ORAHS2025
Abstract Submission

223. Prediction of Diabetic Foot Ulcer Using Bayesian Networks

Invited abstract in session TC-1: Analytics and healthcare management, stream Sessions.

Tuesday, 13:30-15:00
Room: NTNU, Realfagbygget R5

Authors (first author is the speaker)

1. Malavika Krishnakumar
Health Sciences Research, Amrita Vishwa Vidyapeetham
2. Vivek Lakshmanan
Podiatry, Clinical Assistant Professor
3. Georg Gutjahr
Department of Health Science Research, Amrita Institute of Medical Sciences and Research Center, Kochi, Kerala, India
4. Ulla Hellstrand Tang
Department of Prosthetics and Orthotics, Institution of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden
5. Agnetha Folestad
Department of Orthopedics, CapioLundby Hospital, Göteborg, Sweden

Abstract

In European clinical practice, the D-Foot tool (Tang et al., 2017) has been introduced as a structured method for screening diabetic patients and assessing their risk of developing diabetic foot ulcers (DFU). However, its applicability outside Europe has not been explored. This prospective cohort study starts with investigating to what extent D-Foot is applicable in India. We observe that the D-Foot tool can be assessed with substantial intra- and inter-rater agreement, as measured by Cohen’s kappa, and high clinical usability, as measured by the System Usability Scale (SUS). Moreover, the study proposes a general approach using Bayesian network (BN) classifiers to predict patient outcomes such as new-onset DFU, Charcot foot, neuropathy, amputation, and mortality. The predictors include demographic data, clinical history, laboratory parameters, ankle-brachial index (ABI), vibration perception threshold tests (VPT), as well as the items from the D-Foot tool. The BN worked well on the Indian data, as measured by cross-validation, and it provides a promising approach for similar applications in diverse clinical and epidemiological environments.

Keywords

Status: accepted

Back to the list of papers