By Abby Van Voorhees, MD, September 01, 2014
In this month's Acta Eruditorum column, Physician Editor Abby S. Van Voorhees, MD, talks with Alfred Lane, MD, MA, about his recent JAMA Dermatology article, "Characterizing the Relationship Between Free Drug Samples and Prescription Patterns for Acne Vulgaris and Rosacea."
Dr. Van Voorhees: What led you to look at the effect of drug sampling on prescription costs?
Dr. Lane: We lost our ability to use samples in 2004 or 2005 because of the Joint Commission rules requiring documentation of where every sample went. Many academic medical centers had the same experience. We fought tooth and nail to try to keep our samples and then we lost the ability.
The thing I noticed in my practice as a pediatric dermatologist was that after that change when I received referrals from pediatricians, the patients were on generics, but when I received referrals from dermatologists, the patients were always on much more expensive medications — and I realized that those were the drugs that were sampled.
That was the background that made us think we should look at this. We absolutely loved our samples when we had them; we hated to give them up. But a few years later we were delighted we didn’t have to mess with them anymore.
Dr. Van Voorhees: Sampling in medicine has engendered some controversy. What are the known advantages and disadvantages of free drug samples?
Dr. Lane: I can’t think of any advantage other than the income that the pharmaceutical company makes. People talk about giving samples to poor patients, but it turns out that the pharmaceutical companies know you’re not doing that. The pharmaceutical companies know what samples you have received and what prescriptions that you write. Other studies [cited in the JAMA Dermatology article] have shown that samples aren’t used for indigent patients. Dermatologists like to use topicals; they think that testing multiple topicals may help but usually they only give patients one sample because that’s the product they have. And usually there’s so little irritant dermatitis we see in the products we use today that I don’t know that that’s an advantage.
I think that there is another advantage — for the doctor, the doctor’s staff, the doctor’s family. It turns out a lot of samples are used by the physician, the physician’s family, and the staff in their office. That was one of the reasons we lost our ability to use samples — because we had to keep a strict inventory and our inventory kept disappearing because everybody who had access to the closet kept using the samples without recording it.
The disadvantages are the extra societal cost from branded products, which our paper showed; for a lot of the medications that are sampled there’s little data showing that they are better than the generic. There’s not effectiveness data, for example, comparing two separate generics to a branded generic that is a combination of two generics and costs a lot more than buying the two generics separately. There’s little data documenting that that’s a better product.
Dr. Van Voorhees: Has it been known if the availability of sampling alters the prescribing practices of dermatologists?
Dr. Lane: Ours is the first paper that really looks specifically at dermatologists. We chose to look specifically at acne.[pagebreak]
Dr. Van Voorhees: What specifically did you study?
Dr. Lane: We used a large database of hundreds of thousands of prescriptions written by dermatologists and compared it to our clinic database. 2010 was the first year that all of our adult patients in our clinic had their prescriptions written in Epic, our EHR system, so we were able to pull the data out of our EHR. We think the environments are similar. We had a little bit higher Medicare/Medicaid population than what would be seen nationally because more of the poorer patients will go to the academic medical center. Our faculty had been without samples for six or seven years, so just in the way we work we use a lot of generics.
Dr. Van Voorhees: How were prescriptions classified when you looked at these databases?
Dr. Lane: We used the ICD-9 codes for acne (706.1) and acne/rosacea (695.3). We looked at the very first visit and those two ICD-9 codes in the National Disease and Therapeutic Index (NDTI). We looked at what prescriptions were given and then what prescriptions were given with a sample.
Dr. Van Voorhees: How could you be sure that a sample was given to those patients?
Dr. Lane: In our clinic we know there were no samples given because there were no samples. In the NDTI database, two days out of every quarter the physicians record everything they do and check off whether they gave a sample with or without a prescription. So we have detailed data going back to 2000 on prescriptions that were written by dermatologists with or without a sample.
