Risk factors of infected sternal wounds versus sterile wound dehiscence

j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.JournalofSurgicalResearch.com

Risk factors of infected sternal wounds versus sterile wound dehiscence Rose H. Fu, MD,a Andrew L. Weinstein, MD,a Michelle M. Chang, BS,a Michael Argenziano, MD,b Jeffrey A. Ascherman, MD,a and Christine H. Rohde, MD, MPHa,* a

Division of Plastic and Reconstructive Surgery, Department of Surgery, New York-Presbyterian Hospital/Columbia University Medical Center, New York, New York b Division of Cardiac, Vascular, and Thoracic Surgery, Department of Surgery, New York-Presbyterian Hospital/ Columbia University Medical Center, New York, New York

article info

abstract

Article history:

Background: Sterile sternal dehiscence (SSD) and sternal wound infections (SWIs) are two

Received 9 July 2015

complications of median sternotomy with high rates of morbidity. Sternal wound com-

Received in revised form

plications also carry significant economic burden, almost tripling patients’ hospital costs

29 July 2015

and are considered a nonreimbursable “never event” for Medicare. Historically, SDD and

Accepted 31 July 2015

SWI have been recognized as discrete entities, but nonetheless continue to be categorized

Available online xxx

as a singular complication in literature. The purpose of this study was to determine specific patient demographic and perioperative predictors of SSD and SWI.

Keywords:

Materials and methods: An institutional review boardeapproved, retrospective study of 8098

Sternal wound

consecutive patients who underwent cardiac surgery at Columbia University Medical

Infection

Center between January 2008 and December 2013 was conducted. Patients were catego-

Mediastinitis

rized into three groups: no sternal wound complication, SSD, or SWI. Statistical analysis

Dehiscence

was performed using univariate and multivariate logistic regression analysis.

Treatment

Results: Of 8098 patients, there were 73 patients (0.9%) with SSD and 40 (0.5%) with SWI who

Reconstruction

required plastic surgical consultation, debridement, and flap closure. In univariate analysis

Risk factors

of SSD, positive predictors (i.e., “risk” factors) were age >42 years, prior surgery this admission, ‡2 arterial conduits, internal mammary artery (IMA) grafting with or without previous IMA grafting, body mass index (BMI) >30 (obese), CHF, diabetes requiring medication, respiratory failure, and unplanned cardiac reoperation; negative predictors (i.e., “protective” factors) were no arterial conduits and extubation within 24 h. In univariate analysis of SWI, positive predictors were IMA grafting with or without previous IMA grafting, postoperative hematocrit urgent/emergent surgical priority, BMI >30 (obese), cardiac ejection fraction <40%, and respiratory failure; negative predictors were no arterial conduits and elective surgical priority. In multivariate regression, BMI >30, diabetes requiring medication, and respiratory failure were determined to be significant positive predictors of SSD, and IMA grafting with or without prior IMA grafting and respiratory failure were significant positive predictors for SWI; no significant negative predictors were identified. Conclusions: This study found that SSD and SWI have many common significant predictors consistent with findings that increased BMI, use of IMA grafts, poor cardiac reserve, and

* Corresponding author. New York-Presbyterian Hospital/Columbia University Medical Center, Division of Plastic and Reconstructive Surgery, 161 Fort Washington Avenue, Rm. 511, New York, NY 10032. Tel.: þ1 212-342-3707; fax: þ1 212-305-9626. E-mail address: [email protected] (C.H. Rohde). 0022-4804/$ e see front matter ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2015.07.045

2

j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8

postoperative respiratory failure confer increased risk of sternal wound complications. Additionally, this study also found that there were predictors unique to each entity supporting that SSD and SWI may be related but are not a singular entity. Recognition and prevention of significant positive and negative predictors of SSD and SWI may be valuable in preoperative counseling, operative planning, and postoperative management. Although sternal wound complications can be successfully managed by plastic surgical intervention, preventing the development of median sternotomy complications may curb costs incurred by both patients and health care systems. ª 2015 Elsevier Inc. All rights reserved.

1.

