Using ML Models to Help MedTech Companies Predict Recalls Based on Critical Software Failures, Assess Software Change Control, and Forecast Software-related Recall Costs.
Open Access DepositedThis research is primarily looking into medical device recalls due to software-related issues. Modern medical devices dependent on software heavily for precision diagnostics, patient monitoring, and automation. There is a significant trend or variability in medical device recalls due to software-related issues. Failure to predict software-related recalls can result in catastrophic losses to manufactures and also to patient safety. These recalls could potentially result in subsequent financial implications. And any such financial implications will amplify overall healthcare expenditures. The predictability of software-related recalls is dependent on understanding of the key features such as Device Class and Medical Specialty Panels in addition to Recall information that correlates with the target features is highly important. Manufacturers can use such knowledge to better optimize any further consequences. There are several consequences such as downtime, reputation, and financial implications. The existing literature focuses on predictability of medical device recalls but fails to address financial implications such as cost impacts. Most of the current work is done on predictive maintenance and understanding the different categories that lead to medical device recalls. Also, the findings of most of the existing research is limited to the temporal duration and dataset. This research is the only study that leveraged ML modeling techniques beyond the software-recall prediction, and subsequently focusing on financial impacts such as cost impacts primarily with direct costs based on the number of quantities impacted by predicting the unit price of the medical device. Also, the use of larger representative and up-to-date data, this study resulted in better testing of all the four Hypotheses.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.