Potential Risk Factors for Surgical Site Infection after Isolated Coronary Artery Bypass Grafting in a Bahrain Cardiac Centre: A Retrospective, Case-Controlled Study.

OBJECTIVE: The purposes of this study were to determine the incidence of surgical site infections (SSI) and associated risk factors in patients undergoing isolated coronary artery bypass grafting (CABG) in our cardiac center during a 2-year period.
MATERIALS AND METHODS: Retrospective case-control analysis for 80 patients who underwent isolated cardiac surgery CABG. These patients were divided into the SSI study group (n = 40) and the noninfected control group (n = 40). Eight potential perioperative risk variables were compared between the two groups using univariate logistic regression analysis.
RESULTS: Univariate analysis was carried out for eight potential risk factors. The risk factors found to be significant were: Impaired estimated glomerular filtration rate (P = 0.011) and impaired left ventricular ejection fraction (P = 0.015). However, Factors found to have no significant influence on the incidence of SSIs were: Perioperative length of hospital stay (days), urgency of surgery, use of bilateral internal mammary artery (BIMA) grafting, prolonged cardiopulmonary bypass duration, elevated body mass index.
CONCLUSIONS: Patients with comorbidities of impaired renal function and/or impaired left ventricular systolic function are at high risk of developing SSI. There appears to be a relationship between SSIs in CABG patients and impaired renal or LV function (low ejection fraction). CABG with BIMA grafting could be performed safely even in diabetics. Future studies should consider further scrutiny of these and other factors in relation to SSIs in a larger surgical population.

INTRODUCTION

Surgical site infections (SSIs) are either superficial or deep and may involve the organs, or spaces accessed during an operation. The reported incidence of SSIs in coronary artery bypass grafting (CABG) surgery ranges between 0.3% and 8%.[1]

There is a strong suggestion that an impairment of vascular supply of the sternum may be one of the most important factors influencing the incidence of deep sternal wound infection (DSWI). Several studies have studied the risk factors for SSIs including DSWI in cardiac surgery. These risk factors included obesity, diabetes mellitus, chronic obstructive pulmonary disease (COPD), connective tissue disease, steroid use, smoking, peripheral vascular disease and renal insufficiency. In addition, intraoperative factors (e.g., use of bilateral internal mammary arteries [BIMA] grafting, prolonged cardiopulmonary bypass [CPB] duration) and postoperative variables (e.g., prolonged mechanical ventilation, reoperation for bleeding, postoperative transfusions and gastrointestinal, nephrological and respiratory complications) have been shown to be associated with DSWI.[234]

The risk for sternal wound infection (SWI) is increased if cardiac surgery involves internal thoracic arteries grafting and a valve procedure, or use of a ventricular assist device.[56] Leg wound infections at donor sites account for >70% of cases with severe infection following cardiac surgery.[7]

Cardiac SSIs increase the length of hospital stay (LOS) and increase treatment costs in proportion to the severity of the infection. These costs increase by 3.8%, 14.7% and 29.4% in mild, moderate and severe infections respectively.[7]

Treatment is often confounded by the emergence of antibiotic-resistant pathogens and in addition, substantial proportions of these infected patients are elderly and have co-existing medical problems. In the past, such elderly patients with significant comorbidities would not have been considered for surgery.[8] As the population ages, it is reasonable to assume that older and sicker patients will be admitted for surgery, and this will inevitably increase the risk and incidence of SSIs.[9]

Within our institution, the infection rate of postoperative wounds has been under surveillance by the infection control team since 2005. According to our preliminary data of rates of SSIs among isolated CABG patients, this was found to be significantly disproportionate to other regional institutions. As a result, we decided to audit, study and identify likely perioperative risk factors among our patients who have undergone isolated CABG between 2012 and 2013.[10]

The most important step in the management of wound infection is prevention, and this is best done by identifying risk factors. The present study was carried out in our centre to identify the incidence of wound infections following isolated CABG and identify the risk factors that may be associated with SSIs in our center.

