Skin Preparation Before Surgery: Options and Evidence

Abstract

Background:

Despite many advances in surgical asepsis, surgical site infection (SSI) remains a challenging and costly problem. Decontamination of the skin with an antiseptic agent is standard practice before any trans-cutaneous invasive procedure, but the antiseptic agent of choice to best reduce the risk of SSI remains controversial.

Methods:

Review of relevant literature.

Results:

Many randomized controlled trials (RCTs) have evaluated chlorhexidine-based and iodine-based preparation solutions, with and without an alcohol component. Most of these trials are underpowered to detect differences in SSI rates. The largest modern RCT showed that a chlorhexidine-isopropyl alcohol preparation reduced the risk of SSI substantially compared with a povidone-iodine preparation without alcohol in clean-contaminated surgery. Many smaller RCTs have shown that chlorhexidine-isopropyl alcohol is superior to povidone-iodine plus isopropyl alcohol or iodine povacrylex plus isopropyl alcohol in pre-operative skin decontamination; whether or not this translates into lower SSI rates is unknown. A mixed treatment comparison of 10 RCTs concluded that alcohol-based preparations have a 98% probability of reducing the risk of SSI more effectively than aqueous-based preparations. Non-randomized observational studies have generally found no difference in SSI rates among various skin antiseptic preparations.

Conclusions:

Alcohol-based agents are likely superior to aqueous agents. Chlorhexidine may decrease SSI rates compared with povidone-iodine, and chlorhexidine-isopropyl alcohol likely offers better skin decontamination before clean surgery than povidone-iodine plus isopropyl alcohol or iodine povacrylex plus isopropyl alcohol. The quality of the available data is moderate. Rigorous, well-powered RCTs with appropriate treatment comparisons are needed to establish the optimal and most cost-effective pre-operative skin preparation in various operations and wound classifications.

Among all hospital-acquired infections, 14%–22% are surgical site infections (SSIs) [1,2]. In surgical patients, SSIs are the most common nosocomial infection [3]. Additional morbidity, mortality, and cost associated with SSIs vary by location and surgery but are substantial [4]. The economic impact of SSIs in the United States exceeds $10 billion annually; SSIs are associated with an additional $3,000–$60,000 in the cost of a patient's hospital stay [5,6]. In Europe, SSIs add an additional €1.5–€19.1 billion to health care expenditures annually [7]. Surgical site infections are associated with an increased hospital length of stay and with increased mortality [2].

The Role of Pre-Operative Skin Preparation with Topical Antiseptics in Preventing SSIs

Decontamination of the skin with an antiseptic agent is the standard of care prior to any trans-cutaneous invasive procedure [2,3,8]. With the availability of newer antiseptic agents, there is renewed interest in topical antiseptics as a means of preventing SSIs [9]. Whereas skin preparation with a topical antiseptic is unlikely to prevent SSIs caused by flora from the gastrointestinal, urinary, biliary or respiratory systems, it is likely to prevent SSI caused by skin flora [10,11].

The U.S. Centers for Disease Control and Prevention (CDC) divides SSIs into superficial incisional, deep incisional, and organ/space SSIs [12]. Two-thirds of SSIs are superficial [3]. Organ/space infections are more likely to result from technical surgical problems (e.g., anastomotic or organ leaks, accidental entry into contaminated spaces) and less likely to be caused by contamination from skin flora. Deep incisional SSIs can be caused by extension of a superficial or organ/space infection, by contamination with bacteria from organs violated during surgery, or by contamination from skin flora.

The purpose of topical antiseptic preparation of the skin is to reduce the burden of skin flora, thereby reducing the likelihood of SSI caused by these organisms [3,6,13 –15]. The most common organisms isolated from SSIs are listed in Table 1. Topical antiseptic skin preparation is more likely to be effective at reducing SSIs than anti-microbial dressings [16,17], anti-microbial sutures [18,19], and peri-operative use of topical antibiotics [20]. Whereas skin antiseptics have never been compared with placebo experimentally, and doing so today would be unethical, the dramatic decrease in wound suppuration and the associated constitutional symptoms of severe “wound” infections observed by Lister [21] when he first used “carbolic acid” (phenol) as a surgical antiseptic in the mid-1800s make clear that topical preparation of the skin is a vital surgical practice.

Table 1.

Most Common Isolates from Surgical Site Infections [ 11 ]

Pathogen	Proportion of isolates (%)
Staphylococcus aureus	30.0
Coagulase-negative staphylococci	13.7
Enterococcus spp.	11.2
Escherichia coli	9.6
Pseudomonas aeruginosa	5.6
Enterobacter spp.	4.2
Klebsiella pneumonia	3.0
Klebsiella oxytoca	0.7
Acinetobacter baumannii complex	0.6
Other	21.4

Current Options in Antiseptic Skin Preparations

The ideal skin antiseptic agent should kill all bacteria, fungi, viruses, protozoa, and spores on the skin; be nontoxic, hypoallergenic, and nonabsorbable; have residual activity after application; and be safe for repeated use [22]. The three most widely used pre-operative skin preparation agents are chlorhexidine gluconate, povidone-iodine/iodine povacrylex, and isopropyl alcohol, although other options are available (Tables 2 and 3).

Table 2.

