Prevalence and Predictors of Pressure Injuries in Patients with Spinal Cord Injuries using Clinical Data

Lindsay Stern, Atena Roshan Fekr

Confirmation: I have read and agree with the IEEE BHI 2025 conference submission's policy on behalf of myself and my co-authors.

Keywords: Pressure Injury, Spinal Cord Injury, Prediction, Risk Factors

TL;DR: Demographic and clinical data that describe and predict pressure injury development in spinal cord injury patients.

Abstract: Pressure injuries (PIs) are injuries to the skin and/or underlying tissue, typically along the bony prominences of the body, often caused by pressure and shear forces. People with spinal cord injuries (SCIs) have a 25%-85% lifetime risk of developing pressure injuries due to various comorbidities, including neurological and musculoskeletal challenges. Currently, risk assessment scales, such as the Braden Scale or Spinal Cord Injury Pressure Ulcer Scale are used to define the risk of developing PIs. The scales are often used to aid clinicians in developing a care plan for PI prevention. However, there is a lack of consensus on which scale is best to use, often requiring clinical best judgment and thus creating a subjective assessment. Studies have indicated that clinical data, such as levels of hemoglobin, lymphocytes, or creatine level, may offer a more objective assessment for defining PI risk. Albeit useful, these studies often offer a generalized conclusion which does not consider the unique case of patients with SCIs. Therefore, this study explores demographic and clinical profile of individuals with SCIs to identify predictors associated with the development of PIs. The demographic data highlights a decrease in mean age and BMI amongst patients with PIs, but an increase in patients’ length of stay in hospital compared to those without PIs. The clinical data highlights differences in multiple biomarkers, including creatinine, cholesterol, and hemoglobin levels. Therefore, these results highlight the unique demographic and clinical predictors that can be leveraged to build Artificial Intelligent models for early, objective prediction of PIs in patients with SCIs. In turn, this will aid clinicians in developing an appropriate care plan for patients and thus further reduce PIs development.

Track: 4. Clinical Informatics

Registration Id: GLN6VJVVXB6

Submission Number: 344