Relation Between Topical Application of Platelet-Rich Plasma and Vancomycin and Severe Deep Sternal Wound Infections After a First Median Sternotomy

Relation Between Topical Application of Platelet-Rich Plasma and Vancomycin and Severe Deep Sternal Wound Infections After a First Median Sternotomy Baron L. Hamman, MDa, Laura Y. Stout, BBAb, Theodore T. Theologes, MS, PA-Ca, Danielle M. Sass, MPH, CPHc, Briget da Graca, JD, MSc, and Giovanni Filardo, PhD, MPHa,c,d,e,f,* Deep sternal wound infections (DSWIs) are serious complications of sternotomy, leading to increased mortality and costs of care. Topical applications of autologous platelet concentrate and vancomycin have both shown promise in preventing DSWIs. From January 1, 1998, to November 30, 2010, 1,866 patients without previous sternotomy underwent cardiac surgery at the Baylor University Medical Center, Dallas, by a single surgeon who systematically adopted application of a paste containing vancomycin, calcium-thrombin, and platelet-rich plasma (PRP paste) to the edges of sternal wounds before closure in December 2005. A propensity-adjusted logistic regression model employing Firth’s penalized maximum likelihood method was used to assess the association between the use of the PRP paste (intervention) and the incidence of severe DSWI. Eleven patients (0.59%) developed severe DSWIs. All were among the 1,318 patients in the control group (0.83%); no severe DSWIs developed in the 548 patients in the intervention group. Both the unadjusted and adjusted associations between the study intervention and DSWI were statistically significant (unadjusted p value [ 0.021; adjusted p value [ 0.005; adjusted odds ratio [ 0.05, 95% confidence interval 0.01, 0.50). In conclusion, the PRP paste appears to prevent severe DSWIs. Ó 2014 Elsevier Inc. All rights reserved. (Am J Cardiol 2014;113:1415e1419) Deep sternal wound infections (DSWIs) are a rare but potentially life-threatening complication after cardiac surgery. DSWIs occur in 1% to 2% of patients who undergo median sternotomy for cardiac surgery and are associated with significant increases in mortality, hospital length of stay, and costs of care.1e3 Thirty-day mortality is reported to be 7.3% in patients who develop DSWIs compared with 1.6% in those who do not,1 and length of stay and hospital costs for cardiac surgery patients who develop DSWIs have been reported to double and triple, respectively.4 Preventing DSWIs is therefore a high priority for all cardiothoracic surgeons, with implications both for improving patient outcomes and for decreasing costs of care. Topical applications of autologous platelet concentrate and of vancomycin have both shown promise for preventing sternal wound infections.5e8 Although the rate of DSWIs in our practice was lower than many of those reported in 2005, the profound implications of this serious complication made us seek to do even better. Therefore, in a Baylor Heart and Vascular Institute, Baylor University Medical Center, Dallas, Texas; bDepartment of Research, HeartPlace, Dallas, Texas; cInstitute for Health Care Research and Improvement, Baylor Health Care System, Dallas, Texas; dThe Heart Hospital Baylor Plano, Plano, Texas; e Department of Statistical Science, Southern Methodist University, Dallas, Texas; and fDepartment of Infectious Diseases, University of Louisville, Louisville, Kentucky. Manuscript received October 23, 2013; revised manuscript received and accepted December 30, 2013. Grant support was provided by the Discovery Foundation and by the Bradley Family Endowment through the Baylor Health Care Foundation. See page 1418 for disclosure information. *Corresponding author: Tel: (214) 265-3633; fax: (214) 265-3628. E-mail address: giovanfi@baylorhealth.edu (G. Filardo).