Commercially Available Pre-Operative Skin Preparation Agents

Skin preparation agent	Comments
Tincture of iodine	Rarely used since introduction of safer and more effective iodophores.
Triclosan	Found in many household items. Effectiveness as a pre-operative antiseptic skin preparation is unproven.
Parachlorometaxylenol (PCMX)	Used as a pre-operative hand/forearm scrub, but inferior to chlorhexidine and povidone-iodine [45].
Sodium hypochlorite	Excellent chemical disinfectant on non-porous surfaces, but limited effectiveness on skin [46].
Chlorhexidine gluconate	Broad-spectrum activity, persistent skin activity, retains effectiveness on contact with organic material, minimal systemic absorption. Direct instillation may damage inner ear, eyes, and nerve tissue; poorly sporicidal [6,47 –50].
Povidone-iodine	Broad-spectrum activity including sporicidal activity. Diminished effectiveness on contact with organic material, variable persistence on the skin, skin irritant, slow-acting [49].
Iodine povacrylex	Broad-spectrum activity of povidone-iodine in an acrylate copolymer, believed to resist removal by organic materials and improve skin persistence [51].
Isopropyl alcohol	Broad-spectrum, rapid action. Poorly sporicidal, no persistence once evaporated, drying to skin and mucous membranes, highly flammable [47,49].

Table 3.

Commercially Available Formulations of Chlorhexidine, Povidone-Iodine, Iodine Povacrylex, and Isopropyl Alcohol

1. Povidone-iodine aqueous solution in 5%, 7.5%, and 10% solutions (various manufacturers)

2. One-step 10% aqueous povidone-iodine paint (Betadine^® Solution Swabsticks, Purdue Products, Stamford CT)

3. Two-step 7.5% aqueous povidone-iodine scrub and 10% povidone-iodine paint (Scrub Care Skin Prep Tray^®, Cardinal Health, Dublin, OH)

4. Iodine povacrylex (0.7% available iodine) in 74% isopropyl alcohol paint (DuraPrep^®, 3M, St. Paul, MN)

5. Aqueous chlorhexidine in 0.5%, 0.75%, 2% and 4% concentrations (various manufacturers)

6. 2% chlorhexidine in 70% isopropyl alcohol paint (ChloraPrep^®, Cardinal Health)

7. Isopropyl alcohol alone in various concentrations (various manufacturers)

Randomized Controlled Trials Comparing Chlorhexidine (with and without Isopropyl Alcohol) and Povidone-Iodine

Two well-powered RCTs have compared the efficacy of a chlorhexidine-70% isopropyl alcohol preparation solution to povidone-iodine without an alcohol component. Darouiche et al. [23] performed an industry-sponsored prospective multi-center RCT comparing 2% chlorhexidine-70% isopropyl alcohol (ChloraPrep^®, Cardinal Health, Dublin, OH) to a two-stage 7.5% povidone-iodine scrub and 10% povidone-iodine paint without an alcohol component (Scrub Care Skin Prep Tray^®, Cardinal Health). The study included 813 adults undergoing clean-contaminated surgery (including abdominal, thoracic, gynecologic, and urologic procedures); excluded were patients with evidence of infection at the operative site and those who were unlikely to be available for 30-d follow-up after surgery. Investigators were blinded to patient treatment assignments throughout the study. The study was powered at 90% to detect a 50% decrease in the SSI rate from a baseline rate of 14% in the povidone-iodine group. Surgical sites were evaluated once daily during hospitalization by study investigators. After discharge, once-weekly telephone calls were made to patients until the end of the 30-d follow-up period; clinical evaluation was arranged whenever SSI was suspected. The CDC definitions of SSI were used.

Darouiche et al. [23] reported SSI rates of 9.5% in the chlorhexidine-isopropyl alcohol group and 16.1% in the povidone-iodine group. Although there was a substantial reduction in the rate of superficial incisional (4.2% vs. 8.6%, p=0.008) and deep incisional (1% vs. 3%, p=0.05) SSIs in the chlorhexidine-isopropyl alcohol group, there was no statistically significant difference in organ/space SSIs (4.4% vs. 4.5%, p>0.99) or in sepsis caused by SSIs (2.7% vs. 4.3%, p=0.26). This trial did not differentiate the effect of chlorhexidine from that of isopropyl alcohol over povidone-iodine [24].

Sistla et al. [25] performed a prospective, single-center RCT in India comparing 10% povidone-iodine without an alcohol component to 2.5% chlorhexidine-70% ethanol. The study randomized 556 patients undergoing elective open inguinal hernia repair. Only those patients undergoing mesh repair (33%) received pre-operative prophylactic antibiotics. Patients with recurrent or complicated inguinal hernias were excluded. Blinding of study investigators was unclear. The study was powered at 80% to detect a reduction in SSI rate from the presumed baseline of 10% to 3%. How SSI was assessed was unclear but authors stated the CDC definitions of SSI were used. Lost to follow-up were 156 of 556 randomized patients (28.1%).

Sistla et al. [25] reported that 62% of patients underwent suture repair (herniorrhaphy), 33% underwent mesh repair (hernioplasty), and 5% underwent herniotomy. They reported an SSI rate of 9.5% in the povidone-iodine group and 7% in the chlorhexidine-alcohol group (p=0.36).

Paocharoen et al. [26] performed a single center RCT in Thailand comparing SSI and bacterial colonization after a two-stage 4% chlorhexidine-70% isopropyl alcohol preparation to a two-stage povidone-iodine preparation without alcohol, the concentration of which was not reported. The study included 500 patients undergoing surgery with clean, clean-contaminated, and contaminated wounds, and excluded patients with dirty wounds, those taking “immunosuppressive drugs,” and those with a serum albumin concentration less than 3 mg/dL. The types of procedures patients underwent; information on blinding of investigators, patients, or surgeons; and a power calculation were not reported. Surgical incisions were examined twice in the week after surgery, and once weekly for the next 3 wks. An incisional SSI was recorded if a wound drained pus or if the surgeon opened it. No information was reported on organ/space SSIs.