Background

Median sternotomy remains the most commonly used incision in cardiothoracic surgery due to its technical simplicity and excellent exposure of the heart, great vessels, and pulmonary hila. As with any incision, wound complications are risks. In particular, infection involving the sternum or the retrosternal mediastinal space is a serious complication of median sternotomy with its incidence varying widely from 0.5%e5% in the literature, with increased rates of up to 80% reported in special populations such as those who have undergone radiation [1,2]. Given the sheer volume of cardiothoracic surgery requiring median sternotomydmore than 500,000 procedures in 2012, with an expected increase of 30%e50% by 2025, sternal wound complications are significant contributors to morbidity and mortality in these patients [3]. Furthermore, because sternal wound complications have been categorized by Medicare as a nonreimbursable “never event”, the substantial economic burden on health care dollars constitutes a growing financial incentive to better understand the risk factors for these wound complications [4]. Sterile sternal wound dehiscence (SSD) and subsequent sternal nonunion, initially classified together with sternal wound infections (SWIs), have been described and recognized as a distinct entity (Stoney et al. in 1978) [5]. However, a review of the literature shows that these separate entities have continued to be studied together. Risk factors for deep SWI and wound dehiscence as defined by the National Nosocomial Infection Surveillance system of the Centers for Disease Control and Prevention vary from patient comorbidities to intraoperative technical failure and postoperative events [6]. Patient conditions such as obesity, history of radiation, and other comorbidities which inhibit wound healing such as diabetes, steroid use, and preoperative and postoperative malnutrition all portend increased risk of sternal dehiscence and infection [7e12]. Intraoperative factors such as poor sternal wire fixation or paramedian sternotomy can lead to sternal instability with normal chest wall movement during physiological respiration, precipitating sternal nonunion, and overlying wound dehiscence [13]. Other intraoperative factors such as bilateral internal mammary harvest for coronary artery bypass grafts can lead to increased bone ischemia and decreased healing potential [14]. Postoperative factors including decreased cardiac function and protracted positive pressure ventilation have been associated with increased sternal wound complications [13].

In our clinical experience at Columbia University Medical Center (CUMC), patients who present with sterile sternal dehiscence typically report sternal clicking and popping resulting in pain, limiting their activities of daily living. Patients also may have skin and soft tissue dehiscence. These patients do not exhibit symptoms of infection, such as leukocytosis, fever, or purulent drainage. On reoperation, these patients are noted to have sternal nonunion, often with tearing through of sternal wires used for initial sternotomy closure and are microbiologically culture negative. Few studies have specifically examined this group of patients, although some data have shown that risk factors for noninfectious dehiscence include obesity, chronic obstructive pulmonary disease (COPD), and New York Hospital Association Class IV status [15]. In our literature search, there has been no current study separately analyzing SSD versus SWI. We hypothesize that these two distinct categories may have specific risk factors and that recognition of patientspecific risk factors predisposing to sternal dehiscence or wound infection can help in preoperative counseling and intraoperative planning. Identifying risk factors can pinpoint specific vulnerable patient populations, with the goal to institute appropriate preventative or treatment measures to reduce the number of future wound complications.

2.

Methods

2.1.

Study design and data collections

An Institutional Review Board (IRB)-approved retrospective study of all patients who underwent cardiac surgery at CUMC between January 1, 2008 and December 31, 2013 was conducted. Patient records were accessed through the Department of Surgery Helix Database, which lists all surgical procedures and complications, WebCis, an online medical records system, and the Physicians Assistant Cardiothoracic Surgery Clinical Research Project, an IRB-approved clinical database of all cardiothoracic surgery patients. Data were collected on 38 perioperative variables including demographics as well as preoperative, operative, and postoperative factors (Table 1).

2.2.

Patient selection and definitions

All subjects aged 18 y who underwent median sternotomy during cardiac surgery at CUMC between January 1, 2008 and December 31, 2013 were included in the study. Subjects who

j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8

Table 1 e Patient perioperative factors. Preoperative Age >50 y Obesity (BMI >30) Sex (male, female) Race (white, Hispanic, Asian, black, other) Active endocarditis Cardiac ejection fraction <40% Cerebrovascular disease CHF COPD Creatinine >1.2 Diabetes requiring medication Hepatic failure Peripheral vascular disease Previous cardiac surgery Previous organ transplant Prior surgery this admission Shock Unstable Operative Surgical priority (urgent/emergent, elective) Total conduits* Arterial conduits* IMA grafting* Postoperative Bleeding requiring reoperation Extubation 24 h Glucose control protocol Hematocrit Renal failure requiring dialysis Respiratory failure Sepsis or endocarditis Stroke Temperature Unplanned cardiac reoperation *

Reported only for patients who underwent coronary artery bypass graft.

