Effect of Preoperative Skin Preparation Methods on Sternotomy Surgical Site Infection Rates: A Quasi-Experimental Study

Surgical site infections (SSIs) are common and expensive health care–associated infections (HAIs). Approximately 160 000 to 300 000 SSIs occur in the United States every year.¹ The risk of death increased by 2 to 11 times in patients who developed an SSI, and 77% of these deaths were attributed to SSIs.² According to 2017 data of the Turkish National Health Service Associated Infections Surveillance Network (USHİESA), the overall SSI rate was 0.72%, and less than 1% (8194) of the 617 745 cases of HAIs were SSIs, with SSIs being the fourth most common HAI.³ In a meta-analysis on post-median sternotomy sternal SSI rates in Turkey, Ozguler and Ozguler⁴ reported that superficial sternal SSIs, deep sternal SSIs, and mediastinitis rates were 1.71%, 0.95%, and 0.8%, respectively (n= 10   954). In a retrospective review of patients who underwent open heart surgery (n = 115), Pala et al⁵ found that 4 of 10 patients (4.3%) who underwent sternotomy developed superficial sternal SSIs, but no patients developed deep sternal SSIs (mediastinitis).

SSI complicates approximately 1% to 8% of cardiac surgeries.^6,7 Despite this low percentage, these SSIs are life-threatening conditions that decrease quality of life and increase morbidity and mortality rates as well as costs.⁸ Despite recent advances in prevention and perioperative care, SSI remains a pressing concern in cardiac surgery due to its considerable impact on length of hospital stay, morbidity, mortality, and utilization of resources.^9,10 The cost increase due to SSIs in cardiac surgeries can triple hospital charges.¹¹ In a study published in 2014, costs associated with surgical revisions due to wound infections amounted to 60 000 to 90 000 Euros in Germany.¹²

Minimizing or preventing SSI incidence in cardiovascular surgery is critical for patient safety. Knowing the risk factors of SSI, evaluating patients for these risk factors, and taking precautions in the preoperative, operative, and postoperative periods can help reduce the incidence of SSI.⁷ Although it is not possible to prevent all patient risk factors, many risk factors related to the surgical process can be prevented. Thus, understanding all the risk factors that cause SSI development and taking the necessary precautions can help decrease the incidence of these infections.^7,13

Preoperative skin preparation is a surgical intervention, and skin antisepsis is a well-known strategy for reducing SSI.¹⁰ A preoperative skin preparation procedure, including removing hair and a shower/bath for cleaning the skin at the intended site of surgery, should be standardized for all patients. This is consistent with the nonmaleficence principle and can be utilized to prevent SSIs.^14-16

Measures to prevent SSIs have diverse levels of adoption among health care professionals. The levels of application of internationally accepted measures have great variability.¹⁷ Although preoperative skin preparation is a significant factor in preventing SSIs, it has been observed by the authors that nurses in Turkey do not adequately perform preoperative skin preparation and do not inform patients about this topic. The hospital where this study was conducted had no standard procedure for preoperative hair removal and bathing. Although studies^18,19 advise against using razors for hair removal, health care professionals in most of the clinics of the hospital use razors to remove hair. If the surgeon in the cardiovascular surgery clinic orders hair removal on the surgical site before surgery, either the patient or the support staff implement this preparation. Moreover, the health care professionals do not routinely monitor whether patients take a shower before surgery.

International studies have investigated the effect of 2 different preoperative hair removal methods^18-20 and 2 different preoperative bath/shower methods^21-23 on SSIs. However, there is no published research investigating the effect of the combination of hair removal methods and bath/shower substances on SSIs. Therefore, the current study is original in that sense. Moreover, there have been very few experimental studies in the nursing field in Turkey on preventing SSIs.^14,24 Dizer et al¹⁴ evaluated the effect of hair removal with an electric clipper and bathing with chlorhexidine at least twice preoperatively on postoperative SSIs, while Isik et al²⁴ looked into the efficiency of antibacterial suture material for cardiac surgery on SSIs.

This study aimed to determine the effect of different preoperative skin preparation methods on the rate of SSI in patients undergoing sternotomy. The research questions were as follows.

Is there any difference in rates of SSI in patients shaving with an electric clipper or razor blade and taking a shower/bath once with regular soap/shampoo, daphne soap, or chlorhexidine solution?

Is the SSI rate lower in intervention group-1 than in the control group?

Is the SSI rate lower in intervention group-2 than in the control group?

Is the SSI rate lower in intervention group-1 than in intervention group-2?

