Each month, three presenters review an interesting journal article in a conversational manner. These articles involve “hot topics” that affect family physicians or “bust” commonly held medical myths. The presenters give their opinions about the clinical value of the individual study discussed. The opinions reflect the views of the presenters, not those of AFP or the AAFP.
This Month's Article
Symm B, Averitt M, Forjuoh SN, Preece C. Effects of using free sample medications on the prescribing practices of family physicians. J Am Board Fam Med. 2006;19(5):443–449.
Does the distribution of sample medications to patients decrease or increase prescription drug costs?
Bob: In our May 2008 discussion of albuterol (Proventil) versus levalbuterol (Xopenex), we commented that the practice of distributing drug samples costs patients more money in the long run. Some of our readers questioned how free samples could ultimately result in higher costs. This study demonstrates just how this paradox occurs.
What does this article say?
Bob: The study was performed at three family medicine clinics in the Scott & White health care system, an 180,000-member health maintenance organization in central Texas. One of the three clinics (clinic X) provided free samples, whereas the two other clinics (clinics Y and Z) did not. All three clinics had a similar case-mix of patients. The authors selected 25 of the most commonly distributed samples and monitored prescribing data from their own pharmacy data-bank. This allowed the authors to directly compare the prescription medication costs of the physicians in clinic X (those with samples) and the physicians in clinics Y and Z (those without access to samples).
What did they find? Clinic X wrote prescriptions more often for the 25 monitored sample medications than clinics Y and Z. As a result, clinic X least often used recommended formulary drugs, and the patients at clinic X ultimately had the highest cost per average 30-day prescription (Table 1).
|Clinic||Number of physicians||Mean cost|
Should we believe this study?
Bob: I do. The only potential limitation is the possible presence of confounding variables that may have existed among the clinics. For example, clinic X had a younger population of patients and a larger minority population. Could these and other variables that were unaccounted for influence the outcome of this study and invalidate the association between free samples and higher costs? Although this is possible, I would think that a younger population would actually have a lower prescription need than an older population.
One should always be wary of potential confounders; however, I believe they do not threaten the internal validity of this particular study. That is, the association of free samples and increased prescription costs appears to be a true association.
Andrea: The conclusion that free samples result in increased prescription costs makes intuitive sense. Take a “real life” example: a patient in your office is newly diagnosed with hypertension. You go to the drug sample cabinet, and what's there? Some new angiotensin receptor blocker (ARB) or beta blocker, not a generic diuretic. So, the patient leaves with the new ARB, comes back a few weeks later with an improved blood pressure, and now needs a prescription. In this scenario, more than two thirds of surveyed physicians reported that they would continue therapy with the expensive sample medication and not make the switch to a less expensive generic.1
Mark: You point out one of the sad truths regarding samples. One family medicine residency program documented that, when antihypertensive drug samples are present, residents and faculty were less likely to follow Joint National Committee (JNC) prescribing recommendations. But when the samples were removed, adherence to JNC prescribing recommendations dramatically improved.2 Simply put, when free drug samples are available, physicians have a tendency to initiate and continue therapy with the sample drug, even if it is not a recommended first-line agent.
Bob: Some will argue that free drug samples are there for those who cannot afford their medications. If this were the case, why is the drug cabinet barely stocked in my clinic, which is located in a poor, underserved area with a large uninsured and Medicaid population, whereas my colleague's office two miles away, which sees only insured patients, has a drug closet that is overflowing with samples?
A recent study confirmed that patients with health insurance are much more likely to receive free drug samples than those without.3 In fact, the poorest one third of patients were less likely to receive samples than those with incomes at 400 percent of the federal poverty level.3 Although many physicians may believe that samples assist patients with the most need, samples are more a marketing tool than a safety net.
Mark: Let's put the costs of this marketing program in perspective. In 2004, the pharmaceutical industry spent $57.5 billion on marketing. More than 27 percent of this marketing budget was spent on samples ($15.9 billion).4 If this was not an effective marketing tool, do you think this sort of money would be spent?
What should the family physician do?
Bob: Every physician needs to make the decision about whether to distribute free drug samples. It is a personal choice. I would like to commend those physicians and institutions that no longer accept free samples and do not allow detailing by pharmaceutical representatives.
Andrea: Free drug samples drive up costs and cause physicians to use second-line choices. Increased cost and less appropriate prescribing are examples of the “harms” most of us took an oath to avoid when we became physicians. I think we should look a little closer at what we do and decide what is really the right thing for our individual patients, as well as for the overall health care system.
Mark: To read more about the detrimental effects of samples and other drug company “freebies” that ultimately drive up health care costs, I would recommend such resources as “No Free Lunch” (http://nofreelunch.org).
Physicians who distribute free drug samples ultimately contribute to higher overall prescription costs.
When free drug samples are used, they are likely second-line choices (not guideline-recommended).
Always look for confounding variables. These are characteristics that are distributed differently among study groups and that can affect the outcome being assessed.
Internal validity means you have evidence that what you did in the study (i.e., the program) caused what you observed (i.e., the outcome). Internal validity can be threatened by confounding variables.