Paocharoen et al. [26] reported an incisional SSI rate of 2% in the chlorhexidine-isopropyl alcohol group and 3.2% in the povidone-iodine group. They also reported that 14.4% of skin cultures were positive after preparation in the chlorhexidine-isopropyl alcohol group compared with 31.2% after preparation with povidone-iodine. p values were not reported in this trial but the authors state that both of these comparisons were statistically significant. The authors did not comment on other outcomes such as organ/space SSIs, and as with the trial by Darouiche et al. [23], their trial did not differentiate the effect of chlorhexidine from that of isopropyl alcohol over povidone-iodine.

Rodrigues et al. [27] performed a pseudo-randomized single center RCT in Brazil comparing a two-step povidone-iodine plus alcohol preparation (povidone-iodine scrub followed by painting with “10% hydroalcoholic povidone-iodine”) with a two-step chlorhexidine-alcohol preparation (2% chlorhexidine soap scrub followed by painting with “0.5% alcoholic chlorhexidine”). The study included 208 patients undergoing open clean and potentially contaminated operations with subcostal, vertical abdominal, or thoracic incisions. Patients were assigned to one or the other preparation by their order of arrival in the operating room. Use of pre-operative prophylactic antibiotics was not reported clearly, but all patients did not receive prophylactic antibiotics. Patients with immunosuppression and those who were lost to follow-up (one patient) were excluded. A single observer examined patients on the third, seventh, and thirtieth post-operative days; information on blinding was not reported. Surgical site infection was defined as: fever without other apparent cause; pain, heat, swelling, or confluent erythema around and extending beyond the incision; pus at the incision or in the deep soft tissue; pus in an organ or cavity handled during the operation; presence of an abscess; radiologic evidence of infection; positive cultures from a normally sterile site; or spontaneous dehiscence of deep tissues. A power calculation was not reported.

Rodrigues et al. [27] reported SSI rates separately for clean operations and potentially contaminated operations. Thirty-nine patients in the povidone-iodine plus alcohol group and 52 patients in the chlorhexidine-alcohol group underwent clean operations, with SSI rates of 5.1% and 15.3%, respectively (p=0.18). Sixty-three patients in the povidone-iodine plus alcohol group and 51 patients in the chlorhexidine-alcohol group underwent potentially contaminated operations, with SSI rates of 7.9% and 5.8%, respectively (p=0.72).

Only one RCT, conducted in the United Kingdom by Berry et al. [28] in 1982 was powered sufficiently to detect differences in rates of SSIs among patients who underwent chlorhexidine-isopropyl alcohol skin preparation versus those who underwent povidone-iodine plus isopropyl alcohol skin preparation. The skin preparations used were 10% povidone-iodine in alcohol or 0.5% chlorhexidine in “spirit”; the types of solvents and their concentrations were unreported. Their study included 866 adults undergoing elective abdominal, genital, and vein surgery and included clean, clean-contaminated, and contaminated incisions. No exclusion criteria were reported, but all exclusions were reportedly made before allocation to a treatment group. A power calculation was not reported. Surgical incisions were assessed on post-operative days three or four and on the day of discharge; patients discharged before 3 d were assessed at discharge. “Blinding” was accomplished by having staff who had no knowledge of the skin preparation used assess the surgical incisions, but state that “on occasion” incisions were assessed by staff who did know what skin preparation agent had been used. Incisions were categorized as “normal,” “erythematous,” “edematous,” “discharging,” or “purulent” in a standardized format by two observers independently, but definitions of these terms were not reported. An SSI was recorded when both observers reported the incision to be anything other than “normal.”

Berry et al. [28] reported an SSI rate of 9.7% in the chlorhexidine-isopropyl alcohol group compared with 14.8% in the povidone-iodine plus isopropyl alcohol group (p=0.03). Although this RCT was performed in a scientifically rigorous manner, blinding was inconsistent, SSI definitions were non-standard, and follow-up duration was insufficient. Furthermore, other than before colon surgery, where patients received pre-operative oral metronidazole and neomycin, no pre-operative antibiotic prophylaxis was used.

Many smaller RCTs (Ostrander et al. [29], Veiga et al. [30], Cheng et al. [31], Saltzman et al. [32], and Savage et al. [33]) that were not powered to detect differences in rates of SSI used the proxy measure of skin decontamination to compare the effectiveness of various preparations of chlorhexidine-isopropyl alcohol to either povidone-iodine plus isopropyl alcohol or iodine povacrylex plus isopropyl alcohol in clean orthopedic, plastic, podiatric, and spine surgery. Three of the five studies concluded that chlorhexidine-isopropyl alcohol solutions decontaminated the skin more effectively than iodine-isopropyl alcohol solutions, whereas two found no differences between the types of skin preparation. Two of the three that compared the commonly used 0.7% available iodine povacrylex-74% isopropyl alcohol and 2% chlorhexidine-70% isopropyl alcohol agents found that the latter provided superior skin decontamination, whereas the third found no difference between the two.

Another small RCT [34] compared 50 patients undergoing elective transvaginal hysterectomy whose vaginas were prepared with 10% povidone-iodine or 4% chlorhexidine. The chlorhexidine group had fewer positive cultures 30 min after the pre-operative preparation (63% vs. 22%, p=0.003), but subsequent samples obtained prior to the end of the operation demonstrated no substantial differences between the preparations. Table 4 summarizes the patients, operations, interventions, and findings of these RCTs.

Table 4.

