Surgical site infections have been linked to longer hospital stays, higher readmission rates, and increased health care costs.1–3 Approximately 200,000 hysterectomies performed each year in the United States are done through a laparotomy.4 The overall rate of surgical site infection after abdominal hysterectomy has been reported to be approximately 4–6%.5,6 Furthermore, surgical site infections are responsible for one third of the readmissions after hysterectomy.3 Therefore, reduction in surgical site infections after abdominal hysterectomy has become one of the Centers for Medicare & Medicaid Services' targets for quality improvement.7 However, the existing published guidelines for interventions to prevent infections do not routinely specify the choice of antiseptic for skin preparation before abdominal hysterectomy.8,9
Chlorhexidine-alcohol topical antiseptics have been shown to outperform povidone-iodine in preventing surgical site infections related to vascular catheter site insertion.10,11 A recent randomized controlled trial comparing the use of povidone-iodine compared with chlorhexidine-alcohol in adults undergoing open surgery found a significantly lower surgical site infection rate in the chlorhexidine-alcohol group (9.5% compared with 16.1%, P=.004).12 However, the generalizability of this trial in gynecologic surgery is limited because only 10% of the patient population in this trial had undergone hysterectomy. Therefore, we undertook this study to compare preoperative chlorhexidine-alcohol topical antiseptic agent with povidone-iodine in patients undergoing abdominal hysterectomy for benign indications.
MATERIALS AND METHODS
A retrospective cohort study of patients undergoing abdominal hysterectomy from July 2012 to February 2015 in the Michigan Surgical Quality Collaborative (referred to as “the collaborative”) was performed. The collaborative is a Blue Cross and Blue Shield of Michigan-Blue Care Network-funded database voluntarily populated by both academic and community hospitals throughout the state. It includes patients from all insurance payers. At participating hospitals, trained, dedicated nurse abstractors collect patient characteristics, intraoperative processes of care, and 30-day postoperative outcomes from general and vascular surgery and hysterectomy cases. Nurse abstractors use chart reviews, patient phone calls, and occasionally discussion with physicians to ensure complete capture of data regardless of the hospital where subsequent treatments happen. Detailed methods of the registry's data collection have been described previously.13
Women 18 years of age or older undergoing abdominal hysterectomy for a benign indication were included in the study. Patients were excluded if data were missing for preoperative antibiotic administration, skin antiseptic preparation agent information, or postoperative surgical site infection data. Because the focus of this study was routine hysterectomy, we excluded small patient subgroups with a much higher baseline risk of surgical site infection and those with an existing infection before surgery. Specifically, patients with a diagnosis of an open wound or preoperative sepsis as well as those with wound classification categories defined as dirty or contaminated were excluded. Surgeries labeled as emergent operations were also excluded, because this category does not reflect routine abdominal hysterectomy. Patients on chronic steroids were excluded as a result of concerns of immunosuppression. Lastly, patients who did not receive preoperative antibiotics based on the criteria set forth by the American College of Obstetricians and Gynecologists14 and the Surgical Care Improvement Project were excluded. A detailed overview of the cohort development is provided in Figure 1. Michigan Surgical Quality Collaborative data sets provided to the researchers contain no patient, hospital, or health care provider identifier. Therefore, this study met the criteria for “exempt” status by the University of Michigan institutional review board–medical (HUM00073978).
