Although it is possible to quantify the amount of free therapy given out by physicians, uncovering their motivations is tougher.
There are several reasons why a doctor might give a patient samples without a prescription—such as for an initial trial of
efficacy and tolerability or simply to make room in the sample closet. Nevertheless, even if just a faction of these sample-only
therapies represents physicians purposely providing free treatment to patients who do not have prescription coverage then
the industry is contributing a major, virtually unheralded subsidy to the healthcare system of the United States.
This has important implications for companies considering making changes to their current sampling system. Samples represent
a "safety valve" for physicians by providing a way to treat patients who otherwise couldn't afford medication. Putting an
end to sampling could erode, perhaps even destroy, the goodwill accrued by the pharma industry with physicians for providing
this subsidy. On the other hand, the industry is currently getting little or no public credit for this enormous subsidy.
Where is the ROI? Yet industry must still consider how much sampling should be reduced to maximize it return on investment and improve the bottom
line. The first step is to look at where samples go after the sales representative puts them in the closet. Physician distribution
of samples to patients can be organized into one of three major groups:
1) Samples given to newly diagnosed patients with a prescription
2) Samples given to previously diagnosed patients with a prescription
3) Samples given to patients with no prescription
ImpactRx data collected during the first six months of 2004 reveals some interesting patterns. (See "Sample Distribution,")
The most striking finding is that roughly 60 percent of all samples are dispensed without an accompanying prescription—clear
evidence of the magnitude of the subsidy provided by the pharma industry. But even stronger evidence can be found with a close
examination of the level of samples being dispensed to previously diagnosed patients who also receive a renewal prescription.
Although the average across all therapeutic classes in that group is 15 percent, the highest percentages are found in the
chronic conditions: diabetes, high-cholesterol (statins), osteoporosis, and hypertension. These classes constitute an average
27.5 percent of all samples being distributed to previously diagnosed patients.
The data does not distinguish between how much of this behavior is a function of physicians "clearing out their closets" as
a result of over sampling and how much is physicians augmenting lack of, or poor, prescription coverage. Yet the information
gathered in physician focus groups provides strong anecdotal evidence that a large portion of these samples is given to patients
to subsidize their drug therapy.
Also noteworthy is the distribution patterns in the more acute classes. Approximately three-quarters of the oral solid antibiotics
and COX-2 samples are distributed to patients without a prescription. Other classes such as proton pump inhibitors (PPIs),
asthma, and allergic rhinitis approach the two-thirds threshold. Not all of these samples are given strictly to augment holes
in prescription coverage. At least 25 percent of all Americans have no prescription coverage, as reported by National Conference
State Legislatures. Some is simply the cannibalization of potential prescription drug customers.
That leaves only 25 percent of samples to be given to newly diagnosed patients with a prescription—the classic scenario of
new-patient trial that represents the highest potential return on investment. And any new system that will reduce the tendencies
or incentives for physicians to dispense free therapy when the patient could afford a prescription represents a double return
for the industry: a paid prescription and no samples dispensed.
What Next? Because the returns associated with sampling are so poorly understood, analyzing the short and long-term effects of changes
in the sample distribution system is challenging. There are hundreds of possible variants between the current status quo and
a complete end to sampling. Yet one thing is clear: The ROI of the current system is so low that any attempt to change it