A dynamic approach to FMEA–RPN risk classification: comparing with the classical FMEA method

Background Failure mode and effects analysis (FMEA) uses risk priority numbers (RPNs) based on 3 parameters to classify risks. Risk analysis methods use thresholds to determine risk. These class intervals do not account for contextual and system dynamics in high-risk environments such as neonatal intensive care units (NICUs). This study proposes a dynamic classification model that prioritizes risk by integrating contribution percentages and percentile ranks through a decision matrix. Material and Methods This study identified 21 ergonomic failure modes through field observations at a hospital’s NICU, interviews with nurses, and a literature review, evaluated a dynamic classification model. The RPN for each failure mode was calculated, and the risks were classified using classical fixed-threshold FMEA. Then, each failure mode was classified using a dynamic model based on a decision matrix that balances risk intensity by its impact contribution percentage and relative importance percentile rank. Model robustness was assessed via a one-at-a-time sensitivity analysis (±1 perturbations in occurrence [O], severity [S], and detection [D]), and agreement was evaluated using weighted agreement measures. Results According to FMEA, risk factors such as “long shift hours” and “constant exposure to alarm sounds” were classified as medium risk. Yet, in the dynamic model, they were classified as very high risk. The model’s sensitivity was tested by measuring the effect of a change in the FMEA input ratings (O/S/D) on the results. The model was determined to provide high agreement with practitioners’ classifications, with a 93.7% weighted agreement across 63 comparisons between nurse assessments and model results. Conclusions The proposed dynamic framework provides context-sensitive prioritization by integrating each failure mode’s contribution to total risk with its position in the unit-specific RPN distribution. When applied to NICU ergonomic failure modes, it produced a different prioritization than classical fixed-threshold FMEA and offers a transparent, reproducible basis for risk-focused improvement planning in high-risk care settings. Med Pr Work Health Saf. 2026;77(2):89–99