The following demographic data, medical comorbidities, and perioperative data were abstracted: age, body mass index (BMI, calculated as weight (kg)/[height (m)]2), race (categorized as either white or nonwhite), insurance status (Medicare, Medicaid, private, uninsured, missing, or other), functional status (a patient's ability to perform basic daily activities to maintain their health, measured as being either dependent or independent), American Society of Anesthesiologists (ASA) classification score (dichotomized to American Association of Anesthesiology class less than 3 or 3 or greater), diabetes mellitus (defined as requiring oral hypoglycemic agents or insulin), tobacco use (defined as smoking cigarettes, cigars, or a pipe, chewed tobacco, or used marijuana within the past year), alcohol use (defined as greater than two drinks in the 2 weeks before admission), estimated blood loss (in milliliters), operative time (in minutes), intraoperative adhesions (categorized as none, minor [requiring lysis of adhesions less than 45 minutes], and major [requiring lysis of adhesions 45 minutes or greater]), and bowel surgery performed (yes or no). Prophylactic antibiotics were categorized as β-lactam antibiotics or β-lactam alternatives. This categorization was done based on the findings of our previous study from this database, which showed increased odds of surgical site infection with β-lactam alternative antibiotics in hysterectomy.15
We hypothesized that the use of chlorhexidine-alcohol topical antiseptic is associated with a lower likelihood of developing surgical site infection. Therefore, the primary outcome measure of this study was the occurrence of any surgical site infection (superficial and deep or organ space) within 30 days of surgery, as defined by the Centers for Disease Control and Prevention.16
Descriptive analyses of demographics, comorbidities, perioperative data, and postoperative surgical site infection were performed. Univariate analyses were performed to identify factors significantly associated with developing postoperative surgical site infection. χ2 and Fisher exact tests were used for categorical variables and parametric one-way analysis of variance or nonparametric Wilcoxon–Mann–Whitney tests were used for continuous variables wherever appropriate. Factors associated with development of surgical site infections based on previous research or clinical plausibility were included in multivariable analyses to ascertain the independent association of skin preparation agent with the rate of surgical site infection.15 These factors included: patient demographic factors (age and BMI), medical comorbidities (ASA class and diabetes status), perioperative variables (estimated blood loss, surgical time, intraoperative adhesions, and antibiotic categories), and hospital characteristics (bed size and teaching status). A second model was constructed using backward stepwise regression of variables with P<.2 as elimination criteria to verify the results of the first model. To account for clustering of data at the hospital level, we used Huber–Eicker–White robust standard errors, which strengthen our model by controlling for variation between hospital sites.17–19 Propensity score matching was used to further validate the results of logistic regression modeling. For this analysis, we performed one-to-one matching using calipers of 0.001.
Stata 14.0 SE for MacIntosh was used for all analyses.
Of the total 4,259 abdominal hysterectomies included, chlorhexidine-alcohol was used in 70.5% (n=3,005) and povidone-iodine in 29.5% (n=1,254) of surgeries. The overall unadjusted rate of any surgical site infection was 2.9% (95% CI 2.5–3.5; n=124). The unadjusted rate of surgical site infection 2.6% (95% CI 2.1–3.3; n=79) for chlorhexidine-alcohol and 3.6% (95% CI 2.7–4.8; n=45; P=.09) for the povidone-iodine group.
Univariate analysis of the demographic factors, medical comorbidities, perioperative variables, intraoperative variables, and hospital factors associated with the development of surgical site infection is presented in Table 1. The baseline comparison between the patients in the povidone-iodine group and those in the chlorhexidine-alcohol group is presented in Table 2. The chlorhexidine-alcohol group had a higher proportion of patients with several factors associated with surgical site infection development (Table 2). For example, patients receiving chlorhexidine-alcohol were more likely to be older, nonwhite, have a higher BMI and a higher proportion of patients with ASA class 3 or greater, be functionally dependent, higher mean blood loss, and increased mean surgical time, higher proportion of patients with intraoperative adhesions categorized as major, and required bowel surgery. The unadjusted rate of surgical site infection for patients treated with chlorhexidine-alcohol was 2.6% (95% CI 2.1–3.3; n=79) and povidone-iodine was 3.6% (95% CI 2.7–4.8; n=45; P=.09; Table 3).
After adjusting for differences between populations using multivariate logistic regression models, we estimate that patients receiving chlorhexidine-alcohol had 44% lower odds of developing a surgical site infection (adjusted odds ratio [OR] 0.56, 95% CI 0.37–0.84, P=.01) in the first model using clinically plausible variables (Table 4). The area under the curve for this model was estimated to be 0.696. In the second model, using backward stepwise approach, patients receiving chlorhexidine-alcohol had 35% lower odds of developing a surgical site infection (adjusted OR 0.65, 95% CI 0.45–0.93, P=.02). However, this model had a lower area under the curve of 0.66.
Propensity score matching (one to one) yielded 845 patients receiving povidone-iodine skin preparation agent and 808 patients in the chlorhexidine-alcohol group. Propensity-matched cohorts were similar with none of the factors significantly different between the two groups (Table 2). The rate of surgical site infection in the matched cohort was 1.5% (95% CI 0.8–2.6; n=12) for the chlorhexidine-alcohol group and 4.7% (95% CI 3.5–6.4; n=40; propensity matched cohort OR 0.31, 95% CI 0.17–5.6, P<.001; Table 3).
