Background: Certain cardiovascular diseases show diurnal variation, with a higher incidence of myocardial infarction in the morning. Conversely, aortic valve replacement surgery performed in the afternoon provided less myocardial injury and improved patient outcomes. We therefore examined a potential daytime varying effect on surgical site infection (SSI) and 1-year mortality in cardiac surgery.
Methods: Data from the prospective, validated Swiss national SSI surveillance system with a post-discharge follow-up rate >90% from adult patients undergoing cardiac surgery in 16 hospitals were analyzed from 2009 to 2014. Patients operated during nighttime and/or as emergency were excluded. The main exposure was time (morning/afternoon) of surgery. The primary outcome was SSI incidence, defined according to CDC criteria; a secondary outcome was 1-year mortality. We fitted generalized linear and additive models (GAM) to describe daytime varying effects predicting the outcome parameters and adjust for confounding variables.
Results: Of the 16841 surgeries included, 11850 (70%) started between 7am and 12 am while the remaining 4991 (30%) started between 12 am and 4pm. Baseline characteristics of morning versus afternoon surgeries are shown in Figure 1. The overall SSI (including graft excision sites) and 1-year mortality rates were 5.9% and 4.7%, respectively. After adjustment for confounders, afternoon surgery was not associated with lower SSI (OR 1.0, 95% CI 0.99-1.0, p=0.42), or 1-year mortality rate (OR 1.0, 95% CI 1.0-1.01, p=0.24) than morning surgery. A GAM did not detect a relevant daytime-varying effect on SSI (p=0.36, Figure 2), but an increase in 1-year mortality in function of daytime (p=0.02, Figure 3). An adjusted subgroup analysis confirmed increased mortality for incision between 9am and 4pm compared to 7-9am (OR 1.01, 95% CI 1-1.02, p=0.003).
Conclusion: This large study did not find a decreased SSI and/or mortality rate for afternoon surgeries. Therefore, the previous findings of reduced myocardial injury due to afternoon surgery cannot be generalized to these important clinical outcome parameters.
S. P. Kuster, None
N. Troillet, None
T. Carrel, None
F. Eckstein, None
A. F. Widmer, None