Formulary Decisions and Clinical Practice

Dennis P. Scanlon, PhD: The study design, as you’ve mentioned, seems to be very important here, and it seems to me that in some cases, some of these studies were emphasizing the risks in addition to the benefits. So, in hypoglycemia, thinking about the SWITCH studies or cardiovascular safety with the DEVOTE studies, I know earlier, Dr Bloomgarden, you were talking about study design. Are there any broad points with respect to either of those particular studies, or generally, that you think are important as we think about how to evaluate what might be, in some cases, marginal differences?

Zachary Bloomgarden, MD: Well, a huge issue with cardiovascular safety studies is that the nature of such a study is that you must carry it out in people at very high risk of a cardiovascular event. And that means that you’re going to take one-tenth or one-twentieth of the population base, as it were (of people with diabetes), and study them, because they’re the ones who will have enough events. In a reasonable size trial of 5000 or 10,000 people, over 3 years or so, you’ll see a difference, or not see a difference, in events. But those people who have a 2%, 3% annual event rate are actually much more ill than the majority of people we treat with diabetes. And we look at people who develop diabetes and we say to ourselves, “This is a person who is going to have diabetes for the next 20 years. We want them to stay healthy for this very long period far beyond what any clinical trial could ever give us.” So, we’re extrapolating from various data sets that aren’t quite real world but we do our best.

Dennis P. Scanlon, PhD: Dr Snow, at Aetna you work, I’m sure, with a number of pharmacy benefit managers, and have to make collective formulary decisions that, ultimately, coverage decisions are based on. Can you talk a little bit about that process? It’s a little bit of a market negotiation on the one hand, but you’re certainly weighing the scientific evidence, and ultimately are trying to navigate these new discoveries.

Kenneth Snow, MD, MBA: Number 1 is to look at the scientific evidence and what is shown to provide a benefit versus what is more questionable. And then, within that, to make sure that all our members have access to therapy that would be appropriate. But, as you say, once we get past that, there is a negotiation of limiting down the number of options for patients because that does lead to an improvement in cost, both for the member and of course for the payer (who in turn, passes any cost on to those who are paying for insurance, whether it’s a company or an individual). So, the goal is to keep the cost down but make sure that there is an adequate array of agents from a particular member.

Also, I just want to add in that there’s almost no situation where somebody actually has either a clear contraindication or a clear failure with a particular therapy. Let’s make it very simple. If they’re allergic to a particular agent, demonstrating that will allow them to move beyond. They’re not going to be held at that point, because it’s not on formulary. But it’s that initial choice of, if you have 3 agents that are all reasonable options and they’re all available, and there’s no reason to believe that you’re more or less likely to succeed with 1, then you should be directed to the 1 that’s going to be more cost-effective.

Zachary Bloomgarden, MD: I would comment, as a practicing clinician, that I wish I had a better way of knowing in advance what option is “the one,” or which 2 or 3 options are “the ones,” which are approved by that particular person’s plan within the overall umbrella of Aetna (or what have you). It’s so hard for us working in the trenches to actually know.

Kenneth Snow, MD, MBA: There’s no doubt that it is difficult, and I look at this from one payer’s perspective but, putting back on my clinician hat for a little bit, I remember the days of multiple. And so, it’s hard enough for me to keep track of what we cover as 1 payer, let alone what the industry does. I absolutely agree with you that it is a challenge and it is, in a way, unfair that there’s not an easier way to do that. This data, for the most part, is publicly available, and it should be the type of information that can be compounded by a particular search engine where it would put it in a format for the busy clinician.

Dennis P. Scanlon, PhD: Put it in an electronic health record.

Robert Gabbay, MD, PhD, FACP: It sort of is in there, but they’re not terribly accurate. We use that and I’m sure many others do, I think, partly because it’s a fluid process.

Kenneth Snow, MD, MBA: It’s a very fluid process, and so it changes. It can change every year or have small nuanced changes. And even within that, there are different levels of pharmacy benefits that a member may have. They may have insurance with whichever provider, whether it’s us or one of our competitors, They may have insurance through that company. And I was talking to Dr Gabbay about this earlier. You could have 5 patients in a row with the same insurance, and yet, they have 5 different drug plans. And so, even though you know their insurance, you still don’t know what their benefit options are. It is something that computers should be aiding the busy clinician with.

Robert Gabbay, MD, PhD, FACP: Somebody should invent that, you’re right. It’s a solvable problem and it’s mindboggling that it hasn’t happened.

Zachary Bloomgarden, MD: Patients worry so much when they’re changed from one agent to another, even if we truly think they’re the same. So, insulin lispro and insulin aspart, Humalog and NovoLog—almost all endocrinologists are willing to use either at any given time. But, for an individual who’s received one for several years, and then there’s a cost benefit for the company, it’s really quite a dilemma for that patient.