What's New
Announcement
Posted on 01.11.01
CERTs Investigators Discuss Postmarketing Surveillance at Annual Public Health Meeting
By Janet Weiner, MPH
"It is simply not possible to identify all the side effects of drugs before they are marketed." -- Wood AJ et al., N Engl J Med, 1998
The Food and Drug Administration's approval process for new drugs has undergone, and withstood, intense public scrutiny over time. The study of drugs after they reach the market has received far less attention, however. Two CERTs investigators highlighted the importance and challenges of such "postmarketing surveillance" at the annual meeting of the American Public Health Association in Boston on November 13, 2000.
Penn CERTs investigator Sean Hennessy, PharmD, discussed methods for gathering additional information about an approved drug's risks and benefits under real-life conditions. He described the need for rigorously designed studies with approved written protocols.
"Regardless of their design, all postmarketing studies involving human subjects should be conducted by qualified investigators according to a written protocol that has been approved by an institutional review board," Hennessy stated.
He outlined the most common epidemiological designs used in postmarketing surveillance: randomized trials, cohort studies, and case-control studies. Well-designed studies are critical in understanding cause-and-effect relationships between drugs and outcomes, and in identifying risk factors for adverse events, he noted.
Large databases can help researchers assess drug safety after approval, according to Richard Platt, MD, MSc, principal investigator of the HMO Research Network CERTs. He described methods used by the Network to assess drug safety, which include using the records and experience of 6 million members of 8 HMOs.
"Networks of databases have been used to study problems that require large or diverse populations," he said.
But merging data from different systems poses special challenges, Platt noted. These include differences in reliability, coding, and level of detail, as well as the need to protect the confidentiality of medical records. Nevertheless, the multicenter nature of these databases offers new possibilities for postmarketing surveillance programs.