received immediate surgical flap closure by the plastic surgery service at the index cardiac procedure were excluded from the study to reduce confounding variables. Subjects who developed sternal wound complications requiring plastic surgery consultation for wound debridement and reconstructive closure were separated into two categories: sterile sternal wound dehiscence and SWIs. SWIs were diagnosed according to Centers for Disease Control and Prevention guidelines, which requires the presence of at least one of the following: (1) an organism isolated from culture of mediastinal tissue or fluid; (2) evidence of mediastinitis seen during operation; or (3) presence of either chest pain, sternal instability, or fever, and either purulent discharge from the mediastinum, isolation of an organism present in a blood culture, or culture of drainage of the mediastinal area [16]. Consequently, if gross purulence was specified in the operative notes at the time of sternal wound debridement or if intraoperative wound and/or bone cultures grew moderate or large numbers of any organism including known skin contaminants, the wound was considered a SWI. Alternatively, if intraoperative cultures grew “no”, “few,” or “light” number of organism and there was no description of purulence in the operative notes, then the wound was considered a sterile sternal dehiscence.

3

2.3. Surgical preparation, antibiotics, and methods of sternotomy fixation The perioperative protocol for cardiac patients at our institution has been described previously [9]. Preoperative preparation of the patient includes chlorhexidine scrub and nasal bactroban to nares beginning 2 d before surgery. Excess hair is removed with a hair clipper in the operating room. The patient’s skin is disinfected with 0.5% chlorhexidine (5 mg/dL) or povidone-iodine (7.5%). Standard perioperative antibiotic prophylaxis was given intravenously starting 30 to 60 min before skin incisions. The standard method of closure of the sternum after sternotomy at our institution is characterized by five sutures of monofilament surgical steel passed through the sternum approximately 1 cm lateral to the sternotomy cut on each side. If the sternum was unusually narrow, osteoporotic, mishandled with sternum retractors, broken, or a paramedian sternotomy was created, the sternum was fixed with the Robicsek lateral support technique [15]. After sternal fixation, the wound is closed in three layers.

2.4.

Statistical analysis

Data were expressed as percentage, mean, and standard deviation or median and interquartile range. To compare point estimates between sterile sternal dehiscence, SWI, and no sternal wound groups, a series of Fisher exact tests, two-sample independent measures t-test, or Wilcoxon ranksum tests were used with Bonferroni corrections applied to maintain family error rates of a ¼ 0.05. To determine the unadjusted and adjusted significant predictors and their corresponding odds ratios of sterile sternal dehiscence and SWI, univariate logistic regression and multivariate logistic regression analyses, respectively, were conducted. All significant predictors corresponding to P values <0.05 in the univariate analyses were included in the multivariate logistic regression analyses, which were performed using stepwise models with alpha-to-enter and alpha-to-remove of 0.05. All analyses were conducted using the Minitab 17 statistical software package (Minitab Inc., State College, PA).

3.

Results

Of the 8098 patients included in this study, 113 patients (1.4%) developed sternal wounds (Table 2). Stratifying by sternal wound type, 73 patients (0.9%) and 40 patients (0.5%) developed SSDs and SWIs, respectively. Compared to the no sternal wound group, patients who developed SSDs were found to be significantly older (95.6% versus 84.4% were age >50 y; P ¼ 0.005) and more obese (body mass index >30; 54.2% versus 27.2%; P < 0.001) and were more like to have had diabetes requiring medication (53.5% versus 29.6%; P ¼ 0.001), prior surgery this admission (11.0% versus 2.9%; P ¼ 0.002), and unplanned cardiac reoperation (20.4% versus 5.2%; P < 0.001; Table 3). In contrast, patients who developed SWI were more likely to have had a cardiac EF <40% (42.1% versus 23.9%, P ¼ 0.013), an urgent or emergent procedure (65.5% versus 37.8%; P ¼ 0.003), a greater frequency of internal mammary artery (IMA) grafting (60.0% versus 37.0%, P < 0.001), and a

4

j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8

Table 2 e Patient demographics. Demographics

Overall (n ¼ 8098)

No sternal wound (n ¼ 7985)

SSD (n ¼ 73)

SWI (n ¼ 40)

