Use of Drug Sample Medications

To the Editor: I read with interest and some concern the article by Zweifler et al in the 2002 September-October issue of the JABFP (Zweifler J, Hughes S, Schafer S, et al. Are Sample Medications Hurting the Uninsured? J Am Board Fam Pract 2002; 15:361–6). The only conclusion I reach after reading the study is that we still do not know and that this study has added little to our knowledge.

I will state my bias at the onset. I give out samples—lots of them—to my Medicare patients without medication coverage and to my uninsured patients, as well as to some of my insured patients who are on so much medicine as to make even the co-payments prohibitive. I realize doing so is much easier in my private solo practice than it would be in a large clinic. Therein lies my first objection—the data are not necessarily generalizable.

In addition, it appears patients using samples were required to have a follow-up appointment to get more supplies. This financial cost to the uninsured patient would be expected to be prohibitive, even if a sliding scale were offered. A fairer way to compare the effect of using sample medications with outcomes from filling prescription medications would be to allow patients to call in for “refills” of samples without the need to be seen. We do this frequently at my office with good results. Patients are less likely to be noncompliant (and thus have higher blood pressures). The medication outcomes will not be evenly comparable, because some patients are embarrassed to come to the office frequently to pick up their samples, but it makes any comparison fairer.

Finally, I believe we are looking at divergent groups of patients in this study. I would expect samples to be given to the poorest and the sickest patients. Both these groups are more likely to have worse disease. Comparing these patients with patients who have insurance (or with money) is comparing apples with oranges. In fact, the only fair (and useful) way to compare these two groups of patients is to observe the patients taking the samples but to stop providing the samples and write prescriptions instead. A comparison of this nature, where patients act as their own controls, is the only reasonable way to determine whether the provision of samples is the true evil this article leads one to believe.

I understand the concern raised by free samples provided to physicians and their patients. Certainly these samples are of the more costly drugs and ultimately will affect our prescribing habits and the cost to our patients. In fairness, however, there are some patients who truly benefit from samples, and there are ways of using samples, short term, then switching to an equivalent generic medication (eg, switching from Toprol XL to metoprolol, or from Accupril to lisinopril). Before we throw the baby out with the bath water, I believe we need far more solid evidence of harm caused by samples than we gain from this study.