Since HOOKED was published, the relatively sparse literature on the effects of drug samples on physicians' practice has been expanded somewhat. My apologies for having come across these relatively old articles only just recently.
David Miller and colleagues from Wake Forest University had a sort of natural experiment occur in 2000 when a large, mostly resident-run internal medicine clinic moved to a new building which lacked an appropriate storage space for the former drug sample closet, and the managers decided to end the practice of stocking samples. It also happened that they had a good computeratized tracking system for all prescriptions. They tracked prescriptions written in the 9 months before and after the switch in four major medication categories (antihypertensives, oral diabetic agents, peptic ulcer/reflux drugs, and nonnarcotic analgesics) in relation to the patients' insurance status. (Presumably these four classes of drugs were chosen because they included both brand-name and generic drugs, and there was little evidence if any that the brand-name drugs were superior to the generics.) They found that once the samples were gone, no significant changes occurred in the Medicaid patients' prescriptions, but nearly three times as many prescriptions for generic drugs were written for the uninsured patients. They suggested that by prompting docs to give uninsured patients expensive brand-name drugs rather than generics, samples drive up the costs of care overall for that group.
Sarah Cutrona and colleagues at Cambridge Hospital looked at data on 32,681 Americans from the 2003 Medical Expenditure Panel Survey (MEPS). MEPS offers the advantage of being a population-based community survey, so its contents are not dependent on differential access to medical care. The investigators compared whether a patient had received a drug sample with insurance and income level. They found that people who had higher incomes and who had continuous insurance were more likely to have received samples. This, of course, makes sense if you figure that to get a sample, you have to get into the door of a physician's office, and the physician has to have a stock of samples to give you; and those are more likely to occur if you're higher income and/or insured. Cutrona and colleagues conclude that samples are for marketing and do not provide for any sort of effective "safety net" for getting drugs to the indigent or uninsured. Well, duh, but it's nice to have confirmation from another source.
Nothing earthshattering here but a few more pieces of the puzzle to document what we have long thought about so-called "free" samples.
Miller DP, Mansfield RJ, Woods JB, et al. The impact of drug samples on prescribing to the uninsured. South Med J 101:888-893, 2008.
Cutrona SL, Woolhandler S, Lasser KE, et al. Characteristics of recipients of free prescription drug samples: a nationally representative analysis. Am J Public Health 98:284-289, 2008.