If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Cardiothoracic and Vascular Surgery, Örebro University Hospital, Örebro, SwedenFaculty of Medicine and Health, School of Medical Sciences, Örebro University, Örebro, Sweden
Faculty of Medicine and Health, School of Medical Sciences, Örebro University, Örebro, SwedenDepartment of Laboratory Medicine, Clinical Microbiology, and Infectious Diseases, Örebro University Hospital, Örebro, Sweden
Department of Cardiothoracic and Vascular Surgery, Örebro University Hospital, Örebro, SwedenUniversity Health Care Research Centre, Faculty of Medicine and Health, Örebro University, Örebro, Sweden
Department of Cardiothoracic and Vascular Surgery, Örebro University Hospital, Örebro, SwedenFaculty of Medicine and Health, School of Medical Sciences, Örebro University, Örebro, Sweden
Sternal wound infections (SWIs) and aortic graft infections (AGIs) are serious complications after cardiac surgery. Staphylococcus aureus and coagulase-negative staphylococci are the most common causes of SWIs, whereas AGIs are less studied. AGIs may occur from contamination during surgery or postoperative haematogenous spread. Skin commensals, such as Cutibacterium acnes, are present in the surgical wound; however, their ability to cause infection is debated.
Aim
To investigate the presence of skin bacteria in the sternal wound and to evaluate their potential to contaminate surgical materials.
Methods
Fifty patients undergoing coronary artery bypass graft surgery and/or valve replacement surgery at Örebro University Hospital from 2020 to 2021 were included. Cultures were collected from skin and subcutaneous tissue at two timepoints during surgery, and from pieces of vascular graft and felt that were pressed against subcutaneous tissue. The most common bacterial isolates were tested for antibiotic susceptibility with disc diffusion and gradient tests.
Findings
Cultures from skin had bacterial growth in 48% of patients at surgery start and in 78% after 2 h, and cultures from subcutaneous tissue were positive in 72% and 76% of patients, respectively. The most common isolates were C. acnes and S. epidermidis. Cultures from surgical materials were positive in 80–88%. No difference in susceptibility was found for S. epidermidis isolates at surgery start compared with after 2 h.
Conclusion
The results suggest that skin bacteria are present in the wound and may contaminate surgical graft material during cardiac surgery.
Sternal wound infections (SWIs) and aortic graft infections (AGIs) are serious complications after cardiac and aortic surgery. The incidence of SWI has been reported to be 3.5–9.7% [
]. Surgical revision is the gold-standard procedure for treating AGIs; however, the exact surgical management is debated and the surgery is complex and demanding [
The bacteria causing these infections usually originate from patients' skin flora. The most common micro-organisms are staphylococci – both Staphylococcus aureus, including meticillin-resistant S. aureus (MRSA), and coagulase-negative staphylococci (CoNS) [
]. AGI generally occurs from either direct contamination (i.e. during surgery or during the early postoperative period before the sternal wound is completely healed) or from haematogenous spread from a distant locus. However, the importance of the contamination of grafts from the patient's skin microbiome during surgery has been insufficiently studied. In cardiac and aortic surgery patients, the time to evident symptoms of postoperative infection may be delayed up to several years after primary surgery [
]. This delay may be related to the low virulence of many common causative micro-organisms, biofilm production, and the complexity of foreign-body infections.
Antibiotic prophylaxis with β-lactam antibiotics is recommended to prevent SWIs in cardiac surgery. Cephalosporins – with or without the addition of systemic vancomycin – are generally used in cardiac surgery around the world [
]. However, in Sweden, cloxacillin is the most used and recommended prophylactic antibiotic. Usually, no changes in standard prophylaxis are made in cardiac surgery involving vascular graft material. Use of topical disinfection solutions, such as chlorhexidine in ethanol, are always used preoperatively, but decreased susceptibility in Cutibacterium acnes has been reported [
Bacterial recolonization of the skin and wound contamination during cardiac surgery: a randomized controlled trial of the use of plastic adhesive drape compared with bare skin.
