Effect of Drug Sample Removal on Prescribing in a Family Practice Clinic

Hartung et al. examine the impact of a policy discontinuing contact with pharmaceutical representatives and the distribution of free samples in a small group practice. Although overall prescribing of branded drugs did not decrease, the use of previously promoted agents did. The magnitude of effects differed considerably across different therapeutic areas.

Given how much is spent on free samples – more than $16 billion per the author’s report – it is truly remarkable that relatively little is known about their influence on physician behavior. One prior report synthesizing results from twenty-three studies highlighted important analytic limitations that were present in many of the investigations, as well as a variety of unanswered questions (1). These issues will be increasingly salient in the coming decade, as greater concerns about physician-industry interaction as well as a drying up of the pharmaceutical pipeline for small molecule drugs transforms office-based pharmaceutical marketing and promotion.

Debates regarding the benefits and drawbacks of samples reach philosophical tones at times – and the most important thing that can help to inform them is high quality data regarding how samples affect prescription drug use, expenditures and outcomes. This study provides additional information to help inform policy, though with some qualifications. Capturing the effect of a policy change such as the one examined is a difficult task, and accounting for secular trends is crucial. There are potentially very substantial differences in demographic characteristics, health status, and health care utilization between the control group of Medicaid recipients and the study subjects. While these differences need not preclude the use of Medicaid recipients to capture secular trends, quantitative analyses during the baseline (pre- policy) period comparing trends between Medicaid subjects and the study subjects would help to demonstrate their comparability. Also, as the authors’ acknowledge, it is plausible that the measured effects of the policy change may have been greater if observing prescribing patterns of drugs with a recent market debut, rather than therapies that may have already been heavily promoted and sampled (and thus induced brand loyalty on the part of physicians) over a substantial period of time.

Although a number of areas in need of future research have been identified by these and other investigators, one of particular importance is how the effect of sampling may vary or be mediated by patient characteristics such as socioeconomic status and race/ethnicity. This is especially relevant given racial and ethnic disparities in prescription use (2), as well as evidence that the poor may not be the primary recipients of free samples (3) and that samples may paradoxically increase individuals’ long-term out-of-pocket prescription costs (4).

References 1. Groves KEM, Sketris I, Tett SE. Prescription drug samples - does this marketing strategy counteract policies for quality use of medicines? J Clin Pharm Therapeutics. 2003;28:259-271.
2. Qato DM, Lindau ST, Conti RM, Schumm LP, Alexander GC. Racial and ethnic disparities in cardiovascular medication use among older adults in the United States. Pharmacoepidemiol Drug Saf. Aug 2010;19:834-842
3. Cutrona SL, Woolhandler S, Lasser KE, Bor DH, McCormick D, Himmelstein DU. Characteristics of recipients of free prescription drug samples: a nationally representative analysis. Am J Public Health. 2009;98:284-289.
4. Alexander GC, Zhang J, Basu A. Characteristics of patients receiving pharmaceutical samples and association between sample receipt and out-of- pocket prescription costs. Med Care. Apr 2008;46:394-402.

Dima M Qato, PharmD, MPH and G. Caleb Alexander, MD, MS

Competing interests:   None declared

Hartung and colleagues examined the impact of recent restrictions on pharmaceutical industry detailing and sampling of prescribing behavior within smaller, independent practices, using state Medicaid pharmacy claims as control for secular trend.[1] This study contributed to the literature by demonstrating that in aggregate, trends of brand name drug use did not change significantly after policy implementation, and use of promoted agents decreased while nonpromoted branded agents increased moderately.

The study raised several interesting questions noteworthy for further exploration. First, is the observed change or non-change of use of branded and promoted drugs likely to be a response from the demand side (e.g., patients) or supply side (e.g. physicians), or both? One of the key objectives of industry detailing and sampling is to promote awareness of targeted drugs and increase market penetration. One of the potential consequences, if detailing and sampling is successful from the industry perspective, is the continued use of the promoted drugs in the absence of free samples. This is not supported by the evidence provided by Hartung et al. This could be due to the fact that patients who rely on free samples are price sensitive, which overcomes the consideration of “better quality” of promoted drugs. Future research exploring patients’ decision-making process in the trade-off of price versus quality will illuminate the mechanism of patient response to the stoppage of free samples, in particular segmenting patients by their socio-economic status.

In the meantime, physicians have been known to factor in their patient’s insurance status when developing their treatment plans.[2] As well, previous studies have also suggested that sample receipt was greater among those who were younger and those not on Medicaid.[3] How did physicians change the prescribing of promoted drugs in response to the policy change in detailing and free sample distribution? Were physicians using free samples as “gifts” to the patients, or as substitutes to reduce the economic burden on the patients? Further study of physician motivation behind the use of free samples will help explain changes in drug utilization patterns after the policy implementation.

Finally, the study focused on a practice which employs 5 physicians and 1 physician assistant. It would be interesting to know if the change in prescribing patterns is clustered among certain physicians. The knowledge, training, seniority, and practice patterns of different physicians might play a role in the response to the policy change in drug detailing and samples.

