Invited Commentary

Abstract

Das and colleagues [1Das S. Rubio A. Simsic J.M. et al.Bloodstream infections increased after delayed sternal closure: cause or coincidence.Ann Thorac Surg. 2011; 91: 793-798Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar] have presented a retrospective review evaluating the relationship between delayed sternal closure and the incidence of postoperative blood stream infections in neonates undergoing a stage I Norwood. Delayed sternal closure is a management strategy that is widely used for neonatal operations, such that it might be considered the standard of care. However, it is evident from this study and others that there are attendant risks to delayed sternal closure, including sepsis, wound infection, mediastinitis, and potentially, death. In view of the intense scrutiny that hospital-acquired infections have received in recent years, it is important to accurately identify the risk factors associated with postoperative infections after neonatal operations and determine whether they are “cause or coincidence.” This study reviewed data from 2004 to 2008, during which approximately two-thirds of the patients undergoing stage I Norwood had delayed sternal closure and one-third underwent primary closure. This was not a “randomized” process, but rather must have been on the basis of surgeon preference or clinical instinct. The most obvious question is whether this nonrandomized enrollment into two groups could have biased the results of the study. If so, the sicker patients would almost certainly be overrepresented in the group with delayed sternal closure. This could provide an alternative explanation for why 13 of the 14 deaths occurred in this group. It is interesting to note that the mortality rate was not different in patients with (3 of 24) or without (11 of 86) blood stream infection. Thus, the intuitive postulate that delayed sternal closure predisposes to blood stream infection, which then results in poor outcomes, may not be as linear as one would initially conceive. The incidence of a positive blood culture in this study was 22% (24 of 110). One can make the case that this might underrepresent or potentially overrepresent the true incidence of sepsis. If this study had been performed prospectively, there would have been a protocol with prescribed times to acquire blood cultures in addition to those drawn for clinical reasons. The protocol-mandated blood cultures could have conceivably revealed additional instances of subclinical sepsis. Conversely, the data reveal that most of the positive blood cultures grew coagulase-negative Staphylococcus, which could be central catheter colonization or a contaminant. The authors do not make the distinction between blood cultures drawn by site (venipuncture vs central catheter) nor concordance of cultures (central catheter colonization vs sepsis). The current Centers for Disease Control and Prevention guidelines would exclude a single positive blood culture for coagulase negative Staphylococcus, and thus there is the possibility that this study could have overestimated the number of patients who truly had sepsis. Providing support for this speculation is that no correlation was found between positive blood cultures and death. A number of variables may contribute or predispose to the development of a blood stream infection beyond the presence of an open sternum. These would include intraoperative factors such as the duration of the procedure, antibiotic strategy, number of blood transfusions, and blood glucose levels. Additional perioperative factors might include the number of central catheters and duration of access, duration on a ventilator, use of total parenteral nutrition, and the fastidiousness of health care providers in hand washing and other bedside infection control measures. These variables—and likely many others—could confound the association between delayed sternal closure and development of blood stream infections. Given the retrospective model of this study, the authors provided a limited multivariate analysis, which paradoxically demonstrated sternal closure was a risk factor, but the duration of open chest was not. It is evident that the best method to assess the importance of these multiple variables would be to study them prospectively. Finally, it would be interesting to know what lessons Das and colleagues learned by way of their study. Have they changed their practice? If so, have they been effective in reducing their incidence of blood stream infections? This updated information would be of interest to many other centers performing neonatal congenital heart operations because it might better answer the question of whether the relationship between delayed sternal closure and blood stream infections is “cause or coincidence.” Bloodstream Infections Increased After Delayed Sternal Closure: Cause or CoincidenceThe Annals of Thoracic SurgeryVol. 91Issue 3PreviewInfants who undergo major cardiac operations are at risk for developing bloodstream infections which contribute to the morbidity, mortality, and cost of treatment. Determining what factors are associated with this increased risk of infection may aid in prevention. We sought to evaluate the practice of delayed sternal closure after neonatal cardiac surgery to determine its role as a risk factor for postoperative bloodstream infection. Full-Text PDF