**11. Suggested treatment algorithms for DSWI**

While accumulating extensive experience with DSWI management as a referral center for these difficult problems. We and others have formalized protocols for managing mediastinal infection utilizing NPT which allows sternal salvage and improved outcomes in the majority of cases of DSWI [14-16, Figure 5].

means for clearing infection, and nearly 50% of patients in their cohort were managed with vac therapy alone or vac + secondary sternal closure. In our own series, we found that use of NPT for controlling mediastinal infection was an independent predictor of survival on multivariable analysis [unpublished data], and that patients managed with NPT had significantly improved long term survival compared with patients not treated

Fig. 4. Kaplan Meier survival curves for patients with poststernotomy mediastinitis treated with negative pressure therapy (NPT) or by traditional means for controlling mediastinal infection (controls). Patients treated with NPT had significantly improved long-term

While accumulating extensive experience with DSWI management as a referral center for these difficult problems. We and others have formalized protocols for managing mediastinal infection utilizing NPT which allows sternal salvage and improved outcomes in the majority

with NPT [Figure 4].

survival by log rank analysis.

of cases of DSWI [14-16, Figure 5].

**11. Suggested treatment algorithms for DSWI** 

Fig. 5. Suggested algorithm for management of mediastinal infection after cardiac surgery via median sternotomy, with permission from Sjogren et al [15]. Emphasis is placed on gentle sternal debridement, negative pressure therapy to the mediastinum, and closure based on monitored serum C-reactive protein levels.

As noted previously, successful treatment of DSWI begins with recognition of signs and symptoms of sternal infection, which may occasionally be subtle. Sternal wound exploration to ensure a prompt and accurate diagnosis is warranted when signs of mediastinitis are present. This approach also helps to distinguish between superficial and deep sternal infections. At the time of initial sternal exploration, tissues and fluid should be obtained for bacteriologic analysis. Targeted antibiotic therapy for 4-6 weeks duration is prescribed and is determined by the culture results [14, 15]. If the infectious process extends beneath the facial layer, all sternal hardware should be removed. The sternum itself is then gently debrided of grossly devitalized tissue, but wide excision of the sternum is not necessary and may be injurious and counterproductive. Limited sternal debridement is now preferred and good results have been seen with this approach [14, 16, 83, 93, 94]. Negative pressure therapy is then instituted on the opened incision after limited sternal debridement. The polyurethane foam is subsequently changed in the operating room or at the bedside every 2- 3 days. During this time, assessment as to the state of the sternum is made to determine if mediastinal flap repair is required or if secondary sternal closure is possible. Our general approach for determining when the sternum can reliably be reapproximated is based on the state of the sternum after several days of NPT. Wound characteristics precluding secondary

Sternal Wound Complications Following Cardiac Surgery 299

Loop has stated: "prevention and better treatment of sternal wound complications must be a major goal in assuring the highest quality of cardiovascular care…[5]. Although most efforts towards DSWI have focused on the treatment of DSWI, several methods to reduce rates of mediastinitis have been proposed and validated recently. As a result, efforts to prevent mediastinal infection may already be working. For instance, investigators from Boston recently reported on their experience with DSWI between 1992 and 2006, separating analysis into early and late time periods. They noted that DSWI had decreased from 1.57% to 0.88% over the last 5 years of their analysis and attributed the positive findings to adoption of strict glucose control algorithms [40]. Tight glycemic control appears to be effective in significantly reducing rates of DSWI. [30, 103]. In addition, Lazar et al demonstrated improved coronary surgery outcomes with a strategy for strict glucose control (125 – 200 mg/dL) using glucose-insulin-potassium solution, including reduced ischemic events and improved rates of wound infection in a cohort of diabetic patients undergoing CABG [31]. Antimicrobial therapy has also positively impacted rates of DSWI. Most notably, appropriate timing and selection of preoperative antibiotics has been associated with reduced rates of sugical site infection [6, 104]. Furthermore, use of nasal mupirocin in patients undergoing cardiac surgery via median sternotomy eradicates 95 – 100% of *S. aureus* for up to one year postoperatively [105], and sternal wound infections are also reduced by nearly 2/3 in some series with the use of mupirocin in patients colonized with *S.* 

