Antiseptic Skin Agents to Prevent Surgical Site Infection After Clean Implant Surgery: Subgroup Analysis of the NEWSkin Prep Trial

Abstract

Background:

There remains debate as to the ideal skin preparation agent to prevent surgical site infection (SSI) after clean implant surgery. This study is a subgroup analysis of all patients undergoing implant surgery in the NEWSkin prep trial.

Patients and Methods:

The NEWSkin prep trial is a randomized clinical trial (RCT; n = 3300) comparing the three antiseptic skin preparation agents commonly used prior to incisional surgery: chlorhexidine with alcohol (C-Alc), povidone iodine with alcohol (PI-Alc), and aqueous povidone iodine (PI-Aq). All participants who underwent clean incisional surgeries with implants from this trial were selected for analysis. The primary outcome was SSI rate; secondary outcomes were complication rates and re-admission rates.

Results:

A total of 712 patients in the NEWSkinPrep study underwent clean implant surgery between January 2015 and December 2018. Randomization resulted in the following distribution: C-Alc, 238; PI-Alc, 243; and PI-Aq, 231. Median age of participants was 63 years and 41% were female. Surgical site infection rates were: C-Alc, 10.13%; PI-Alc,11.52%; and PI-Aq, 11.26%. Povidone iodine with alcohol did not appear to be non-inferior to C-Alc (mean difference, 1.39%; 95% confidence interval [CI], −4.17 to 6.95) and PI-Alc was not superior to PI-Aq (mean difference, 0.27%; 95% CI, −5.45 to 5.99; p = 0.9271). There were no differences seen in secondary outcomes between groups.

Conclusions:

These results indicate a similarity between PI-Alc and C-Alc and that PI-Alc was not superior to PI-Aq. Because this study was not powered to assess these differences in clean cases with implants, additional studies are needed to assess these agents.

Surgical site infection (SSI) is considered one of the most significant post-operative complications because of its negative impacts on health and cost consequences^1,2 and it remains the second most common nosocomial infection despite improvements in operative practices and sterilization techniques.² The incidence of SSI varies depending on wound classification, geographic location, and patient demographics.^3–8

Skin preparation using an antiseptic solution prior to incision is performed routinely, because skin contamination is the main source of infection. Many studies have examined and compared the effectiveness of antiseptic agents in reducing the rate of SSI, however, debate still remains as to the most appropriate agent.^9–12 An emerging body of evidence suggests that the combination of chlorhexidine and alcohol is the gold standard for skin antisepsis prior to surgery to reduce the rate of infection.^10–12

The recommendations of the World Health Organization (WHO)¹³ and the National Institute for Health and Care Excellence (NICE)¹⁴ is that chlorhexidine-alcohol based antiseptic solutions be used for surgical site skin preparation. These recommendations are controversial, however, with some members of the WHO working group suggesting they were made prematurely based on limited existing evidence and underpowered studies.^15,16 This gap in evidence was also highlighted in the latest Cochrane meta-analysis on pre-operative skin antiseptics for clean surgeries.¹⁷ It concluded that more research was required: to determine whether chlorhexidine or povidone-iodine was the better solution and whether the addition of alcohol to these solutions resulted in benefit.

The group of patients most at risk of significant and sometimes catastrophic complications of SSIs are those patients undergoing implant surgery; we sought to answer the questions highlighted by the Cochrane meta-analysis in this patient group.

The NEWSkin Prep Trial was a combined non-inferiority (PI-Alc vs. C-Alc) and superiority (PI-Alc vs. PI-Aq) trial for all incisional surgery. This article examines the complete subset of patients in the NEWSkin Prep study who underwent clean surgery with implants.

Patients and Methods

Study design

The NEWSkin Prep trial was an investigator-initiated, three-arm, prospective, randomized clinical trial (RCT) involving patients undergoing incisional surgery. The full trial methodology and results are reported elsewhere,¹⁸ however, the trial involved two separate hypotheses to compare three arms: the non-inferiority (followed by potential superiority) of PI-Alc compared with C-Alc and the superiority of PI-Alc compared with PI-Aq. These same hypotheses were used in this study to compare those patients who underwent clean implant surgery in the larger trial.

