Darouiche: Chlorhexidine–Alcohol versus Povidone–Iodine for Surgical-Site Antisepsis
Darouiche was a multicenter, patient- and assessor-blinded randomized superiority trial of a chlorhexidine–alcohol scrub versus an aqueous povidone–iodine scrub and paint in 849 adults (of 897 randomized) undergoing clean-contaminated surgery at six US university-affiliated hospitals, with surgical-site infection within 30 days as the primary outcome. Infection occurred in 9.5% with chlorhexidine–alcohol and 16.1% with povidone–iodine (relative risk 0.59, 95% CI 0.41–0.85; P = 0.004).[1]
- Design and population
- Multicenter, patient- and assessor-blinded randomized superiority trial. Adults undergoing clean-contaminated colorectal, small-intestinal, gastroesophageal, biliary, thoracic, gynecologic or urologic surgery at six US university-affiliated hospitals; 849 of 897 randomized patients were in the intention-to-treat analysis.
- Arms
- Chlorhexidine–alcohol scrub (2% chlorhexidine gluconate, 70% isopropyl alcohol) versus aqueous 10% povidone–iodine scrub and paint
- Report and follow-up
- 2010 · 30 days after surgery; weekly telephone contact after discharge
| Endpoint | Results | Effect | Finding |
|---|---|---|---|
| Any surgical-site infection within 30 days (CDC criteria), intention-to-treat | 39/409 (9.5%) vs 71/440 (16.1%) | RR 0.59; CI 0.41–0.85; P 0.004 | Met |
Harms: Adverse events in at least 5% of patients in either group: 228/409 (55.7%) vs 256/440 (58.2%); Serious adverse events in more than 1% of patients in either group: 72/409 (17.6%) vs 70/440 (15.9%); Preparation-related adverse events (pruritus, erythema or both around the wound): 3/409 (0.7%) vs 3/440 (0.7%)
Design
Adults 18 or older were eligible for colorectal, small-intestinal, gastroesophageal, biliary, thoracic, gynecologic or urologic operations performed under controlled conditions without substantial spillage or unusual contamination. Patients with an allergy to chlorhexidine, alcohol or iodophors, infection at or next to the operative site, or an anticipated inability to complete 30-day follow-up were excluded. Enrollment ran from April 2004 to May 2008. Randomization was 1:1, stratified by hospital, with computer-generated numbers and no blocking.[1]
Patients and the site investigators who diagnosed infection did not know the assignment; the operating surgeon learned it only after the patient reached the operating room. Infection was defined by CDC criteria and classified as superficial incisional (skin and subcutaneous tissue, excluding stitch abscesses), deep incisional (fascia and muscle) or organ-space. Patients were examined at least daily in hospital, at discharge and at follow-up, and were telephoned weekly through day 30.[1]
The sample size assumed a baseline infection rate of 14% with povidone–iodine falling to 7%, which required about 430 evaluable patients per group for 90% power at a two-sided alpha of 0.05. The intention-to-treat population required randomization and application of the assigned preparation (which required surgery); the per-protocol population also required clean-contaminated surgery and completed follow-up. All P values were two-tailed and were not adjusted for multiple testing.[1]
The trial was investigator-initiated, with its first author acting as sponsor. One author employed by the manufacturer of both preparations (Cardinal Health) contributed to the design and manuscript but not to data collection or analysis; several other authors reported research grants or fees from the company.[1]
Results
Among 409 chlorhexidine–alcohol and 440 povidone–iodine patients in the intention-to-treat analysis, any surgical-site infection occurred in 39 (9.5%) and 71 (16.1%) (relative risk 0.59, 95% CI 0.41–0.85; P = 0.004). The per-protocol analysis of 813 patients gave similar results, and time to infection was longer with chlorhexidine–alcohol (log-rank P = 0.004).[1]
By infection type, superficial incisional infection occurred in 17 (4.2%) versus 38 (8.6%) (relative risk 0.48, 95% CI 0.28–0.84; P = 0.008) and deep incisional infection in 4 (1.0%) versus 13 (3.0%) (relative risk 0.33, 95% CI 0.11–1.01; P = 0.05). Organ-space infection was equal at 18 (4.4%) versus 20 (4.5%) (relative risk 0.97, 95% CI 0.52–1.80), and sepsis from surgical-site infection occurred in 11 (2.7%) versus 19 (4.3%) (relative risk 0.62, 95% CI 0.30–1.29; P = 0.26).[1] The authors attributed the overall benefit to the incisional infections, most of which are caused by gram-positive skin flora.[1]
In subgroup analyses that the trial was not powered to compare, abdominal surgery was followed by infection in 37/297 (12.5%) versus 63/308 (20.5%) (95% CI for the absolute difference −13.9 to −2.1 percentage points), and nonabdominal surgery in 2/112 (1.8%) versus 8/132 (6.1%) (95% CI −7.9 to 2.6). The interaction between treatment and surgery type was not significant (P = 0.41). Among urologic operations, infection occurred in 0/26 versus 3/35 (8.6%); among gynecologic operations, in 0/42 versus 1/40 (2.5%).[1]
Surgery and treatment delivery