Figure 2 highlights the increase in surgical site infection as a function of proportion of hysterectomies using povidone-iodine skin preparation at each hospital site. Hospitals with a higher rate of povidone-iodine use had higher rates of surgical site infection.
In patients undergoing abdominal hysterectomy in a statewide collaborative, preoperative skin preparation with chlorhexidine-alcohol is associated with lower odds of developing surgical site infection compared with povidone-iodine (adjusted OR 0.56). This finding is important for several reasons. Abdominal hysterectomy remains common in the United States with 200,000 open abdominal hysterectomies performed per year. Therefore, evidence-based interventions to reduce surgical site infections are urgently needed. Second, surgical site infections after abdominal hysterectomy will be a metric included in the Hospital Acquired Condition Reduction Program score, which will affect hospital reimbursements.20 Third, surgical site infections are the single most significant cause for readmissions after hysterectomy.3 Finally, surgical site infections have been shown to be associated with a significant increase in morbidity and mortality after surgery as well as a negatively effect on patient quality of life of patients.21 Although major perioperative guidelines specify the choice of antibiotics to prevent infections, to our knowledge, there are no current established national guidelines that indicate the choice of preoperative topical skin antisepsis before abdominal hysterectomy. Our findings, along with the previous randomized controlled trials in the field of general surgery, provide a basis for recommending chlorhexidine-alcohol for abdominal hysterectomy.12
Our results are consistent with previous studies comparing povidone-iodine with chlorhexidine-alcohol. Levin et al22 reported on a study comparing chlorhexidine-alcohol with povidone-iodine in gynecologic surgeries. Overall reduction in surgical site infections with the implementation of chlorhexidine-alcohol (from 14.6% to 4.5%) was noted in this study. Although the baseline rates of surgical site infections were much higher than previously reported rates in the United States, this study supported the use of chlorhexidine-alcohol instead of povidone-iodine. Darouiche et al12 reported the results of their randomized controlled trial and concluded that chlorhexidine-alcohol was superior to povidone-iodine in preventing surgical site infections (rate of 16% compared with 9.5%, P=.004, respectively) in clean-contaminated surgeries. However, their study included only 40 gynecologic patients in each arm; given the overall low rate of surgical site infection after hysterectomy (compared with other procedures like bowel surgery), these results needed further investigation in patients undergoing hysterectomy. Similarly, a comprehensive systematic review found an overall risk reduction of 32% with the use of chlorhexidine-alcohol compared with povidone-iodine in clean-contaminated surgery.23
The mechanism of action and the superiority of chlorhexidine-alcohol seem to be related to its persistent activity with rapid onset of action related to drying of alcohol and the ability of chlorhexidine to remain active even when the field is soiled with blood or serum.24 However, it is unclear whether the superiority of chlorhexidine-alcohol is the result of the alcohol base or the chlorhexidine component itself. In cesarean deliveries, Tuuli et al25 reported that the use of chlorhexidine-alcohol for preoperative skin antisepsis resulted in a significantly lower risk of surgical site infections when compared with iodine in alcohol, indicating that chlorhexidine gluconate may help decreases the rate of surgical site infection. Further studies comparing alcohol-based preparations are needed to answer this question, albeit a randomized study may not be feasible as a result of the necessity of a relatively large sample size.
The alcohol-based skin preparations have their limitations. Before evaporation, a significant fire risk exists and significant injuries to the patient and staff have been reported.26,27 In addition, these agents cannot be used on mucosal surfaces. Therefore, aqueous preparations have a role in vaginal hysterectomy as well as in vaginal preparation during abdominal hysterectomy. Although the isopropyl alcohol and chlorhexidine gluconate pose a very small risk of combustion at temperatures provided by the electrosurgical devices, the overall qualitative benefits from a reduction of surgical site infections in the chlorhexidine-alcohol group make this agent a superior choice. Strategies to ensure adequate waiting time after skin preparation during timeout to prevent surgical fires are recommended.28,29 Allergies to chlorhexidine-alcohol have been reported in the literature. Although the precise estimates of chlorhexidine-alcohol allergy remain unknown, one study estimated that skin patch testing revealed positive reactions in 2% of the patients tested. Further studies should make note of both severe allergies as well as skin reactions in patients in the chlorhexidine-alcohol group.30
This study has many notable strengths. The collaborative is a collective of Michigan hospitals throughout the state with a common goal of surgical quality improvement through improved patient care and reduction of health care costs. The collaborative focuses on standardized data collection and nurse abstractors who attend regular seminars and are audited frequently. Although other data sets are retrieved from billing and administrative data that does not include postdischarge data or standardized definitions, the collaborative uses the Centers for Disease Control and Prevention definition of surgical site infections and uses a 30-day endpoint to capture any complications for both inpatient and outpatient procedures. Lastly, we used a multivariate logistic regression model to account for multiple patient and perioperative factors such as type of prophylactic antibiotic and estimated blood loss. These factors are known predictors of surgical site infections and are not readily available in administrative data sets.31
Limitations in our study include the lack of data on agents used during the preparation of the vagina for hysterectomy. Variation in the use of povidone-iodine, saline solution, or 2% chlorhexidine vaginal preparation could affect the conclusion of our study. In addition, there could be several factors such as level of operator experience, variation in surgical technique, and the presence of a trainee that can affect the incidence of surgical site infections. Lastly, because this is a retrospective evaluation, there is a potential of remaining bias as a result of unmeasured confounders.