]. In a previous study, we found that the addition of benzylpenicillin to cloxacillin prophylaxis was associated with a reduction in deep infections caused by CoNS and polymicrobial infections due to CoNS and C. acnes. Moreover, with the addition of benzylpenicillin, the total incidence of deep SWI decreased from 4.7% to 1.7% [
]. Since CoNS do not usually display susceptibility to benzylpenicillin due to the presence of the mecA and/or the blaZ gene, this finding warrants further research to analyse and better understand the colonization of the surgical wound by skin commensals.
The aim of the present study was to investigate the presence of skin bacteria in the surgical wound during cardiac surgery and to evaluate the ability of these bacteria to contaminate the vascular grafts implanted during routine aortic surgery procedures.
Methods
Patients
This study was approved by the Swedish Ethical Review Authority in the region of Uppsala (reference number 2020-02282). Written informed consent was obtained from all patients. Fifty patients who underwent cardiac surgery at the Department of Cardiothoracic and Vascular Surgery, Örebro University Hospital, Sweden, were enrolled in the study from September 2020 to September 2021. Inclusion criteria were coronary artery bypass graft surgery (CABG), valve replacement, or a combination of these via sternotomy and not meeting any exclusion criteria. Exclusion criteria were reoperation, aortic graft surgery, active infection or antibiotic treatment at the time of surgery, immunosuppressive treatment at the time of surgery, skin lesion or skin disease, and any antibiotic prophylaxis other than the routine prophylaxis stated below.
Antibiotic prophylaxis
Patients undergoing CABG surgery received 2 g of cloxacillin and 3 g of benzylpenicillin intravenously 10 min before skin incision, with additional doses of 2 g of cloxacillin every 2 h until wound closure. Patients undergoing valve replacement surgery or a combination of valve replacement and CABG received the same initial prophylaxis with 2 g of cloxacillin and 3 g of benzylpenicillin 10 min before skin incision. Every 2 h, additional doses of 2 g of cloxacillin and 1 g of benzylpenicillin were administered.
Sampling procedure
All patients showered preoperatively using 4% chlorhexidine soap (Descutan®; Fresenius Kabi AB, Uppsala, Sweden) at the ward on the evening before and on the morning of surgery. The patients washed their whole body, including their hair, twice during each shower. Each patient received clean patient clothes and was transported in a bed with clean bedding to the operating room. Any chest hair was shortened using an electric cutter the night before surgery. After disinfection with 0.5% chlorhexidine in 70% ethanol (Fresenius Kabi AB, Halden, Norway), the chest and abdomen were draped according to local routine (Procedure Pac; Mölnlycke Health Care AB, Karvina, Czech Republic), not including plastic adhesive draping of the sternal area. Sterile surgical gowns (Mölnlycke Health Care AB, Samutprakarn, Thailand) and sterile indicator gloves (Mölnlycke Health Care AB, Selangor, Malaysia) were worn by the surgical team.
Samples were obtained during surgery in a standardized manner by the operating room nurse or an additional surgical assistant. Samples of the skin and subcutaneous tissue were taken using rayon swabs (Copan Italia S.p.A., Brescia, Italy), which were rubbed on a 2 cm2 area on the caudal part of the sternal wound for 15 s. The first sample of the skin was taken immediately before the incision, and the first sample of subcutaneous tissue was taken as soon as possible after the incision. Identical sampling was repeated 10 min after the second dose of antibiotics, that is, 2 h after the first skin incision. At this second timepoint, a 2 × 1 cm piece of aortic polyester graft (Gelweave™; Vascutek Ltd, Terumo, UK) and a 2 × 1 cm piece of polytetrafluoroethylene (PTFE) felt (Bard® PTFE Felt; Bard Peripheral Vascular, Inc., Tempe, AZ, USA) were pressed against the subcutaneous tissue for 15 s. PTFE felt is commonly used in aortic surgery for the reinforcement of suture lines in fragile tissues. Each piece of graft and felt was divided into two equal parts using sterile scissors; the parts were then placed in separate bottles containing the broth described below. Samples from the gloves of the operating surgeon were taken after ∼2 h of surgery. While still gloved, the surgeon moved three fingers of their right hand in a circular motion on a blood agar plate (Columbia Blood Agar Base; Oxoid) supplemented with 6 ± 1% defibrinated horse blood) for 15 s. The procedure was repeated on an FAA plate (4.6% (w/v) (LAB 90 Fastidious Anaerobe Agar; LAB M, Heywood, Bury, UK), supplemented with 5% horse blood).
