133 Cardiac surgery in Influenza-Like-Illness season 6 Baseline season (n=13,440) ILI season (N=12,994) P value Aortic valve surgery 5.6 (3.5 to 9.6) 5.7 (3.6 to 9.3) 0.87 CABG 5.1 (3.4 to 8.3) 5.1 (3.4 to 8.3) 0.69 CABG + 1 valve 7.4 (4.5 to 13.6) 7.4 (4.5 to 14.1) 0.39 CABG + 2 valves 11.1 (6.1 to 22.4) 17.0 (6.3 to 28.3) 0.12 Cardiovascular surgery, other 6.3 (3.6 to 18.8) 5.9 (3.1 to 17.8) 0.50 Congenital defect repair 4.7 (3.0 to 9.2) 5.2 (3.2 to 9.3) 0.74 Mitral valve surgery 6.2 (3.8 to 10.6) 6.4 (4.0 to 10.8) 0.26 Pericardiectomy 6.1 (3.6 to 11.8) 5.9 (3.4 to 12.5) 0.86 Pulmonary valve surgery 4.0 (2.8 to 6.4) 5.6 (2.7 to 8.4) 0.24 Tricuspid valve surgery 5.8 (3.4 to 13.9) 9.0 (6.1 to 19.8) 0.04 Tumor removal, intracardiac 5.5 (3.1 to 9.1) 4.7 (2.9 to 11.4) 0.85 Two valves 8.3 (5.0 to 17.1) 7.9 (4.8 to 17.0) 0.52 Supplementary Table 4. Median (IQR) duration of mechanical ventilation (hours) Academic Non-academic P value ACEF II risk score 2.1 (1.5 to 2.7) 1.9 (1.3 to 2.6) <0.0001 APACHE IV predicted mortality 1.4 (0.4 to 3.4) 1.2 (0.4 to 3.0) <0.0001 Supplementary Table 5. Median (IQR) ACEF II risk score and APACHE IV predicted mortality split by hospital type Sensitivity analyses For the Cox regression analysis, using duration of IMV as the outcome, results were similar, with only the 2014-2015 season no longer significant. In 2015-2016 and 2017-2018, IMV was slightly shorter in the ILI season compared to baseline season (HR 1.21; 95% CI 1.11-1.33 and HR 1.24; 95% CI 1.15-1.34, respectively). No significant difference was found in the other years: 2013-2014 (HR 1.11; 95% CI 0.98-1.26), 2014-2015 (HR 1.13, 95% CI 0.96-1.34), 2016-2017 (HR 0.99, 95% CI 0.91-1.07) and 2018-2019 (HR 1.01; 95% CI 0.93-1.10) (Supplementary Figure 1). For in-hospital mortality, the OR was 1.67 (95% CI 1.15-2.43, P value 0.008), a result very similar to the analysis that included the APACHE IV predicted mortality as a predictor. Finally, the results from the P/F ratio analysis including and excluding the APACHE IV predicted mortality as a predictor were almost identical (Supplementary table 6).
RkJQdWJsaXNoZXIy MTk4NDMw