Bundled approaches for surgical site infection prevention: the continuing quest to get to zero.

Abstract

Although awell-timed operation can enhance both duration andqualityof life, postoperative complications, including surgical site infections (SSIs), can offset some of the benefits of the operation. In addition to functional loss and mortality, SSIs are associated with increased length of stay and higher health care costs.1,2 SSIs are the most common reason for readmission after surgery,3 accounting for nearly 20% of unplanned readmissions.Among infectingorganisms,Staphylococcusaureusconsistently ranksas themostprevalentpathogen associatedwith SSIs, particularly after cardiac and orthopedicprocedures.4SaureusSSIs canbedevastating, especially when a prosthetic device is involved. Although the results vary, several prior investigations suggest that perioperative use of topical mupirocin along with chlorhexidine gluconate bathing amongpatients known tobe colonized with methicillin-susceptible S aureus (MSSA) or methicillin-resistant S aureus (MRSA) may reduce the risk of SSIs.5,6 The evidence for benefit ismost robust for cardiac and orthopedic procedures.7 Yet decolonization protocols are not consistently applied in US hospitals.8 In this issueof JAMA, Schweizerandcolleagues9 report the results of the Study to Optimally Prevent SSIs in Select Cardiac and Orthopedic Procedures (STOP-SSI), a multicenter, pragmatic study that examined the implementation of a bundled intervention of screening, decolonization, and targeted antimicrobial prophylaxis to prevent complex S aureus SSIs (deep incisional or organ space infections) among patients undergoing cardiac surgery andhip or knee arthroplasties. The specific interventions were simple and included screening and then decolonizing individuals who were positive for either MSSA orMRSA using intranasal mupirocin and chlorhexidine bathing, both for 5 days. In addition, the standardsurgicalprophylaxis regimenwasmodifiedto includevancomycin (along with the usual firstor second-generation β-lactam) for individuals known to be colonized with MRSA. The study took place in 20 hospitals across the United States and included 42 534 unique operations, consisting of 10 833 cardiac procedures and 31 701 hip or knee arthroplasties. Use of the bundlewas associatedwith adecrease in complex S aureus SSIs. Overall, 101 complex S aureus SSIs occurred after 28 218 operations during the preintervention period compared with 29 SSIs after 14 316 operations during the intervention; mean SSI rate per 10 000 operations decreased from36 for thepreinterventionperiod to 21 for the intervention period; rate ratio (RR), 0.58 (95% CI, 0.37-0.92). Although the absolute difference of 15 infections per 10 000 operations seemsmodest, eachcomplexSSIprevented is clinically meaningful. For the individual patient, development of a serious SSI after cardiac or orthopedic surgeryusually translates into months of parenteral antibiotics, additional surgical procedures, and extended inpatient and subacute care facility stays. The lengthy recovery can negate any benefit provided by the original operation. In a substantial portion, S aureus SSI will contribute to death.10,11 Given the considerable clinical and economic consequences of SSIs, some have suggested that the goal for the US health care system is zero tolerance of adverse events.12 The study by Schweizer et al has a number of notable strengths. For example, the inclusion of patients undergoing emergency or urgent operations, a population recognized as at high risk for SSI, improves the generalizability of the findings. In addition, the primary study outcome was limited to complex S aureus SSIs, eliminating much of the subjectivity of infection surveillance. Even though surveillance practices variedamongparticipatinghospitals, complicatedSaureusSSIs arenot clinically subtle andcanbe identifiedeasily byany surveillance system. The pragmatic nature of this investigation is reflected in reported bundle compliance; even after a 3-month phase-in period,bundleadherencewas83%,withonly39%implementing all components of the bundled intervention (fully adherent). The authors observed a “dose-response” association for bundle adherence; rates of complex SSIs decreased significantly among patients in the fully adherent group (RR, 0.26 [95% CI, 0.10-0.69]) but not in the partially adherent or nonadherent groups. Although a matter of speculation, improved adherence to thebundlemayproducebetter results in other settings,particularly if baseline infection rates arehigher than those reported in this study. Moving forward, efforts to promote andmaintain adherence topreventionprotocolswill remain important. Leveraging the potential of the electronic medical record through the use of automated reminders and checklists is one approach. Adherence rateswere particularly low in urgent and emergency operations in which nearly half of surgeons and sites were not adherent. Given the experience reportedbySchweizer et al, universal decolonizationand routine use of vancomycin may be a suitable alternative for urgent operations. The authors tested 36 infecting isolates for evidence of high-level resistance to mupirocin or chlorhexidine gluconate; 1 isolate demonstrated high-level mupirocin resistance, and none carried qac genes. Although these results are Related article page 2162 Opinion