None of these perioperative risk factors have been studied previously in our population undergoing isolated CABG. Our pool of patients originates from a general Middle Eastern population, known to have a high prevalence of diabetes and obesity, a sedentary lifestyle with a lack of exercise, which represents a change of lifestyle following the discovery of oil in the region.[1011121314]

MATERIALS AND METHODS

Definitions of Infection

Sternal SSI was defined according to the SSI criteria of the US Centers for Disease Control and Prevention. Sternal infections occurring within 30 days after surgery can include the following types: (1) Superficial incisional (infection above the sternum with no bony involvement); (2) deep incisional (infection involving the sternum); and (3) organ/space (site-specific infection such as mediastinitis).[815]

Leg SSI was defined as redness, swelling, increased pain, excessive bleeding or discharge at the incision site among the patients who had undergone CABG. All SSI cases were diagnosed by attending physicians and confirmed by the nosocomial infection control committee. Patients who did not have any SSI formed the control group.

Study Design

From January 2012 to December 2013, 357 isolated CABG procedures were performed at our cardiac center. Totally, 40 postoperative patients diagnosed with an SSI (including sternal SSI, leg SSI and double SSI [both sternum and leg]), in accordance with the CDC criteria for SSI surveillance, were selected randomly. These 40 patients formed our study group (SSI) (n = 40, group I). This group was matched according to age, sex, nature of procedure and timing within the above study period with a control group (non-SSI) of 40 postoperative patients who did not suffer from an SSI (non-SSI group: n = 40, group II).

The eighty selected CABG patients' data were collected and analyzed retrospectively. Eight potential risk variables were compared between groups I and II.

Potential Risk Factors

Eight possible perioperative risk factors were analyzed and included the following: A prolonged LOS (LOS by days) which was arbitrarily taken as beyond a 30 days stay, a previous diagnosis of diabetes mellitus, impaired estimated glomerular filtration rate (eGFR < 60 ml/min) taken as a sign of renal insufficiency, urgency of surgery (i.e. surgery done within 24 h of diagnosis of surgical coronary artery disease), the use of BIMA for grafting, impaired ejection fraction (EF) including moderate and severe (EF% <45%), prolonged CPB duration taken as >2 h and an elevated body mass index (BMI) that is, >25.

Data Analysis

The risk factors for infection were assessed by univariate analysis. Discrete variables were assessed using Chi-squared analysis or Fisher's exact test. Variables were assessed using two tailed Student's t-test. All variables suggested by the univariate analysis were entered into a stepwise binary logistic regression analysis model. The chosen level of significance was 5%. All analysis was performed using the Statistical Package for the Social Sciences (SPSS) 19.0. IBM Corporation.

RESULTS

All of the 80 patients who enrolled during the study period underwent isolated CABG only.

Among the eight potential risk factors studied, the factors that had significant differences between the SSI study group and non-SSI control group were an impaired eGFR (P = 0.011, odds ratio [OR]: 3.8) and an impaired EF% (P = 0.015, OR: 5.1) [Tables 1 and 2].

Table 1

Association of exposure with SSI

Table 2

Potential preoperative risk factor for cardiac surgical infection for underwent CABG, 2012-2013*

Patient's LOS (days), urgency of surgery, BIMA grafting, prolonged CPB duration and an increased BMI had no significant influence on the incidence of wound infection [Table 1].

Of the 40 patients in the SSI group, 22 were male, and 18 were female that was equivalent to the non-SSI group. The mean age for the SSI group was 59 versus 61 for the non-SSI group. There was no significant statistical difference in the age and sex of both groups.

Only 2 patients in the SSI group stayed in the hospital beyond 30 days. None of the non-SSI group had a prolonged LOS. The difference was statistically insignificant.

Thirty-five patients (87%) in the SSI group were previously diagnosed with diabetes mellitus (type 1 or type 2) while 30 patients (75%) in the non-SSI group were diabetic. Hence, diabetes was not found to be a preoperative predictor of SSIs (P - 0.156).

Urgent isolated CABG was performed on 6 patients (15%) in the SSI group compared to only 4 patients in the non-SSI group. There was no statistically significant difference with a P = 0.499 and an OR: 1.5.

Seventeen patients (42%) in the SSI group had a prolonged CPB time of more than 2 h duration while 13 (32%) of patients in the non-SSI group had a prolonged CPB time. Again, prolonged CPB duration was not found to be an intraoperative predictor of SSI (P - 0.356).

Bilateral internal mammary harvesting was performed on 6 patients in the SSI group while 11 patients in the non-SSI group had bilateral mammary harvesting performed. Bilateral mammary harvesting was not found to be a risk factor for SSI with a P - 0.1/OR: 2.1.

