ASSOCIATION OF SYSTEMIC INFLAMMATORY REACTION AFTER CARDIAC SURGERY WITH INCREASED 30-DAY MORTALITY: A MACHINE LEARNING APPROACH FOR RISK PREDICTION
Bibliographic record
Abstract
Background and Aim: Systemic inflammation (SIRS) characterizes the postoperative course of cardiac surgery, worsening patient outcomes in its most relevant form. We investigated the impact of SIRS on 30-day mortality and developed a machine-learning model for SIRS prediction. Methods: A retrospective evaluation of patients who underwent cardiac surgery from 2016- 2020 in a single hospital was performed. SIRS was assessed 12 hours post-surgery using the ACCP/SCCM criteria. A multivariate logistic model identified SIRS predictors. SIRS-positive patients were matched 1:1 with SIRS-negative ones. The effect of SIRS on mortality was investigated within the propensity-matched cohort. Baseline Risk (BRM) and Procedure- adjusted Risk (PARM) Random Forest Models were trained, optimized, and calibrated. Models’ performance was assessed via cross-validation (CV), using the AUC as a benchmark metric. Results: A total of 1908 patients were included. SIRS incidence was 28.7%, and 30-day mortality was 4.6%. Propensity scoring matched 483 patient pairs. SIRS was significantly associated with mortality (OR 2.77; 95%CI 1.40-5.47, p=0.003). SIRS mediated a proportion of its intraoperative predictors’ impact on mortality: 24.3% for anemia, 9.9% for vasopressors, 4.0% for hyperlactatemia (p<0.001). The BRM produced an AUC 0.77±0.04 in the 5-fold CV and an AUC 0.73 (95%CI 0.70-0.85) on the test set, whereas the PRM culminated in an AUC 0.81±0.02 in the CV and 0.82 (95%CI 0.76-0.85) on the test-set (DeLong p<0.001). Conclusion: Clinically-defined SIRS after cardiac surgery is associated with 30-day mortality. Machine learning allows to predict SIRS effectively, and paves the way for research exploring targeted interventions to prevent/mitigate its negative effects.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.005 | 0.007 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".