Comparison Groups and Main Findings of Randomized (and One Pseudo-Randomized) Controlled Trials

Lead author (year)	Number of subjects (patients or limbs)	Type(s) of surgery	Interventions	Findings
Rodrigues (2013) [27] (pseudo-randomized)	208	Open clean and potentially contaminated surgery with subcostal abdominal, vertical abdominal, or thoracic incisions	Two-stage povidone-iodine scrub and 10% hydroalcoholic povidone-iodine paint vs. Two-step 2% chlorhexidine soap scrub and 0.5% alcoholic chlorhexidine paint	5.1% and 7.9% SSI rate in povidone-iodine plus alcohol group in clean and potentially contaminated surgery, respectively vs. 15.3% (p=0.18 vs povidone-iodine plus alcohol) and 5.8% (p=0.72 vs. povidone-iodine-alcohol) SSI rate in chlorhexidine-alcohol group in clean and potentially contaminated surgery, respectively
Darouiche (2010) [23]	813	Elective clean-contaminated surgery, including abdominal, thoracic, gynecologic, and urologic surgery	Two-stage 7.5% povidone-iodine scrub and 10% paint (no alcohol) vs. 2% chlorhexidine-70% isopropyl alcohol (ChloraPrep^®)	16.1% SSI rate in povidone-iodine group vs. 9.5% SSI rate in chlorhexidine-isopropyl alcohol group p=0.004
Sistla (2010) [25]	556	Elective open inguinal hernia repair	10% povidone-iodine (no alcohol) vs. 2.5% chlorhexidine-70% ethanol	9.5% SSI rate in povidone-iodine group vs. 7% SSI rate in chlorhexidine-alcohol group p=0.36
Paocharoen (2009) [26]	500	Clean, clean-contaminated, and contaminated surgery	2-stage povidone-iodine (no alcohol) vs. Two-stage 4% chlorhexidine-70% isopropyl alcohol	3.2% incisional SSI rate in povidone-iodine group vs. 2% incisional SSI rate in chlorhexidine-isopropyl alcohol group p value not reported, but stated to be significant
Berry (1982) [28]	866	Elective clean, clean-contaminated, and contaminated operations, including abdominal, genital, and vein surgery	10% povidone-iodine-alcohol (concentration unreported) vs. 0.5% chlorhexidine in spirit (concentration unreported)	14.8% SSI rate in povidone-iodine plus alcohol group vs. 9.7% SSI rate in chlorhexidine-isopropyl alcohol group p=0.03
Ostrander (2005) [29]	250	Elective clean foot and ankle surgery	Iodine povacrylex (0.7% available iodine)-74% isopropyl alcohol vs. 3% chloroxylenol vs. 2% chlorhexidine-70% isopropyl alcohol	Positive cultures from 65% of hallux samples, 45% of toe samples, and 23% of anterior tibia samples (controls) in iodine povacrylex-isopropyl alcohol group vs. Positive cultures from 30% of hallux samples, 23% of toe samples, and 10% of anterior tibia samples (controls) in chlorhexidine-isopropyl alcohol group Hallux: p<0.01 Toes: p<0.05
Veiga (2008) [30]	250	Elective clean plastic surgery	10% povidone-iodine-isopropyl alcohol (concentration unreported) vs. 0.5% chlorhexidine-isopropyl alcohol (concentration unreported)	7.9 CFUs on hypertonic mannitol agar and 17.6 on blood agar in povidone-iodine-isopropyl alcohol group vs. 2.7 CFUs on hypertonic mannitol agar and 7.8 on blood agar in chlorhexidine-isopropyl alcohol group Hypertonic mannitol: p=0.006 Blood agar: p=0.014
Cheng (2009) [31]	50	Elective clean forefoot surgery	1% povidone-iodine-23% isopropyl alcohol vs. 0.5% chlorhexidine-70% isopropyl alcohol	67 CFUs before preparation, 4 after in povidone-iodine-isopropyl alcohol group vs. 58 CFUs before skin preparation, 4 after in chlorhexidine-isopropyl alcohol group p=not significant
Saltzman (2009) [32]	150	Elective clean shoulder surgery, mostly arthroscopic	Iodine povacrylex (0.7% available iodine)-74% isopropyl alcohol vs. 2% chlorhexidine-70% isopropyl alcohol	19% grew positive cultures in iodine povacrylex-isopropyl alcohol group vs. 7% grew positive cultures in chlorhexidine-isopropyl alcohol group p=0.01
Savage (2012) [33]	100	Elective lumbar spine surgery	Iodine povacrylex (0.7% available iodine)-74% isopropyl alcohol vs. 2% chlorhexidine-70% isopropyl alcohol	6/50 positive cultures after preparation, 32/50 positive cultures after wound closure in iodine povacrylex-isopropyl alcohol group vs. 0/50 positive cultures after preparation, 34/50 positive cultures after wound closure in chlorhexidine-isopropyl alcohol group After preparation: p=0.24 After closure: p=0.22
Culligan (2005) [34]	50	Elective transvaginal hysterectomy	10% povidone-iodine (no alcohol) vs. 4% chlorhexidine (no alcohol)	63% positive cultures after 30 min, 36% after 90 min, 50% after 150 min in povidone-iodine group vs. 22% positive cultures after 30 min, 14% after 90 min, 14% after 150 min in chlorhexidine group After 30 min: p=0.003 After 90 min: p=0.12 After 150 min: p=0.17

CFUs=colony forming units; SSI=surgical site infection.

Meta-Analyses

A meta-analysis by Dumville et al. [22] for the Cochrane Collaboration analyzed 13 RCTs with 2,632 total patients that compared pre-operative skin preparation agents in elective clean surgery. Eleven different comparisons were made in these 13 studies; five studies compared iodine-containing preparations with chlorhexidine-containing preparations. A mixed treatment comparison analysis concluded that alcohol-containing preparations had the highest probability of effectiveness for reducing the risk of SSI. The authors reported a 78% probability that 4% chlorhexidine scrub-70% isopropyl alcohol was the best preparation for preventing SSIs, followed by povidone-iodine plus isopropyl alcohol with a 16% probability of being the best preparation. However, the authors emphasized that the available evidence is of low quality and at uncertain risk of bias. The only outcome included in the analysis of Dumville et al. [22] is SSI, and the analysis was limited to studies of clean surgery.

