Reporting rules for adverse events, unanticipated problems differ slightly

IRBs and investigators often confuse the two

IRBs often struggle with decisions regarding the reporting of adverse events and unanticipated problems, and the recent increases in IRBs’ workloads do not help the situation, experts say.

One of the biggest regulatory problems facing IRBs is dealing with the volume of information they receive and efficiently and effectively determining data items that require reporting vs. items that don’t, notes Carol Weil, JD, compliance oversight coordinator and public health analyst with the Office of Human Research Protections (OHRP) in Rockville, MD.

"Different agencies and sponsors have different requirements, so it’s difficult for IRBs to advise researchers of what needs to be reported to whom," she says.

For instance, the term "adverse event" does not appear in the Department of Health and Human Services (HHS) regulations that pertain to IRBs, Weil explains. "Under HHS rules, the requirement is for any unanticipated problems, and this includes adverse events and nonadverse events."

Nonetheless, IRBs and investigators continue to think primarily in terms of adverse events, Weil adds.

That term is found solely in the device and drug regulations, and it pertains to investigational test articles, says Jeffrey A. Cooper, MD, MMM, deputy director of the Association for the Accreditation of Human Research Protection Programs Inc. (AAHRPP) of Washington, DC. "It is not in the IRB and informed consent regulations, so if there’s no test article, such as in a social-science study, then you can have an unanticipated problem, but no adverse event," he adds.

However, there is an overlap between the two, and this helps to confuse IRBs about reporting requirements.

"This area is so difficult because there are so many different perspectives," says Ada Sue Selwitz, MA, director of the Office of Research Integrity and an adjunct associate professor of behavioral sciences at the University of Kentucky in Lexington.

If IRBs only had to deal with HHS regulations pertaining to unanticipated problems, then it would be far simpler to set up effective systems and policies, she says.

"But in addition to unanticipated problems, there are industry sponsors governed by the Food and Drug Administration [FDA], and their regulations for adverse event reporting are dramatically different from the requirements for IRBs," Selwitz adds.

Add to this mix the fact that sponsors also are regulated by the FDA, and they may have their own interpretations of what the FDA requires them to report, and they in turn order investigators to handle adverse events a certain way, she notes.

Because of the confusion that adverse events and unanticipated problems reporting may cause IRBs, it’s probably a good idea for IRBs to clarify their policies and procedures regarding the reporting of these items. Here are some suggestions for improving an IRB’s adverse event/unanticipated problem reporting:

• Clearly define adverse events and unanticipated problems. "It’s important that organizations make their investigators understand this concept of unanticipated problems and how this involves risk to subjects and others," Cooper says. "Unanticipated problems are major things that happen that are unexpected and present some change to the risk-benefit ratio of the research."

The University of Kentucky has written a draft policy that outlines what adverse events and unanticipated problems entail and how the investigator needs to report these events, Selwitz says.

For example, such a policy might say that it applies to any problem or adverse event that affects the rights, welfare, and safety of subjects and which could involve physical, social, and psychological risk or involve an inappropriate breach of confidentiality or invasion of privacy.

A policy also might differentiate between unanticipated problems, unexpected adverse experiences, and unanticipated adverse device effect, defining each and including examples.

"What’s a risk or a problem?" Weil says. "You can read into the language the notion that if someone breaks a fingernail, then that’s not a risk and not a problem."

Still, if there’s a fairly innocuous problem that happens to everyone involved in the trial, then it might be irrelevant to human subjects protection and to the research, but it’s still important to know, she adds.

Another aspect to consider is that despite an institution’s best intentions to define and differentiate between adverse events and unanticipated problems, it’s likely investigators will continue to use the term "adverse event" to refer to all of the above, Selwitz notes.

"We’re dealing with a constituency of investigators who think in terms of adverse events, so we have to balance these different perspectives and requirements," she says.

• Provide examples of unanticipated problems. One example of an unanticipated problem, which occurred during the course of one research study, is that a researcher’s computer had been stolen from her office, Weil says.

The computer contained identifiable data, which could pose a risk of harm to human subjects, she adds.

"That was the kind of thing that you wouldn’t see as a problem [directly] with the research, but it was reported as an unanticipated problem," Weil says.

Another example of an unanticipated problem is when an adverse event that is expected in a small number of research participants suddenly appears in a greater number of subjects. While this event was anticipated on a small scale, it became an unanticipated problem when it began to occur on a broader scale, she explains.