Study sample and design. This quasi-experimental study investigated the effect of different preoperative skin preparation methods on the prevention of SSIs in patients undergoing sternotomy. The research was carried out in the cardiovascular surgery clinic of a university hospital between January and December 2016. The study sample was calculated from the total number of male patients who had undergone sternotomy (N = 202) between January 1, 2015, and December 1, 2015. The study population consisted of 120 male patients who underwent sternotomy between those dates. According to the sample calculation whose universe is known, the required sample size was 81 patients, with a margin of error of α = .05 and a confidence interval of 1-α = .95. A total of 96 patients participated (34 patients in the control group [CG], 31 patients in intervention group-1 [IG-1], and 31 patients in intervention group-2 [IG-2]) (Figure 1).

The inclusion criteria were 1) age older than 18 years, 2) having body hair in the sternum area, 3) undergoing elective sternotomy for the first time, 4) class 1 (clean) and class 2 (clean-contaminated) wounds, 5) no mental or communication difficulties, 6) no allergies to chlorhexidine, and 7) voluntary participation. The exclusion criteria were 1) malignancy, 2) revision surgery, 3) patient receiving immunosuppressive therapy, 4) using antibiotics, and 5) a history of infection 1 week before the study.

Data collection tools. This study employed a 1) personal data and preoperative information form, 2) postoperative information form, and 3) SSI assessment form based on available literature.^20,23 The personal data and preoperative information form included the following variables: patient age, educational status, smoking habits, body mass index, chronic diseases, primary diagnosis, and length of preoperative hospital stay. The postoperative information form was designed to collect surgical and postoperative status variables, including type of surgery, surgery duration, amount of blood transfusion, drain used, vital signs, use of antibiotics, length of stay in the intensive care unit, and length of postoperative hospitalization. The SSI assessment form was prepared based on the Centers for Disease Control and Prevention (CDC).¹⁵ The following wound assessment yes/no variables were collected: localized swelling at the wound site, tenderness, pain, erythema, warmth/increased temperature, purulent drainage, wound dehiscence, and deliberate opening of an incision by a surgeon for purposes of diagnosis, taking cultures from the incision, culture results, or diagnosis of infection.

Pilot testing. To ensure the study was implemented with minimum difficulties, the researcher had conducted a pilot study with 10 participants. Each participant received an informed consent form, the survey questionnaire, and an evaluation form for the survey. Following the pilot study, some modifications and amendments were made to the survey. These were mostly related to the design of the survey. The participants from the pilot study were excluded from the main study.

Control group. Patients in the CG received the routine skin intervention of shaving body hair with a razor blade; this was done by support staff. After shaving, the patients were told to take a bath/shower by the support staff; however, it was not checked whether patients did so.

Intervention group. In the intervention group, as part of the preoperative skin preparation procedure, body hair removal in the sternal area was carried out by the researcher (EK) using an electric clipper with a disposable cartridge the night before surgery. After removal of body hair, patients were educated (average of 5 minutes) by the researcher about taking a shower/bath with olive oil–containing daphne soap (IG-1 group) or using a 50-mL 2% chlorhexidine solution (IG-2 group). Patients were also provide with a brochure that contained information about taking a shower/bath with the respective cleansing agents. After the educational intervention, patients took a shower/bath using the assigned cleaning agent.

Data collection. Data were collected via face-to-face interviews with patients who agreed to participate in the study in accordance with informed consent. Data were also collected from the medical records. To minimize contamination across groups, the data were collected first from the control group (in March, April, May, and June), then from the IG-1 group (July, August, and September), and lastly from the IG-2 group (October, November, and December). The study was conducted at 1 center to minimize infection risks due to the surgical procedure variations. The forms used were similar for the control and intervention groups. Sternum incisions of all patients were monitored for SSI variables 4, 15, 30, and 90 days following surgery.

Study process. Patients who met the inclusion criteria completed the personal data and preoperative information form during the preoperative period. Interventions were then implemented in the IG-1 and IG-2 groups (Figure 2). The postoperative information form and SSI inspection form were completed by the researcher at the clinic, and a routine sternum swab cultures were collected from all patients on postoperative day 4. The SSI inspection form was also completed by the researcher when patients returned for routine care on postoperative days 15 and 30 at the outpatient cardiovascular surgery polyclinic. Because patients ceased attending the clinic after day 30, all 90-day data were obtained over the telephone. To determine whether any signs of wound infection developed between days 30 and 90, the researcher called the patient and asked the questions in the SSI inspection form. Patient answers were recorded as yes/no.

