Post-Marketing Drug Safety: Keep Patients Your Priority

The December 1, 2004, issue of JAMA contains several articles by prominent authors in the area of drug safety and regulation.^1–5 They highlight the need to be mindful of our profession's long-accepted first principle: “First, do no harm.” To me, the “short messages” include a need for greater investment in drug-safety mechanisms and greater caution about relationships between physicians and the pharmaceutical industry.

Psaty et al. present information about cerivastatin, marketed as Baycol (Bayer) and withdrawn in 2001, and call for an additional regulatory body to monitor drug safety.¹ Strom discusses the current adverse drug reaction (ADR) reporting system and recommends that resources be added to the existing structure at the Food and Drug Administration (FDA).² In spite of attempts to present a balanced view of a complicated topic, I am left with many doubts, since even some of the facts of the case seem disputed. In any case, we must draw what conclusions we can and translate them into patient care.

The principal thrust of the report by Psaty et al. is that, after the release of cerivastatin on February 18, 1998, the dataset of ADRs grew, but important analyses of these data remained internal Bayer Corporation information.^1,5 As Piorkowski notes, Bayer modified the label of cerivastatin five times during the 3 years that the drug was available to try to improve its safety before ultimately withdrawing it from the market.⁴

Psaty et al. summarize changes in the drug regulatory environment aimed at more rapid approval of beneficial new drugs.¹ They conclude that this approach “relied increasingly on the pharmaceutical industry to conduct its own post-marketing safety evaluation.” They raise the concern that “a pharmaceutical company's appraisal of suspected ADRs may be influenced by economic considerations,” and call for legislation to “mandate and provide adequate support for independent reviews and analysis of post-marketing data.” Other authors have provided additional commentary on this subject, including an FDA scientist, David Graham, MD, MPH, in high-profile Congressional testimony.^6–8 But many issues remain unresolved, while our patients look to us for guidance.⁹

In my opinion, the pendulum has swung too far toward the tendency to use new drugs quickly without building in adequate safety mechanisms and questioning their added value. From the economic standpoint, it is difficult to know whether a company that keeps a potentially hazardous product on the market in the interest of short-term economic gains has a net win or loss when that product is recalled or removed. However, physicians should not have any doubt about the mission of drug companies—to make money for their shareholders. Yet many physicians act in a way that makes them beholden to drug companies, as if “innocent ignorance” is possible.

Make no mistake—pharmaceutical companies have provided astonishing advances in medical care, but they are under intense pressure to recoup the expenses of drug development. Their marketing (for example, AstraZeneca spent $122 million on advertising of Crestor from January to September 2004⁹) often includes payments to physicians who believe in the drug to serve as opinion leaders. Decisions driven by business interests may not advance medical practice as effectively as one would hope.¹⁰ For example, do we really need ten ACE inhibitors, seven angiotensin II blockers, or six statins marketed for the same indications? Although regulatory reform may improve drug safety, it is not the only step we need to take. Emphasis, through regulation, on the safety of an individual drug does not guarantee safer drug use. Patients and physicians must have the ability to track total medication use and the time and tools to monitor drug appropriateness and safety. Although we have tools for this, we need to use them more effectively, and we need better ones.

As the drug-approval process in the U.S. is accelerated, more of the post-marketing safety assessment of a drug will fall on the U.S. health care system, and we are ill-prepared to deal with it. Patients and physicians have a “presumption of safety” for new drugs. New drug use is driven by physicians hoping for a new advance for their patients, and by patients who receive directly marketed information that is known for its distortions. Prescribing decisions are fraught with problems, including distortion of benefits based on relative risk changes, inappropriate use of surrogate end points while extrapolating clinical benefits, inappropriate presumptions of a “class effect” with respect to drug efficacy and safety, use of medications for non-FDA-approved indications, and incomplete prescription risk assessment or safety monitoring.¹ For example, among a group of Medicare patients hospitalized with heart failure and diabetes in 1998, 7% received a prescription for metformin, 7% for a thiazolidinedione, and 14% for either drug. Two years later, in a similar cohort, metformin use was 11%, thiazolidinedione was 16%, and either drug was 24%.¹¹ Yet the FDA warns against both practices. The troglitazone hepatotoxicity story showed that physicians often do not adhere to recommended safety-monitoring schedules. In the Netherlands, 23% of patients prescribed rosuvastatin had risk factors for myopathy, and 37% of patients had it prescribed as the initial drug.¹² Yet this drug has no proven benefit on clinically important outcomes.

It is time for physicians and their patients to resist drug-company marketing and to have a more measured approach to medication use—new and existing. Ensuring drug safety is a labor-intensive process that often conflicts with the ever-quickening pace of medical practice. It is much quicker to simply prescribe a medication than to take the time to review the true risks and benefits of a drug with a patient, especially when that patient comes primed by drug-company marketing. Yet, investment in cognitive services by those who pay for health care is declining. This must change if we want to ensure drug safety. Computer-aided clinical-decision support tools benefit hospitalized patients, but most drug use occurs in the outpatient setting, which, with the exception of highly managed care systems, is extremely fragmented and does not promote drug safety. Although process improvements by the FDA may improve drug safety, it will take much more to improve the safety of drug use, and we owe it to our patients to advocate for these changes. ▪

Letters to the Editor

Our goal is to provide the latest news and practical information about preventing and treating diabetes, obesity, and cardiovascular disease. Your comments and feedback are welcome. Send email to docnews{at}diabetes.org.

Visit DOC News online at www.diabetes.org/docnews, where you'll find current and archived stories plus other useful features.