Participants

The NEWSkin Prep trial was conducted in two major university affiliated hospitals in Newcastle. The ethical approval was obtained through the Hunter New England regional ethics review board (HNEHREC 13/12/11/3.02). Participants were allocated randomly to one of three antiseptic preparation groups: C-Alc, PI-Alc, and PI-Aq. All patients who underwent clean surgery with implants in the NEWSkin Prep trial were analyzed in this study. Implants were defined as permanent non-biologic foreign bodies and included: joint replacement, hernia mesh, metallic rods, plates and screws, vascular grafts, and cardiac valves.

Interventions and procedures

The antiseptic skin preparation agents compared were chlorhexidine gluconate 0.5% weight/volume with ethanol 70% volume/volume (Perrigo, Sydney, Australia) (C-Alc); povidone-iodine 10% weight/volume with ethanol 70% volume/volume (Pfizer, Sydney, Australia) (PI-Alc); and povidone-iodine 10% weight-volume (Pfizer, Australia) (PI-Aq). The C-Alc arm was changed to chlorhexidine gluconate 2% weight/volume with isopropyl alcohol 70% volume/volume (Halmed, Melbourne, Australia) partway through the trial because pf supplier problems in Australia. Protocol change (ANZCTR) and ethics approval (HNEHREC) were undertaken to reflect this change in the main trial.¹⁸ Staff were trained in the correct use and application of antiseptic agents.¹⁸ Detection of SSI was performed by patient reporting, photos of the wound, review of general practitioner (GP) databases, review of hospital and regional network databases, and a formal point of contact at 30 days and 90 days after surgery by a blinded research nurse via means of a questionnaire.

Outcomes

The primary end point was the occurrence of any SSI within 90 days of surgery as per the Centers for Disease Control and Prevention (CDC) criteria.¹⁹ Secondary end points were total surgical complication rates (Clavien-Dindo²⁰) and re-admission rates within 90 days.

Statistical analysis

Sample size of the main trial was driven by two comparisons explained elsewhere.¹⁸ Categorical variables were reported with frequencies and percentages; continuous variables were reported with mean and standard deviation (if normally distributed) or median with minimum and maximum values (if non-normally distributed).

All outcomes were assessed in a modified intention to treat analysis (patients who had no surgical incision were excluded from analyses). In this study, categorical outcomes (the occurrence of any SSI, total surgical complication rates (Clavien-Dindo²⁰), re-admission rates within 90 days) were analyzed using χ² or Fisher exact p values. Statistical analyses were programmed using STATA, version 17 (StataCorp LLC, College Station, TX, USA).

Results

Between January 2015 and December 2018, 3,301 patients underwent incisional surgeries within the NEWSkin Prep Trial. Of these, 712 (21.6%) patients underwent clean incisional surgeries with implants. One patient died in the operating room. The patients in the three study groups were comparable with respect to demographic characteristics, co-existing illnesses, and risk factors for infection (Table 1).

Table 1.

