Failure mode and effects analysis: an empirical comparison of failure mode scoring procedures

To empirically compare 2 different commonly used failure mode and effects analysis (FMEA) scoring procedures with respect to their resultant failure mode scores and prioritization: a mathematical procedure, where scores are assigned independently by FMEA team members and averaged, and a consensus procedure, where scores are agreed on by the FMEA team via discussion.

A multidisciplinary team undertook a Healthcare FMEA of chemotherapy administration. This included mapping the chemotherapy process, identifying and scoring failure modes (potential errors) for each process step, and generating remedial strategies to counteract them. Failure modes were scored using both an independent mathematical procedure and a team consensus procedure.

Almost three-fifths of the 30 failure modes generated were scored differently by the 2 procedures, and for just more than one-third of cases, the score discrepancy was substantial. Using the Healthcare FMEA prioritization cutoff score, almost twice as many failure modes were prioritized by the consensus procedure than by the mathematical procedure.

This is the first study to empirically demonstrate that different FMEA scoring procedures can score and prioritize failure modes differently. It found considerable variability in individual team members' opinions on scores, which highlights the subjective and qualitative nature of failure mode scoring. A consensus scoring procedure may be most appropriate for FMEA as it allows variability in individuals' scores and rationales to become apparent and to be discussed and resolved by the team. It may also yield team learning and communication benefits unlikely to result from a mathematical procedure.