126 Chapter 6 The season in which cardiac surgery was performed had no significant effect on the P/F ratio on any of the days (Table 3). Variables Estimate (95% CI) P Value Baseline season 1 ILI season, P/F day 1 −2.7 (−9.5 to 4.0) .43 ILI season, P/F day 3 15.8 (−8.2 to 39.8) .20 ILI season, P/F day 7 −10.2 (−74.8 to 54.4) .75 Table 3. Multivariable linear regression analysis. Multivariable linear regression analysis with PaO2 to fraction of inspired oxygen (P/F) ratio on three different days as outcome variables; and age, sex, body mass index, Creatinine, and Ejection Fraction II risk score, type of surgery, Acute Physiology and Chronic Health Evaluation IV predicted mortality, year of surgery, influenzalike illness (ILI) season, Charlson comorbidity index, intensive care unit occupancy rate, and hospital type (academic/nonacademic) as predictor variables The differences in duration of mechanical ventilation between groups for different types of surgery can be seen in Supplemental Table 4. When adding a Bonferroni correction (α = 0.05, m = 12), setting the significance threshold at 0.0042, there were no differences in hours of mechanical ventilation for a specific type of surgery. The median Age, Creatinine, and Ejection Fraction II risk score and APACHE IV predicted mortality was higher among patients who underwent surgery in an academic hospital compared with patients who underwent surgery in nonacademic hospitals (Supplemental Table 5). A sensitivity analysis excluding the APACHE IV score as a predictor variable in the regression analyses was performed and showed similar results (Supplementary Figure 1, Supplemental Table 6). Discussion In a large Dutch ICU database, we found that patients who underwent elective cardiac surgery in the ILI seasons were at increased risk of in-hospital mortality compared with patients who underwent cardiac surgery during the baseline seasons. However, we found neither an increase in the duration of mechanical ventilation in hours nor a decrease in the P/F ratio at day 1, 3, or 7 in the ILI seasons compared with baseline seasons. Earlier research investigating the relationship between viral respiratory disease and outcomes after cardiac surgery also found seasonal differences in outcomes. A study looking at outcomes from more than 16,000 cardiac surgery patients found higher odds of in-hospital mortality for patients who underwent surgery in the winter season compared with other seasons11. One retrospective study in adults, evaluating a group of patients who underwent cardiac surgery between 2009 and 2011, found a higher prevalence of ARDS and longer duration of IMV in influenza season compared with a season with low prevalence of ILI5. Although the ILI season and winter season often overlap, they are not the same. Meteorologic seasonal variety in postoperative
RkJQdWJsaXNoZXIy MTk4NDMw