Reference:
1. Hartung DM, Evans D, Haxby DG, Kraemer DF, Andeen G and Fagnan LJ. Effect of drug sample removal on prescribing in a family practice clinic. Annals Fam Med. 2010;8:402-409.
2. Meyers DS, Mishori R, McCann J, Delgado J, O’Malley AS, Fryer E. Primary care physicians’ perceptions of the effect of insurance status on clinical decision making. Ann Fam Med. 2006;4:399-402.
3. Alexander GC, Zhang JX, Basu A. Characteristics of patients receiving pharmaceutical samples and association with sample use and out-of-pocket prescription costs. Med Care 2008, 46(4):394-402.

Competing interests:   None declared

Hartung et al conducted a sophisticated analysis of prescribing practices before and after prohibiting sampling and detailing in a single clinic. Limitations of the study are noted by the authors and the reviewer Tumerman, including looking at data from a single independent pharmacy and potential confounding related to insurance type. The authors do not present data on changes in prescribing by insurance status. They found that while promoted drug costs decreased slightly, this was more than offset by increases in other branded medications. In their Introduction the authors note potential concerns about pharmaceutical marketing practices including impact on prescribing behavior and ethical issues. The authors also comment on medication adherence in their Discussion. Another cause for concern is the potential for decreased control of chronic conditions in patients receiving sample medications. 1 Because this was a longitudinal study, the impact of other broader influences could not be measured. It is possible that the impact of the steps taken by the study group was offset by changes in pharmacy practices in response to decreased access to physicians in office settings, such as the growing use of direct to consumer marketing and prime time medication advertising. One lesson may be that if we want to reduce the influence of marketing on physician decision making we may need to focus more on the establishment of evidence-based, cost-effective formularies.

1 Zweifler J, Hughes S, Schafer S, Garcia B, Grasser A, Salazar L. Are Sample Medicines Hurting the Uninsured? J Am Board Fam Pract 2002;15:361– 6.

Competing interests:   None declared

Dr Hartung's recent article on the affect of eliminating sampling on prescribing habits of family physicians represents a good effort at providing some data to an often emotionally laden debate. Having had the experience of spearheading a system wide effort at eliminating detailing and sampling within a large Midwestern university based department of Family Medicine, I have some thoughts and concerns about this study that I would like to share.

First, I wonder about using the Medicaid population as the control. This population already tends to receive a different profile of prescriptions, based on restrictions imposed by active pharmaceutical management, compared to other patients groups. They tend to have higher generic prescription use and selective brand use based on preferred pricing contracts.

Secondly, I wonder about the personal relationship between the pharmaceutical representatives and the providers in this setting. How carefully was the elimination of detailing, not just sampling, enforced? Our experience showed that even after samples were eliminated, certain physicians continued to see representative informally, despite policies indicating that this was not acceptable. This is particularly a difficult issue in smaller communities, where representative are often patients of, and friends with, physicians and staff.

I also believe that there may have been some bias introduce by picking the one pharmacy that was used. It is quite possible that the patients that used this pharmacy have a different profile than those that use the other pharmacies in town. This may be especially true if there is a price differential between the pharmacies. One would have to assume that the other pharmacies are chain “big box’ pharmacies. Four dollar per month pricing of frequently used prescriptions may be a draw for patients in lower socioeconomic groups.

In addition, a different picture may have developed if only new prescriptions for patients not having received these classes of medications in the past had been included in the study. It appears that chronically refilled medications were included and it is quite possible that providers are less motivated to change a chronic medication to a generic or less costly medication versus using them when starting a patient on a new medication.

Finally, old habits die hard. Studying a group of providers that have been in the habit of prescribing one way for years, and then expecting changes to occur over a relatively short period of time may be unrealistic.

Beyond the measurable effect on prescribing habits, there are other concerns to consider when thinking about the relationship between pharmaceutical representatives and physicians. Many physicians have relied too heavily on pharmaceutical representatives for their on going education about new drugs. To believe that this is an appropriate and safe resource for education about the safety and efficacy of a product is sure folly. One only needs to consider the relatively early recalls of several drugs that were marketed and sold to patients and physicians alike, as the next greatest wonder pill, to understand the hazard inherent in this type of “continuing education”.

Also not addressed is the issue of safe sampling which must include labeling and patient education and the ability to track medications dispensed.

In the end, my verdict is that practice life, appropriateness of prescribing and patient care are all benefited by the removal of the practice of entertaining pharmaceutical representatives and samples in our clinics. And to answer the most quoated and perhaps compelling reason to sample(although often used as a smoke screen), we have found better and safer ways to help our truly pharmaceutically indigent patients.

Marc Tumerman MD

Competing interests:   None declared

Our 3 physician office has been drug sample and drug rep free for over twenty years. It is interesting that some of the changes in prescribing patterns occurred immediately but were fairly small. I wonder if the 18 month observation window might have influenced this. It would be interesting for the authors to keep collecting data and see if more changes occur over time. Of course, Direct to Consumer advertising also greatly influences prescribing patterns, but that was not the object of this study. I applaud the authors for performing and publishing this much needed study. I look forward to a new era in medicine in which physicians prescribing patterns are influenced less by pharmaceutical marketing and more by patients needs.

Competing interests:   None declared