The SSI diagnoses of patients with symptoms of infection were made by their physicians using the criteria outlined by the Centers for Disease Control and Prevention.¹⁵ According to these diagnostic criteria, SSIs are classified as superficial incisional SSI (occurs within 30 days after surgery and involves only skin and subcutaneous tissue of the incision), deep incisional SSI (occurs 30 to 90 days after surgery and involves deep soft tissues of the incision; eg, fascial and muscle layers), and organ/space SSI (occurs 30 to 90 days after surgery and involves any part of the body deeper than the fascial/muscle layers that is opened or manipulated during the surgical procedure).¹⁵

Data analysis. The descriptive findings were presented as mean ± standard deviation (SD) for the continuous data and with frequency and percentage for categorical data. The normality assumptions were controlled by the Shapiro-Wilk test. Categorical data were analyzed by Pearson chi-square or Fisher’s exact test. The one-way analysis of variance test was used to compare continuous data (eg, age and body mass index) among groups. Statistical analysis was performed using IBM SPSS Statistics for Windows, Version 23.0 (IBM Corp.). Two-sided P values < .05 were considered statistically significant.

Ethical considerations. Approval was obtained from the Ethical Committee of the Akdeniz University Faculty of Medicine Clinical Research Ethics Committee, confirming that this study followed international standards and the principles adopted in the World Medical Association Declaration of Helsinki (approval no. 124, August 2015). Regarding the protection of medical and personal data, the Helsinki Declaration Principles were complied with regarding respect for human dignity. Written patient consent was obtained via informed consent forms.

The study sample (Table 1 [continued]) consisted of 96 male patients: control group (n = 34; 35.4%), IG-1 (n = 31; 32.3%), and IG-2 (n = 31; 32.3%). The control, IG-1, and IG-2 groups had a mean age of 62.18 ± 9.43, 58.65 ± 10.52, and 58.58 ± 12.03 years, respectively, which was statistically insignificant (P = .298). The CG, IG-1, and IG-2 groups did not differ by smoking duration (36.12 ± 11.47, 29.57 ± 13.48, and 29.81 ± 10.80 years, respectively; P = .425), period of smoking cessation (106.79 ± 132.53, 164.92 ± 142.58, and 142.88 ± 96.92 months, respectively; P = .656), diabetes mellitus duration (12.25 ± 9.13, 8.48 ± 8.58, and 6.95 ± 4.13 years; P = .764), hypertension duration (7.58 ± 5.68, 7.75 ± 6.18, and 8.00 ± 6.78 years; P = .741), and chronic obstructive pulmonary disease duration (3.00 ± 0.71, 12.00 ± 15.56, and 1.00 ± 0.0 years; P = .878).

The CG, IG-1, and IG-2 groups were similar in terms of education level (primary school: 41.2%, 61.3%, and 51.6%; secondary school: 8.8%, 12.9%, and 12.9%; high school: 14.7%, 9.7%, and 12.9%; university: 35.3%, 16.1%, and 22.6%; P = .617), body mass index (< 25 kg/m²: 35.3%, 35.5%, and 45.3%; 25–30 kg/m²: 55.9%, 48.4%, and 35.5%; > 30 kg/m²: 8.8%, 16.1%, and 19.2%; P = .510), smoking status (nonsmoker: 23.5%, 9.7%, and 16.2%; current smoker: 20.6%, 35.5%, and 29.0%; former smoker: 55.9%, 54.8%, and 54.8%; P = .529), and chronic disease (present: 64.7%, 67.7%, and 74.2%; absent: 35.3, 32.3, and 25.8%; P = .704). The IG-1 group had a higher rate of diabetes mellitus (45.2%) than the CG (23.5%) and IG-2 (35.5%) groups. However, the difference was not statistically significant (P = .184).

There was no significant difference among the CG, IG-1, and IG-2 groups regarding length of preoperative hospital stay (4.97 ± 4.75, 6.26 ± 4.86, and 6.39 ± 4.91 days, respectively; P = .759) postoperative hospital stay (7.85 ± 2.45, 8.29 ± 3.63, and 7.68 ± 1.68 days; P = .979), stay in the intensive care unit (3.47 ± 1.33, 3.42 ± 1.65, and 4.00 ± 1.55, days; P = .320), and surgery (4.47 ± 1.06, 4.87 ± 1.31, and 3.97 ± 1.06 hours; P = .335). The control group (mean, 12.09 days) received antibiotics longer than the IG-1 group (mean, 9.42 days) and the IG-2 group (mean, 10.16 days). However, the difference was not statistically significant (P = .087). There was no significant difference in blood transfusion units (10.29 ± 4.05, 9.52 ± 3.38, and 10.00 ± 3.82; P = .141) and drainage tube placement duration (2.85 ± 1.10, 2.68 ± 1.01, and 2.84 ± 1.00; P = .374) among the groups (Table 2).

