03 · What You Need to Know
A serious incident creates several parallel responsibilities
Continue protecting the participant after the acute event
Immediate care is the first priority when a participant becomes ill or injured during research. Once the acute situation has stabilized, the participant may still require follow-up.
The research team should follow the protocol and applicable clinical arrangements concerning outcome monitoring, additional evaluation, communication with appropriate healthcare professionals, interruption or discontinuation of the research intervention, and any other necessary safety measures.
Do not let administrative reporting create the illusion that the participant's case is closed.
Preserve facts before reconstructing the story from memory
A serious incident may later be reviewed by investigators, sponsors, safety monitors, IRBs or RECs, institutions, regulators, insurers, or others. Accurate contemporaneous information matters.
Document what happened, when it happened, what research activities preceded it, what was observed, what the participant reported, what actions were taken, who was contacted, and what the participant's outcome was as information becomes available.
Preserve relevant source records, device or equipment information, medication or product details, procedural records, communications, and other materials required by the study and applicable rules.
Separate observation from interpretation. “Participant lost consciousness three minutes after the procedure” is a fact. “The procedure caused the loss of consciousness” is a causal assessment that may require further evidence.
Do not wait for perfect information before making time-sensitive reports
Serious safety events often begin with incomplete information. A final diagnosis may be unavailable, causality may remain uncertain, and the participant's outcome may still be evolving.
That does not necessarily justify delaying required notification. The appropriate approach is generally to make the initial report according to the applicable timeframe with the information available, then provide follow-up information as required.
The relevant pathway depends on what kind of event is being reported and to whom. Sponsor SAE reporting, IRB or REC unanticipated-problem reporting, institutional reporting, and regulatory safety reporting can have different criteria and deadlines.
Classify the incident without collapsing all safety concepts together
A serious incident may require several separate assessments.
| Assessment |
Question |
| Adverse event |
Does the occurrence meet the applicable AE definition? |
| Seriousness |
Does it meet specified serious adverse event criteria? |
| Relatedness |
Could research participation or a research procedure have caused or contributed to it? |
| Expectedness |
Is the event consistent with what was already known and described? |
| Unanticipated problem |
Is it unexpected, related or possibly related, and evidence of greater risk than previously recognized under the applicable framework? |
| Protocol deviation or noncompliance |
Did failure to follow the approved protocol or applicable requirements contribute to the incident? |
| Study-level safety signal |
Does the incident alter what is known about risks for other participants? |
A single incident can fall into several of these categories, but the categories are not synonyms.
Investigate what happened without prematurely deciding who is to blame
A useful investigation asks how the incident occurred and what allowed it to occur. Beginning with “Who made the mistake?” can prematurely narrow the analysis.
A medication error, for example, might involve incorrect preparation by one person. It might also involve ambiguous labeling, inadequate separation of products, confusing protocol instructions, weak verification procedures, poor training, or several contributing factors.
The purpose of a safety investigation is not to avoid individual accountability when it is warranted. It is to avoid mistaking the most visible human action for the entire causal explanation.
Look for root and contributing causes
Depending on the incident, the investigation may examine protocol design, staff training, workload, delegation, equipment, environmental conditions, eligibility decisions, communication, data or safety monitoring, investigational product handling, consent procedures, and whether existing controls functioned as intended.
Useful questions include: What should have prevented this? Did that safeguard exist? Was it followed? Could a reasonable person have misunderstood the procedure? Has anything similar happened before? Could it happen again?
The resulting corrective action should correspond to the cause identified. “Remind staff to be careful” is not a particularly persuasive solution to a recurring system problem.
Implement urgent protection before the investigation is complete when necessary
Some incidents expose an immediate hazard that cannot reasonably wait for the entire investigation and amendment process.
OHRP recognizes that investigators may initiate protocol changes before IRB approval when necessary to eliminate apparent immediate hazards to participants. Other proposed changes generally require review and approval before implementation.
An urgent protective action might involve stopping a hazardous procedure, quarantining a potentially problematic product, temporarily halting enrollment, increasing monitoring, or taking another measure appropriate to the identified hazard.
Watch Out
Emergency protective action and permanent protocol amendment are not the same thing. Act when necessary to remove an apparent immediate hazard, but document what was done, notify the appropriate oversight bodies, and obtain the required review for continuing changes.
Ask whether the incident changes risk for other participants
The affected participant is only one part of the safety analysis. A serious incident can reveal that people currently enrolled or not yet enrolled face a risk that was previously underestimated.
OHRP states that unanticipated problems generally warrant consideration of substantive protocol or consent changes or other corrective actions. Examples include changing eligibility criteria, adding monitoring, suspending enrollment, suspending procedures in current participants, revising consent documents, and informing already enrolled participants about newly recognized risks.
The threshold for changing the entire study after a safety problem depends on what the incident reveals rather than on seriousness alone.
Current participants may need new information
If the incident produces significant new findings relevant to willingness to continue, participants may need to be informed.
OHRP describes informed consent as an ongoing process and notes that significant new findings developed during research that may relate to willingness to continue should, when appropriate, be provided to participants. Changes in protocol design or risk can require supplementation or repetition of consent.
The communication should be understandable. Participants need to know what has changed, how it affects them, and what choices are available, not merely that “an amendment has been approved.”
Pausing a study can protect participants while uncertainty is resolved
A serious incident does not always require permanent termination. Sometimes the most proportionate response is a temporary pause.
Enrollment may stop while the cause is investigated. A particular procedure may stop while routine safety follow-up continues. Current participants may need continued intervention or monitoring because abruptly stopping would itself create risk.
OHRP's current IRB-procedure guidance specifically recognizes that suspension or termination decisions should consider participants already enrolled and, where applicable, orderly termination or transfer.
This is why “stop the study” should not be treated as a single operational action. Researchers and oversight bodies need to determine which activities should stop, which should continue for participant protection, and what participants should be told.
Corrective action should prevent recurrence, not merely document regret
OHRP recommends that reports of unanticipated problems describe changes to the protocol or other corrective actions taken or proposed. Current IRB written-procedure guidance similarly identifies documentation of steps taken to prevent recurrence as relevant to review.
Correction
Addresses the immediate problem, such as replacing mislabeled materials or correcting an erroneous instruction.
Corrective and preventive action
Addresses why the problem occurred and changes the system so the same or similar failure is less likely to recur.
Depending on the cause, meaningful action might include redesigning a process, adding independent verification, changing eligibility, modifying monitoring, retraining staff with demonstrated competency, revising materials, improving supervision, or changing the protocol itself.
Responsibility and blame should not be confused
A serious incident may involve an investigator, research staff member, sponsor, institution, healthcare team, equipment provider, or several parties. Their responsibilities can differ.
The question of who is responsible for research-related injury or harm includes participant care, reporting, financial arrangements, oversight, and potentially legal issues. A safety investigation should establish facts and contributing causes rather than casually assigning legal responsibility.
Close the loop
A serious incident should eventually produce a documented answer to several questions: What happened? What happened to the participant? What caused or contributed to the incident? Who was notified? What was changed? Were current participants informed when necessary? Did the corrective actions work?
Without that final step, a research team can become very good at filing incident reports while remaining surprisingly talented at recreating the same incident.