65.6  15.7

65.4  15.7

66.2  10.4

66.6  12.9

27.5  5.6

27.5  5.6

30.7  6.2

29.7  7.3

64.0% 36.0%

64.0% 36.0%

74.0% 26.0%

52.5% 47.5%

72.1% 8.8% 3.5% 10.7% 4.9%

72.2% 8.8% 3.5% 10.7% 4.9%

64.4% 12.3% 5.5% 13.7% 4.1%

70.0% 5.0% 0 17.5% 7.5%

Age Overall (mean  SD) BMI Overall (mean  SD) Sex Male Female Race White Hispanic Asian Black Other

lower postoperative hematocrit (26.9  3.9 versus 28.5  4.6; P ¼ 0.014) compared to the no sternal wound group. Additionally, both the SSD and SWI groups were found to have a greater median number of total conduits and median number of arterial conduits and more frequent postoperative respiratory failure (47.2% versus 18.6%; P < 0.001) than the no sternal wound group. No significant differences between the groups with respect to sex or race were seen. Univariate logistic regression analysis of SSD revealed eight positive predictors and two negative predictors to be significant (Table 4). Positive predictors included obesity (3.15 [1.98e5.03]), congestive heart failure (1.72 [1.06e2.78]), diabetes requiring medication (2.59 [1.34e5.01]), prior surgery this admission (4.02 [1.91e8.49]), 2 arterial conduits (2.05 [1.21e3.46]), IMA grafting (1.84 [1.16e2.92]), respiratory failure (3.86 [2.24e6.66]), and unplanned cardiac reoperation (4.63 [2.28e9.38]); negative predictors included no arterial conduits (0.43 [0.21e0.85]) and extubation 24 h (0.57 [0.34e0.94]). When adjusted in multivariate regression with stepwise model builder that included all predictors with P values <0.05,

the significant predictors were obesity (2.14 [1.12e4.15]), diabetes requiring medication (2.59 [1.34e5.01]), and respiratory failure (5.02 [2.58e9.74]). Univariate logistic regression analysis of SWI determined five positive predictors and three negative predictors to be significant (Table 5). Positive predictors included obesity (2.17 [1.16e4.05]), cardiac EF <40% (2.32 [1.21e4.42]), urgent and/or emergent case (3.10 [1.44e6.69]), IMA grafting (2.54 [1.35e4.79]), and respiratory failure (2.93 [1.36e6.34]); negative predictors included no arterial conduits (0.43 [0.21e0.85]), elective case (0.32 [0.15e0.69]), and postoperative hematocrit (0.92 [0.86e0.99]). The significant predictors on multivariate analysis were IMA grafting (3.24 [1.30e8.05]) and respiratory failure (3.89 [1.56e9.73]). Finally, there were no significant differences in death or survival rates found between SSD, SWI, and no sternal wound groups (Table 6). Of note, the rate of death for patients who did not develop a sternal wound compared with that for patients who developed an SSD in both the overall (P ¼ 0.057) and hospital groups (P ¼ 0.032) did not achieve

Table 3 e Significant perioperative factors in SSD and SWI groups compared to no sternal wound. Perioperative factor Preoperative Age >50 y Obesity (BMI >30) Cardiac EF <40% Diabetes requiring medication Prior surgery this admission Operative Elective case Urgent/emergent case Total conduits, median (IQR) Arterial conduits, median (IQR) IMA grafting Postoperative Hematocrit (mean  SD) Respiratory failure Unplanned cardiac reoperation *

Overall

No sternal wound

Sterile sternal dehiscence

SWI

84.5% 27.5% 24.0% 29.9% 3.0%

84.4% 27.2% 23.9% 29.6% 2.9%

95.9%* (P < 0.001) 54.2%* (P < 0.001) 25.7% 53.5%* (P ¼ 0.002) 11.0%* (P ¼ 0.002)

85.0% 45.0% 42.1%* (P ¼ 0.013) 39.1% 7.5%

61.9% 38.1% 1 (0e3) 0 (0e1) 37.3%

62.2% 37.8% 0 (0e3) 0 (0e1) 37.0%

48.9% 51.1% 2 (0e3)* (P ¼ 0.003) 1 (0e2)* (P ¼ 0.005) 52.1%

34.5 %* (P ¼ 0.002) 65.5%* (P ¼ 0.002) 2 (0e3)* (P ¼ 0.010) 1 (0e2)* (P ¼ 0.010) 60.0%* (P < 0.001)

28.5  4.7 19.1% 5.4%

28.5  4.6 18.6% 5.2%

27.4  4.0 47.2%* (P ¼ 0.010) 20.4%* (P ¼ 0.008)

26.9  3.9* (P ¼ 0.014) 40.7%* (P < 0.001) 15.4%

Denotes statistical significance compared to no sternal wound group.

