E-DRUG: please be a little moderate.... (3)
--------------------------------------------
Dear e-druggers~
I respect the positions of prescribers described in this post and offer some comments regarding risks. Although most approval processes dictate that drugs must be proven to be "safe and effective" before entering the market place (and this is indeed the expectation of the general public), studies relied on for approvals are designed to answer questions of efficacy, not safety.
It would be impossible to test drugs for all potential safety concerns prior to marketing. Because of this, our drug systems rely almost completely on post-marketing observational and clinical studies to identify and define important safety issues that would be difficult if not impossible to quantify in a clinical setting (where individual practitioners may not see or follow enough numbers to identify risks unless they are easily detected and occur with sufficient frequency).
"The literature" may be a poor source of risk information. For example, the rate of intrathecal granulomas associated with implanted spinal pumps has been referred to as rare in the literature (consisting of numerous reviews and few, if any, epi studies). But the literature fails to adequately distinguish that the represented numbers are reporting rates, not true incidence rates. In this case, patients may (and likely do) die before the event is detected (rarely do autopsies include an MRI which would be required to detect). Regardless of whether or not the rate is considered "rare," the seriousness of this event is such that it clearly necessitates that clinicians know and communicate this potential outcome to patients, including the unknown incidence.
Pharmaceuticals are necessary and essential tools in therapeutics; the usefulness of which depends on a delicate balance of anticipated benefits for an individual patient versus known and unknown risks. Benefit information may be gleaned from clinical trials, "detailers" and/or the prescriber/patient experience, but what about risk information? For physicians who are required to act as "learned intermediaries" for their patients, this becomes problematic. Ideally, epidemiologic studies would identify, define and quantify important risks associated with drugs, and this information would be transparent to both prescribers and patients. Currently, clinicians and patients rely on incomplete and often biased information on risks from all sources available, including the medical literature and the press.
Sarah Sellers, PharmD MPH
ssellers@jhsph.edu