Microbiology
The swabs and imprints were subcultured in a standardized manner on blood agar plates (Columbia II Agar 3.9% w/v; Oxoid) supplemented with 6% defibrinated horse blood (SVA, Uppsala, Sweden) and incubated aerobically at 36 °C for 48 h and on FAA plates (Lab M) anaerobically at 37 °C for 5 days.
The samples of the aortic grafts and PTFE felts were incubated in enrichment broth (2.97% (w/v) Fastidious Anaerobic Broth, Lab M, supplemented with 1% (w/v) d-glucose) for 7 days. If no growth was detected, the broth was subcultured on an FAA plate (Lab M) for 5 days in an anaerobic environment at 37 °C. Bacterial growth was determined to the species level by means of matrix-assisted laser desorption/ionization time-of-flight mass spectrometry with a Microflex LT and Biotyper 3.1 (Bruker Daltonik, Bremen, Germany). The obtained isolates of CoNS and Cutibacterium sp. were further stored in a preservation broth (trypticase soy broth with 0.3% yeast extract and 29% horse serum) at 80 °C.
Antibiotic susceptibility testing for staphylococci was performed using the disc diffusion test for cefoxitin (30 mg), fusidic acid (10 mg), erythromycin (15 mg), clindamycin (2 mg), trimethoprim/sulfamethoxazole (25 mg), gentamicin (10 mg), norfloxacin (10 mg), ciprofloxacin (5 mg), and rifampicin (5 mg) (all antibiotic discs from Oxoid, Basingstoke, UK), with a 0.5 McFarland bacterial suspension in 0.85% NaCl on Mueller Hinton II agar 3.8% w/v plates (BD Diagnostic Systems, Sparks, MD, USA). After 16–20 h of incubation at 35 °C, the zone diameters were determined.
Antibiotic susceptibility testing by determining the minimum inhibitory concentration (MIC) for oxacillin, cefuroxime, benzylpenicillin, and cefazoline was performed using a gradient test for the C. acnes isolates at the start of surgery to determine the efficacy of each antibiotic agent. MIC was determined by means of a gradient test (Etest; bioMérieux, Marcy l’Etoile, France) on FAA plates (LAB M) with 0.5 McFarland suspensions of bacteria in NaCl and incubation at 36 °C in an anaerobic atmosphere for 24 h. Antibiotic susceptibility testing was performed and breakpoints applied according to the European Committee on Antimicrobial Susceptibility Testing (http://www.eucast.org) recommendations.
Statistical analysis
Descriptive statistics were used and are shown as percentages, standard deviations, means, and medians. Differences between subgroups were calculated using Fisher's exact test. Two-tailed P < 0.05 was considered statistically significant. Wilcoxon's signed rank test was used to compare MIC values for different antibiotic agents for the C. acnes isolates at the start of the surgery. Microsoft Excel® version 1902 was used for tables and figures. IBM SPSS® Statistics version 27 was used for data analysis and figures.
In a previous similar study at our centre performed in 2013, when benzylpenicillin was not part of the routine antibiotic prophylaxis, 54% of the patients undergoing open-heart surgery had subcutaneous bacterial growth. The sample size in the present study was chosen to detect a 50% decrease in positive bacterial samples compared with that study. To have a power of 80% (α = 0.05) to show a reduction from 54% to 27%, each group would need 48 patients. The historical population that the present study aimed to compare with consisted of 67 patients. Based on this calculation, it was decided to include 50 patients in the present study to allow for some dropouts.
