Evidence Hub

Development of an emergency department risk score for in-hospital mortality in elderly patients with urinary tract infection

Tzu-Heng Cheng, Wei Lu, Chen-Bin Chen, Chen-June Seak, Chieh-Ching Yen

Journal of the Formosan Medical Association = Taiwan yi zhi, 2025

2,391 participants · 5 citations

Abstract

BACKGROUND

Urinary tract infection (UTI) is among the most prevalent bacterial infections globally, particularly impacting elderly populations in emergency departments (EDs). This study aimed to develop and validate an ED-specific risk score predicting in-hospital mortality for elderly patients with UTI.

METHODS

Using data from the MIMIC-IV-ED database, we enrolled patients aged 65 or older with confirmed UTI admitted to wards or ICUs. Variables were selected through stepwise logistic regression using Bayesian Information Criterion in the derivation cohort (70 %), and dichotomized by optimal cut-offs identified by Youden's index. A point-based scoring system was developed from regression coefficients. Internal validation was performed on a separate validation cohort (30 %) assessing discrimination (AUC), calibration, and clinical utility using decision curve analysis.

RESULTS

Among 2391 patients, 160 (6.7 %) experienced in-hospital mortality. The EARL-UTI score included five variables: age >83 years, Emergency Severity Index 15.7 %, albumin 2.4 mmol L-1. Scores ≥4 predicted a mortality rate of 29.9 % compared to 4.5 % in lower-scoring groups, with sensitivity and specificity of 59.6 % and 93.1 %, respectively, and a negative predictive value of 95.5 %. Validation cohort performance remained consistent (AUC 0.73), and the model outperformed established risk scores.

CONCLUSIONS

The EARL-UTI score provides an easily implemented bedside tool for rapid identification of elderly…

How strong is the evidence?

Evidence Hub reads this as a cohort study, done in people. Its evidence level is 3 of 5 (Moderate · trial / cohort / secondary analysis). Trials without a large randomized design, cohorts and secondary analyses give moderate evidence. Reported size: 2,391 participants or records.

How the five levels of evidence work. The level is assigned automatically from the abstract and can be wrong, so check the paper.

Read the paper

Related studies

Go deeper in Evidence Hub

Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.

Open this study Sign up free