# Key predictors of 30-day mortality in non-cardiac surgery: development, key risk factors, and calibration of a risk model

## TL;DR

This study identifies key risk factors for 30-day mortality in non-cardiac surgery patients and develops a predictive model to help improve patient care.

## Contribution

The study introduces a new nomogram model for predicting 30-day mortality in non-cardiac surgery patients.

## Key findings

- Preoperative ventilatory support, postoperative sepsis, coma, myocardial infarction, and cardiac arrest were significant predictors of 30-day mortality.
- The nomogram showed good calibration and strong discrimination for predicting mortality risk.
- The model can help identify high-risk patients early for targeted interventions.

## Abstract

Thirty-day mortality remains a critical indicator of perioperative safety and quality of care in patients undergoing non-cardiac surgery. Identifying predictive factors for 30-day mortality is critical for crisis mitigation and improving patient outcomes. This study aimed to identify key predictors of 30-day mortality in non-cardiac surgery patients and develop a predictive model to aid clinical decision-making.

This retrospective observational study analyzed data from patients aged ≥18 years who underwent non-cardiac surgery at a tertiary care hospital between January and August 2022. Data on demographics, comorbidities, intraoperative variables, and postoperative complications were collected. Multivariate logistic regression identified independent predictors of 30-day mortality, and a nomogram was constructed to facilitate individualized risk assessment.

Among 7,528 patients, 76 (1.0%) died within 30 days. Independent predictors of mortality included preoperative ventilatory support (odds ratio [OR] 8.76; 95% confidence interval [CI]: 3.58–21.42; p < 0.001), postoperative sepsis (OR 5.86; 95% CI: 2.23–15.37; p < 0.001), coma (OR 18.56; 95% CI: 6.81–50.55; p < 0.001), myocardial infarction (OR 26.35; 95% CI: 4.46–155.63; p < 0.001), and cardiac arrest (OR 57.9; 95% CI: 18.22–184.05; p < 0.001). The nomogram demonstrated good calibration at lower risk levels, with slight overestimation at higher probabilities. The area under the receiver operating characteristic curve indicated excellent discrimination.

The developed nomogram serves as a useful tool for perioperative risk stratification in non-cardiac surgery. Incorporating both preoperative and postoperative factors, aids in early identification of high-risk patients and supports targeted clinical interventions. External validation and refinement of the model are recommended for broader applicability.