5

j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8

Table 4 e Univariate and multivariate logistic regression analyses of sterile sternal dehiscence. Parameter

P value

Univariate regression Positive predictors Obesity (BMI >30) CHF Diabetes requiring medication Prior surgery this admission 2 arterial conduits IMA grafting Respiratory failure Unplanned cardiac reoperation Negative predictors No arterial conduits Extubation 24 h Multivariate regression Positive predictors Obesity (BMI >30) Diabetes requiring medication Respiratory failure

<0.001 0.029 0.006 0.002 0.011 0.010 <0.001 <0.001

OR (95% CI)

3.15 1.72 2.59 4.02 2.05 1.84 3.86 4.63

(1.98e5.03) (1.06e2.78) (1.34e5.01) (1.91e8.49) (1.21e3.46) (1.16e2.92) (2.24e6.66) (2.28e9.38)

0.013 0.030

0.43 (0.21e0.85) 0.57 (0.34e0.94)

0.025 0.006 <0.001

2.14 (1.12e4.15) 2.59 (1.34e5.01) 5.02 (2.58e9.74)

statistical significance at the Bonferroni-corrected alpha level.

4.

Discussion

As the incidence of cardiovascular heart disease rises with the aging population, even a small percentage of sternal wound complicationsd0.5%e5% in literaturedcan amount to a significant health care burden given the sheer volume of median sternotomies performed each year. Although both SSD and infected sternal wounds are managed surgically in similar procedures, each has been recognized as a unique complication: deep SWIs and mediastinitis carry a mortality rate of up to 50%, whereas SSD patients generally fare better, with their primary complaints related to morbidity and reduction in quality of life, with chest wall discomfort and

Table 5 e Univariate and multivariate logistic regression analyses of SWI. Parameter Univariate regression Positive predictors Obesity (BMI >30) Cardiac EF <40% Urgent/emergent case IMA grafting Respiratory failure Negative predictors No arterial conduits Elective case Hematocrit Multivariate regression Positive predictors IMA grafting Respiratory failure

P value

OR (95% CI)

0.018 0.014 0.003 0.004 0.009

2.17 2.32 3.10 2.54 2.93

(1.16e4.05) (1.21e4.42) (1.44e6.69) (1.35e4.79) (1.36e6.34)

0.013 0.003 0.022

0.43 (0.21e0.85) 0.32 (0.15e0.69) 0.92 (0.86e0.99)

0.009 0.008

3.24 (1.30e8.05) 3.89 (1.56e9.73)

Table 6 e Discharge information. Discharge disposition Died Overall Operating room Hospital Survived Hospital discharge 30 d

Overall (%)

No sternal wound (%)

SSD (%)

SWI (%)

5.3 0.2 5.1

5.3 0.2 5.1

11.0 0 11.0

2.5 0 2.5

94.7 91.1

94.7 91.0

89.0 95.2

97.5 100.0

pulmonary dysfunction limiting their activities of daily living. Our study is the first published study to attempt to clarify distinct risk factors which may separate SSD from its infected counterpart. In our retrospective review of over 8000 patients, there were statistically significant demographic and perioperative differences between patients who did not suffer postoperative wound complications with those who did suffer either SSD or SWI. Perhaps not surprisingly, in univariate analysis, the demographic and perioperative factors between the SSD and SWI cohorts did not significantly differ, suggesting that similar predisposing risk factors may contribute to the development of any wound complication. The sacrifice of the internal mammary arteries for coronary artery bypass grafting has been established as a risk factor for sternal wound complications in existing literature, and our study continues to support this notion; IMA grafting is a significant risk factor for SWI in the present study [5,7e13]. The internal mammary arteries contribute significant blood supply to the sternum and its sacrifice results in sternal ischemia. The resulting vascular insufficiencies can blunt oxygen delivery to and curb debris clearance from a fresh surgical incision leading to impaired wound healing. In our SSD cohort, although the use of IMA grafts was not a significant risk factor, obesity was a particularly strong predictor for sterile dehiscence, given its statistical significance in both the univariate and multivariate models. Obesity has long been associated with an increased incidence of surgical complications, including atelectasis, mortality, thrombophlebitis, wound dehiscence, and wound infections [17]. Its specific detrimental effect on wound healing exists through a myriad of mechanisms, including local tissue ischemia. From a microscopic perspective, adipose tissue is also relatively avascular; as adipose tissue increases in obesity, angiogenesis does not proceed proportionally [21]. Adiposity, in conjunction with diabetes, which contributes to microvascular disease, compounds local vascular insufficiency, predisposing to wound complications. Structurally, adipose tissue is composed of lobules each supplied by a terminal capillary. Disruption of an end capillary results in the fat necrosis of the entire lobule, increasing dead space and predisposing to seroma formation [18e20]. Several studies have also demonstrated that adipose tissues exhibit altered immune mediators and extracellular matrix remodeling proteins, which may prolong the inflammatory stage of wound healing. From a macroscopic perspective, we propose that the sheer weight of increased soft tissue can add undue mechanical

