Validation of the PreOperative Score to predict Post-Operative Mortality (POSPOM) in Dutch non-cardiac surgery...

Area of Science:

• Anesthesiology and Perioperative Medicine
• Surgical Outcomes Research
• Health Services Research

Background:

• Standardized risk assessment tools are crucial for identifying patients at elevated risk of postoperative complications and mortality.
• The PreOperative Score to predict Post-Operative Mortality (POSPOM) is a tool used for risk stratification in surgical patients.
• Validation of POSPOM in diverse surgical populations is essential for its clinical utility.

Purpose of the Study:

• To validate the POSPOM score for predicting in-hospital mortality in a large cohort of non-cardiac surgery patients.
• To assess the performance of POSPOM in predicting various postoperative complications.
• To evaluate the discriminatory and calibration performance of POSPOM.

Main Methods:

• Analysis of data from the control cohort of the TRACE study involving 2490 non-cardiac surgery patients.
• Post-hoc calculation of POSPOM scores and comparison with observed in-hospital mortality.
• Assessment of discrimination using receiver operating characteristic curves (C-statistics) and calibration using calibration plots.

Main Results:

• Observed in-hospital mortality was 0.5%, while POSPOM predicted 1.3%.
• POSPOM demonstrated strong discrimination for in-hospital mortality (C-statistic = 0.86).
• Prediction of postoperative complications showed poor to fair discrimination, varying by severity, and poor calibration with overestimation of mortality.

Conclusions:

• POSPOM exhibits strong discriminatory power for in-hospital mortality but suffers from poor calibration, overestimating risk.
• The score's performance in predicting postoperative complications is limited, though it improves with complication severity.
• Further refinement or alternative tools may be needed for accurate risk prediction in non-cardiac surgery.