Results
Fifty patients were included in the study. The mean age of the study population was 70 ± 8 years, and the majority of the patients were male (84%). The baseline characteristics of the study population are presented in Table I. The most common surgical procedure was CABG alone (N = 27), followed by CABG and aortic valve replacement combined (N = 10).
Table IBaseline characteristics of included patients (N = 50) sampled during open cardiac surgery
a 2 g of cloxacillin + 3 g of benzylpenicillin 10 min before skin incision, 2 g of cloxacillin every 2 h until wound closure.
b 2 g of cloxacillin + 3 g of benzylpenicillin 10 min before skin incision, 2 g of cloxacillin + 1 g of benzylpenicillin every 2 h until wound closure.
Table II shows the bacterial growth in the samples obtained from the skin and subcutaneous tissue of the sternal wound immediately before and after skin incision, respectively, and after 2 h and contamination from surgical materials. The most common bacterium in cultures of the skin and subcutaneous tissue was C. acnes, whereas cultures from contaminated surgical materials were most commonly S. epidermidis. The percentage of positive cultures from the skin increased from 48% at skin incision to 78% after 2 h of surgery (P < 0.01).
Table IIBacterial growth presented as number of positive cultures out of the 50 study patients
Several different bacterial isolates may be present in the same sample. Other CoNS include S. capitis, S. saccharolyticus, S. warneri, S. lugdunensis, S. haemolyticus, and S. hominis.
a Gelweave™ Grafts, Vascutek Ltd, Terumo, UK.
b Bard® PTFE Felt, Bard Peripheral Vascular, Tempe, Inc., AZ, USA.
No difference in bacterial recolonization was found between patients undergoing CABG in which cloxacillin with only one dose of benzylpenicillin was administered and patients undergoing valve replacement surgery in which additional doses of benzylpenicillin were administered throughout the surgery. Significantly more cultures were positive for C. acnes in males compared with females in the subcutaneous tissue at surgery start (P = 0.02) and after 2 h (P < 0.01). Male patients also had significantly fewer negative cultures after 2 h (P < 0.01) (Table III).
Table IIIComparison of bacterial growth in male and female patients
The MIC50 value for benzylpenicillin was 0.008 mg/L, which was lower than the MIC50 values for oxacillin at 0.125 mg/L (P < 0.01), cefuroxime at 0.064 mg/L (P < 0.01), and cefazoline at 0.250 mg/L (P < 0.01) (Figure 1).
Figure 1Minimum inhibitory concentration values determined by gradient test for 48 isolates of Cutibacterium acnes sampled on skin or subcutaneous tissue at the start of the operation. KZ, cefazoline; MIC, minimum inhibitory concentration; OX, oxacillin; PG, benzylpenicillin; XM, cefuroxime.
The disc diffusion test of the isolates of S. epidermidis showed no difference in susceptibility patterns when isolates from the start of surgery were compared with isolates obtained after 2 h and cultures from surgical materials (Figure 2). The number of isolates that were resistant to tested antibiotics were few. The ten S. epidermidis isolated at the start of the surgery were all susceptible to cefoxitin, gentamicin, rifampicin, trimethoprim/sulfamethoxazole, and norfloxacin but resistant to clindamycin in 1/10, erythromycin 3/10, and fusidic acid 2/10, respectively. Following 2 h of surgery, the isolated S. epidermidis cultures (N = 26) were resistant to clindamycin (N = 1), erythromycin (N = 2), fusidic acid (N = 4), and rifampicin (N = 1), respectively.
Figure 2Median and range of zone diameter from disc diffusion test for various antibiotics of all Staphylococcus epidermidis isolated at start of surgery (N = 10) compared to isolates sampled after 2 h of surgery (N = 26). CN10, gentamicin; DA, clindamycin; E, erythromycin; FD, fusidic acid; FOX, cefoxitin; NOR, norfloxacin; RD, rifampicin; SXT, sulfamethoxazole/trimethoprim. ∗Extreme outliers, values >3rd quartile + 3× interquartile range, or <1st quartile – 3× interquartile range.