Dr. Van Voorhees: What results did you find? What did you notice when you compared the earlier time point with the later time point?
Dr. Lane: We looked at the patient costs for the first visit for acne — which we chose because we thought if people were already on another medication that would bias the study — in 2010. Patients seen in an academic medical center had a prescription cost at the pharmacy of $200. Patients who were seen in an office that used samples, their average first prescription cost was $465, so more than twice the amount.
We tried to figure out what was driving this and where it was going. From the NDTI database we were able to look at other specialties, and it turns out that the frequency of other specialties writing prescriptions with samples was going down from 2001 to 2010, while the number of dermatologists writing prescriptions with samples was going up. This showed that the sampling has had an impact on dermatologists in that they are using more and more samples over the last 10 years. When we looked at what they were it turned out that most of them were so-called branded generics. That would be a generic drug that has a new formulation that now becomes a brand name that is much more expensive than a common generic.
Dr. Van Voorhees: Is it possible that this was intentional and that there’s more preference for the features of these branded or branded generic products in some practices?
Dr. Lane: The difference is intentional in that physicians who have samples in their clinics don’t just have samples; they have pharmaceutical representatives coming in to see them and talk to them and their staffs to try to get them to use their medications. The AMC doesn’t have samples and usually doesn’t have as many pharmaceutical rep visits. I’ve always had an open door; I would see any rep if they scheduled an appointment. Once we stopped giving samples, within a year they never came by. They would call and say they wanted to leave samples; I would tell them we didn’t accept samples and they would never show up for their appointment. The marketing process is directed at those who use samples.[pagebreak]
Dr. Van Voorhees: Do you think there are true differences between the branded generics and the generics?
Dr. Lane: That’s the question: Are there truly differences? The branded generics are often tested against placebo, not against the same combination as generics. So a true study looking at the comparative effectiveness of the branded generic versus the generic isn’t done. You’d need a large difference that I think you’re unlikely to find; cost effectiveness and comparative effectiveness analysis just isn’t done for those products.
Dr. Van Voorhees: Using the data from 2010 from your database, was there any way to filter for the use of coupon cards in those scripts that were written?
Dr. Lane: We tried to look at this. Before we did this study the first author and myself went to five or 10 dermatology offices to see what samples they had. With many samples there were also those cards. The problem is that if I write for a brand-name drug or a branded generic, the coupons overcome the difference that the insurance companies have imposed in the co-pay for generics versus branded drugs. It helps pay the $40 difference for the first prescription, and then once the patient starts to use the drug the patient will pay the difference themselves. We didn’t look at that but the coupons are designed to get the patient to eventually pay that difference. But the other problem is that the cost to the insurer for the branded drugs is significantly higher. So somewhere between $6 and $16 billion is the cost of sampling to the health care system, and it doesn’t add benefit, or at least it isn’t proven to provide a better product to patients.
We didn’t look at patient satisfaction. If it turns out that giving samples leads to higher patient satisfaction that would induce physicians to give samples. But a lot of the press coming out of this paper is patients saying they had no idea that they were being manipulated to buy more expensive things. I think if the public becomes more aware of the cost to them and the system, the impact of me giving you a sample on your perception of me as a better doctor may go away.
Dr. Van Voorhees: Is there any takeaway conclusion you’d like to be sure the practicing dermatologist should consider as a result of this paper?
Dr. Lane: I think the practicing dermatologist needs to personally consider, are you being manipulated by the pharmaceutical companies? They have a lot of knowledge about your practice. Is that manipulation worth it to your patient to pay more for medications that may be more expensive but may not be proven to be more beneficial?
Dr. Lane is professor of dermatology and pediatrics at Stanford University. His article appeared in the May 2014 issue of JAMA Dermatology; JAMA Dermatol. 2014;150(5): 487-493. doi:10.1001/jamadermatol.2013.9715.
Editor’s note: The opinions expressed in this column are those of the study author.