Deep Sternal Wound Infection and Risk Analysis - What Do We Really Know?

We read with great interest the critique of Losanoff et al. [1] regarding our article on risk factor analysis for deep sternal wound infection (DSWI) following cardiac surgery [2]. Their comments, however, fail to be relevant. The paper was approved for publication by the journal in early 2001. It was therefore impossible for us to relate to the work of Losanoff’s group that was published in 2002 [3] [4]. One’s own scientific work should certainly be promoted in ways other than launching unjustified requests and making false assumptions.

The current problem of risk factor analysis was clearly illustrated in Table 1 of our article. Numerous studies on thousands of patients reaching different conclusions have already been published, illustrating the multifactorial cause of deep sternal wound infections. One accepted risk factor in most studies is diabetes. As already discussed in our paper, the Furnary group clearly demonstrated that the risk of DSWI in diabetics can be substantially lowered with a strict perioperative regimen of monitoring and glucose levels control [5].

The use of both internal thoracic arteries (ITAs) has been identified as an independent risk factor in several studies [6], but some authors did not see an increased risk of mediastinitis. Our data do indeed support the idea that the use of pedicled bilateral ITAs is an independent risk factor for sternal wound infection.

This risk factor seems to be controllable by using a skeletonizing takedown technique as discussed in our article. After changing our policy to a skeletonizing technique, the infection rate in this subgroup has been reduced like in other institutions [7]. These data will be published soon. This discussion illustrates the difficulties analyzing clinical events depending on many unrelated variables.

We believe - in contrast to Losanoff - that more than 9,000 patients comprise a patient cohort large enough for conclusions analyzing more than 19 important risk factors. We concentrated our analysis by intention on this literature based factors. Soft factors such as use of bone wax, excessive use of diathermy, osteoporosis and ”suboptimal sternal closure“ were not included in our analysis, as they are ill-defined and subjective by nature. Depending on specific local factors, especially those ”soft“ risk factors identified by one group may not be relevant in another institution.

Losanoff failed to recognize our detailed perioperative standard protocol with details regarding sternal closure techniques; otherwise, there would be no discernable reason for his request for more detailed information on this subject. In addition, the definition of a ”suboptimal sternal closure“ and how it is monitored at his institution is lacking. In our opinion, the way to avoid ”suboptimal sternal closure“ would generally be to institute a proper policy for sternotomy closure.

The underlying problem of DSWI is multifactorial as described. Most prevention guidelines basically address only a few factors based on experience in individual institutions. We believe that it would be difficult to reach a universal guideline. After having evaluated the DSWI data in our institution, two factors were changed in our perioperative protocol - the principal use of a skeletonized ITA and the principal increase (from 6 - 7 to 8) of sternal wires twisted together in pairs [8] used for sternotomy closure. We therefore think that those studies are extremely helpful regarding quality control and are also very helpful in reducing the incidence of DSWI in single institutions.

The high mortality rates in patients requiring more than twice revisions of sternal structures are established in the literature [9]. Treatment of DSWI was not the focus of the paper, so this information was not given in detail. In all patients, a stable fixation with the known techniques was applied [10]. A recent improvement was made by using the Ley prosthesis [11], which will hopefully reduce the high mortality in DSWI further.

References

• 1 Losanoff J E, Richman B W, Jones J W. Risk analysis of deep sternal wound infection and mediastinitis in cardiac surgery.  Thorac Cardiovasc Surg. 2002;  50 (6) 385
• 2 Gummert J F, Barten M J, Hans C. et al . Mediastinitis and cardiac surgery - an updated risk factor analysis in 10,373 consecutive adult patients.  Thorac Cardiovasc Surg. 2002;  50 (2) 87-91
• 3 Losanoff J E, Richman B W, Jones J W. Disruption and infection of median sternotomy: a comprehensive review.  Eur J Cardiothorac Surg. 2002;  21 (5) 831-839
• 4 Losanoff J E, Jones J W, Richman B W. Primary closure of median sternotomy: techniques and principles.  Cardiovasc Surg. 2002;  10 (2) 102-110
• 5 Furnary A P, Zerr K J, Grunkemeier G L, Starr A. Continuous intravenous insulin infusion reduces the incidence of deep sternal wound infection in diabetic patients after cardiac surgical procedures [see comments].  Ann Thorac Surg. 1999;  67 (2) 352-360
• 6 Ridderstolpe L, Gill H, Granfeldt H, Ahlfeldt H, Rutberg H. Superficial and deep sternal wound complications: incidence, risk factors and mortality.  Eur J Cardiothorac Surg. 2001;  20 (6) 1168-1175
• 7 Gurevitch J, Paz Y, Shapira I. et al . Routine use of bilateral skeletonized internal mammary arteries for myocardial revascularization.  Ann Thorac Surg. 1999;  68 (2) 406-411
• 8 Baskett R J, MacDougall C E, Ross D B. Is mediastinitis a preventable complication? A 10-year review.  Ann Thorac Surg. 1999;  67 (2) 462-465
• 9 Levi N, Olsen P S. Primary closure of deep sternal wound infection following open heart surgery: a safe operation?.  J Cardiovasc Surg (Torino). 2000;  41 (2) 241-245
• 10 Robicsek F, Daugherty H K, Cook J W. The prevention and treatment of sternum separation following open heart surgery.  Coll Works Cardiopulm Dis. 1977;  21 61-63
• 11 Astudillo R, Vaage J, Myhre U, Karevold A, Gardlund B. Fewer reoperations and shorter stay in the cardiac surgical ward when stabilising the sternum with the Ley prosthesis in post-operative mediastinitis.  Eur J Cardiothorac Surg. 2001;  20 (1) 133-139