Technical details of the median sternotomy incision and closure almost certainly impact the likelihood for DSWI postoperatively. Baskett et al have argued that assiduously following technical details and proper surgical and aseptic techniques can also dramatically reduce rates of poststernotomy infection [42]. They emphasize the importance of accurate reapproximation of the sternal halves and caution against the use of bone wax to gain sternal hemostasis [42]. Since sternal instability is often evoked as one mechanism contributing to the development of sternal infection, evaluation of the most effective sternal closure methods has been undertaken. For example, Schimmer et al compared standard closure techniques by transsternal or peristernal wiring with techniques using additional lateral wire reinforcement in the method described by Robicsek in a cohort of 815 high-risk patients [108]. There were no differences observed in the rates of sternal dehiscence or superficial or deep sternal wound infections, but they did show that more sternal wires placed for closure was associated with significantly reduced rates of DSWI [108]. Others have emphasized that rigid sternal closure techniques are preferred, particularly in those considered at high risk for sternal wound complications such as dehiscence and infection. These techniques are widely used in most surgical practices that incorporate osteotomy incisions. In fact, cardiac surgery is now the only discipline not routinely repairing osteotomies with rigid plating techniques [109]. Lee et al recently reported their experience with titanium plate fixation of the sternum in 750 patients at high risk for sternal wound complications, noting 97.6% freedom from sternal infection or dehiscence [110]. Levin et al have also introduced another form of rigid sternal closure as an alterntive to wire circlage and highlighted advantages to this approach [111]. Although rigid sternal closure techniques have not been compared prospectively with wire circlage, cadaveric studies have shown rigid plate fixation techniques to be superior to wire circlage by providing increased stiffness to the wound closure and less lateral displacement of the sternal halves [112].

**13. Can mediastinitis be prevented?** 

*aureus* [106, 107].

sternal reapproximation include multiple transverse sternal fractures, poor bone stock, costosternal separation, or the requirement for such extensive sternal debridement that reapproximation of the sternal halves is not feasible. Gustafsson en et al have advocated use of serum C-reactive protein levels to guide wound closure timing [95, Figure 6]. C-reactive protein levels less than 70 mg/L corresponded with successful sternal reapproximation [95]. Successes have also been reported with sternal plating as a treatment for the fractured sternum [11, 96], but our approach has been conservative in this regard since any residual infectious process could contaminate the implanted hardware [91].

Numerous clinical advantages for DSWI management protocols incorporating NPT have been observed, many centered upon the sternal stabilization achieved when vacuumassisted clousure is engaged [70]. The sternal stability afforded by NPT improves pain compared with open packing or other approaches for addressing the infected mediastinum [88]. In addition, the stabilized sternum yields several pulmonary benefits, the first of which is the ability to successfully separate from mechanical ventilation. This promotes earlier and more effective patient mobilization and prevents the patient from being confined to bed, where other complications common to DSWI therapy are often incurred [70]. Negative pressure therapy also improves ventilation and overall pulmonary function and leads to more effective chest physiotherapy [70, 97]. Importantly, no deleterious hemodynamic effects of NPT have been documented although this has been speculated [70].

### **12. Complications of treating mediastinitis**

It has been estimated that approximately 15% of patients develop recurrent infection [98]. In the experience of Bapat et al, this has included recurrent infection with the same organism associated with the original sternal infection [99], and we speculate that if the polyurethane foam required for NPT is not adequately inserted with each dressing change, small, isolated spaces may arise within the wound that can become superinfected. We recently reported rates of recurrent wound complications associated with various mediastinal flap coverages. For example, muscle flap repair of the treated mediastinum, consisting predominantly of pectoralis muscle flaps, was associated with increased rates of recurrent wound complications such as hematoma, seromas, and recurrent infection [14]. Conversely, use of NPT prior to definitive repair of the sternal wound defect was associated with increased rates of successful secondary sternal closure without the need for any flap transfer, and with shortened length of hospital stay after definitive repair. Excellent results have been reported elsewhere when NPT is incorporated into management protocols for DSWI [65].

Petzina et al recently reported a 7.2% rate of "major complications" associated with NPT for DSWI in a cohort of 69 patients. Most complications were bleeding-related [100]. On the other hand, cardiac function and hemodynamics appear to be stable during NPT to the open mediastinal wound [101, 102].

When NPT is used, caution should be exercised with regard to the length of therapy. In our own experience, prolonged use of NPT leads to a "frozen" mediastinum, making subsequent closure by vascularized flaps or other technique difficult to perform and places the cardiac structures at risk for injury during subsequent sternal repair. Others have noted the similar difficultiess [99]. In such cases, continued application of NPT to closure by secondary intent may be the best therapeutic option rather than to place the mediastinal structures at risk for injury during attempted flap repair.
