, and therefore the decision to intervene or not in these cases is sometimes more difficult. Nonetheless, it is a decision that should be made in consultation with the family and should not be driven by concern about scorecarding of the outcome. Whether outcomes reporting is to the public or to payers or regulators, these “compassionate cases” are few enough to make statistical adjustment of risk for physicians difficult. This is why these cases were removed from public reporting in New York State after 2008. However, there is a necessity to generate a clear definition of the “compassionate case” patient and a reliable way to ensure that it is adhered to. If some cases are to be excluded from reporting on a national level in response to the concern about risk avoidance, a broad consensus about the definition of such cases will be required. Discussions are ongoing at the Interventional Council and the Board of Governors of the American College of Cardiology.
On the other end of the provider risk avoidance spectrum is scorecarding for the appropriateness of therapy, and now the meaningful use of therapy. These concepts imply that there are therapies that can be done but should not be done because of lack of demonstrated superiorities or because their “value” in monetary terms is not worth the outcome achieved. Whereas the risk avoidance of high-risk patients leading to excess mortality is viewed as a restriction in the access to care, the intervention in cases not considered as appropriate is viewed as overuse of therapies. The appropriate use criteria continue to evolve but are not yet perfect. It is designed as a quality improvement tool but has now become another driver of risk avoidance, the risk of financial penalty. Concern has been expressed that cases not labeled “appropriate” may have reimbursement denied by payers. Whether this comes to be or not, the concern is there and has led some practices to upcoding so that the number of patients not having the appropriate designations continues to decrease.
There is no doubt that rules will influence behavior. Some will be constructive changes—that is what quality initiatives are supposed to be about. But some changes will be cosmetic with the reporting adjusted to fit the rules. Do these risk avoidance measures have to continue to dominate decision making? Can we return to the primary question—how to reduce the risk for the patient? Transparency is demanded and the days of turning a blind eye to the selection of therapies or the outcomes achieved are over. We will always have record keeping, and it will be necessary if true quality is to be maintained and improved. Most physicians would support reporting if all data is considered and the playing field is level. The incentives, however, need to be reordered. If reimbursement is to be tied to outcome and to correct selection of therapy, the methods of assessing these must be adequate. Administrative databases do not contain the information necessary to perform risk adjustment. They should not be used for determining rewards or penalties for physicians or hospitals. Adequately adjusted clinical data is necessary. As long as fee-for-service dominates health care financing, the incentives will drive increased utilization. To understand whether it is appropriate, we need to know more than who receives the therapy and their outcomes; we also need to know who does not get therapy and their outcomes. Administrative databases are accumulating large amounts of data, but, remember, they are not able to reflect proper risk adjustment. We will need much more comprehensive clinical data and perhaps a change in the incentives. I often wonder, do our colleagues in Canada or Europe have their conversations in the doctors’ lounge dominated by “risk avoidance”?
- 2016 American College of Cardiology Foundation