Another meta-analysis by Lee et al. [35] included 9 RCTs with 3,614 patients, analyzed skin culture results as well as SSIs, and included trials that examined clean and clean-contaminated surgery. Meta-analysis of the seven studies comparing SSI after chlorhexidine preparation (with or without isopropyl alcohol) and after povidone-iodine preparation (with or without alcohol) showed that chlorhexidine decreased the risk of SSI by 36% compared with povidone-iodine (p<0.0001). The authors reported that the evidence was of moderate quality in assessing this outcome. Meta-analysis of the four studies comparing skin culture results after the same preparations showed that chlorhexidine decreased the risk of positive skin cultures by 56% compared with povidone-iodine (p<0.00001). The authors reported moderate-quality evidence for this outcome as well, but also reported potential publication bias in the culture data. They found no differences in the effectiveness of chlorhexidine-based preparations based on their different concentrations of chlorhexidine.

Lee et al. [35] also performed a cost-effectiveness analysis, using their findings regarding the efficacy of chlorhexidine-isopropyl alcohol and the operative experience of the Hospital of the University of Pennsylvania (a tertiary/quaternary care center in Philadelphia, PA) as a model. This analysis showed that the net cost savings threshold for 2% chlorhexidine-70% isopropyl alcohol occurred at an incremental risk reduction of 10%. In other words, if 2% chlorhexidine-70% isopropyl alcohol was at least 10% more effective than povidone-iodine in prevention of SSIs, then using the chlorhexidine-isopropyl alcohol preparation would save money by preventing costly SSIs, despite its greater cost of the chlorhexidine product.

Data from Non-Randomized Studies

Non-randomized studies have also compared skin preparation with povidone-iodine (with or without isopropyl alcohol) to chlorhexidine (with or without isopropyl alcohol) [36 –41]. We used the Methodological Index for Non-Randomized Studies (MINORS) [42] to evaluate the methodological quality of these studies. The MINORS is a validated scoring system designed specifically to assess the methodological soundness of non-randomized or observational surgical trials. Comparative observational studies are rated on 12 separate criteria, each of which is scored 0 (not reported), 1 (reported but inadequate), or 2 (reported and adequate), for a maximum, ideal score of 24. Table 5 lists the MINORS criteria and rates the observational studies available in the literature. We discuss in detail below the two trials with the highest MINORS rating.

Table 5.

MINORS [ 42 ] Criteria for Observational Studies Comparing Chlorhexidine with or without Alcohol to Povidone-Iodine with or without Alcohol

MINORS criteria	Hakkarainen ^a (2014) [36]	Young (2013) [37]	Amer-Alshiek (2013) [38]	Menderes (2012) [39]	Levin (2011) [40]	Swenson ^a (2009) [41]
Clearly stated aim	2	2	2	2	2	2
Inclusion of consecutive patients	0	2	2	2	0	2
Prospective collection of data	2	1	1	1	0	2
End points appropriate to aim of study	2	1	1	1	1	2
Unbiased assessment of study endpoint	1	1	1	1	1	1
Follow-up period appropriate to aim of study	1	0	0	1	1	2
Lost to follow-up <5%	2	0	0	0	0	0
Prospective calculation of the study size	0	0	0	2	0	2
Adequate control group	2	2	2	2	2	2
Contemporary groups	2	1	1	1	1	1
Baseline equivalence of groups	2	2	2	1	2	1
Adequate statistical analysis	2	2	2	2	2	2
Total score	18	14	14	16	12	19

0: not reported; 1: reported but inadequate; 2: reported and adequate

These studies had the highest MINORS scores and are discussed in detail in the results section.

Swenson et al. [41] compared three skin preparations: A two-step povidone-iodine scrub with isopropyl alcohol between povidone-iodine applications, a one-step chlorhexidine-isopropyl alcohol combined scrub, and a one-step iodine povacrylex plus isopropyl alcohol combined scrub. The study included 3,209 adults undergoing general surgical procedures, and was divided into three temporal periods: in period 1 the first preparation was used on all patients, in period two the second was used on all patients, and in period three the third was used on all patients. They found the lowest rate of SSI, 3.9%, occurred in period three, compared with 6.4% in period 1 and 7.1% in period two.

This study was designed prospectively but was a non-randomized single-center experience. Period three had significantly fewer patients with sepsis, and significantly shorter operative times, which may account for its lower rate of SSI. This study is the outlier in the data comparing chlorhexidine-based and iodine-based preparations in that it finds chlorhexidine-alcohol to be inferior to povidone-iodine and iodine povacrylex.

Hakkarainen et al. [36] published recently the results of an analysis based on the Comparative Effectiveness Translational Network and the Surgical Care and Outcomes Assessment Program (SCOAP) in the state of Washington. They followed 7,669 patients at 47 hospitals undergoing elective surgery prospectively from January 2011 to June 2012. Patients were considered to have SSI if the SCOAP data indicated SSI, if the data abstractor discovered that wound edges were reopened, an abscess was drained, or a drainage procedure was undertaken. Patients were assessed for SSI only while admitted to the hospital; a subgroup analysis was done of patients who underwent elective colorectal procedures and had a hospital length of stay greater than 10 d. In the main analysis and this subgroup analysis the authors found no differences in rates of SSIs between patients prepared with chlorhexidine or povidone-iodine, or between patients prepared with or without isopropyl alcohol. Table 6 summarizes the comparison groups and main findings of these two observational studies.

Table 6.