DISCUSSION

Surgical site infections are a manifestation of an imbalance between microbial growth and host's defenses. The Surgical stress response imposes an impairment of these defenses.[16]

Loop et al., described several risk factors for sternal wound complications in cardiac surgery. Bilateral internal mammary harvesting, diabetes, obesity, blood transfusion and operative time were considered significant risk factors for sternal wound complications.[17] Other authors had described other risk factors for SSIs in cardiac surgery, with conflicting findings.

Preoperative hospital admission duration, antibiotic prophylaxis use, surgical urgency, reoperation, surgical time, CPB duration, amount of blood transfused, postoperative blood loss, chest re-exploration, rewiring of a sterile sternal dehiscence, duration of mechanical ventilation and days of treatment in the intensive care unit were described as other perioperative factors contributing to the development of SSIs.[18]

Our cardiac center had performed 377 isolated CABG procedures over the designated study period of which, 67 patients developed an SSI. This was an incidence rate of 17%. This elevated incidence of infection prompted us to perform this study in an attempt to analyze perioperative surgical and patient risk factors that might contribute to this rate.

Following a literature review, we decided to study eight variables [Table 2] in our cohort of patients who originate from a population, known to have a high prevalence of prediabetes, diabetes mellitus, obesity and pursues a modern, sedentary lifestyle.[10] The Bahraini population has become increasingly modernized, over the last 40 years, resulting in a transformation from an active lifestyle to one that lacks physical activity, sunlight exposure and has acquired unhealthy dietary patterns. These social factors have led to a higher prevalence of chronic obesity, insulin resistance, prediabetes, and type 2 diabetes.[1112131419]

We found that an impaired renal function and/or impaired left ventricular ejection fraction (LVEF) are statistically significant patient characteristics for acquiring SSIs. Interestingly, other risk factors like diabetes, bilateral mammary harvesting and an elevated BMI were found to be statistically insignificant risk factors.

Sakamoto et al. had previously concluded that patients in a poor perioperative condition, that is, in a poorly perfused state and requiring hemodynamic supportive devices, were more likely to develop DSWI.[20] We believe, having an impaired LVEF could be considered a marker for poor tissue perfusion in the perioperative period. Since our study did not identify those patients who had had the aid of hemodynamic support devices, we think these patients would be included in our cohort that had a severely impaired EF. A moderate or severely impaired EF can lead to a state of generalized poor perfusion, which would hinder wound healing. Hence, we correlate our findings with those of Sakamoto et al.

Renal failure was found by some authors, to be a significant risk factor for mediastinitis and hemorrhage after cardiac surgery and isolated CABG.[62021] We studied patients with an impaired preoperative estimated GFR, that is, chronic kidney disease (CKD) stage 2 or more; an eGFR < 60 ml/min. We found this to be a statistically significant risk factor. CKD impairs immunity and the healing process through hyperuremia, presence of anemia of chronic disease, previous multiple blood product transfusions and a long-term indwelling dialysis catheter. Such catheters can harbor microorganisms that become a source of infection.

Several authors have identified obesity/overweight (BMI > 25) as a major risk factor.[62223] Surprisingly, our study found that obesity and being overweight were not statistically significant risk factors. This could be because of the high prevalence of obesity in both our study and control groups (87% and 72% respectively) which is a reflection of its high prevalence in the general population.[10111213] In Bahrain, a study by Hubail and Culligan showed that the prevalence of a BMI ≥25.0 kg/m2 was 56.4% in males and 79.7% in females among the general Bahrain population.[19] Obesity is a known modifiable risk factor for coronary artery disease. The high prevalence of obesity and the fact that our study is a study of coronary artery disease might have led to a selection bias. Both of these factors might have affected our study finding regarding obesity.