6

j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8

stress on chest closure, leading to increased movement between the sternal halves hindering sternal union, and disrupting epithelialization and angiogenesis of healing soft tissues, increasing the risk of wound breakdown. The macroscopic concept of increased mechanical stress leading to sternal wound complications is not new. Currently, there is literature to support that comorbidities such as COPD may predispose to sternal dehiscence because of its mechanical forces on sternal closure through increasing chest wall motion [15]. Celik et al. retrospectively compared 842 patients (328 with moderate to severe COPD, 514 without COPD), all of whom received the conventional figure-of-eight peristernal closure technique and found that those with COPD suffered significantly higher rates of sternal dehiscence than those without COPD (7.9% versus 1.2%) with a higher prevalence of sternal dehiscence in those patients with more severe COPD (3.5% of those with FEV1 60%e70% versus 45.4% among those with FEV1 of 30%e40%). They also prospectively randomized 221 patients with moderate to severe COPD who received the Robicsek reinforced osteosynthesis technique into two groupsdthose who received a thoracic support vest (Posthorax) and those who did not. Results showed a significant decrease in postoperative sternal dehiscence and infections with the use of a thoracic support vest (1% versus 11.5%). Although COPD confers increased risk to a variety of surgical complications, the decreased incidence of dehiscence seen with institution of the Posthorax vest supports our hypothesis that mechanical stress, days or weeks after median sternotomy, can lead to sternal instability; respiratory movements or cough can lead to loosening of sternal wires, causing sternal halves to separate leading to dehiscence. Chronic obstructive pulmonary disease did not emerge as a significant risk factor for sternal wound complications in our analysis, and this may be due to our study limitation where severity of COPD was not categorized. In the above study by Celik et al., only moderate to severe COPD patients were included. Thus, the significance may be dampened in our cohort if a larger percentage of our patients with COPD had mild disease. Our analysis found postoperative respiratory failure to be a significant risk factor for both SSD and SWI. Positive pressure ventilation may increase mechanical forces on sternal closure. Positive pressure ventilation introduces stress on the chest wall uniquely different from normal physiological negative pressure breathing. Increased respiratory forces may contribute to micromotion across sternotomy halves. There are no present studies quantifying chest wall movement during mechanical ventilation versus physiological breathing. This may be an area of future study and may lend credence to our hypothesis. Nevertheless, sternal stability is established as a key factor in prevention of sternal dehiscence and infection. The best technique to achieve stable sternal closure, however, is controversial, and there currently exists no unified algorithm for sternal closure, especially in high-risk patients. Conventional closure of the sternum includes six or more wire sutures passed transsternally or peristernally either in single sutures or in a figure-of-eight technique, with literature showing equivocal efficacy in achieving sternal stability [22]. Some surgeons endorse the Robicsek technique, using

continuous, parasternal atraumatic wire sutures which may achieve greater sternal stability but the literature remains divided. Celik et al. treated every patient in their prospective cohort with the Robicsek closure, but sternal wound complication rates were not decreased compared to their retrospective controls who received standard closure (11.5% versus 7.9%) [23]. Similarly Schimmer et al., in their prospective randomized multicenter trial of 815 consecutive patients, found no significant differences in sternal dehiscence or infection between the Robicsek technique and the single wire suture technique [24]. However, many other surgeons, such as Molina and Sharma, have found success using the Robicsek technique, especially in high-risk populations [25,26]. Finally, there is growing support for rigid sternal plate fixation for improved sternal stability as compared with wires [27e29]. Raman et al. advise rigid sternal plating in high-risk patients for primary closure of the sternum after cardiac surgery. Enhanced sternal stability may improve sternal healing and decrease the incidence of sternal wound dehiscence and infections. In their study, Raman et al. found a decrease in deep SWIs in those who received rigid plating compared with those who were closed with wire (0% versus 14.8%). This may be an operative consideration for those patients who have multiple high-risk comorbidities such as obesity and COPD. However, rigid fixation has its own caveats: more difficulty on emergent reentry, increased cost of plating systems, and potential danger of drilling around the heart and great vessels. To circumvent the risks for internal stabilization techniques, external stabilization devices such as the Posthorax vest have gained some favor as mentioned above. Further support for the use of supportive garments includes the prospective randomized multicenter trial performed by Gorlitzer et al., where 1351 of 2539 patients were randomized to receive a thoracic support vest. Patients did not differ significantly in preoperative demographic factors or comorbidities. They found that use of a posthorax supportive vest decreased deep sternal wound complications significantly (1.04% versus 2.27%). This study did not specify closure technique and did not differentiate between SSD versus deep SWI. However, this study does support the hypothesis that the Posthorax vest, or any external stabilization device, can aid in sternal stability in the postoperative period, decreasing healing difficulties. In their subgroup analysis, those patients who refused to wear thorax support vest were at higher risk for deep SWIs, supporting our hypothesis that increased sternal instability and subsequent sternal dehiscence can predispose to sternal infection complications.