0002-9149/14/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjcard.2013.12.046

December 2005, Baylor University Medical Center started encouraging surgeons in our practice to apply a paste consisting of vancomycin, calcium-thrombin, and plateletrich plasma (PRP paste) to the edges of sternal wounds before closure. One surgeon in our practice systematically adopted this change, whereas others chose to wait for evidence of its clinical and cost-effectiveness. Here, we describe the impact of this change in practice on the incidence of severe DSWIs in that single surgeon’s patient population. Methods All consecutive patients (n ¼ 1,866) who underwent cardiac surgical procedures with full sternotomy and who had not previously undergone coronary artery bypass graft, valve surgery, or other procedures requiring sternotomy, performed by one surgeon from January 1998 to November 2010, were included in this study; among them, 548 (29.37%) received the PRP paste and 1,318 (73.63%) were historical controls. All surgeries were performed at the Baylor University Medical Center in Dallas, the 1,026bed flagship hospital of the Baylor Health Care System. Institutional review board approval was obtained for a retrospective analysis of patient records. All patients received prophylactic antibiotics in accordance with national guidelines for the prevention of surgical site infections.9 Antibiotics were started within 1 hour of incision and discontinued 48 hours of surgery. Before 2006, antibiotic selection was based on surgeon choice, and patients usually received vancomycin or ceftazidime. In late 2006, with an increasing focus on standardization of care, cefazolin became the antibiotic of choice. www.ajconline.org

1416

The American Journal of Cardiology (www.ajconline.org)

Table 1 System for classifying sternal wound infections by severity Category

Description

1 2 3 4

Suture granuloma or suture reaction or wire reaction Bone click or unstable bone Skin dehiscence or skin-only superficial wound infection Deep tissue infection from the subcutaneous to the bone but no bone involvement Infection including bone involvement Infection including bone involvement plus septicemia

5 6

In December 2005, Baylor University Medical Center started encouraging surgeons in our cardiothoracic surgery group to apply a “paste” to the edges of the sternum just before closure. In aiming to produce a paste with a consistency that would allow accurate and reproducible application to the sternal edge without waste, we arrived at the following recipe: (1) combine 5 ml of 10% calcium chloride solution (100 mg/ml; 1.4 mEq of calcium per ml) with 5,000 IUs of topical thrombin (bovine origin), (2) draw 0.6 ml of the resulting suspension and combining it with 3 ml of PRP, (3) add the mixture to 2 g of vancomycin hydrochloride powder, and (4) mix the powder and the liquid together so that it forms a paste-like consistency. The PRP was produced using a Harvest SmartPReP 2 system,10 which generates an average platelet concentration 4.8 greater than that found in unprocessed, anticoagulated whole blood. We processed 60 ml blood draws to produce w7 ml of PRP with platelet concentrations in the range of 1.5 to 2 million platelets/ml, which is considered optimal for stimulating angiogenesis.11 All the patients undergoing surgery between December 2005 and November 2010 received the “PRP paste” in addition to the routine antibiotic prophylaxis. For most of these patients, the PRP was prepared using the SmartPReP 2 APCþ system manufactured by Harvest Technologies Corp.12 Patients who underwent surgery between January 1998 and November 2005 were considered controls, given that they received only the routine antibiotic prophylaxis. All patients were assessed for the development of sternal wound infections, in hospital, during follow-up office visits, and via telephone calls for 12 months after discharge. Identified infections were classified according to the system we (B.H. and T.T.) developed, listed in Table 1. Infections falling in categories 1 to 4 were considered superficial and not included in this analysis; categories 5 and 6 constituted the severe DSWIs that were of interest here. Clinical and nonclinical data on the patients’ risk factors were collected and entered into the Society of Thoracic Surgeons Adult Cardiac Surgery Database.13 Data abstraction, collection procedures, and variable definitions used for the Adult Cardiac Surgery Database are standardized and have been described elsewhere.13 The key Society of Thoracic Surgeons risk of operative mortality14 factors was considered: age, gender, body surface area (m2), race, ethnicity, diabetes mellitus, infectious endocarditis, renal failure, creatinine, chronic lung disease, hypertension, peripheral vascular disease, cerebrovascular disease, myocardial