In this study, skin commensals – predominantly C. acnes and S. epidermidis – were present in the sternal wound during cardiac surgery. Bacterial contamination was already present at the start of the surgery, and was more profound after 2 h. In addition, common surgical materials, such as vascular graft and PTFE felt, were easily contaminated by short contact with subcutaneous tissue or skin.
The fact that the surgical wound is not sterile and is predominantly contaminated by the patients' own skin bacteria is in accordance with previous studies [
Bacterial recolonization of the skin and wound contamination during cardiac surgery: a randomized controlled trial of the use of plastic adhesive drape compared with bare skin.
Wound contamination in cardiac surgery. A systematic quantitative and qualitative study of the bacterial growth in sternal wounds in cardiac surgery patients.
]. However, the fact that these bacteria are likely to contaminate surgical materials such as vascular grafts during the surgical procedure has not been previously reported, to the best of our knowledge. Both S. epidermidis and C. acnes can produce biofilm [
]. In cardiac surgery, it is debated whether postoperative infections that occur months or years after the operation are caused by contamination during primary surgery or are a result of haematogenous spread. The results of this study imply that it is possible for surgical materials to be contaminated at the initial surgical procedure and that the perioperative antibiotic prophylaxis is not sufficient to prevent colonization and completely eradicate the bacteria and their biofilm formation.
No difference in bacterial growth was found in comparison with a previous similar study at our centre performed in 2013, when benzylpenicillin was not routinely administered perioperatively [
Bacterial recolonization of the skin and wound contamination during cardiac surgery: a randomized controlled trial of the use of plastic adhesive drape compared with bare skin.
]. In the previous study, identical skin samples displayed the growth of CoNS in 6/67 (9%) patients at the start of the operation and in 15/63 (23.8%) patients after 2 h. C. acnes was present in 34/67 (50.7%) skin samples at surgery start and in 28/63 (44.4%) samples after 2 h. Cultures from subcutaneous tissue after 2 h were positive for CoNS in 4/63 (6.3%) patients and for C. acnes in 34/63 (54%) patients [
Bacterial recolonization of the skin and wound contamination during cardiac surgery: a randomized controlled trial of the use of plastic adhesive drape compared with bare skin.
Theoretically, benzylpenicillin may still reduce the capacity of bacteria to survive and cause a clinical infection. In a previous study, the addition of a single dose of benzylpenicillin to cloxacillin prophylaxis was associated with a reduction in deep SWIs [
]. In the present study, C. acnes was commonly found in the surgical wound and on implantable foreign materials. The role of C. acnes in postoperative infections in cardiac surgery has been debated, but it is recognized as a pathogen of foreign material infections in other surgical areas, such as shoulder arthroplasties [
]. Our study confirms previous research showing that C. acnes is present in the sternal wound, and there is growing evidence that it can cause postoperative SWIs [
Incidence, microbiological findings, and clinical presentation of sternal wound infections after cardiac surgery with and without local gentamicin prophylaxis.
]. Moreover, there might be synergistic mechanisms between C. acnes and other microbes, for example, related to its ability to form biofilm. The determination of MIC50 via a gradient test of C. acnes isolates showed lower MIC50 values for benzylpenicillin compared with cefazoline, cefuroxime, and oxacillin, suggesting that the addition of benzylpenicillin to cloxacillin could be more effective for preventing C. acnes infections than the presently recommended prophylaxis. Several studies have also shown significant reduction of C. acnes after preoperative topical application with benzoyl peroxide in shoulder surgery [
]. The use of benzoyl peroxide in cardiac surgery has not been studied.
The present study also showed a significant difference in bacterial growth between male and female patients. Cultures from tissue were more often negative in female patients (63–75%) than in male patients (14–21%, P < 0.01–0.03), whereas male patients were colonized by C. acnes to a higher extent (45–81%) than female patients (13%) throughout the operation (P < 0.01–0.02). Female patients also had more negative cultures from the graft (38% vs 7%, P = 0.04) and gloves (50% vs 14%, P = 0.04) than male patients. C. acnes resides in sebaceous glands, which are mostly found on the face, scalp, and upper thoracic area and are found in higher numbers in male patients [
The resistance patterns between the isolates present at the beginning of surgery and the isolates found after 2 h and on surgical materials were similar. This finding contradicts the understanding that the subsequent colonization of multidrug-resistant, nosocomial bacteria during the surgical procedure plays an important role under normal circumstances; instead, it suggests continuous recolonization from the patients' own flora.
