A Gradient Risk Sampling Method For Performance Inspection and Preventive Maintenance

Clinical engineering departments traditionally determine if equipment should undergo safety and performance inspection (SPI) and preventive maintenance (PM) by using variations of a risk assessment method described by Fennigkoh and Smith. These methods use the sum of categorical risk measures to group equipment into two levels. Equipment with risk below a specified critical risk is excluded from the equipment management plan and no (or 0%) SPI and PM is required. Equipment with risk at or exceeding the critical risk is included in the plan and a considerable fraction (typically greater than 90%) must undergo SPI and PM. The 0% and >90% criteria, traditionally regarded as PM completion ratios, are reinterpreted to be sampling fractions. The traditional methods are shown to be a limiting case of a simplified logistic function describing SPI and PM value and sampling fraction as a function of equipment risk. This leads to a risk based multi-logistic sampling model that includes a perspective of organizational benefit and risk. A simplified sampling plan that includes traditionally excluded equipment is discussed and results of a pilot study are presented. The results suggest that an appropriate value and risk based sampling plan can form the basis for an organizational value based clinical equipment management plan. Suggestions for improving equipment management plans by expanding the model and improving the definition and measurement of risk are provided. Index under:acceptance sampling; clinical equipment management; logistic regression; management, clinical equipment; performance inspection; preventive maintenance; risk analysis