Comparison Groups and Main Findings of Observational Comparative-Effectiveness Studies with MINORS [ 42 ] Score of Eighteen or Greater

Study first author (year)	Comparison groups	Findings
Swenson (2009) [41]	Two-step povidone-iodine scrub w/isopropyl alcohol	6.4% SSI rate in povidone-iodine plus isopropyl alcohol group
	vs.	vs.
	Iodine povacrylex-74% isopropyl alcohol	3.9% SSI rate in iodine povacrylex-isopropyl alcohol group
	vs.	vs.
	2% chlorhexidine-70% isopropyl alcohol	7.1% SSI rate in chlorhexidine-isopropyl alcohol group
		Comparing chlorhexidine/isopropyl alcohol to both other groups: p=0.001
Hakkarainen (2014) [36]	Povidone-iodine (no alcohol)	6% SSI rate in povidone-iodine group
	vs.	vs.
	Iodine povacrylex-isopropyl alcohol	5.3% in iodine povacrylex-isopropyl alcohol group
	vs.	vs.
	Chlorhexidine (no alcohol)	4% SSI rate in chlorhexidine group
	vs.	vs.
	Chlorhexidine-isopropyl alcohol	5.3% SSI rate in chlorhexidine-isopropyl alcohol group
		No significant differences between comparison groups after adjustment for other risk factors

SSI=surgical site infection.

Discussion

There are several possible explanations for the contradiction between the data from the well-powered RCTs and well-conducted prospective but non-randomized observational studies performed in the United States. First, RCTs often involve dedicated, blinded data collectors, strict definitions, standardized processes, and typically follow patients for longer periods of time. For this reason, they tend to find a greater rate of superficial and deep incisional SSIs than observational studies that depend on administrative data or data extracted from clinical documentation. Swenson et al. [41] and Darouiche et al. [23] reported approximately the same proportion of superficial/deep incisional SSIs to organ/space SSIs, whereas Hakkarainen et al. [36] and Paocharoen et al. [26] did not report this proportion or data with which it could be calculated. Second, as Hakkarainen et al. [36] discussed, results from RCTs are often not reproducible in the less tightly controlled conditions of real-world clinical practice. This may be because unknown or unmeasured confounders and selection bias exist in the observational data, and which are controlled for by randomization. Third, it may be that the intervention in the RCT is not the same intervention that is implemented in routine practice. For example, it is unlikely that pre-operative skin preparation throughout the state of Washington is carried out according to exactly the same protocol used in the Darouiche et al. [23] study. Fourth, an RCT may not be generalizable because it studied a restricted patient population with pre-specified inclusion and exclusion criteria.

Hakkarainen et al. [36] cited a study by Tschudin-Sutter et al. [43] as another possible explanation for the discrepancy between the randomized and non-randomized data: the latter authors found no correlation between positive skin cultures after preparation with povidone-iodine plus isopropyl alcohol and subsequent SSI in 1,005 consecutive patients undergoing vascular, visceral, cardiac, thoracic, orthopedic, neurologic, and gynecologic surgery; 91.3% of cases were classified as Class I/clean wounds. They reported that 5.5% of patients with bacterial growth on culture after skin preparation developed SSIs, whereas 3.99% of patients without growth developed SSIs (p=0.64). However, in this study, the majority of SSIs (56.1%) were organ/space infections, whereas 31.7% were deep incisional and only 12.2% were superficial. This is nearly the inverse of the historically observed ratio, wherein two-thirds of SSIs are superficial or deep incisional SSIs and one-third are organ/space SSIs [3]. With such findings it is not surprising that persistent skin flora after skin preparation would correlate poorly with SSI.

Although underpowered, the studies by Sistla et al. [25] and Rodrigues et al. [27] raise important differences between the patient populations under study. Surgical site infection is the most common health care-associated infection in hospitals in the developing world, with SSI rates 1.9–3.5 times greater per surgical procedure than in high-income countries [44]. There are many potential reasons for this difference: resource constraints; inadequate health infrastructure; poorer nutritional status, housing, access to clean water, and hygiene of patients; shaving of skin hair instead of clipping; lack of pre-operative antiseptic showering; more restrictive use of pre-operative antibiotic prophylaxis; longer pre-operative and post-operative hospital length of stay; and the greater prevalence of smoking all likely play a role. It may be the case that patients with the economic and social advantages of living in high-income countries undergoing operations in advanced health care systems will experience a lower SSI rate when a more effective skin preparation agent is utilized, but this benefit may not carry over to patients in the developing world who have many more risk factors for SSIs.

Limitations and Conclusions

Overall, we agree with Lee et al. [35] that the best available data on pre-operative skin preparation are only of moderate quality. This area of research is characterized by trials that are underpowered to detect rare events such as SSIs, and instead have utilized the plausible but unproven proxy measure of reduction of positive skin cultures. Whereas most trials used the CDC classification for SSIs [12], some of the older trials did not. The only modern well-powered RCT to date compared chlorhexidine-isopropyl alcohol with povidone-iodine without an alcohol component, and was sponsored by a manufacturer of skin preparations.

Future research in this area must include well-designed prospective RCTs with sufficient power to detect differences in rates of SSI in a variety of surgical procedure groups. Given the high and increasing volume of surgical procedures performed worldwide and the major expense of SSIs, governments and private payers should fund such studies in order to establish clearly what preparations are most effective for preventing SSIs across the gamut of surgical procedures, and are most cost-effective overall. Care should also be taken when interpreting results from the cited studies. As Maiwald and Chan [24] discussed, studies comparing combined chlorhexidine-isopropyl alcohol preparations with povidone-iodine preparations without alcohol have been cited as proof of the superior efficacy of chlorhexidine, a misinterpretation of the actual treatment comparison made.