Baseline Patient Characteristics and Risk Factors for SSI

Characteristic	Chlorhexidine (2%, 0.5%) and 70% alcohol (C-Alc) (n = 237)	Povidone-iodine (10%) and 70% alcohol (PI-Alc) (n = 243)	Povidone-iodine (10%) aqueous (PI-Aq) (n = 231)
Age, median, y (min–max)	61 (19–95)	65 (18–92)	63 (21–89)
Female gender, no. (%)	91 (38)	100 (41)	100 (43)
Body mass index, median (min–max)	30 (18–59)	29 (17–61)	29 (18–50)
Smoking status
Daily	34 (14)	40 (16)	39 (17)
Irregular and weekly	3 (1.3)	1 (0.4)	2 (0.9)
Ex-smoker	89 (38)	99 (41)	91 (39)
Never smoked	112 (47)	103 (42)	99 (43)
Diabetes mellitus, no. (%)	36 (15)	33 (14)	47 (20)
ASA
1	21 (9)	21 (8.8)	25 (11)
2	117 (50)	131 (55)	109 (47)
3	83 (36)	81 (34)	80 (35)
4	12 (5.2)	7 (2.9)	17 (7.4)
Immunosuppressive medications, no. (%)	15 (6.4)	20 (8.5)	18 (8)
Pre-operative radiotherapy, no. (%)	2 (0.9)	2 (0.9)	1 (0.4)
Surgical mix, no. (% of specialty mix)
Cardiac (n = 30)	10 (33.3)	7 (23.3)	13 (43.3)
Hernia (n = 200)	72 (36)	67 (33.5)	61 (30.5)
Neurosurgery (n = 190)	62 (32.6)	64 (33.7)	64 (33.7)
Orthopedics (n = 247)	80 (32.4)	85 (34.4)	82 (33.2)
Implant (n = 45)	14 (31.1)	20 (44.4)	11 (24.4)
Intra-operative blood transfusion, no. (%)	4 (1.7)	2 (0.8)	0 (0)

ASA = American Society of Anesthesiologists.

Surgical site infection occurred in 26 of 231 (11.26%) patients in the PI-Aq group; in 28 of 243 (11.52%) patients in the PI-Alc group; and in 24 of 237 (10.13%) patients in the C-Alc group. An omnibus test across all three groups showed no significant difference for SSI (p = 0.88), or for any of the secondary outcomes (Table 2).

Table 2.

Outcomes by Treatment Arm

Characteristic	Class/statistic	Chlorhexidine (2%,0.5%) and 70% alcohol (C-Alc) (n = 237)	Povidone-iodine (10%) and 70% alcohol (PI-Alc) (n = 243)	Povidone-iodine (10%) aqueous (PI-Aq) (n = 231)	Total	p
Post-operative SSI	No	213 (89.87%)	215 (88.48%)	205 (88.74%)	633 (89%)	0.875
	Yes	24 (10.13%)	28 (11.52%)	26 (11.26%)	78 (11%)
	Missing	0	0	0	0
Re-admission	No	223 (95.71%)	228 (95.40%)	220 (96.07%)	671 (96%)	0.937
	Yes	10 (4.29%)	11 (4.60%)	9 (3.93%)	30 (4.3%)
	Missing	4	4	2	10
Post-operative complications	No	178 (75.11%)	181 (74.79%)	182 (78.79%)	541 (76%)	0.529
	Yes	59 (24.89%)	61 (25.21%)	49 (21.21%)	169 (24%)
	Missing	0	1	0	1

SSI = surgical site infection.

Povidone iodine with alcohol did not appear to be non-inferior to C-Alc with respect to the overall rate of SSI (mean difference, 1.39%; 95% CI, −4.17 to 6.95) because this difference did not meet the pre-specified margin of non-inferiority of 4% (upper 95% CI limit is 6.95%). Comparison between PI-Alc and C-Alc revealed no significant difference in the secondary outcomes of re-admission rates and post-operative complications (Table 3).

Table 3.

Efficacy Outcomes: Povidone-Iodine With Alcohol Versus Chlorhexidine With Alcohol

Outcome	Povidone-iodine (10%) and 70% alcohol (PI-Alc) (n = 243)	Chlorhexidine (2%,0.5%) and 70% alcohol (C-Alc) (n = 237)	PI-Alc–C-Alc Mean difference (95% CI)
SSI, no. (%)	28 (11.52%)	24 (10.13%)	1.39% (−4.17 to 6.95)
Re-admission within 90 d, no. (%)	11 (4.60%)	10 (4.29%)	0.31% (−3.41 to 4.03)
Post-operative complications, no. (%)	61 (25.21%)	59 (24.89%)	0.31% (−7.45 to 8.07)

CI = confidence interval; SSI = surgical site infection.

In the second comparison, PI-Alc was not superior to PI-Aq with respect to the overall rate of SSI (mean difference, 0.27%; 95% CI, −5.45 to 5.99; p = 0.9271). There were also no significant differences in the secondary outcomes of re-admission rates and post-operative complications (Table 4).

