Tips on Unanticipated Problems and Adverse Event Reporting

The IRB requires human subject researchers to inform the IRB of problems that arise while conducting approved studies. Two kinds of problems are relevant to the IRB’s responsibilities: unanticipated problems and adverse events.

“Unanticipated problems involving risks to subjects or others” are mentioned but not defined in 45 CFR part 46. The federal Office of Human Research Protections (OHRP) defines an unanticipated problem as an unfavorable incident or occurrence that meets all three of the following conditions (http://www.hhs.gov/ohrp/policy/advevntguid.html):

  1. the incident is unexpected in terms of its severity, nature, or frequency, relative to the research project described in the protocol or the subject population;
  2. the incident is possibly related to participation in the research;
  3. the incident places research subjects or others at a greater risk of harm – whether physical, psychological, economic, or social – than was previously recognized. 

Generally, unanticipated problems necessitate substantive changes to research procedures and informed consent documents. Other corrective actions, such as notification of research subjects, may also be appropriate.

The following examples of unanticipated problems are modeled on two of the sample cases provided in the guidance document linked above:

  1. a behavioral investigator collects individually identifiable information about illicit drug use by surveying college students. The data are stored on a laptop without encryption, but the laptop is then stolen from the investigator’s car.  The incident qualifies as an unanticipated problem because it is unexpected, related to the research subjects’ participation, and puts the research subjects at greater risk of social and psychological harm. 
  2. as a result of a processing error by a technician, a research participant is accidentally given a much higher dose of an experimental drug than was described in the approved protocol. Even if the subject experiences no ill effects, the incident still qualifies as an unanticipated problem. It was unexpected, the direct result of research participation, and the mistake exposed the research subject to additional risk of physical harm, even if no harm actually ensued.

By contrast, “adverse events” are not even mentioned in 45 CFR part 46.  OHRP defines an adverse event as an unfavorable medical (physical or psychological) occurrence that is at least temporally associated with the subjects’ participation in the research (http://www.hhs.gov/ohrp/policy/advevntguid.html). Thus some, but not all incidence may be categorized as both an unanticipated problem and an adverse event.  For example, some adverse events may be expected, such as when research subjects with a particular disease experience severe symptoms. These types of adverse events would not also be considered unanticipated problems specifically because the problem was expected.  Moreover, some unanticipated problems are not medical and thus are not adverse events. Also, an unfavorable medical occurrence might be temporally associated with research participation – i.e., it occurs while the study is ongoing – without being potentially related to the research.

Knowing the difference between unanticipated problems and adverse events is more important for the IRB than for researchers.  The distinction is important because as is described at 45 CFR 46.103(a), the IRB must report unanticipated problems to OHRP, but is not required to report adverse events. USM policy requires, however, that researchers report both unanticipated problems and adverse events to the IRB. Both should be disclosed as soon as is reasonably possible, so as to help the IRB mitigate negative effects of unfavorable events. Since the distinction is more pertinent to the IRB’s responsibilities than the responsibilities of researchers, USM uses a common form to report both unanticipated problems and adverse events.