What is an Unanticipated Problem, and how does it differ from adverse events and deviations?

Strictly speaking the category of "unanticipated problem" fits a broader category of events that involve risk to subjects or others. You should review our Unanticipated Problem definitions, which also includes some examples.

To recap: unanticipated problems are (a) unexpected in nature/severity/frequency in relation to either the research itself or the population being studied, (b) related or possibly related to the protocol, AND (c) suggests an increase in risk to subjects or others.

Once you define the problem you can use our Unanticipated Problem Reporting Guide to help you determine if you need to report the event.

Adverse events and deviations may be specific kinds of unanticipated problems (UPR). Examples include:

Three subjects enrolled in a phase I trial of a study drug all experience organ failure that was not previously associated with the drug or the patient's underlying condition, and requires hospitalization to treat. This is clearly an adverse event for each of these three subjects, and the event is (a) unexpected (wasn't previously associated with the condition or drug), (b) related or possibly related to the study, and (c) suggests an increase in risk to others (if the drug is administered to others with the condition they might suffer similar organ failure). As such this adverse event is ALSO an unanticipated problem. Please note: you only need to submit this as an Adverse Event, because as such it is a specific kind of UPR.

A subject is administered the incorrect dose of study medication, even if there is no immediate indication of harm to the subject. This deviation from the protocol is obviously an unexpected event that is directly related to the study and suggests an increase in risk to subjects (either increased likelihood of side effects or decreased efficacy, depending on the error). As such this deviation is ALSO an unanticipated problem.

The UPR form is most likely to be used for events that are not subject specific. Among the examples that fit this are: loss of a computer with identifiable research information on it; interim findings by a DSMB or similar body that the risks of the research indicate the study should end early; contamination of drug processing equipment; etc.