The choice of preoperative skin antiseptic given before abdominal hysterectomy is an important factor in development of surgical site infections. Chlorhexidine-alcohol skin antiseptic resulted in an overall decreased likelihood of surgical site infections compared with povidone-iodine when adjusted for patient and perioperative factors. Given an overall low rate of surgical site infection after hysterectomy, conducting a clinical trial will require a very large sample size and may be prohibitively expensive. Findings from this retrospective study of patients undergoing abdominal hysterectomy along with randomized controlled trial data from the general surgery literature should prompt the development of guidelines to focus on specifying the chlorhexidine-alcohol-based skin antiseptic before abdominal surgery.
1. Schweizer ML, Cullen JJ, Perencevich EN, Vaughan Sarrazin MS. Costs associated with surgical site infections in Veterans Affairs hospitals. JAMA Surg 2014;149:575–81.
2. Young MH, Washer L, Malani PN. Surgical site infections in older adults: epidemiology and management strategies. Drugs Aging 2008;25:399–414.
3. Merkow RP, Ju MH, Chung JW, Hall BL, Cohen ME, Williams MV, et al. Underlying reasons associated with hospital readmission following surgery in the United States. JAMA 2015;313:483–95.
4. Wright JD, Herzog TJ, Tsui J, Ananth CV, Lewin SN, Lu YS, et al. Nationwide trends in the performance of inpatient hysterectomy in the United States. Obstet Gynecol 2013;122:233–41.
5. Lake AG, McPencow AM, Dick-Biascoechea MA, Martin DK, Erekson EA. Surgical site infection after hysterectomy. Am J Obstet Gynecol 2013;209:490.e1–9.
6. Colling KP, Glover JK, Statz CA, Geller MA, Beilman GJ. Abdominal hysterectomy: reduced risk of surgical site infection associated with robotic and laparoscopic technique. Surg Infect (Larchmt) 2015;16:498–503.
7. U.S. Department of Health and Human Services. U.S. efforts to reduce healthcare-associated infections. Available at: https://www.cms.gov
/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Value-Based-Programs/HAC/Hospital-Acquired-Conditions.html. Retrieved June 9, 2017.
8. Bratzler DW, Dellinger EP, Olsen KM, Perl TM, Auwaerter PG, Bolon MK, et al. Clinical practice guidelines for antimicrobial prophylaxis in surgery. Surg Infect (Larchmt) 2013;14:73–156.
9. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR. Guideline for prevention of surgical site infection, 1999. Centers for Disease Control and Prevention (CDC) Hospital Infection Control Practices Advisory Committee. Am J Infect Control 1999;27:97–132.
10. Chaiyakunapruk N, Veenstra DL, Lipsky BA, Saint S. Chlorhexidine compared with povidone-iodine solution for vascular catheter-site care: a meta-analysis. Ann Intern Med 2002;136:792–801.
11. Mimoz O, Lucet JC, Kerforne T, Pascal J, Souweine B, Goudet V, et al. Skin antisepsis with chlorhexidine-alcohol versus povidone iodine-alcohol, with and without skin scrubbing, for prevention of intravascular-catheter-related infection (CLEAN): an open-label, multicentre, randomised, controlled, two-by-two factorial trial. Lancet 2015;386:2069–77.
