Andrew L. Pecora, MD, CPE: You can Google for free, but it’s really not for free because you get 22 commercials. That is how Google makes its money and there’s no law that says Google cannot advertise. [On the other hand, by law,] the pharmaceutical industry, which does have large margins, [has] very severe limitations on how they can advertise and transfer money. I think it’s now $10. If you’re a doctor and you spend $10.12, you’re on the list. From your perspective, how do we get these things aligned so there’s a rational macro business model that makes sense?
Rena M. Conti, PhD: So I think what you’re alluding to is that there’s this kind of fixed cost of producing healthcare that providers need to invest in, and then, there is the marginal cost of production of producing health for patients. And the question is, where’s the money going to come from for doing all of this fixed-cost investment?
I would argue that hospitals and other providers need to fundamentally rethink what they invest in for very high fixed cost. So, for example, it used to be the case 10 years ago or 15 years ago that every hospital in America was running to go buy their own MRI machine. Or the new one is all those robotic surgery machines that do certain types of cancer treatment. There is no reason why every hospital in America, or every provider in America, needs their own MRI or robotic surgery lab. It used to be that you made money off of those types of fixed investments, but it may not be [the case] anymore.
So, instead, I think we’ll see providers move toward investments that really do pay off in the long-term and that aren’t really used to block out competitors from doing or keeping up with competitors. Instead, it will be really tailored to managing the healthcare of their own sets of patients and their unique qualities. That may be very different for practices in an urban environment relative to practices in a rural environment. There may be more fixed costs for rural providers than there are for urban providers. But some of that cost is going to have to be defrayed from payment by insurers—both government and other payers as well. And some of it’s going to have to come from private sector investing.
Andrew L. Pecora, MD, CPE: Patient engagement remains a crucial aspect of healthcare reform, and meaningful use requirements by the Centers for Medicare and Medicaid Services advocate for patient access to their electronic health data via patient portals. Would you describe new payment models as being patient-centric?
Rena M. Conti, PhD: Again, I think that in the rush to put in all of these new payment models and get providers [started] toward taking risk, or at least reporting quality, there can be this tendency to lose sight of patients and their families. And I very much hope that the patient comes back into the center of the picture here.
Again, I believe, strongly, that certain types of risk models for certain types of cancer patients are actually going to bring patients more into the center of care. So, for example, I believe that medical care models, or medical home-type models, have great promise in taking care of patients in a holistic way—in [terms of] meeting both their care needs for cancer treatment, but also in taking care of diabetes and other types of illnesses they may be suffering from, and also in dealing with transportation costs and taking care of some of the other support that they need. So I think providers that are already doing these models are talking about the great success that they’re having in reorienting care toward patients. It may not be what’s being measured currently, but we’re going to get there.
Exploring Racial, Ethnic Disparities in Cancer Care Prior Authorization Decisions
October 24th 2024On this episode of Managed Care Cast, we're talking with the author of a study published in the October 2024 issue of The American Journal of Managed Care® that explored prior authorization decisions in cancer care by race and ethnicity for commercially insured patients.
Listen
New AI Tool Identifies Undiagnosed PNH in Health Records
October 30th 2024The machine learning model shows promise in detecting paroxysmal nocturnal hemoglobinuria (PNH) by assessing electronic health records (EHR) data, potentially transforming the diagnostic landscape for rare diseases.
Read More