Neutrosophic Statistical Regression Models for Predicting the Incidence of Nosocomial Infections in Post-Trauma Patients

Abstract

The rationale for predicting nosocomial infections in post-trauma patients is based upon the ability to assess all relevant and subjectivity factors effectively for each individual. This research is currently relevant to public health because nosocomial infections are one of the most stubborn factors plaguing patients and health facilities alike. Therefore, it is timely and appropriate to be able to predict accurate and generalizable findings—yet the current literature does not utilize a predictive approach that considers the uncertainty and imprecision when clinicians encounter such clinical data. Therefore, this dissertation aims to fill that gap with a statistical-neutrosophic regression model which inherently adjusts for the vagueness/uncertainty of such clinical data. Thus, through the application of novel statistical-neutrosophic regression techniques on a blend of relevant factors stemming from clinical and demographic data relating to post-trauma patients, predictive output will reveal the potential for such nosocomial infections to occur. The results demonstrate the validity of statistical-neutrosophic models to acknowledge the complexities of clinician data to determine the best predictive outcomes. This dissertation adds theoretically to the body of knowledge regarding medical prediction opportunities using statistical-neutrosophic techniques and practically, as clinicians/managers can use this dissertation relative findings to predict potentially at-risk individuals sooner than later to better organize resource management. Ultimately, this addition to the literature enhances healthcare quality just a fraction more while simultaneously reducing the chance of having adverse hospital-induced infections.