The paper trail

ILLUSTRATION: NORM BENDELL

A primer to the FDA audit process for clinical trials

If you are involved in managing, conducting, or sponsoring a clinical trial, three fears can cause you to lose sleep: low enrollment, poor results, and the possibility of an FDA audit. The first worry is manageable, the second uncontrollable, and the third, well . . . the third is utterly terrifying. This is not to imply that you are trying to hide anything, or even that you have something to hide. But rather, the notion that the governmental body charged with determining your drug’s commercial fate wants to audit your trial is a legitimate, albeit sometimes irrational, horror.

The FDA conducts audits for various reasons but always with the same goal: to ensure the safety of research participants and guarantee the accuracy of each project’s clinical data. Specifically, because the FDA sets the rules under which U.S.-based research is conducted and ultimately must render a verdict on the outcome of every trial, it is crucial that compliance with the regulations (including patient safeguards) and proper collection of the data be ensured by the agency itself.

The worst-case consequences of a negative FDA review are well known to investigators: disqualification from research practice and possible jail sentencing. In truth, these outcomes are rare and result from criminal investigations of actual fraud (see “Investigators gone bad”, April 2001). The more likely disciplinary actions range from a figurative slap on the wrist to restrictions that limit participation in future projects.

But the FDA can audit more than just the conduct of researchers. Investigational Review Boards (IRBs), the sponsoring pharmaceutical companies, and the contract research organizations (CROs) that they often hire to monitor and manage their trials are all prime targets. So what triggers an FDA audit? What are they looking for? And what do you really have to fear if you are one of the managers, monitors, coordinators, investigators, reviewers, or research sponsors?

Study conduct
Most audits occur at investigative sites. But the FDA lacks the financial and personnel support necessary to audit every site during every trial phase. In fact, small studies are often never reviewed, and some products are approved entirely without audits.

FDA auditors conduct one of two different investigational site reviews: study- or investigator-oriented. In the former, the auditor seeks verification that a specific clinical study was conducted properly and that the data produced by the site was accurate and complete. In the latter audit, the FDA is concerned about a researcher’s general practice. Thus, for study-oriented reviews, the conduct associated with just one protocol is considered, whereas for investigator-oriented audits, several research studies may be reviewed.

There is no mechanism that automatically selects a site for audit. However, investigator-oriented reviews generally occur only when the FDA has cause to suspect a particular researcher’s conduct. (This may be the result of serious concerns reported by the site’s monitor or coordinator, a CRO, the trial’s sponsor, or even a study participant.) Study-oriented audits are triggered by more subtle factors, however, including the following.

Patient enrollment. Particularly in Phase III trials, the highest-enrolling sites can be suspect because, statistically, the research findings are skewed in favor of those sites’ results.

Patient retention. Large numbers of “screen failures” can indicate a problem with the way a site identifies and screens its participants. Similarly, sites that have unusually high patient dropout rates are curious and sometimes trigger audits.

Adverse events. When large numbers of severe adverse events occur at only one or two sites in a multisite trial, a red flag is raised. (Serious events that occur at numerous sites often trigger a review of the entire study.)

Trial importance. Pivotal studies (e.g., individual trials from which data will be used to make significant claims) and the investigators who conduct them are ripe for review.

The process
Researchers are never caught off-guard by an audit because the FDA notifies them in advance. In addition, to help with preparations, a pharmaceutical sponsor (or CRO if there is one) will send its own auditors to review an investigational site before the FDA’s visit. This serves, of course, as a means of identifying and correcting problems in advance of the more formal audit.

On arrival at the site, a regulatory auditor will identify himself or herself and provide a written notice of inspection. The auditor will then proceed to review the case report forms and all supporting materials for some, or all, of the patients enrolled in a specific trial (or in several trials, if the audit is investigator-oriented). The focus of the audit is to determine:

• how study-related tasks were delegated (and to whom),
• where they were performed,
• how the data was recorded,
• how accurately the data was captured,
• how the drug accountability was maintained,
• how the study was monitored, and
• what other, ongoing reviews—including the follow-up of discontinued patients and the reporting of adverse events to the IRB—are still occurring. In the case of an investigator-oriented audit, this process is simply more detailed.

In general, audits tend to identify similar problems at all research centers. The most common issues found during FDA reviews include inaccurate or incomplete drug accountability logs, errors in adverse event and concomitant medication reporting, and errors in the transcription of source documentary evidence into study-related records. Two other common findings are that review pages, contained at the end of patient visit charts, have been signed in advance of the visits and that FDA Form 1572, which identifies those conducting a particular clinical trial at a specific location, has not been updated.

At the conclusion of an audit, the FDA inspector discusses any findings of concern with the principal investigator. Commonly, FDA Form 483, which summarizes the observations, is also completed. On the basis of an evaluation of the reported findings, one of three letters is then sent to the investigational site: a notice that no significant observations were reported, a letter that identifies deviations that may or may not necessitate a response or corrective action from the investigator, or a warning letter that notifies the site of serious violations that require immediate rectification.

CROs, IRBs, and sponsors
As stated earlier, other parties associated with the clinical research process may be subjected to FDA review. Monitors who routinely oversee investigational study conduct, coordinators who assess patients and chart their trial-related data, and even project managers and pharmaceutical sponsors may have their records audited. (It bears mentioning that clinical laboratories, drug manufacturing facilities, statistical databases, and animal-based research sites may be investigated; however, these areas lie outside the scope of this article.)

Besides investigational centers, the most audited body in clinical research is the IRB, because it is charged with the responsibility of maintaining patient safety. One IRB may oversee many trials, affecting a staggering number of patients.

An FDA inspector conducts on-site reviews, similar to investigational site audits, of IRB practices, procedures, and records. The inspector carefully scrutinizes the IRB’s considerations and approvals of patient informed consents and protocol documents, its physical makeup (membership requirements are specified in the regulations), and its voting procedures.

The outcome of an IRB audit is also quite similar to that of an investigational review: The auditor presents the IRB chairperson with an observational summary and may include FDA Form 483. After an evaluation of the findings is complete, any necessary corrective actions are placed in writing and sent to the IRB, which then may be required to respond in kind.

Sanctions
Administrative actions may be taken against any person or agency reviewed by the FDA. In all cases, one of three actions is possible: a consent agreement, a disqualification, or a criminal prosecution. The most severe of these punishments, criminal prosecution, is also the least used and is almost exclusively reserved for fraudulent investigators.

In the case of disqualifications, the offender’s right to participate in clinical research is withdrawn. But disqualifications are rare; an extensive framework exists for contesting such an action, and, at any time, the accused may accept a consent agreement from the FDA. Thus, a decision to enter into a consent agreement occurs most commonly. In a consent, the offending person (or body) agrees to meet certain FDA requirements. In general, these mandates fall into one of two categories: restrictions on future participation in research trials, or complete separation from research for a period of time. The restrictions imposed for physicians range from agreeing to conduct trials only under the supervision of an FDA-approved investigator all the way to being barred from participating in any trial for 3–10 years.

In reality, IRBs are occasionally required to respond to FDA concerns in writing, but as with CROs and pharmaceutical companies, the threat of substantive sanctions is chilling enough to guarantee corrective action and diligent self-monitoring. Further, monitors and coordinators are almost never the targets of FDA investigations—their supervisors are the “bigger fish”. Principal investigators, in fact, represent the biggest fish. They often believe that they can take greater chances without being caught and are, after all, the controlling factor in how accurately clinical data reflects actual patient outcomes and how well research patients are treated. Thus, the FDA’s focus on them is particularly justified and serves to protect patients nationwide.