RESUMO
PURPOSE: Culture of extracted drains or epicardial pacing wires is an easy and noninvasive method for detecting mediastinitis after open-heart surgery, although studies on its sensitivity and specificity are limited. We, therefore, investigated the usefulness of this approach for diagnosing mediastinitis. METHODS: We retrospectively studied the culture results of drains and epicardial pacing wires extracted from 3308 patients. Prediction models of mediastinitis with and without culture results added to clinical risk factors identified by a logistic regression analysis were compared. RESULTS: The incidence of mediastinitis requiring surgery was 1.89% (n = 64). Staphylococcus was the causative bacterium in 64.0% of cases. The sensitivity, specificity, and positive and negative predictive values of positive culture results were 50.8%, 91.8%, 10.7%, and 99.0%, respectively. Methicillin-resistant Staphylococcus aureus had the highest positive predictive value (61.5%). A multivariate analysis identified preoperative hemodialysis (OR 5.40 [2.54-11.5], p < 0.01), long operative duration (p < 0.01), postoperative hemodialysis (OR 2.25 [1.01-4.98], p < 0.05), and positive culture result (OR 10.2 [5.88-17.7], p < 0.01) as independent risk factors. The addition of culture results to pre- and postoperative hemodialysis and a lengthy operative time improved the prediction of mediastinitis. CONCLUSIONS: A culture survey using extracted drains and epicardial pacing wires may provide useful information for diagnosing mediastinitis.
