Reliability Analysis Techniques Applied to Highly Complex Medical Equipment Maintenance

EasyChair Preprint 10776

Abstract

The objective of this study was to apply reliability concepts in order to define adequate maintenance strategies for a group of highly complex equipment (Computerized Tomography) in the state of Rio de Janeiro. Methods: Equipment were located in 43 health units (private and public) Brazil. Failures claimed by customers and those observed by maintenance operators were identified. These data were collected considering all CT scanners and for each of the four brands present in the database. A FMEA was performed in order to identify failure risks and priorities thus allowing to rank and prioritize the failures detected by the analysis. Each brand of equipment was thus classified according to its risk criticality (high-medium-low); and for each criticality, the most appropriate maintenance strategy (mitigation action) was defined. Additionally, with the observed times between failures, the maintenance indicators MTTF (Mean Time to Failure), MTTR (Mean Time to Repair) and MTBF (Mean Time Between Failure) were calculated. Results: About 200 repairment requests from 2017 and 2018 had their information digitized. The most recurrent complained failure was “Failure of System Initialization, followed by “Generator failure” and “table malfunction”. Two of these could be classified as needing “predictive maintenance”, while the third (system initialization failure) was assigned “corrective maintenance). Conclusion: Reliability methods allowed for the definition of optimized maintenance strategies, with a resulting reduction in maintenance costs and an increase in equipment availability to patients.

Keyphrases: CT Scanner, Clinical Engineering, FMAE, Maintenance, Reliability

@booklet{EasyChair:10776,
  author    = {Marina Rezende and Rogério Santos and Fernando Coelli and Renan Almeida},
  title     = {Reliability Analysis Techniques Applied to Highly Complex Medical Equipment Maintenance},
  howpublished = {EasyChair Preprint 10776},
  year      = {EasyChair, 2023}}