Incidence, predictors and validation of risk scores to predict postoperative mortality after noncardiac vascular surgery, a prospective cohort study. (January 2020)

• Record Type: Journal Article
• Authors: Reis, Pedro; Lopes, Ana Isabel; Leite, Diana; Moreira, João; Mendes, Leonor; Ferraz, Sofia; Amaral, Tânia; Abelha, Fernando
• Is Part Of: International journal of surgery. Volume 73(2020)
• Page Start: 89
• Page End: 93
• Publication Date: 2020-01
• Subjects: Vascular surgery -- Mortality -- Surgical Apgar -- Charlson Comorbidity Index (CCI) -- Vascular-Physiological and Operative Severity Score for the enUmeration of Mortality and Morbidity (V-POSSUM) -- Preoperative Score to Predict Postoperative Mortality (POSPOM); Surgery -- Periodicals; Surgical Procedures, Operative -- Periodicals; 617.005
• DOI: 10.1016/j.ijsu.2019.12.010
• Languages: English
• ISSNs: 1743-9191

Abstract

Abstract: Background: Noncardiac vascular surgery (VS) patients have comorbidities that increase the risk of death after surgery. Assessing that risk is important to allocate the necessary resources and improve quality of care. We aimed to evaluate the incidence and predictors of 30-day post-operative mortality (POM) after VS and compare the performance of existing risk scores. Materials and methods: Prospective cohort study including consecutive patients submitted to elective VS at a tertiary university hospital. We collected patients' demographics/perioperative data and calculated Surgical Apgar, age-adjusted Charlson Comorbidity Index (CCI), Vascular-Physiological and Operative Severity Score for the enUmeration of Mortality and Morbidity (V-POSSUM) and Preoperative Score to Predict Postoperative Mortality (POSPOM). We performed multivariate logistic regression to assess independent factors with Odds Ratio (OR) and 95% confidence interval (CI) calculation and Cox-regression for time-to-event analysis. We tested the predictive ability of the scores using the area under ROC curve (AUROC). Results: POM was 6.2% (n = 19/306), not different from expected by V-POSSUM (6.5%) or POSPOM (5.6%). Post-operative myocardial infarction (MI) and acute kidney injury (AKI) were associated with higher POM (OR 4.8, p = 0.011 and OR 5.4, p = 0.001, respectively). On multivariate analysis, Chronic kidney disease (CKD) (OR 4.0, p = 0.021), Age (OR 1.1, p = 0.002), Peripheral arterial disease (PAD) (OR 8.0, p = 0.006), intra-operative red blood cells (RBC) Transfusion (OR 1.9, p < 0.001) and Atrial fibrillation (OR 8.4, p = 0.002) were considered independent predictors of POM (CAPTA score). The AUROC of our model was 0.882, better V-POSSUM (0.858), POSPOM (0.784), CCI (0.732) or Surgical Apgar (0.649). Conclusion: Observed POM was similar to predicted by V-POSSUM or POSPOM. Age, PAD, CKD, atrial fibrillation and intraoperative RBC transfusion were independent risk factors for POM. Score V-POSSUM performed better than POSPOM, CCI or Surgical Apgar. Highlights: Vascular surgery may have significant perioperative mortality. We evaluated incidence, predictors and compared risk scores to predict POM. POM was 6.2% (n = 19/306), not different from expected by V-POSSUM or POSPOM. Age, PAD, CKD, atrial fibrillation and RBC transfusion were risk factors for POM. Score V-POSSUM performed better than POSPOM, CCI or Surgical Apgar.