Table 4.

Efficacy Outcomes: Povidone-Iodine With Alcohol Versus Povidone-Iodine Aqueous

Outcome	Povidone-iodine (10%) and 70% alcohol (PI-Alc) (n = 243)	Povidone-iodine (10%) aqueous (PI-Aq) (n = 231)	Difference (95% CI)	p
SSI, no. (%)	28 (11.52%)	26 (11.26%)	0.27% (−5.45 to 5.99)	0.9271
Re-admission within 90 d, no. (%)	11 (4.60%)	9 (3.93%)	0.67% (−2.99 to 4.33)	0.7192
Post-operative complications, no. (%)	61 (25.21%)	49 (21.21%)	3.99% (−3.6 to 11.59)	0.3040

CI = confidence interval; SSI = surgical site infection.

Discussion

In our report of the full study population, we found that PI-Alc was non-inferior to Cl-Alc, easily meeting the non-inferiority threshold of 4%; this included clean-contaminated, clean, contaminated, and dirty subgroups. There were 1,506 patients in the clean subgroup of whom 721 also had an implant. In this implant subgroup, PI-Alc was not shown to be non-inferior to C-Alc with respect to SSI rates. There were also no differences seen in complication rates and re-admission rates between the three skin preparation agents. The point estimates in this subgroup were virtually identical to the entire group, indicating that the lack of non-inferiority is likely because of the loss of power from the smaller sample size, rather than a true difference in effect in this subgroup.

The rates of SSI seen in this study are relatively high compared to the literature. There are a couple of potential explanations for this. The mean body mass index (BMI) of >29 is certainly high in comparison to other RCTs. The other factor is the mode of follow-up. Participants were followed up to three months with telephone calls at 30 and 90 days and questionnaires designed to pick up SSI. In addition, all notes from GPs and hospitals as well as networks were analyzed for re-admission or care data in order to ensure that patients who may have had subsequent treatment elsewhere were not missed. This systematic follow up regime probably meant that fairly minor SSIs were picked up compared to using just symptomatic SSIs presenting for treatment. This is supported by the fact that more than 70% of the SSIs in this analysis were identified after discharge, the re-admission rate for SSI alone was extremely low, and no implant infections occurred during the study period (90 days).

There are limited data in the form of RCTs on skin antisepsis for implant surgery. Most of the RCTs performed have been on clean surgery rather than selectively implant surgery. Two RCTs have compared C-Alc and PI-Aq after clean surgery on the shoulder (n = 150)²² and on inguinal hernia (n = 556)²³ with neither showing any difference in SSI rates. Only a minority of the patients in these trials had implants.^21,22 Two RCTs have compared C-Alc and PI-Alc after clean surgery on the foot and ankle (n = 127)²³ and a range of operations (n = 500)²⁴ with neither showing any difference in SSI rates.^23,24 It is unclear from either of these two trials how many patients in the cohort had implant surgery. The latest Cochrane review on skin antiseptics pre-operatively for clean surgery also revealed no benefit associated with the addition of alcohol to solution (6 RCTs; n = 1,400) in reduction of SSI rates: relative risk (RR), 0.77; 95% CI 0.51–1.17.¹⁷

Given the small number of clinical trials conducted on clean surgery and the paucity of data regarding clean implant surgery the information in this trial is vital. The SSI rates of C-Alc (10.13%), PI-Alc (11.52%), and PI-Aq (11.26%) seen in this trial are not substantially different clinically one from the other and are very similar to the point estimates for the entire NEWSkin Prep trial.¹⁸ This highlights further the apparent lack of difference between preparations and enables surgeons and healthcare providers to make informed choices about their use. It would appear, based on these figures and the absence of implant infections seen in this trial, that decisions on the use of antiseptic agents could be based around cost of the agents and potential risk associated with them.

The main limitation of this study lies in the fact it represents a subgroup analysis and is therefore underpowered to determine non-inferiority based on the chosen margin of 4%. This is reflected in the wide confidence intervals seen in the study. Another limitation is the heterogeneity of surgical interventions ranging from joint replacements through to cardiac valve replacements. Despite these limitations it represents, to our knowledge, the largest available form of randomized evidence on implant surgery and as such provides valuable information both for clinical use and in designing more specific trials on implant surgery in the future.

