nically diverse populations. The Ingenix site has access to some laboratory data in addition to claims data, and the HMO Research Network and the Kaiser Foundation Research Institute sites have access to electronic medical records. Study designs for the contracted studies often are presented to the Drug Safety and Risk Management Advisory Committee or involve other outside experts through the special government employee mechanism for review and comment as a form of scientific peer review.

The funding for the cooperative agreement program is severely limited and the program has always been small. In 1985, the funding level was $1.2 million; since then, resources have varied. Despite inflation in the interim, funding for FDA drug safety cooperative agreements reached a low of $900,000 in 2000 (personal communication, Gerald Dal Pan, FDA, March 30, 2006). In fiscal year (FY) 2006, funding for FDA drug safety contracts totals only $1.6 million, and it is scheduled to decrease to $900,000 in FY 2007. According to an ODS annual report, the contract program in 2004 supported five feasibility7 studies and three in-depth studies, but in FY 2006 the program will have sponsored feasibility studies for two drug safety questions and will not have sufficient funds to execute one high-priority in-depth study fully—on the cardiovascular risks posed by drugs prescribed for attention deficit hyperactivity disorder (ADHD) (IOM Staff Notes, 2005–2006). In contrast, a similar program funded by the Centers for Disease Control and Prevention (CDC) to study safety problems associated with vaccines, the Vaccine Safety Datalink (VSD), included data on more than 7 million people covered by eight managed-care organizations. CDC supported the VSD with $13 million and eight full-time staff persons in FY 2004 (Davis, 2004).

FDA also works with the CERTs that have access to large healthcare databases, including the HMO Research Network and the Department of Veterans Affairs (VA). CDER has internal access to the General Practice Research Database (GPRD)8 and to proprietary databases9 that house extensive information on drug use. Access to the GPRD was expected to provide valuable information to CDER for drug safety purposes, but ODS/OSE has struggled to get sufficient computer resources and staff trained to use it. Four full-time safety evaluators now work with those databases, and two staff epidemiologists work part-time with them in their research.10 CDER staff presented their first findings from the GPRD at the 2006 summer meeting of the International Society of Pharmacoepidemiology.

7

A feasibility study involves preliminary assessments of whether a database contains sufficient exposures or outcomes in appropriate populations to answer the study question.

8

A computerized database of longitudinal medical records from primary care practices in the United Kingdom and a source of data for many epidemiologic studies around the world.

9

Verispan, LLC; IMS Health; and Premier.

10

Personal Communication, G. Dal Pan, FDA (ODS/OSE), 2006.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement