arrhythmia as a surrogate marker of efficacy, but although the drug reduced arrhythmia, it also increased cardiac death (NHLBI, 2005).

Additional information on the safety of marketed drugs may come from large, automated databases. FDA purchases access to some of those databases as a resource for pharmacoepidemiologic studies designed to test hypotheses, particularly those arising from AERS. (Chapter 4 contains a more detailed discussion of relevant programs and agreements with academic research institutions.) FDA also obtains information from IMS Health, a provider of market research services to the pharmaceutical and health care industries. Among the services obtained from IMS Health are the National Disease and Therapeutic Index, which provides data on diagnoses, patients, and treatment patterns; Integrated Promotional Services, which measures professional and consumer promotional activity in the pharmaceutical industry; and the National Prescription Audit, which tracks pharmaceutical products dispensed in retail, mail-order, and long-term care channels.

Trained staff and (often expensive) supportive technology are typically needed to use some of those databases fully. CDER’s limited resources for such activities have precluded taking full advantage of their potential contribution to understanding the safety of approved drugs (see further discussion in Chapter 4).

Resources also severely constrain their external research program. DDRE has about 18 epidemiologists. They work with the safety officers (who are also referred to as safety evaluators and generally are pharmacists) on assessments, determining for example the background risk of a condition to determine whether reported rates may be above expected levels. They also oversee agency-sponsored epidemiologic research.

As CDER receives new information related to a drug’s safety profile, it makes risk assessments and determines how risks can best be managed. For monitoring purposes, every marketed drug is assigned to a safety evaluator, usually a pharmacist in DDRE. Generally, one safety evaluator oversees all drugs in a class, such as statins, so he or she tends to work consistently with a specific OND division that handles those drugs. AE reports on a drug are automatically forwarded by e-mail to the appropriate safety evaluator and to the OND reviewer with responsibility for that drug.

ODS/OSE employs about 25 safety evaluators, and each receives about 500–800 reports a month to monitor, including some that are designated as “serious” on the basis of criteria established by FDA. The committee was told that safety evaluators now have less time than before to keep up with their inboxes as they are spending more time on OND consultations, in developing complex postmarket risk assessments, and in such activities as