From the available data we conclude that skin preparation should include an alcohol component, because alcohol is inexpensive, safe, and effective, and the network analysis performed by Dumville et al. [22] showed alcohol-based preparations to have the highest likelihood of being most effective at reducing surgical site infections, whether the alcohol is combined with chlorhexidine or with povidone-iodine. Chlorhexidine-isopropyl alcohol preparations may be more effective at preventing SSIs than povidone-iodine alone in clean-contaminated surgery, and chlorhexidine-isopropyl alcohol is likely a more effective skin-decontaminating agent than either povidone-iodine plus isopropyl alcohol or iodine povacrylex plus isopropyl alcohol in clean surgery. Finally, despite its greater cost, chlorhexidine-isopropyl alcohol may offer a cost savings over povidone-iodine through reduction of costly SSIs, shortened hospital stays, and decreased need for further interventions and treatments. However, strong conclusions about cost effectiveness of different skin preparation agents cannot be made based on the available data, and will depend on local factors as well as the relative efficacy of different skin preparation agents.

Footnotes

Author Disclosure Statement

Dr. Feroze Sidhwa has no conflicts of interest. Dr. Kamal M.F. Itani has received research funds to his institution from Merck, Sanofi, and Dr. Reddy's laboratories. He was a research consultant for Irrimax.

References

Magill

, Edwards

, Banberg

, et al. Multistate point-prevalence survey of health care-associated infections. N Engl J Med, 2014; 370:1198–1208.

Surgical site infection. In: NICE Clinical Guideline. London: National Health Service. National Institute for Health and Clinical Excellence, Office of Public Sector Information, 2008.

Mangram

, Horan

, Pearson

, et al. Guideline for prevention of surgical site infection, 1999. Hospital Infection Control Practices Advisory Committee. Infect Control Hosp Epidemiol, 1999; 20:250–278.

Perencevich

, Sands

, Cosgrove

, et al. Health and economic impact of surgical site infections diagnosed after hospital discharge. Emerg Infect Dis, 2003; 9:196–203.

Won

, Wong

. Surgical site infection. In: Mayhall

(ed): Hospital Epidemiology and Infection Control. Philadelphia: Lippincott Williams & Wilkins, 2012.

Anderson

, Kaye

, Classen

, et al. Strategies to prevent surgical site infections in acute care hospitals. Infect Control Hosp Epidemiol, 2008; 29(Suppl 1):S51–S61.

Leaper

, van Goor

, Reilly

, et al. Surgical site infection—A European perspective of incidence and economic burden. Int Wound J, 2004; 1:247–273.

Conner

, Blanchard

; Association of Operating Room Nurses. Perioperative Standards and Recommended Practices. Denver: Association of periOperating Registered Nurses, 2011.

Leaper

. Surgical-site infection. Br J Surg, 2010; 97:1601–1602.

10.

Owens

, Stoessel

. Surgical site infections: Epidemiology, microbiology and prevention. J Hosp Infect, 2008; 70:3–10.

11.

Hidron

, Edwards

, Patel

, et al. NHSN annual update: Antimicrobial-resistant pathogens associated with healthcare-associated infections: Annual summary of data reported to the National Healthcare Safety Network at the Centers for Disease Control and Prevention, 2006–2007. Infect Control Hosp Epidemiol, 2008; 29:996–1011.

12.

Horan

, Gayness

, Martone

, et al. CDC definitions of nosocomial surgical site infections, 1992: A modification of CDC definitions of surgical wound infections. Am J Infect Control, 1992; 20:271–274.

13.

Krizek

, Robson

. Evolution of quantitative bacteriology in wound management. Am J Surg, 1975; 130:579–584.

14.

Garibaldi

, Skolnick

, Lerer

, et al. The impact of preoperative skin disinfection on preventing intraoperative wound contamination. Infect Control Hosp Epidemiol, 1988; 92:109–113.

15.

Moen

, Noone

, Kirson

. Povidone-iodine spray technique versus traditional scrub-paint technique for preoperative abdominal wall preparation. Am J Obstet Gynecol, 2002; 187:1434–1437.

16.

Mueller

, Krebsbach

. Impact of an antimicrobial-impregnated gauze dressing on surgical site infections including methicillin-resistant Staphylococcus aureus infections. Am J Infect Control, 2008; 36:651–655.

17.

, Litton

. Use of chlorhexidine-impregnated dressing to prevent vascular and epidural catheter colonization and infection: A meta-analysis. J Antimicrob Chemother, 2006; 58:281–287.

18.

Galal

, El-Hindawy

. Impact of using triclosan-antibacterial sutures on incidence of surgical site infection. Am J Surg, 2011; 202:133–138.

19.

Rozzelle

, Leonardo

, Li

. Antimicrobial suture wound closure for cerebrospinal fluid shunt surgery: A prospective, double-blinded, randomized controlled trial. J Neurosurg Pediatr, 2008; 2:111–117.

20.

McHugh

, Collins

, Corrigan

, et al. The role of topical antibiotics used as prophylaxis in surgical site infection prevention. J Antimicrob Chemother, 2011; 66:693–701.

21.

Lister

. Antiseptic principle in the practice of surgery. Br Med J, 1867; 2:9–12.

22.

Dumville

, McFarlane

, Edwards

, et al. Preoperative skin antiseptics for preventing surgical wound infections after clean surgery. Cochrane Database Syst Rev, 2013; 3:CD003949.

23.

Darouiche

, Wall

Jr , Itani

, et al. Chlorhexidine-alcohol versus povidone-iodine for surgical-site antisepsis. N Engl J Med, 2010; 362:18–26.

24.

Maiwald

, Chan

ESY

. The forgotten role of alcohol: A systematic review and meta-analysis of the clinical efficacy and perceived role of chlorhexidine in skin antisepsis. PLoS One, 2012; 7:e44277.

25.

Sistla

, Prabhu

, Sistia

, et al. Minimizing wound contamination in a ‘clean’ surgery: Comparison of chlorhexidine-ethanol and povidone-iodine. Chemotherapy, 2010; 56:261–267.

26.