We conducted an extensive literature review with respect to diabetes and the incidence of SSIs.[24] We found that the diabetes is one of the major risk factors for post CABG SSIs. The increased infection rate in diabetes has been attributed to the impairment of neutrophil chemotaxis, phagocytosis, adherence plus the glycosylation of collagen matrix proteins - all of which lead to weakened antibacterial defenses and delayed wound healing.[14] We studied the preoperative diagnosis of diabetes regardless of type and control, anticipating a correlation between being diabetic and acquisition of an SSI. However, our results showed no correlation between a preoperative diagnosis of diabetes and CABG SSI. This again was surprising, and we assume this would be because of the high prevalence of diabetes in both our study and control groups (87.5% and 75%, respectively). Studies of diabetes in Bahrain indicate prevalence rates of 25.8% in males and 36.4% in females. These are considered to be among the highest in the world.[19] In addition, we practice tight intraoperative and postoperative glucose control which may have contributed to this lack of correlation. Hence, we suggest that further regional studies should focus rather on the preoperative control of diabetes, for example correlating the preoperative level of HBA1C with CABG SSIs.

The role of surgical urgency as a risk factor for developing SSI in cardiac surgery is controversial. Sakamoto et al. studied surgical urgency and found it to be a significant factor for DSWIs,[20] while Ku et al.,[25] could not identify such a correlation. We looked at surgical urgency as a risk factor, defining it for isolated CABG as the performance of CABG within 24 h of diagnosis of coronary artery disease that required surgical intervention and/or unscheduled CABG performed out of normal working hours in our center. About 15% of our study group were done as an emergency versus 10% in the control group. Our univariate analysis could not show a correlation between urgency of surgery and SSI. This may be due to that our preoperative preparation for these urgent cases is very similar to that of elective CABG, that is, bathing, chest hair shaving, and iodine preparation. In addition, preoperative hemodynamic stabilization with assist devices (e.g., Intra-Aortic Balloon Pump (IABP)) and medical management allows time for adequate preoperative preparation. Another possible explanation is our use of an identical antibiotic prophylaxis regimen for all cardiac surgical cases, elective or urgent.

Kouchoukos et al.[26] and Grossi et al.[27] reported that the use of bilateral mammary grafting in isolated CABG did significantly increase the incidence of SWIs; attributed to the diminished blood supply to the sternum resulting in impaired healing. Saso et al.[28] found a reduction in SWIs in isolated CABGs with skeletonized bilateral internal mammary harvest rather than pedicled harvests. The practice in our center is pedicled dissection of internal mammary harvests. We found no significant relationship between bilateral mammary harvesting and SSIs. We did not specifically address deep versus superficial SWIs but rather SSIs in general. Another limitation is the univariant correlation analysis performed by our study that did not investigate the concordance of diabetes and bilateral mammary harvesting on SSI incidence.

Cardiopulmonary bypass causes immunosuppression. The lungs' role of macrophageal scavenging is bypassed, in addition to the release of immunosuppressive immunomodulators.[29] We studied the relationship between CPB duration and SSIs, we arbitrarily considered 2 h to represent a prolonged duration of exposure to CPB. Sakamoto et al. and Minohara et al. studied CPB duration and SWIs and found no relation.[2030] Our study concurs with their findings. About 42% of our study group had a prolonged CPB, with 32% in the control group. This was found to be insignificant. Our explanation; that CPB-induced immunomodulation might be event-related rather than time-dependent. CPB triggers the complement cascade and activates cytokines like C3a and TGF-β1.

Some studies have shown a relationship between prolonged perioperative hospital stay and SSIs.[313233] We looked at prolonged perioperative hospital stay (>30 days) and found only 2 patients from our study group had a prolonged hospital stay. However, controversy remains regarding the role of a prolonged perioperative hospital stay.

CONCLUSION

In our Middle Eastern population, we found that a poor preoperative clinical condition manifested as a moderately or severely impaired left ventricular and/or renal impairment were significant preoperative risk factors for acquiring SSIs in patients undergoing isolated CABG.

Interestingly, obesity, diabetes, bilateral mammary harvesting, prolonged CPB time, urgency of surgery and length of perioperative hospital stay were found to be statistically insignificant risk factors in this small study.

We recognize limitations in our study that include the small sample size and the univariant analysis of individual known risk factors. We did not investigate the possibility of the concordance of these factors. Our study also omitted the investigation of other known risk factors, e.g., blood transfusions, smoking and other comorbidities like COPD and hypertension.

We suggest that future regional studies should carry out multivariant analysis on these and other risk factors in a larger cohort. We postulate that several factors do play a role in the acquisition of post CABG SSIs and possibly have an additive/cumulative effect on the incidence of SSIs.