The groups did not differ by primary diagnosis (coronary artery disease: 82.4%, 83.9%, and 80.6%; valve disease: 14.7%, 16.1%, and 9.7%; other: 2.9%, 0.0%, and 9.7%; P = .377) and type of operation (CABG: 79.5%, 83.9%, and 80.6%; valve repair: 14.7%, 12.9%, and 9.7%; CABG plus valve repair: 2.9%, 3.2%, and 0.0%; other: 2.9%, 0.0%, and 9.7%; P = .543). None of the surgery-related variables were significantly different among groups.

Evaluation of the sternal incision area as related to SSI diagnostic criteria is presented in Table 3 for postoperative days 4, 15, and 30. On day 4, the groups had similar rates of swelling, tenderness, pain, purulent drainage, wound dehiscence, and infection doubt by the surgeon. On day 4, the rate of erythema was highest in the control group (73.5%) and lowest in the IG-2 group (16.1%) (P < .001).

On day 15, the groups had similar rates of swelling, pain, erythema, local warmth/increased temperature, purulent drainage, surgical incision opening by the surgeon, and infection doubt by the surgeon. On day 15, tenderness was more common in the control group (41.2%) than in the IG-1 (25.8%) or IG-2 groups (9.7%) (P = .015).

On day 30, there was no significant difference in tenderness, pain, erythema, purulent drainage, wound dehiscence, and infection doubt by the surgeon between the groups. In general, the prevalence of swelling, tenderness, pain, local warmth/increased temperature, purulent drainage, wound dehiscence, and surgical incision opening by the surgeon were similar among the groups. In general, the rate of erythema was highest in the control group (73.5%) and lowest in the IG-2 group (22.6%) (P < .001).

Although not presented in Table 3, at the 90-day follow-up, 2 patients in the IG-1 group self-reported that they went to the doctor because of minimal drainage in the sternal incision site and started antibiotic treatment (approximately between days 40 and 45).

Finally, sternal SSI occurred in 10.4% of all study patients; 8.8% of the CG patients, 12.9% of the IG-1 patients, and 9.7% of the IG-2 patients developed an SSI (P > .05) (Table 4). No microorganisms were detected in the swab cultures of 4 of 10 patients with SSI. Staphylococcus aureus and coagulase-negative staphylococci were isolated in 1 patient and 5 patients, respectively.

Figure 3, Figure 4, and Figure 5 show the most common symptoms in the CG, IG-1, and IG-2 groups. Wound dehiscence occurred in 2 patients in the IG-2 group, and the wound was deliberately reopened by the surgeon in 1 CG patient. Three (3) CG patients, 2 IG-1 patients, and 3 IG-2 patients were diagnosed with SSI by the patient’s physician. Two (2) patients in IG-1 self-reported the diagnosis of SSI on day 90 via telephone.

SSIs are significant complications that result in increased antibiotic use, costs, length of hospital stay, and morbidity and mortality rates.^11,25 In this study, the symptoms of potential sternal incision infection of patients in the CG, IG-1, and IG-2 were similar among groups. The majority of potential syptoms of infections developed on postoperative day 4 and started to decrease on day 15. The development of tenderness, pain, and erythema in most patients is indicative of the inflammatory phase of epithelialization,²⁶ and the authors believe that was reflected in the current study.

Incidence rates for sternal SSIs between 0.5% and 3% have been reported for patients who had undergone heart surgery.⁶ Si et al⁸ reported that the rate of SSI in patients who had undergone coronary artery bypass graft was 11.7%, and 2.3% of those infections were superficial incision SSIs. Superficial incisional sternal wound infection rates were reported as 0.5% to 0.8% in a consensus statement in 2016.⁷ In the current study, the overall infection rate of all patients was 8.3%, and 100% of these infections were superficial incision SSIs. It is possible that deep incisional and organ/site SSIs did not develop in the patients in our study due to the use of antibiotics for a greater duration than generally recommended by the World Health Organization.²⁷

There are recommendations about the removal of body hair during preoperative skin preparation to improve surgical care and prevent SSIs.^16,21 It has been stated that body hair should not be removed unless it is required and, if body hair removal is required, this should be done using appropriate methods.^18,19 In addition, taking a shower/bath using soap or antiseptic solution in the preoperative stage is recommended because it reduces bacteria on the skin, which is a significant risk factor for sterile surgical wounds.^19,21 Although there are studies that report that taking a bath using an antiseptic solution before surgery decreases the infection rate,^14,19,28 there are other studies that state it has no effect.^29,30 In our study, no statistically significant difference in SSI rates was seen among the CG, IG-1, and IG-2 groups.