4.1.

Limitations

Even though our study represents the largest published population of consecutive median sternotomy patients at a single institution to date, the incidence of wound complications remains low and thus, some risk factors which did not reach significance in our study could potentially approach significance if placed in a larger patient pool. Further analysis from pooled data with other centers may be warranted. In addition, smoking status and steroid use were not analyzed in our study because of insufficient data; these factors are known to

j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8

significantly impact wound healing and may alter the significance if one cohort had a significantly higher percentage of smokers or chronic steroid users. Inherent to retrospective studies, temporal relationships are difficult to assess. That is, if SSD and SWI exist on a continuum, our study would not capture the evolution of a wound complication as patients were categorized into SSD and SWI based on their presentation at a singular time point. Further stratification of time of presentation with type of wound complication may be helpful in clarification of differences between SSD and SWI.

5.

Conclusion

This study represents the largest published cohort of consecutive median sternotomy patients to date at a single institution and aims to examine if SSD and SWIs have distinct risk factor profiles. Our analysis found that although there are unique differences, SSD and SWIs do have a great degree of commonality, suggesting that although they may be distinct entities, these two entities may be related. Identification of risk factors for these sternal wound complications may help identify patients at risk and may help guide operative planning and postoperative management to decrease incidence of median sternotomy wound complications.

Acknowledgment Thanks to Tianna Umann for valuable assistance throughout the data collection. Authors’ contributions: The study was designed and coordinated by C.H.R., J.A., and M.A.; as principal investigator, C.H.R. oversaw and provided guidance throughout all aspects of the study from conception to manuscript submission. C.H.R., R.H.F, and M.M.C. were responsible for acquisition of data and in conjunction with author A.L.W., undertook data analysis and review. All authors collaborated in data interpretation. R.H.F. and A.L.W. drafted the initial manuscript with critical revisions from the remainder of the authors.

Disclosure The authors have no conflicts of interest to declare in relation to the content of this article.

references

[1] Agarwal JP, Ogilvie M, Wu LC, et al. Vacuum-assisted closure for sternal wounds: a first-line therapeutic management approach. Plast Reconstr Surg 2005;116:1035. discussion 1041e3. [2] Ascherman JA, Patel SM, Malhotra SM, Smith CR. Management of sternal wounds with bilateral pectoralis major myocutaneous advancement flaps in 114 consecutively treated patients: refinements in technique and outcomes analysis. Plast Reconstr Surg 2004;114:676.

