| Resumo: |
Artigo: (Risk score for prolonged mechanical ventilation in patients undergoing coronary artery bypass grafting) Objective: To construct a risk score model for prolonged mechanical ventilation (PMV) in patients undergoing coronary after bypass grafting (CABG). Methods: An observational, retrospective and historical cohort study, with 4.165 patients submitted to coronary artery bypass grafting, between January 1996 and December 2016. The duration of ≥12 hours on mechanical ventilation defined as prolonged. The multiple logistic regression was used to evaluate the predictor variables. The modeling database, divided in 2/3 of the sample (2.746 patients) generated the preliminary model and its performance was tested in the validation database, divided in 1/3 of the sample (1.419 patients). The final risk score was validated in the total database and the accuracy of the model was tested by performance statistics. Results: The incidence of PMV was 18.8% (783). The variables associated with prolonged PMV were: age ≥65 years (OR=1.91; [CI95%=1.62-2.24]; p<0.001); urgent/emergency surgery (OR=2.79; [CI95%=2.09-3.73]; p<0,001); obesity (OR=1.49; [CI95%= 1.21-1.84]; p<0.001); chronic renal failure (OR=1.98; [CI95%=1.61-2.44]; p<0.001); chronic obstructive pulmonary disease (OR=1.43; [CI95%=1.16-1.76];p<0,001) and extracorporeal circulation time ≥120 minutes (OR=1.75; [IC95%=1.42-2.16]; p<0.001). The area under de the ROC curve was 0.66 (IC95%=0.64-0.68), the chi-square was 3.38 and the correlation coefficient of r=0.99 (p<0.000). Conclusion: The preoperative predictor variables (age ≥65 years, urgent/emergency surgery, obesity, chronic renal failure and chronic obstructive pulmonary disease) and perioperative (extracorporeal circulation time ≥120 minutes) were associated with PMV and allowed the constructed of risk score classified in low, medium, high and very high. |
|---|