A systematic review by Tanner et al¹⁹ reported that in 3 studies comparing body hair removal with a razor blade and an electric clipper, body hair removal with a razor blade carried a significantly greater SSI rate than the electric clipper (risk ratio, 2.09; 95% CI, 1.15-3.80). A literature review by Jose and Dignon¹⁸ found that electric clippers and depilatory cream caused fewer SSIs than shaving with a razor. Contrary to these findings and similar to our findings, in a study on body hair removal in the male genital area, there were no statistically significant differences between the use of an electric clipper and a razor blade with regard to SSI development.²⁰ Cowperthwaite and Holm³¹ reported that cleaning the surgical site with a razor blade would cause skin wounds of different levels.

Studies on this topic have found different results.^29,27 The World Health Organization states that it remains unknown if having a bath before surgery or taking a shower with antimicrobial soap is more effective in reducing SSI than common soap.¹³ In our study, no statistically significant differences were found among the use of common soap/shampoo, daphne soap with olive oil, and chlorhexidine solution as part of preoperative skin preparation. Similar to our results, there are other studies that report that chlorhexidine is not effective in preventing infections.^23,30,32 However, Dizer et al¹⁴ reported that using electric clippers for body hair removal and bathing with chlorhexidine at least twice reduced the incidence of SSIs. This result supports the findings of international studies that report that using chlorhexidine solution for preoperative skin preparation prevents infections.^28,29

In the current study, no microorganisms were detected in 4 of 10 patients with SSI. In the remaining 6 patients, S aureus and coagulase-negative staphylococci were isolated in 1 patient and 5 patients, respectively. Yavuz et al³³ stated that coagulase-negative staphylococci were the most prevalent microorganisms that reproduced in SSIs following open heart surgery. Other studies have found that coagulase-negative staphylococci are 1 of the 5 most prevalent infection microorganisms observed in hospitals.^33,34 On the other hand, in many studies S aureus is reported as the pathogen isolated in most cases.^8,9,29 The most frequently reported pathogen in SSIs is S aureus (30.4%) followed by coagulase-negative staphylococci (11.7%), Escherichia coli (9.4%), and Enterococcus faecalis (5.9%).²⁷ Despite its low reliability, swab culture is a commonly used method in hospitals due to its noninvasive nature.

This study had several limitations. First, the data collection phase was 1 year. Second, participants were recruited from the cardiovascular surgery clinic of only 1 university hospital in Turkey. Third, patients were given daphne soap with olive oil and chlorhexidine solution but actual bathing was not observed. Fourth, although the preoperative skin preparation was the same for all patients, 4 different surgeons performed the procedures. Fifth, sample size calculations were not performed. As a result, the study was underpowered to reveal statistically significant differences among the groups in SSI incidence.

This quasi-experimental study conducted among 96 male patients who were undergoing sternotomy found no statistically significant difference in the effect of preoperative skin preparation methods on SSIs among the CG, IG-1, and IG-2 groups. For preoperative bathing, the authors recommend that patients use natural, affordable, and easily accessible soaps (eg, olive oil laurel soap) that do not contain chemicals and do not cause allergies. Future robust randomized controlled trials with large samples are recommended to compare these methods to determine an optimal and affordable preoperative skin preparation method.

The authors thank the patients who participated in this study.

Dr Karazeybek is an assistant professor, Faculty of Nursing, Akdeniz University, Antalya, Turkey. Dr Erbasan is a professor, Faculty of Medicine, Akdeniz University, Antalya, Turkey. Dr Şenol Çelik is a professor, Faculty of Nursing, Koc University, Istanbul, Turkey. Address all correspondence to: Ebru Karazeybek, PhD, RN, Faculty of Nursing, Akdeniz University, Dumlupınar Boulevard Campus, 07058, Antalya, Turkey; tel: +905052519968; fax: +90(242) 2261469; email: ekarazeybek@gmail.com.

This work was supported by Research Fund of Hacettepe University (project no. THD-2016-8049).

None disclosed