infarction, time from last myocardial infarction to surgery, atrial fibrillation, tobacco use, congestive heart failure, previous pace/ implantable cardioverter defibrillator, previous percutaneous coronary intervention, preoperative angina, ejection fraction, left main disease, surgery status, type/ complexity of surgery, and preoperative use of an intra-aortic balloon pump. Means, SDs, and percentages were calculated to describe the study cohort. Given the very low incidence rate of the outcomes, a logistic regression model employing Firth’s penalized maximum likelihood method15 with the study intervention as the sole independent variable was used to assess the unadjusted association between the intervention and severe DSWI. To account for possible confounders of this association, a propensity score approach was employed.16 A logistic regression model was fit to estimate the likelihood of receiving the study intervention. Covariates for this logistic model included established risk factors identified by the Society of Thoracic Surgeons,17 as well as other clinical and demographic factors—these are listed in Table 2. All continuous predictors were modeled with restricted cubic splines.18e20 Missing data were present in only 4 variables included in the propensity model (body surface area ¼ 0.05%, creatinine ¼ 2.25%, ejection fraction ¼ 13.71%, status ¼ 0.2%) and were accounted for via multiple imputation using predictive mean matching (a total of 200 imputations were performed for each missing value).21,22 Estimates from the resulting propensity model were then used to adjust the effect of the study intervention on DSWI in a logistic regression employing Firth’s penalized maximum likelihood method.15 All analyses were performed using the R software, version 2.15.1.23 Results Characteristics of the study population are listed in Table 2. The incidence of all wound infections, by category, is listed in Table 3. Overall, 11 patients (0.59%) developed severe DSWIs (categories 5 and 6)—all these events were among the 1,318 patients in the control group (incidence 0.83%). The incidence in the intervention group (548 patients) was zero. All severe DSWIs developed within 4 months of surgery. Both the unadjusted and adjusted associations between the study intervention and the severe DSWI were found to be significant (unadjusted p value ¼ 0.021; adjusted p value ¼ 0.005; adjusted odds ratio ¼ 0.05, 95% confidence interval 0.01, 0.50). Discussion We compared the incidence of severe DSWIs before and after instituting use of a PRP paste as standard practice in patients who underwent cardiac surgical procedures. After adjustment for key risk factors, patients who had surgery before implementation of the intervention were 18 times more likely to experience severe DSWI than those who had surgery postimplementation. Our results are consistent with previous reports of topical application of vancomycin or autologous platelet gel to sternal incisions in cardiac surgery. A 1989 randomized controlled trial in 416 cardiac surgery patients found application of topical vancomycin decreased incidence of

Miscellaneous/Preventing Severe Deep Sternal Wound Infections

1417

Table 2 Characteristics of 1,866 patients who underwent CABG, valve, or CABG þ valve surgery at the Baylor University Medical Center (Dallas, Texas) between 1998 and 2010 Characteristic

Age (yrs) Men Women BSA (m2) White Black Asian Pacific Islander Native American Other/unknown Hispanic Diabetes mellitus Infectious endocarditis Renal failure Creatinine (mg/dl) Chronic lung disease Hypertension Peripheral vascular disease Cerebrovascular disease Myocardial infarction Time from last MI to surgery (h) 6 >6 but <24 24 Atrial fibrillation Smoker Never Former Current Heart failure Previous pace/ICD Previous PCI Preoperative angina Ejection fraction Left main disease Operation status Elective operation Nonelective operation CABG Valve CABG þ valve Preoperative IABP

Intervention Status

p Value*

Control (n ¼ 1,318, 70.6%)

Intervention (n ¼ 548, 29.4%)

64.1  11.6 66.8 33.2 2.0  0.3 79.1 10.1 1.1 0.1 0.2 9.3 5.2 34.7 1.6 7.4 1.2  1.1 14.6 68.6 18.8 12.6 38.8

64.9  11.7 69.5 30.5 2.0  0.2 81.0 12.2 0.7 0.4 0.4 5.3 7.7 33.0 2.6 7.7 1.3  1.0 16.1 75.2 14.6 13.0 41.6