The surgeons' gloves were contaminated in 80% of the cases despite all surgeons using double gloving. Although theoretically appealing, it has previously been hard to prove that double gloving is effective in reducing surgical cross-infection [
]. Our study indicates that just the contact with the surgical wound might be a more important source of contamination of gloves than the skin of the surgeons' hand. Irrespective of the source of contamination, changing gloves immediately before handling vascular prostheses may be effective in reducing AGIs, but this remains to be investigated.
There are several limitations to this study. Emergency surgery was an exclusion criterion. However, it is possible that the cultures of these patients would have had a slightly different bacterial spectrum due to deviation from the routine preoperative wash procedures using chlorhexidine. By excluding this group, it is possible that this study underestimates the bacterial contamination of the sternal wound.
The pieces of graft and PTFE felt that were used for sampling were divided with scissors that had previously been used during surgery, which may have constituted a risk of contamination of the surgical materials. However, if a graft had been used in surgery, the same scissors would have been used to cut the graft; thus, this study appropriately reflects the normal surgical conditions. In this study, the samples taken with rayon swabs served as a more objective description of the bacterial colonization of the surgical wound, while the cultures from surgical material taken at the same timepoint represent the contamination of the graft and PTFE felt by normal handling during surgery.
Samples from the surgeons' gloves were collected 2 h after the start of surgery; however, the exact time of sampling differed by up to 31 min. This irregularity occurred because it was important to collect the sample at the most convenient time during the surgery. No data are available on whether or when the surgeon changed gloves. The isolates present on the gloves were similar to the isolates found on the patients' skin and subcutaneous tissue, but it is difficult to draw firm conclusions from these samples.
The size of the study population limits the possibility of subgroup analyses. The sex distribution in the study group was skewed, with 84% male patients. In general, study populations in cardiac surgery research have a greater ratio of male patients, usually 75–85% [
]. To better understand differences between the sexes regarding bacterial colonization and pathobiology, further studies are needed.
The findings of this study suggest that skin bacteria, such as S. epidermidis and C. acnes, which probably originate from the patient, are present in the sternal wound and may contaminate surgical graft material during cardiac surgery.
Acknowledgements
We thank the laboratory staff at the Department of Laboratory Medicine, Clinical Microbiology, and Infectious Diseases at Örebro University for their skilful technical assistance with the microbiological analyses presented in this paper. We are also indebted to H. Hildingsson and S. Bratt for their invaluable help during the collection of cultures and administration of the study.
Conflict of interest statement
Ö. Friberg reports having received compensation for consultancy services for AbbVie Inc. B. Söderquist is a member of an advisory board at Advanz Pharma. B. Söderquist also gave a lecture about prosthetic joint infections at Correvio Pharma Corp. in 2019, receiving financial compensation. C. Wistrand reports receiving compensation by Mölnlycke for partaking in panel discussions. N. Sandström declares no conflicts of interest.
Funding sources
This work was supported by Nyckelfonden, Örebro County Council [grant no. OLL-934809].
References
Ridderstolpe L.
Gill H.
Granfeldt H.
Ahlfeldt H.
Rutberg H.
Superficial and deep sternal wound complications: incidence, risk factors and mortality.
Bacterial recolonization of the skin and wound contamination during cardiac surgery: a randomized controlled trial of the use of plastic adhesive drape compared with bare skin.
Wound contamination in cardiac surgery. A systematic quantitative and qualitative study of the bacterial growth in sternal wounds in cardiac surgery patients.
Incidence, microbiological findings, and clinical presentation of sternal wound infections after cardiac surgery with and without local gentamicin prophylaxis.
00107-X&id=gr1.jpg){kind=link}
00107-X&id=gr2.jpg){kind=link}