FDA audits of clinical studies: policy and procedure.

The Food and Drug Administration has been conducting audits of the work of clinical investigators since 1962, and has done so on a routine basis since 1977. Since 1977 the work of approximately 2,800 drug investigators has been subject to inspection; approximately 500 of these have been "for cause," including 49 performed outside of the United States. The routine audits are done on studies important to the FDA's approval of therapeutic claims, and are thus not targeted at the investigator, but at the validity of the study. Nevertheless, problems with consent were found in 54%; inadequate drug accountability in 25%; nonadherence to the protocol in 26%; inadequate and inaccurate study records in 21%; problems with institutional review board notification/approval in 13%; nonavailability of study records in 3%; and inappropriate delegation of authority in 4%. Ten percent of routine inspections revealed deficiencies of frequency and/or magnitude sufficient to warrant follow-up activity. We believe that this figure represents the only available assessment of the incidence of scientific misconduct. Of interest is a decrease to 6% in the past few years. "For cause" inspections may be either investigator-targeted or study-targeted. The former have resulted in regulatory action in 100 cases and successful criminal prosecution in 16. The latter have resulted in nonapproval or delay of approval of drugs for marketing. A subset of study-targeted inspections is the foreign inspections performed on studies important to U.S. approval of claims of efficacy of 24 drugs.(ABSTRACT TRUNCATED AT 250 WORDS)