7

[3] Etzioni DA, Starnes VA, Chapter 2: the epidemiology and economics of cardiothoracic surgery in the elderly. 1 ed. Cardiothoracic Surgery in the Elderly. 2011: Springer-Verlag New York. [4] CMS Never Events Research Report. Real life solutions. Available at: www.aha-solutions.org/.../pdf.../rlsolutions2014-0603-wp-cmsnever.pdf. Updated June 3, 2014. Accessed March 24, 2015. [5] Stoney WS, Alford WC Jr, Burrus GR, Frist RA, Thomas CS Jr. Median sternotomy dehiscence. Ann Thorac Surg 1978;26: 421. [6] Vymazal T, Hora´cek M, Durpekt R, Hladı´kova´ M, Cvachovec K. Is allogeneic blood transfusion a risk factor for sternal dehiscence following cardiac surgery? A prospective observational study. Int Heart J 2009;50:601. [7] Hazelrigg SR, Wellons HA Jr, Schneider JA, Kolm P. Wound complications after median sternotomy. Relationship to internal mammary grafting. J Thorac Cardiovasc Surg 1989; 98:1096. [8] Zacharias A, Habib RH. Factors predisposing to median sternotomy complications. Deep vs superficial infection. Chest 1996;110:1173. [9] Loop FD, Lytle BW, Cosgrove DM, et al. J. Maxwell Chamberlain memorial paper. Sternal wound complications after isolated coronary artery bypass grafting: early and late mortality, morbidity, and cost of care. Ann Thorac Surg 1990; 49:179. discussion 186e7. [10] Cayci C, Russo M, Cheema FH, et al. Risk analysis of deep sternal wound infections and their impact on long-term survival: a propensity analysis. Ann Plast Surg 2008;61:294. [11] Lazar HL, Fitzgerald C, Gross S, Heeren T, Aldea GS, Shemin RJ. Determinants of length of stay after coronary artery bypass graft surgery. Circulation 1995;92(9 Suppl l): II20. [12] Roy MC. Surgical-site infections after coronary artery bypass graft surgery: discriminating site-specific risk factors to improve prevention efforts. Infect Control Hosp Epidemiol 1998;19:229. [13] Shafir R, Weiss J, Herman O, Cohen N, Stern D, Igra Y. Faulty sternotomy and complications after median sternotomy. J Thorac Cardiovasc Surg 1988;96:310. [14] Cabbabe EB, Cabbabe SW. Surgical management of the symptomatic unstable sternum with pectoralis major muscle flaps. Plast Reconstr Surg 2009;123:1495. [15] Olbrecht VA, Barreiro CJ, Bonde PN, et al. Clinical outcomes of noninfectious sternal dehiscence after median sternotomy. Ann Thorac Surg 2006;82:902. [16] Garner JS, Jarvis WR, Emori TG, Horan TC, Hughes JM. CDC definitions for nosocomial infections, 1988. Am J Infect Control 1988;16:128. [17] Pierpont YN, Dinh TP, Salas RE, et al. Obesity and surgical wound healing: a current review. ISRN Obes 2014;2014: 638936. [18] Cottam DR, Mattar SG, Barinas-Mitchell E, et al. The chronic inflammatory hypothesis for the morbidity associated with morbid obesity: implications and effects of weight loss. Obes Surg 2004;14:589. [19] Markman B. Anatomy and physiology of adipose tissue. Clin Plast Surg 1989;16:235. [20] Gottrup F, Firmin R, Rabkin J, Halliday BJ, Hunt TK. Directly measured tissue oxygen tension and arterial oxygen tension assess tissue perfusion. Crit Care Med 1987;15:1030. [21] Khan T, Muise ES, Iyengar P, et al. Metabolic dysregulation and adipose tissue fibrosis: role of collagen VI. Mol Cell Biol 2009;29:1575. [22] Khasati N, Sivaprakasam R, Dunning J. Is the figure-of-eight superior to the simple wire technique for closure of the sternum? Interact Cardiovasc Thorac Surg 2004;3:191.

8

j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8

[23] Celik S, Kirbas A, Gurer O, Yildiz Y, Isik O. Sternal dehiscence in patients with moderate and severe chronic obstructive pulmonary disease undergoing cardiac surgery: the value of supportive thorax vests. J Thorac Cardiovasc Surg 2011;141: 1398. [24] Schimmer C, Reents W, Berneder S, et al. Prevention of sternal dehiscence and infection in high-risk patients: a prospective randomized multicenter trial. Ann Thorac Surg 2008;86:1897. [25] Molina JE, Lew RS, Hyland KJ. Postoperative sternal dehiscence in obese patients: incidence and prevention. Ann Thorac Surg 2004;78:912. discussion 912e7.

[26] Sharma R, Puri D, Panigrahi BP, Virdi IS. A modified parasternal wire technique for prevention and treatment of sternal dehiscence. Ann Thorac Surg 2004;77:210. [27] Song DH, Lohman RF, Renucci JD, Jeevanandam V, Raman J. Primary sternal plating in high-risk patients prevents mediastinitis. Eur J Cardiothorac Surg 2004;26:367. [28] Pai S, Gunja NJ, Dupak EL, et al. In vitro comparison of wire and plate fixation for midline sternotomies. Ann Thorac Surg 2005;80:962. [29] Cicilioni OJ Jr, Stieg FH 3rd, Papanicolaou G. Sternal wound reconstruction with transverse plate fixation. Plast Reconstr Surg 2005;115:1297.