Paocharoen

, Mingmalairak

, Apisarnthanarak

. Comparison of surgical wound infection after preoperative skin preparation with 4% chlorhexidine [correction of chlohexidine] and povidone iodine: A prospective randomized trial. J Med Assoc Thai, 2009; 92:898–902.

27.

Rodrigues

, Simoes Mde

. Incidence of surgical site infection with pre-operative skin preparation using 10% polyvidone-iodine and 0.5% chlorhexidine-alcohol. Rev Col Bras Cir, 2013; 40:443–448.

28.

Berry

, Watt

, Goldacre

, et al. A comparison of the use of povidone-iodine and chlorhexidine in the prophylaxis of postoperative wound infection. J Hosp Infect, 1982; 3:55–63.

29.

Ostrander

, Botte

, and Brage

. Efficacy of surgical preparation solutions in foot and ankle surgery. J Bone Joint Surg Am, 2005; 87:980–985.

30.

Veiga

, Damasceno

, Veiga-Filho

, et al. Povidone-iodine versus chlorhexidine in skin antisepsis before elective plastic surgery procedures: A randomized controlled trial. Plast Reconstr Surg, 2008; 122:170e–171e.

31.

Cheng

, Robertson

, St Mart

, et al. Quantitative analysis of bacteria in forefoot surgery: A comparison of skin preparation techniques. Foot Ankle Int, 2009; 30:992–997.

32.

Saltzman

, Nuber

, Gryzlo

, et al. Efficacy of surgical preparation solutions in shoulder surgery. J Bone Joint Surg Am, 2009; 91:1949–1953.

33.

Savage

, Weatherford

, Sugrue

, et al. Efficacy of surgical preparation solutions in lumbar spine surgery. J Bone Joint Surg Am, 2012; 94:490–494.

34.

Culligan

, Kubik

, Murphy

, et al. A randomized trial that compared povidone iodine and chlorhexidine as antiseptics for vaginal hysterectomy. Am J Obstet Gynecol, 2005; 192:422–425.

35.

Lee

, Agarwal

, Lee

, et al. Systematic review and cost analysis comparing use of chlorhexidine with use of iodine for preoperative skin antisepsis to prevent surgical site infection. Infect Control Hosp Epidemiol, 2010; 31:1219–1229.

36.

Hakkarainen

, Dellinger

, Evans

, et al. Comparative effectiveness of skin antiseptic agents in reducing surgical site infections: A report from the Washington State Surgical Care and Outcomes Assessment Program. J Am Coll Surg, 2014; 218:336–344.

37.

Young

, Knepper

, Vigil

, et al. Sustained reduction in surgical site infection after abdominal hysterectomy. Surg Infect, 2013; 14:460–463.

38.

Amer-Alshiek

, Alshiek

, Almog

, et al. Can we reduce the surgical site infection rate in cesarean sections using a chlorhexidine-based antisepsis protocol?. J Matern Fetal Neonatal Med, 2013; 26:1749–1752.

39.

Menderes

, Ather Ali

, Aagaard

, et al. Chlorhexidine-alcohol compared with povidone-iodine for surgical-site antisepsis in cesarean deliveries. Obstet Gynecol, 2012; 120:1037–1044.

40.

Levin

, Amer-Alshiek

, Avni

, et al. Chlorhexidine and alcohol versus povidone-iodine for antisepsis in gynecological surgery. J Womens Health, 2011; 20:321–324.

41.

Swenson

, Hedrick

, Metzger

, et al. Effects of preoperative skin preparation on postoperative wound infection rates: A prospective study of 3 skin preparation protocols. Infect Control Hosp Epidemiol, 2009; 30:964–971.

42.

Slim

, Nini

, Forestier

, et al. Methodological index for non-randomized studies (MINORS): Development and validation of a new instrument. ANZ J Surg, 2003; 73:712–716.

43.

Tschudin-Sutter

, Frei

, Egli-Gany

, et al. No risk of surgical site infections from residual bacteria after disinfection with povidone-iodine-alcohol in 1,014 cases: A prospective observational study. Ann Surg, 2012; 255:565–569.

44.

Allegranzi

, Bagheri Nejad

, Combescure

, et al. Burden of endemic health-care-associated infection in developing countries: Systematic review and meta-analysis. Lancet, 2011; 377:228–241.

45.

CareFusion. PCMX 12 mL Scrub Brush. www.carefusion.com/medical-products/infection-prevention/skin-preparation/surgical-trays-brushes-bulk-solutions/PCMX.aspx. (Last accessed April 17, 2014).

46.

Macias

, Arrequin

, Munoz

, et al. Chlorhexidine is a better antiseptic than povidone iodine and sodium hypochlorite because of its substantive effect. Am J Infect Control, 2013; 41:634–637.

47.

Boyce

, Pittet

. Guideline for hand hygiene in health-care settings. MMWR Morb Mort Wkly Rep, 2002; 51:1–45.

48.

Hidalgo

, Dominguez

. Mechanisms underlying chlorhexidine-induced cytotoxicity. Toxicol In Vitro, 2001; 15:271–276.

49.

Larson

. APIC guidelines for handwashing and hand antisepsis in health care settings. Am J Infect Control, 1995; 23:251–269.

50.

Maki

, Alvarado

, Ringer

. Prospective randomised trial of povidone-iodine, alcohol, and chlorhexidine for prevention of infection associated with central venous and arterial catheters. Lancet, 1991; 338:339–343.

51.

3M.Corporation. 3M DuraPrep Surgical Solution—Commonly Asked Questions. http://multimedia.3m.com/mws/mediawebserver?mwsId=66666UgxGCuNyXTtoXf6O8s6EVtQEcuZgVs6EVs6E666666–&fn=DuraPrep FAQ.pdf (Last accessed September 1, 2014).