Hemodialysis Outcomes Score In Trauma (HOST): A novel and easy model for predicting death in patients receiving pre-injury hemodialysis

• Published in: The American journal of surgery 2025-04, Vol.242, p.116176, Article 116176
• Main Authors: Stettler, Gregory R., Carroll, Hannah L., Roeber, Heidi L., Avery, Martin D., Arafeh, Mohamed-Omar S., Russell, Gregory B., Hoth, J Jason, Mowery, Nathan T., Nunn, Andrew M.
• Format: Article
• Language: English
• Subjects: Hemodialysis; Kidney failure; Trauma
• ISSN: 0002-9610; 1879-1883; 1879-1883
• DOI: 10.1016/j.amjsurg.2024.116176

We sought to construct a simple clinical mortality prediction model in trauma patients that required pre-injury hemodialysis: Hemodialysis Outcomes Score in Trauma (HOST). Trauma patients on pre-injury hemodialysis admitted between July 2013 to December 2021 were reviewed. Univariate and multivariable analysis was used to determine independent predictors of mortality and construct the HOST score. There were 663 patients identified as receiving pre-injury hemodialysis. Most patients were male (54.6 ​%), suffered a blunt mechanism (97.4 ​%), and were severely injured (median ISS 21). Mortality at 28-days for patients receiving pre-injury hemodialysis was 6.8 ​% compared to 4.8 ​% in injured patients that did not require pre-injury hemodialysis (p ​= ​0.03). Multivariate logistic regression identified GCS, HR, and hematocrit to be associated with 28-day mortality. HOST may serve as a tool with readily accessible input variables that is able to predict 28-day mortality. •Injured patients on pre-injury hemodialysis are more likely to die than patients that do not.•Hemodialysis Outcomes Score in Trauma (HOST) includes readily accessible variables of HR, GCS, and hematocrit.•HOST is a good predictor of 28-day mortality.