A nomogram model for predicting risk factors and the outcome of skin ulcer

Wound healing is a complex process, and numerous factors affect the healing of skin ulcers.

In order to identify the factors associated with wound healing, it is necessary to establish a visualized predictive model for evaluating the risk factors of patients with skin ulcers and to validate its effectiveness.

A retrospective observational study was conducted on 453 patients with skin ulcers admitted to the Dermatology ward of the Army Medical Center (Daping Hospital) in Chongqing, China, from January 2011 to July 2022. The nomogram was formulated according to a multivariate logistic regression analysis identifying seven potential predictors of prognosis, including age, area, pre-admission course, etiology, diabetes, medical treatment, and self-medication. This nomogram model was validated by bootstrap internal validation (1000 replicated samplings).

Logistic regression analysis showed that age, skin ulcer area, pre-admission course, etiology, comorbidity of diabetes, medical treatment, and self-medication were independently related to skin ulcer prognosis. These indicators were utilized to develop nomogram models. The predictive ability for skin ulcer prognosis was 0.814 based on the area under the curve values. The calibration curve showed a close match between the actual and predicted probabilities. Decision-making analysis demonstrated the clinical application value of this nomogram.

The prediction nomogram developed in this study exhibits good accuracy in predicting the risk factors of skin ulcers and provides an objective tool for clinical staff to assess and target the risk factors concerning the prognosis of skin ulcers.