Priority assessment of procuring medical equipment in Turkish hospitals using input-weighted fuzzy logic architecture

Abstract

Medical devices, which are very expensive in general, are frequently used in all stages of the healthcare. However, as similar to every electronic device, they have a particular lifetime. Over the time, devices may lose their functions and give continuous malfunctions with a need to renew. In addition, outdated technology treats the hospital budget by consuming more energy and requiring more maintenance (and repair) costs. Therefore, hospitals need to be re-equipped with recent high-technology devices by the time. However, several factors may challenge determining the priorities of the medical devices properly. Today, many decision support systems are available and very useful in various fields of clinics. Hence, a decision support system may also be helpful to assist prioritizing renewal requests of medical devices. In this study, an input-weighted fuzzy logic model was designed and implemented using the Matlab programming environment to help clinical engineers in prioritizing the clinicians’ renewal requests. The proposed fuzzy logic model accepts four main criteria of technical service features, financial features, hospital features, and device features by combining twenty sub-criteria from user inputs. 12 field experts obtained these features, feature weights to construct main criteria, fuzzy membership functions, rule base repository, and fuzzification/defuzzification methods of the proposed system. The performance of the developed model was evaluated with a total of 50 authority decisions given in 5 different hospitals. As a result, the proposed input-weighted fuzzy logic model and the field experts reached the same procurement priorities. By utilizing the analysis of variance (ANOVA) test, we found no statistically significant difference between experts’ decisions and simulation results. In addition, the experts reported the efficiency of the proposed system. By comparing the existing studies, we conclude that our approach provides a pioneer model and a Matlab application to determine the procurement priorities for medical devices. However, although the achieved results are impressive, we note that the model needs to be supported by other studies that cover more experienced participants before clinical common use. In the present form the proposed model and the application can give an idea to clinical engineers and managers who are responsible to the procurement processes in clinics. This study has a potential to start studies to develop national-wide procurement decision support systems, in our opinion.