0.5 2.1 97.4 9.3

0.5 1.0 98.5 10.6

63.1 14.3 22.6 25.9 1.6 18.4 54.2 47.3  15.5 14.4

79.6 11.7 8.8 41.1 4.6 21.9 58.2 47.8  15.4 22.1

87.8 12.2 74.2 12.8 13.1 9.4

55.5 44.5 62.6 20.8 16.6 4.4

0.540 0.255 0.850 0.016

0.040 0.520 0.189 0.847 0.530 0.434 0.005 0.033 0.819 0.275 0.372

0.390 <0.001

<0.001 <0.001 0.095 0.113 0.360 <0.001 <0.001 <0.001 <0.001

Data are presented as percentage, or mean  SD. BSA ¼ body surface area; CABG ¼ coronary artery bypass graft; IABP ¼ intra-aortic balloon pump; ICD ¼ implantable cardioverter defibrillator; MI ¼ myocardial infarction; PCI ¼ percutaneous coronary intervention. * Fisher’s exact test.

sternal wound infections from 3.6% to 0.5%,8 and a more recent study in 36 patients investigating the impact of topical application to the sternal incision on serum levels of vancomycin in cardiac surgery patients reported that no patient developed a sternal wound infection.24 Topical application of platelet gels has similarly been reported to reduce incidence of wound infections in cardiac surgery patients.5,7 One recent study found that application of platelet-rich and platelet-poor plasma during closure of chest and leg wounds in patients with coronary artery bypass graft reduced the odds of a surgical wound infection by 93%6;

however, another found a nonsignificantly higher incidence of DSWI in high-risk patients treated with platelet-rich plasma.25 Although the general consistency of results showing that the reduced incidence of DSWI with application of platelet gels supports their effectiveness, the magnitude of the effect is hard to judge, partly because of differences in how sternal wound infections are defined,26e28 in the standard prophylaxis used (which not all the published reports clearly specify6,7,29) and in the composition of the PRP produced by different preparation methods.30

1418

The American Journal of Cardiology (www.ajconline.org)

Table 3 Incidence of any sternal wound infection among the 1,866 patients who underwent CABG, valve, or CABG þ valve surgery at the Baylor University Medical Center (Dallas, Texas) between 1998 and 2010 by category of infection Sternal Wound Infection Category

None 1. Suture granuloma or suture reaction or wire reaction 2. Bone click or unstable bone 3. Skin dehiscence or skin-only superficial wound infection 4. Deep tissue infection from the subcutaneous to the bone but no bone involvement 5. Infection including bone involvement 6. Infection including bone involvement plus septicemia

Intervention Status Control (n ¼ 1,318, 70.6%)

Intervention (n ¼ 548, 29.4%)

1,302 (98.8) 1 (0.1)

546 (99.6) 2 (0.4)

Total (n ¼ 1,866, 100.0%) 1,848 (99.0) 3 (0.2)

0 (0.0) 0 (0.0)

0 (0.0) 0 (0.0)

0 (0.0) 0 (0.0)

4 (0.3)

0 (0.0)

4 (0.2)

9 (0.7) 2 (0.1)

0 (0.0) 0 (0.0)

9 (0.5) 2 (0.1)

Data are presented as n (%). CABG ¼ coronary artery bypass graft.