IRBs also would need to report an unanticipated problem in the rare case of an investigator being arrested for an unrelated crime. There was one example of this scenario, and the police wanted to go through the files of the researcher, which would have breached confidentiality of research subjects, Weil says.

"An unanticipated problem can arise from almost any sort of information that the IRB might receive," Cooper notes. "Unanticipated problems could be unplanned changes, deviations, exceptions, or violations."

While most deviations, exceptions, and violations are minor and trivial, having no impact on risk-benefit ratio, there could be some that indicate there is a problem with the study’s design, which may need to be modified because it involves unanticipated problems for subjects and others, he says.

For example, in the AbioCor Implantable Replacement Heart clinical trials, one patient was unable to return to his home after experiencing a stroke and breathing difficulties because his home did not have the infrastructure necessary for the necessary medical equipment.

Once this unanticipated problem arose, it would be important for the IRB to modify the informed consent so that future subjects would not be confronted with that problem, Cooper notes. "These things happen all the time, and they are things that people just didn’t think about," he says. "The most important thing is that people recommend these as issues that are reported to the IRB."

• Devise a good strategy and guidelines for reporting of unanticipated problems. IRBs and institutions might offer investigators a series of instances in which unanticipated problems are reported and offer them a flowchart or guidelines for where, when, and how to report these events. (See flowchart.)

"We’re developing an unanticipated problems/adverse events reporting policy for our investigators that specifies what problems and events they need to report to the IRB, and when and how," Selwitz says. "The other challenge for IRBs is that it’s critically important to know what the IRB’s responsibility is for reporting these issues to either the FDA or OHRP."

The first step to developing a reporting strategy and guidance is to assess all regulations and policies that impact researchers and IRBs, she suggests.

Then harmonize and streamline the institution’s policy based on the nature of the institution’s research programs, Selwitz says.

Developing the policy will take time and work. The University of Kentucky had a committee of eight people who worked on the new policies for more than a year, she reports. "We took at least five or six months assessing the different requirements," Selwitz adds.

• Be familiar with OHRP’s guidance on this subject. "It’s important to be familiar with the regulations that deal with unanticipated problems and also to be familiar with the guidance OHRP has published on unanticipated problems," Cooper says. "It can be found on the OHRP’s findings and guidance document on-line."

HHS has rules requiring that institutions have policies in place for ensuring that adverse events and unanticipated problems are promptly reported," Weil says. "That means there ought to be a written policy somewhere so investigators know the time frame for reporting and so that the IRB knows which institutional officials they need to report to and how that information gets up to OHRP."

• Offer guidance on when it is necessary to require a modified consent form. "We ask IRBs to make a determination of whether or not it’s important to amend the consent form, either because it’s a risk that’s not mentioned at all, or because it’s identified as one that’s expected to have a lesser occurrence than what happened," Weil says.

For instance, suppose 80% of subjects have headaches, and headache is only listed as a rare risk, then the consent form would need to be modified to show that a headache is a common problem, she explains.

On-line flowchart

Jeremy Wood, PhD, communications consultant and founder of, has created an on-line SAE flowchart called SAEtool ( The free program is designed to help users determine whether and how FDA regulations may apply. Each screen offers the user a series of questions or prompts, such as "My question is related to: Drug, Biologic, Medical Device." The user selects the appropriate response and is sent to the next appropriate screen. For example, if you were to select Drug, the screen to follow would ask you to select from among the following the scenario that applies to your situation:

  • an Investigational New Drug covered by an IND (Investigational New Drug application);
  • postmarketing reporting of adverse drug experiences;
  • a prescription drug marketed for human use without an approved new drug application;
  • none of the above.

Once the user has selected all of the elements that apply to his or her situation, the tool will advise the user whether an adverse event has occurred and/or needs to be reported to federal entities.

Though the tool may be useful to investigators, IRBs still must have their own policies in place regarding reporting adverse events and unexpected problems. "The FDA Investigational New Drug regulation (21CFR 312.66) requires the investigator to report promptly to the IRB all unanticipated problems involving risk to human subjects . . .’" says John Isidor, JD, CEO of Schulman Associates IRB in Cincinnati. "Accordingly, it is imperative for the investigator to understand and comply with the adverse event reporting policy of the reviewing IRB."