It may not be time to clean out the sample closets

In a 2014 editorial, Katz concluded, relying heavily on evidence from a study of the relationship between free drug samples and prescribing patterns for acne and rosacea, that it is time to get rid of samples from dermatology offices (1,2). I am not convinced that such a conclusion is well supported by the evidence.

The acne/rosacea study had claimed that the leading drug written for acne by dermatologists nationally, Epiduo, had a generic equivalent in adapaline-benzoyl peroxide. While both adapaline and benzoyl peroxide were available individually as generic products, there was at the time no generic adapaline-benzoyl peroxide combination product. There is no evidence to support the idea that 2 separate agents are equivalent to a single product that includes both agents; there is evidence that they are not equivalent. A controlled clinical trial compared adherence to the combination product Ziana (clindamycin-tretinoin) to its two components separately. Adherence to acne treatment was better with the single combination product than with two separate agents (3).

Much of the argument against samples was based on the assumption that the care provided in a single academic center that does not use samples should be considered the gold standard by which to judge the practice of other dermatologists. Acne guidelines support treating with a combination of two drugs, a topical retinoid and a topical benzoyl peroxide (4). Is it better practice to prescribe these as two separate agents when better adherence can be achieved with a single branded product? Perhaps in some situations it is, perhaps in others it isn’t. Parents who recognize the potential for poor adherence in their teenagers might prefer a single combination product rather than try to apply two separate agents. When offering treatment to patients, offering both options and letting patients choose may be appropriate. Some patients may prefer the less complex, higher cost treatment regimen, while other may prefer the more complex, lower cost regimen. We don’t have enough information to judge whether the quality of acne care was better at the academic center or in the national sample. The data are not reported in a fashion that would allow us to judge whether the academic center in the study is prescribing only a single agent or if two agents are prescribed separately. If the academic center was following guidelines but not offering patients the choice of a combination product, then using them as the “gold standard” of optimal care may not be warranted. It may be that the insurance status of patients at the academic center differs systematically from patients seen nationally such that at the academic center prescribing multiple generics truly is in patients’ best interest, while prescribing a combination product is optimal for insured patients seen nationally.

Other reasons to support the notion that we should eliminate samples are specious at best. The claim that sampling circumvents the involvement of pharmacists does not seem compelling. Do we have evidence that pharmacists are providing critically important input to patients with respect to topical acne or other dermatologic therapy? Even if pharmacists are extensively counseling our patients, sampling does not typically circumvent the pharmacist, as patients are still getting a prescription from the pharmacist. Moreover, for the few patients who do get a full course of therapy in the form of samples, it is likely they received the samples because they could not afford to purchase the medication and therefore would not have had the benefit of the medication at all, much less the involvement of a pharmacist in their choice of treatment.

Is the claim that sampling will cause excessive use of new medications that are hazardous supported by evidence? Katz et al give the example of topical calcineurin inhibitors and the black box warning that appeared after the products’ launch. Notably, while pointing to the black box warning, Katz et al didn’t cite any data to suggest topical calcineurin inhibitors are actually hazardous. The black box warning is about lymphoma, but the data do not support an increased risk of lymphoma or other malignancies with topical calcineurin inhibitors (5–7). On the contrary, topical calcineurin inhibitors are safer than topical corticosteroids, have lower oncogenic risk, and are not associated with the atrophy, telangiectasias, and even potential HPA axis suppression associated with topical corticosteroids. Topical calcineurin inhibitors can cause stinging, so letting the patient try a sample of them before purchasing them is particularly helpful.

The cost estimates in the acne/rosacea article may be misleading. From the societal perspective, they don’t account for the discounts that many insurers negotiate. From the patient’s perspective, they don’t account for the discounts patients may receive with coupon cards and other mechanisms. Just as they do not offer samples, it may be that the academic center isn’t permitted to offer patients the discounts. For better or worse, those discounts likely do heavily influence prescribing patterns, but that’s another story. The cost estimates also do not take into account the massive increases in generic prices we’ve seen in the past few years (8).

In addition to the other potential benefits to patients, a controlled trial found that having acne patients sample treatment at the office visit may increase adherence to that treatment (9). Katz et al’s suggestion that sampling’s benefits can be achieved with Crisco is not yet supported by evidence. A major barrier to adherence is fear (10), and applying Crisco to patients may not get them past the fear of applying a drug. Moreover, many products can’t be demonstrated adequately using Crisco, including foams, sprays, shampoo, and tape formulations. Moreover, the sample of Crisco will not show patients if the medication is tolerable for them; patients who may have discovered intolerability had they had a sample may, in a sample-free world, end up buying expensive medications that they cannot use.

We applaud the desire to see patients get the best care at the lowest cost. The motivation is good, but we aren’t sure the available evidence supports the idea that emptying dermatologists’s closets of samples promotes cost-effectiveness. Larger, randomized controlled trials of sampling in dermatology may be informative. Having practiced in both sample-full and sample-empty worlds, we don’t see that eliminating samples has changed our prescribing behavior. If a generic was appropriate, we wrote for the generic whether we gave patients samples or not. We, too, want to give our patients the best possible care at the lowest possible cost. Other doctors may or may not be different in that regard, and it is too soon to rush to judgment. Our greatest disappointment in getting rid of those samples has been the times when patients have asked for samples, and we have not been able to meet their need. Perhaps the authors are right that sampling drives prescribing behavior in dermatology (so far, only association has been demonstrated), but it may be that sampling encourages use of better drugs. Even if the authors are right that samples can sometimes encourage the use of a less than ideal treatment, it may be that a better solution would be to add samples of generic medications to our sample closets rather than losing the benefits of the samples we currently have.

Disclosure statement

Dr. Feldman has received extensive research, speaking and consulting support from many pharmaceutical companies.