Limitations of our study include the use of historical controls. This approach, enabling comparison within a single surgeon’s practice, was chosen over comparison with contemporary surgeries by other surgeons in the practice, to minimize problems related to variation in technique, and standard practices for antibiotic prophylaxis and wound care between surgeons and hospitals. It does, however, limit the generalizability of our results, and the intervention group may have benefited from advances in surgical, nursing, and wound care techniques over the more than decade of cardiac surgery covered by this study. Furthermore, there was a change in the antibiotics routinely used for prophylaxis of surgical site infections in the practice that occurred nearly simultaneously with the introduction of the PRP paste. We therefore cannot determine whether the reduction observed in DSWIs was due to the change in prophylactic antibiotics, the use of the PRP paste, or the combination of these 2 factors. Also, as in any observational study, we cannot rule out effects due to unobserved differences between comparison groups. This limitation is, however, mitigated by our having adjusted for the key Society of Thoracic Surgeons risk factors for operative mortality, which overlap substantially with those identified for DSWIs (increasing age, diabetes mellitus, chronic lung disease, congestive heart failure, gender, obesity, and smoking7,25), as well as obesity (as measured by body surface area), and type/complexity of surgery. The discrepancy between the results observed here and the lack of any significant effect reported in the recent randomized controlled trial testing platelet-rich plasma for DSWI prevention in high-risk patients25 suggests that our results should be interpreted with caution, particularly with regard to the incremental value of including PRP in the topical paste. However, the trial in question was small (n ¼ 196) and, given the rarity of DSWIs, may not have been adequately powered to detect a difference. Moreover, our results may underestimate the effectiveness of the intervention: a greater proportion of the intervention-period surgeries were more complex procedures (fewer isolated coronary artery bypass graft

and elective operations), in which there is a greater risk of wound infection. Had data on operation length been available, we may have been able to adjust more effectively for this change in practice, but that adjustment would likely have strengthened our results. Disclosures An educational grant was provided by Terumo Cardiovascular Systems Corporation, which manufactures the SmartPrep 2 System used to produce the platelet-rich plasma in this study. 1. Cayci C, Russo M, Cheema FH, Martens T, Ozcan V, Argenziano M, Oz MC, Ascherman J. Risk analysis of deep sternal wound infections and their impact on long-term survival: a propensity analysis. Ann Plast Surg 2008;61:294e301. 2. de Lissovoy G, Fraeman K, Hutchins V, Murphy D, Song D, Vaughn BB. Surgical site infection: incidence and impact on hospital utilization and treatment costs. Am J Infect Control 2009;37:387e397. 3. Toumpoulis IK, Anagnostopoulos CE, Derose JJ Jr, Swistel DG. The impact of deep sternal wound infection on long-term survival after coronary artery bypass grafting. Chest 2005;127:464e471. 4. Graf K, Ott E, Vonberg RP, Kuehn C, Haverich A, Chaberny IF. Economic aspects of deep sternal wound infections. Eur J Cardiothorac Surg 2010;37:893e896. 5. Trowbridge CC, Stammers AH, Woods E, Yen BR, Klayman M, Gilbert C. Use of platelet gel and its effects on infection in cardiac surgery. J Extra Corpor Technol 2005;37:381e386. 6. Khalafi RS, Bradford DW, Wilson MG. Topical application of autologous blood products during surgical closure following a coronary artery bypass graft. Eur J Cardiothorac Surg 2008;34:360e364. 7. Englert SJ, Estep TH, Ellis-Stoll CC. Postoperative surgical chest and leg incision sites using platelet gel: a retrospective study. J Extra Corpor Technol 2008;40:225e228. 8. Vander Salm TJ, Okike ON, Pasque MK, Pezzella AT, Lew R, Traina V, Mathieu R. Reduction of sternal infection by application of topical vancomycin. J Thorac Cardiovasc Surg 1989;98:618e622. 9. Bratzler DW, Hunt DR. The surgical infection prevention and surgical care improvement projects: national initiatives to improve outcomes for patients having surgery. Clin Infect Dis 2006;43:322e330. 10. Terumo Cardiovascalar Group. HarvestÒ PRP with SmartPrepÒ 2 system. 11. Giusti I, Rughetti A, D’Ascenzo S, Millimaggi D, Pavan A, Dell’Orso L, Dolo V. Identification of an optimal concentration of platelet gel for promoting angiogenesis in human endothelial cells. Transfusion 2009;49:771e778.

