The OIG found that the FDA does not have a mechanism to identify all clinical trials and institutional review boards (IRBs). Moreover, it lacks a comprehensive database for tracking its inspections of clinical trials. “Data limitations hinder the FDA’s ability to ensure that participants are protected from unreasonable risks,” said HHS Inspector General Daniel R Levinson.
The report focuses on inspections carried out by the Office of Regulatory Affairs (ORA) as part of the FDA’s Bioresearch Monitoring Program (BiMo) - an important mechanism for protecting human subjects once a trial is underway. The agency inspected only 1% of clinical trial sites and few IRBs from 2000-05, the OIG concluded after reviewing multiple data sources. Of the site visits, 75% were surveillance inspections, which usually target completed trials to verify data quality.
The OIG found a lack of consistency in determining action needed following site visits between ORA inspectors and staff in the three Centers that handle drug, device and biologics registration. Notably, the Center for Drug Evaluation and Research (CDER) downgraded 68% of recommendations of the most serious inspection classification category ‘official action indicated’ to ‘voluntary action indicated’, which neither requires the site formally to address violations nor makes the classification publicly available. In other examples, Center reviewers identified violations where a BiMo investigator had found none; the CDER reclassified 26% of the inspections that ORA identified as ‘no action indicated’.
Previous OIG reports have found similar weaknesses. In 1998, a series of reports concluded that IRBs lacked the time and expertise sufficiently to monitor research, and a 2000 report found that data integrity concerns, rather than human subject protection, drove FDA’s oversight of clinical investigators.
The OIG made the following recommendations to the FDA to improve its oversight of clinical trials:
· Develop a comprehensive internal database of all clinical trials
· Create a registry of IRBs
· Create a cross-centre database to allow complete tracking of FDA inspections
· Seek legal authority to provide oversight that reflects current clinical trial practice
· Establish a mechanism to provide feedback to FDA district office staff on their inspection reports and findings.
The FDA concurred with four of the recommendations, but did not address the provision of feedback regarding inspection reports and findings. The agency pointed out that BiMo inspections make up only one part of its efforts to ensure protection of human trial subjects; it views the protocol review before a trial commences as the most important step in protecting subjects.
Recently, the FDA has taken steps to improve its BiMo inspection processes. In 2004, it created the BiMo Steering Committee – since renamed the Human Subject Protection/Bioresearch Monitoring Council - to review processes and make recommendations for improving them. The Council issued new guidance on clinical trial conduct and policy last year.
The ORA employs about 1,300 investigators, who conduct a total of 22,000 inspections annually. About 200 investigators have received specialised training to conduct BiMo inspections. BiMo investigators work out of all 19 of the ORA’s district offices.
