TL;DR
This paper introduces "CLASHED," a novel risk prediction model for myocardial injury after non-elective surgery in elderly patients. CLASHED significantly outperforms the established Revised Cardiac Risk Index (RCRI) in identifying patients at risk, offering a more accurate tool for pre-operative assessment in this vulnerable population.
What Problem Does This Paper Address?
As global life expectancy increases, so does the number of elderly individuals undergoing surgery. These patients are at a higher risk of complications, with myocardial injury being a significant contributor to peri-operative mortality and prolonged hospital stays. Current pre-operative risk assessment tools, like the Revised Cardiac Risk Index (RCRI), are not always effective, particularly for elderly patients undergoing urgent, non-elective procedures. This gap leaves a critical need for a more precise and tailored risk prediction model to identify and manage patients at risk of myocardial injury after non-cardiac surgery (MINS).
Methods
The researchers conducted a retrospective study on 370 elderly patients (over 65) who underwent non-elective surgery. They collected demographic data, medical history, and performed physical examinations, electrocardiography, and echocardiography. Crucially, they analyzed laboratory results, including cardiac biomarkers.
To develop their new model, they employed multivariable analysis to identify independent predictors of MINS. This led to the creation of the "CLASHED" model, which incorporates parameters such as creatinine, lymphocyte count, aortic regurgitation, stroke history, hemoglobin, ejection fraction, and D-dimer. The model's predictive performance was then rigorously evaluated using Receiver Operating Characteristic (ROC) curve analysis and compared against the RCRI. Internal validation was performed using bootstrapping to ensure the model's reliability.
Datasets
The study utilized a dataset of 370 elderly patients who underwent non-elective surgery at a tertiary hospital. This dataset was collected retrospectively from archival records. No new benchmark datasets were introduced.
Key Insights and Ideas
- Elderly patients are a high-risk group: The study reinforces that elderly patients undergoing non-elective surgery face significantly higher risks of adverse outcomes, including MINS.
- RCRI limitations: The established RCRI, while useful, shows limitations in accurately predicting MINS in this specific, vulnerable population, highlighting the need for a more specialized tool.
- Biomarkers and clinical factors are key: The CLASHED model's success stems from integrating a combination of readily available clinical parameters and biomarkers, demonstrating their collective power in risk stratification.
- D-dimer's importance: The inclusion of D-dimer as a significant predictor is noteworthy, suggesting its role in assessing thrombotic risk and its association with myocardial injury in this context.
- Aortic regurgitation matters: Moderate-to-severe aortic regurgitation emerged as a strong predictor, underscoring the importance of cardiac structural assessment.
Key Results
- Prevalence of MINS: 13% (48 out of 370 patients) developed MINS.
- CLASHED Model Performance: The CLASHED model achieved an Area Under the Curve (AUC) of 0.788.
- RCRI Performance: The RCRI achieved an AUC of 0.611.
- Superiority of CLASHED: The CLASHED model was found to be significantly superior to the RCRI in predicting MINS (p < 0.05).
- Independent Predictors: Key independent predictors of MINS identified were creatinine, lymphocyte count, moderate-to-severe aortic regurgitation, stroke history, hemoglobin, LV ejection fraction, and D-dimer.
Limitations and Future Work
The authors acknowledge several limitations:
- MINS Definition Variability: There isn't a universally accepted definition of MINS for non-elective patients, which can affect comparability across studies.
- Troponin Measurement Timing: Troponin was measured within the first 3 days post-surgery or only upon clinical suspicion, potentially missing asymptomatic MINS cases.
- Study Cohort Heterogeneity: Combining three different types of surgery (hip fracture, neurosurgical, and acute abdominal) might have introduced minor limitations.
- Single-Center Study: The research was conducted at a single geographical location, limiting generalizability.
Future Work: The authors suggest that the CLASHED model should be validated in larger, multi-center studies to confirm its effectiveness and widespread applicability. They also emphasize that this model, using easily obtainable parameters, is well-suited for elderly patients undergoing major non-elective surgery.
Summary updated