Miscellaneous/Preventing Severe Deep Sternal Wound Infections 12. Harvest Technologies Corp. SmartPReP 2 APCþ. 13. Ferguson TB Jr, Dziuban SW Jr, Edwards FH, Eiken MC, Shroyer AL, Pairolero PC, Anderson RP, Grover FL. The STS National Database: current changes and challenges for the new millennium. Committee to establish a National Database in Cardiothoracic Surgery, the Society of Thoracic Surgeons. Ann Thorac Surg 2000;69: 680e691. 14. Shahian DM, O’Brien SM, Filardo G, Ferraris VA, Haan CK, Rich JB, Normand SL, DeLong ER, Shewan CM, Dokholyan RS, Peterson ED, Edwards FH, Anderson RP. The Society of Thoracic Surgeons 2008 cardiac surgery risk models: part 1ecoronary artery bypass grafting surgery. Ann Thorac Surg 2009;88:S2eS22. 15. Heinze G. A Comparative investigation of methods for logistic regression with separated or nearly separated data. Stat Med 2006;25: 4216e4226. 16. D’Agostino RB Jr. Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group. Stat Med 1998;17:2265e2281. 17. Hsiao WC, Stason WB. Toward developing a relative value scale for medical and surgical services. Health Care Financ Rev 1979;1: 23e38. 18. Filardo G, Hamilton C, Hamman B, Grayburn P. Obesity and stroke after cardiac surgery: the impact of grouping body mass index. Ann Thorac Surg 2007;84:720e722. 19. Filardo G, Hamilton C, Hamman B, Ng HK, Grayburn P. Categorizing BMI may lead to biased results in studies investigating inhospital mortality after isolated CABG. J Clin Epidemiol 2007;60: 1132e1139. 20. Filardo G, Hamilton C, Hamman BL, Ng HKT, Grayburn P. Continuous exposures in epidemiologic studies: the effect of body mass index on risk of acute renal failure following coronary artery bypass graft surgery. Am J Epidemiol 2006;163:S231.

1419

21. Little R, An H. Robust likelihood-based analysis of multivariate data with missing values. Stat Sinica 2004;14:949e968. 22. Shahian DM, Blackstone EH, Edwards FH, Grover FL, Grunkemeier GL, Naftel DC, Nashef SA, Nugent WC, Peterson ED. Cardiac surgery risk models: a position article. Ann Thorac Surg 2004;78: 1868e1877. 23. The R Project for Statistical Computing. www.r-project.org. Accessed February 20, 2014. 24. Lazar HL, Barlam T, Cabral H. The effect of topical vancomycin applied to sternotomy incisions on postoperative serum vancomycin levels. J Card Surg 2011;26:461e465. 25. Dorge H, Sellin C, Bury MC, Drescher A, Seipelt R, Grossmann M, Danner BC, Schoendube FA. Incidence of deep sternal wound infection is not reduced with autologous platelet rich plasma in highrisk cardiac surgery patients. Thorac Cardiovasc Surg 2013;61: 180e184. 26. Horan TC, Gaynes RP, Martone WJ, Jarvis WR, Emori TG. CDC definitions of nosocomial surgical site infections, 1992: a modification of CDC definitions of surgical wound infections. Infect Control Hosp Epidemiol 1992;13:606e608. 27. El Oakley RM, Wright JE. Postoperative mediastinitis: classification and management. Ann Thorac Surg 1996;61:1030e1036. 28. Pairolero PC, Arnold PG. Management of infected median sternotomy wounds. Ann Thorac Surg 1986;42:1e2. 29. Vang SN, Brady CP, Christensen KA, Allen KR, Anderson JE, Isler JR, Holt DW, Smith LM. Autologous platelet gel in coronary artery bypass grafting: effects on surgical wound healing. J Extra Corpor Technol 2007;39:31e38. 30. Everts PA, Brown Mahoney C, Hoffmann JJ, Schonberger JP, Box HA, van Zundert A, Knape JT. Platelet-rich plasma preparation using three devices: implications for platelet activation and platelet growth factor release. Growth Factors 2006;24:165e171.