The chlorhexidine–alcohol arm was scrubbed with an applicator containing 2% chlorhexidine gluconate and 70% isopropyl alcohol (ChloraPrep), using more than one applicator when the field exceeded 33 by 33 cm. The povidone–iodine arm was scrubbed and then painted with an aqueous 10% povidone–iodine solution (Scrub Care Skin Prep Tray).[1]
Of 1003 patients screened, 897 were randomized (431 and 466); 22 and 26 did not have surgery, leaving 409 and 440 in the intention-to-treat analysis. Twenty-five patients had clean rather than clean-contaminated surgery, 4 dropped out and 7 died before day 30, which left 391 and 422 in the per-protocol analysis. Background care was balanced: every patient received systemic antibiotic prophylaxis within 1 hour before incision, 51.7% versus 48.9% received postoperative antibiotics, mean operative time was 3.0 hours in each group and 26.7% versus 27.0% showered preoperatively with hospital-specific products.[1]
Abdominal operations made up 72.6% versus 70.0% of cases (colorectal 45.5% versus 43.4%) and nonabdominal operations 27.4% versus 30.0% (thoracic 10.8% versus 13.0%, gynecologic 10.3% versus 9.1%, urologic 6.4% versus 8.0%).[1]
Adverse events
Adverse events in at least 5% of patients occurred in 228/409 (55.7%) versus 256/440 (58.2%), and serious adverse events in 72/409 (17.6%) versus 70/440 (15.9%). Three patients in each group (0.7%) had pruritus or erythema around the wound attributed to the preparation; no serious event was judged related to it, and there was no operating-room fire or chemical skin burn.[1]
Seven patients died within 30 days: four (1.0%) in the chlorhexidine–alcohol arm, none with surgical-site infection, and three (0.7%) in the povidone–iodine arm, all from sepsis due to organ-space infection.[1]
Limitations
The comparison changed the antiseptic, the vehicle and the application method together: an alcohol-based chlorhexidine applicator against an aqueous povidone–iodine scrub and paint. The trial therefore cannot separate the contribution of chlorhexidine from that of alcohol.[1]
Because surgeons learned the assignment in the operating room, intraoperative care was not masked. The 849-patient intention-to-treat population omitted 48 randomized patients who did not have surgery. The deep incisional result sat at the margin of significance (P = 0.05; relative-risk interval 0.11–1.01), the secondary and subgroup P values were unadjusted, and only 61 operations were urologic.[1]
In conversation
Same setting ChEETAh (2022)
The WHO guideline's evidence review of 17 randomized trials, which included this one, found moderate-quality evidence that alcohol-based chlorhexidine reduced infection compared with aqueous povidone–iodine (odds ratio 0.65, 95% CI 0.47–0.90), and low-quality evidence for alcohol-based chlorhexidine compared with alcohol-based povidone–iodine (odds ratio 0.58, 95% CI 0.42–0.80).[2]
Guidelines and approvals
- WHO (Global Guidelines for the Prevention of Surgical Site Infection, 2nd edition, 2018), section 4.7. The panel recommends alcohol-based antiseptic solutions based on chlorhexidine gluconate for surgical site skin preparation in patients undergoing surgical procedures . The guideline's systematic review of 17 randomized trials of antiseptic agents included this trial.[2]
Report history
| Published | Report | Cutoff | Headline results | Citation |
|---|---|---|---|---|
| 2010-01-07 | Initial report · primary | — | Any surgical-site infection within 30 days (CDC criteria), intention-to-treat: 39/409 (9.5%) vs 71/440 (16.1%); RR 0.59; CI 0.41–0.85; P 0.004 Met Superficial incisional infection within 30 days (skin and subcutaneous tissue, excluding stitch abscesses): 17/409 (4.2%) vs 38/440 (8.6%); RR 0.48; CI 0.28–0.84; P 0.008 Deep incisional infection within 30 days (fascia and muscle): 4/409 (1.0%) vs 13/440 (3.0%); RR 0.33; CI 0.11–1.01; P 0.05 Organ-space infection within 30 days: 18/409 (4.4%) vs 20/440 (4.5%); RR 0.97; CI 0.52–1.8; P >0.99 Sepsis from surgical-site infection within 30 days: 11/409 (2.7%) vs 19/440 (4.3%); RR 0.62; CI 0.3–1.29; P 0.26 Any surgical-site infection after abdominal (colorectal, biliary, small-intestinal, gastroesophageal) surgery; absolute difference in percentage points (chlorhexidine–alcohol minus povidone–iodine): 37/297 (12.5%) vs 63/308 (20.5%); difference -8; 95% -13.9–-2.1; P 0.009 Any surgical-site infection after nonabdominal (thoracic, gynecologic, urologic) surgery; absolute difference in percentage points (chlorhexidine–alcohol minus povidone–iodine): 2/112 (1.8%) vs 8/132 (6.1%); difference -4.3; 95% -7.9–2.6 Any surgical-site infection after urologic surgery: 0/26 (0%) vs 3/35 (8.6%); — | Darouiche RO, Wall MJ Jr, Itani KMF, et al. "Chlorhexidine–Alcohol versus Povidone–Iodine for Surgical-Site Antisepsis." *N Engl J Med.* 2010;362(1):18-26. doi:10.1056/NEJMoa0810988 DOI |
References
1. Darouiche RO, Wall MJ Jr, Itani KMF, et al. "Chlorhexidine–Alcohol versus Povidone–Iodine for Surgical-Site Antisepsis." N Engl J Med. 2010;362(1):18–26. doi:10.1056/NEJMoa0810988
2. World Health Organization. Global Guidelines for the Prevention of Surgical Site Infection. 2nd ed. Geneva: World Health Organization; 2018. Section 4.7.