In summary, based on the results of this study, there does not appear to be an advantage associated with any of the three commonly used skin preparation agents when used for clean implant surgery and choice should be based on a combination of cost, potential risk, and local availability.

Footnotes

Acknowledgments

The authors thank the Hunter New England Local Health District (HNELHD), New South Wales Health, Australia for providing the following grants: John Hunter Charitable Trust grant, Hunter Cancer Research Alliance grant, Hunter New England Health Research and Translation Centre grant and Cancer Institute NSW funding. We thank all the surgeons involved in the study: Abson S, Balogh Z, Bendinelli C, Chen S, Christie J, Draganic B, Ferch R, Gani J, Hansen M, Hardstaff R, Heer M, Hunt J, Ius Y, Jaaback K, James A, Karihaloo C, Koshy A, Majid A, Martin P, McKenzie S, McManus B, Mejia R, Nicholson D, O'Neill C, O'Sullivan R, Otten G, Pockney P, Regalo B, Schmid B, Seah P, Shah K, Singh T, Smith S, Spittaler P, Tawari S, Thiruchelvam D, Tiu A, Wills V, Wright T, Young J.

Authors' Contributions

Conceptualization: Smith, Attia, Gani. Protocol development: Smith,

Supervision: Smith, Attia, Gani. Methodology: Smith, Attia, Gani. Validation: Smith, Carroll, Lott, Hampton. Writing—original draft: Smith, AbuHassanain.

Writing—review and editing: Smith, AbuHassanain, Attia, Carroll, Lott, Gani. Hampton. Data curation: AbuHassanain, Carroll, Lott. Formal analysis: AbuHassanain. Project administration: Carroll, Lott. Ethics application: Carroll.

Resources: Carroll, Lott. REDCap database creation and updating: Carroll, Lott, Hampton. Patient follow-up: Carroll, Lott, Hampton.

Funding Information

We received grants from Australian government not for profit organisations for this RCT: Hunter New England Local Health District (HNELHD), New South Wales Health, Australia provided a John Hunter Charitable Trust grant, Hunter Cancer Research Alliance (NSW, Australia) provided a grant and Cancer Institute (NSW, Australia) provided Research and Translation Centre grant and funding.

Author Disclosure Statement

None declared.

References

Badia

, Casey

, Petrosillo

et al. Impact of surgical site infection on healthcare costs and patient outcomes: A systematic review in six European countries. J Hosp Infect, 2017; 96(1):1–15; doi: 10.1016/j.jhin.2017.03.004

Perencevich

, Sands

, Cosgrove

, et al. Health and economic impact of surgical site infections diagnosed after hospital discharge. Emerg Infect Dis, 2003; 9(2):196–203; doi: 10.3201/eid0902.020232

Mukagendaneza

, Munyaneza

, Muhawenayo

et al. Incidence, root causes, and outcomes of surgical site infections in a tertiary care hospital in Rwanda: A prospective observational cohort study. Patient Saf Surg, 2019; 13:1–8; doi: 10.1186/s13037-019-0190-8

Onyekwelu

, Yakkanti

, Protzer

, et al. Surgical wound classification and surgical site infections in the orthopaedic patient. J Am Acad Orthop Surg Glob Res Rev, 2017; 1(3):e022; doi: 10.5435/JAAOSGlobal-D-17-00022.