12. Darouiche RO, Wall MJ Jr, Itani KM, Otterson MF, Webb AL, Carrick MM, et al. Chlorhexidine-alcohol versus povidone-iodine for surgical-site antisepsis. N Engl J Med 2010;362:18–26.
13. Hendren S, Fritze D, Banerjee M, Kubus J, Cleary RK, Englesbe MJ, et al. Antibiotic choice is independently associated with risk of surgical site infection after colectomy: a population-based cohort study. Ann Surg 2013;257:469–75.
14. Antibiotic prophylaxis for gynecologic procedures. ACOG Practice Bulletin No. 104. American College of Obstetricians and Gynecologists. Obstet Gynecol 2009;113:1180–9.
15. Uppal S, Harris J, Al-Niaimi A, Swenson CW, Pearlman MD, Reynolds RK, et al. Prophylactic antibiotic choice and risk of surgical site infection after hysterectomy. Obstet Gynecol 2016;127:321–9.
16. Mu Y, Edwards JR, Horan TC, Berrios-Torres SI, Fridkin SK. Improving risk-adjusted measures of surgical site infection for the national healthcare safety network. Infect Control Hosp Epidemiol 2011;32:970–86.
17. King G, Roberts ME. How robust standard errors expose methodological problems they do not fix, and what to do about it. Polit Anal 2015;23:159–79.
18. White HA. Heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica 1980;48:817–38.
19. Everitt BS, Rabe-Hesketh S, editors. Maximum likelihood estimation. Handbook of statistical analyses using stata. 4th ed. Boca Raton (FL): Chapman & Hall/CRC; 2007.
20. Centers for Medicaid & Medicare Services. CMS to Improve quality of care during hospital inpatient stays. Available at: http://http://www.cms.gov
/Newsroom/MediaReleaseDatabase/Fact-sheets/2014-Fact-sheets-items/2014-08-04-2.html. Retrieved July 2, 2015.
21. Badia JM, Casey AL, Petrosillo N, Hudson PM, Mitchell SA, Crosby C. Impact of surgical site infection on healthcare costs and patient outcomes: a systematic review in six European countries. J Hosp Infect 2017;96:1–15.
22. Levin I, Amer-Alshiek J, Avni A, Lessing JB, Satel A, Almog B. Chlorhexidine and alcohol versus povidone-iodine for antisepsis in gynecological surgery. J Womens Health (Larchmt) 2011;20:321–4.
23. Noorani A, Rabey N, Walsh SR, Davies RJ. Systematic review and meta-analysis of preoperative antisepsis with chlorhexidine versus povidone-iodine in clean-contaminated surgery. Br J Surg 2010;97:1614–20.
24. Brown TR, Ehrlich CE, Stehman FB, Golichowski AM, Madura JA, Eitzen HE. A clinical evaluation of chlorhexidine gluconate spray as compared with iodophor scrub for preoperative skin preparation. Surg Gynecol Obstet 1984;158:363–6.
25. Tuuli MG, Liu J, Stout MJ, Martin S, Cahill AG, Odibo AO, et al. A randomized trial comparing skin antiseptic agents at cesarean delivery. N Engl J Med 2016;374:647–55.
26. Rogers ML, Nickalls RW, Brackenbury ET, Salama FD, Beattie MG, Perks AG. Airway fire during tracheostomy: prevention strategies for surgeons and anaesthetists. Ann R Coll Surg Engl 2001;83:376–80.
27. Weber SM, Hargunani CA, Wax MK. DuraPrep and the risk of fire during tracheostomy. Head Neck 2006;28:649–52.
28. Bonnet A, Devienne M, De Broucker V, Duquennoy-Martinot V, Guerreschi P. Operating room fire: should we mistrust alcoholic antiseptics? Ann Chir Plast Esthet 2015;60:255–61.
29. Culp WC Jr, Kimbrough BA, Luna S, Maguddayao AJ. Operating room fire prevention: creating an electrosurgical unit fire safety device. Ann Surg 2014;260:214–7.
30. Abdallah C. Perioperative chlorhexidine allergy: is it serious? J Anaesthesiol Clin Pharmacol 2015;31:152–4.
31. Anderson DJ, Podgorny K, Berríos-Torres SI, Bratzler DW, Dellinger EP, Greene L, et al. Strategies to prevent surgical site infections in acute care hospitals: 2014 update. Infect Control Hosp Epidemiol 2014;35:605–27.