thebench.bio #1 Ecosystem for the Bio-Tech Industry
All CAPA questions

Question 12 of 20 Intermediate Troubleshooting

How would you investigate a deviation attributed to human error?

Short answer

Treat human error as the start of the investigation, not the end. Classify it as a slip, lapse, mistake or violation, then look for the conditions that made it likely, such as design, workload, procedure and environment, because EU GMP expects a human error conclusion to be justified.

Why

EU GMP Chapter 1 says that where human error is suspected or identified, the conclusion should be justified after making sure process, procedural or system errors have not been overlooked. That turns "operator error" into a hypothesis that needs evidence. The UK Health and Safety Executive gives a useful structure.

Errors are unintended, while violations are deliberate departures from a rule. Slips and lapses happen in familiar tasks, like pressing the wrong button or forgetting a step, and cannot be eliminated by training, but better design can reduce them. Mistakes are wrong decisions made in good faith, often in new situations or where training was poor, and good procedures and training help.

Violations are usually well meant shortcuts where equipment or tasks are poorly designed, and fixing them needs understanding why the rule was bypassed. HSE also lists factors that make errors more likely, such as poor design, distraction, time pressure, workload, competence, morale, noise and communication. A good investigation interviews the person without blame, observes the task, checks whether others make the same error, and compares the procedure as written with the task as done.

The type of error then points to the action. Retraining someone who slipped during a routine step will not prevent the next slip.

How to work it through

  1. Secure the facts with a blame-free interview soon after the event.
  2. Observe the task as it is really done and compare it with the written procedure.
  3. Classify the failure as a slip, lapse, mistake or violation.
  4. Identify performance influencing factors such as layout, workload, labeling, shift timing and interruptions.
  5. Check whether other people have made the same error.
  6. Choose actions matched to the error type and record why system causes were or were not found.

An example

An operator adds 2.5 kg of an excipient instead of 25.0 kg. The interview shows he read the scale correctly but entered the target from a batch record where the decimal sat near a fold. Observation finds the weighing station has no link to the batch record and the line was short-staffed that night.

A search finds two similar near misses on other shifts. This is a slip encouraged by document design and workload, so actions are an electronic weigh-and-dispense check and a redesigned form, not retraining alone.

Where the answer stops holding

Interviews depend on trust, and a punitive culture will hide the real story. Classification is sometimes uncertain, especially between mistakes and violations. Some errors remain unexplained, and the honest conclusion then states the most likely causes rather than inventing certainty.

A strong answer covers

  • Human error as a hypothesis needing justification
  • Slip, lapse, mistake or violation
  • Performance influencing factors
  • Observation and comparison with the written procedure
  • Action matched to error type

Common mistakes

  • Closing the investigation with operator error and retraining.
  • Interviewing in a way that feels like discipline, so people stop reporting.
  • Treating a well-meant workaround as misconduct without asking why it exists.