Thu

, Dibley

, Ewald

, et al. Incidence of surgical site infections and accompanying risk factors in Vietnamese orthopaedic patients. J Hosp Infect, 2005; 60(4):360–367; doi: 10.1016/j.jhin.2005.02.006

Barie

. Surgical site infections: Epidemiology and prevention. Surg Infect, 2002; 3(Suppl 1):s9–21; doi: 10.1089/sur.2002.3.s1-9

Zhang

, Wang

, Chen

et al. Incidence and risk factors of surgical site infection following colorectal surgery in China: A national cross-sectional study. BMC Infect Dis, 2020; 20(1):1–11; doi: 10.1186/s12879-020-05567-6

Lubega

, Joel

, Justina-Lucy

. Incidence and etiology of surgical site infections among emergency postoperative patients in mbarara regional referral hospital, South Western Uganda. Surg Res Pract, 2017; 6365172;doi: 10.1155/2017/6365172

Aworinde

, Olufemi-Aworinde

, Fehintola

, et al. Antiseptic skin preparation for preventing surgical site infection at caesarean section. Open J Obstet Gynecol, 2016; 6:246–251.

10.

Darouiche

, Wall

Jr , Itani

, et al. Chlorhexidine–alcohol versus povidone–iodine for surgical-site antisepsis. N Engl J Med, 2010; 362(1):18–26; doi: 10.1056/NEJMoa0810988

11.

Tuuli

, Liu

, Stout

MJ et al

. A randomized trial comparing skin antiseptic agents at cesarean delivery. N Engl J Med, 2016; 374(7):647–655; doi: 10.1056/NEJMoa1511048

12.

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(1):55–63; doi: 10.1016/0195-6701(82)90031-7

13.

Allegranzi

, Bischoff

, de Jonge

et al. New WHO recommendations on preoperative measures for surgical site infection prevention: An evidence-based global perspective. Lancet Infect Dis, 2016; 16(12):e276–e87; doi: 10.1016/S1473-3099(16)30398-X

14.

UK NGUT. Surgical site infections: Prevention and treatment. updated April 2019. Available from: www.nice.org.uk/guidance/ng125. [Last accessed: August 2023].

15.

Maiwald

, Widmer

. WHO's recommendation for surgical skin antisepsis is premature. Lancet Infect Dis, 2017; 17(10):1023–1024; doi: 10.1016/S1473-3099(17)30448-6

16.

Sidhwa

, Itani

. Skin preparation before surgery: Options and evidence. Surg Infect, 2015; 16(1):14–23; doi: 10.1089/sur.2015.010

17.

Dumville

, McFarlane

, Edwards

, et al. Preoperative skin antiseptics for preventing surgical wound infections after clean surgery. Cochrane Database Syst Rev, 2015;(4)CD003949; doi: 10.1002/14651858.CD003949.pub4

18.

ANZCTR. Effect of surgical skin preparations on post-operative skin infections in patients having surgical procedures 2015 [updated 2019. Available from: https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=366083. [Last accessed: August 2023].

19.

Berríos-Torres

, Umscheid

, Bratzler

DW et al

. Centers for Disease Control and Prevention guideline for the prevention of surgical site infection, 2017. JAMA Surg, 2017; 152(8):784–791; doi: 10.1001/jamasurg.2017.0904

20.

Clavien

, Barkun

, De Oliveira

, et al. The Clavien-Dindo classification of surgical complications: five-year experience. Ann Surg, 2009; 250(2):187–196; doi: 10.1097/SLA.0b013e3181b13ca2

21.

Saltzman

, Nuber

, Gryzlo

, et al. Efficacy of surgical preparation solutions in shoulder surgery. J Bone Joint Surg Am, 2009; 91(8):1949–53; doi: 10.2106/JBJS.H.00768

22.

Sistla

, Prabhu

, Sistla

, et al. Minimizing wound contamination in a 'clean' surgery: comparison of chlorhexidine-ethanol and povidone-iodine. Chemotherapy, 2010; 56(4):261–7; doi: 10.1159/000319901

23.

Bibbo

, Patel

, Gehrmann

, et al. Chlorhexidine provides superior skin decontamination in foot and ankle surgery: a prospective randomized study. Clin Orthop Relat Res, 2005; 438:204–8; doi: 10.1097/01.blo.0000167832.47464.22

24.

Paocharoen

, Mingmalairak

, Apisarnthanarak

. Comparison of surgical wound infection after preoperative skin preparation with 4% chlorhexidine and povidone iodine: a prospective randomised trial. J Med Assoc Thai, 2009; 92(7):898–902.