The prospective single-arm observational study included in the systematic review was excluded from the meta-analysis.
Meta-analysis includes findings of studies of surgical site infections with sterile vs nonsterile gloves after outpatient procedures. RR indicates relative risk. The size of the diamond represents the 95% CI of the overall assessment. If the diamond is wide, there is less confidence that it is accurate; however, if the diamond is narrow and small, then there is a higher confidence that the final analysis is accurately representing the truth.
eTable 1. Study Quality Characteristics of Randomized Clinical Trials (According to Cochrane Collaboration Guidelines)
eTable 2. Study Quality Characteristics of Observational Studies (According to the Newcastle-Ottawa Scale)
eTable 3. Summary of Included Studies
eFigure 1. Incidence of Primary Outcome
eFigure 2. Subgroup Meta-analyses
Brewer JD, Gonzalez AB, Baum CL, Arpey CJ, Roenigk RK, Otley CC, Erwin PJ. Comparison of Sterile vs Nonsterile Gloves in Cutaneous Surgery and Common Outpatient Dental ProceduresA Systematic Review and Meta-analysis. JAMA Dermatol. 2016;152(9):1008–1014. doi:10.1001/jamadermatol.2016.1965
Whether the use of sterile vs nonsterile gloves in outpatient cutaneous procedures affects the rate of postoperative wound infection is unknown.
To explore rates of surgical site infection (SSI) with the use of sterile vs nonsterile gloves in outpatient cutaneous surgical procedures.
This systematic review and meta-analysis identified studies from Ovid MEDLINE (1946 to present), Ovid Cochrane Central Register of Controlled Trials (1991 to present), Ovid EMBASE (1988 to present), EBSCO Cumulative Index to Nursing and Allied Health Literature (1980 to present), Scopus (1996 to present), and Web of Science (1975 to present).
Studies with information on sterile vs nonsterile gloves in outpatient surgical procedures were retrieved. Only randomized clinical trials and comparative studies were included for final analysis.
Data of trial design, surgery characteristics, and outcomes from published manuscripts and unpublished data were independently extracted.
Main Outcomes and Measures
Randomized clinical trials were considered high quality if randomization, allocation concealment, blinding, and follow-up completeness were appropriate. Relative risk and 95% CIs were derived for postoperative wound infections.
Fourteen articles met eligibility and inclusion criteria for systematic review; they included 12 275 unique patients who had undergone 12 275 unique outpatient procedures with sterile or nonsterile gloves and had follow-up regarding SSI. With the exclusion of 1 single-arm observational study of 1204 patients, 11 071 patients from 13 studies remained in the meta-analysis. Of these, 228 patients were documented as having postoperative SSI (2.1%), including 107 of 5031 patients in the nonsterile glove group (2.1%) and 121 of 6040 patients in the sterile glove group (2.0%). Overall relative risk for SSI with nonsterile glove use was 1.06 (95% CI, 0.81-1.39).
Conclusions and Relevance
No difference was found in the rate of postoperative SSI between outpatient surgical procedures performed with sterile vs nonsterile gloves.
Many health care specialties provide outpatient cutaneous surgical procedures. Gloves were first introduced to surgical practice more than a century ago, mostly for hygiene purposes. After that, gloves were mainly used for other reasons, including the prevention of hand dermatitis.1 During the past few decades, the use of surgical gloves has become standard practice to prevent postoperative wound infections or surgical site infection (SSI). However, whether the use of sterile vs nonsterile gloves makes a difference in the development of postoperative SSIs in the setting of cutaneous and minor outpatient surgical procedures remains unclear.2,3
Several studies2,4- 18 have investigated nonsterile vs sterile gloves in the setting of Mohs micrographic surgery (MMS), outpatient dental procedures, and laceration repair. Studies of dental procedures4,12,13,15,19- 22 have not found a difference between sterile vs nonsterile gloves in adverse outcomes, including SSI. Other studies7,10,11,18 have found that sterile vs nonsterile gloves do not make a difference in SSI in the setting of MMS, but 1 study6 demonstrated that heightened infection control practices could decrease the occurrence of SSI from 2.5% to 0.9%. In another study of 3491 outpatient surgical procedures, those that were more complex and not considered simple excisions had an SSI rate of 14.7% with nonsterile gloves compared with 3.4% with sterile gloves (P = .001); no difference in SSI was noted in procedures categorized as simple excisions (1.7% and 1.6% with nonsterile and sterile gloves, respectively).17
Given the many outpatient minor and cutaneous surgical procedures that occur worldwide, and given the current heightened consciousness about appropriate management of health care resources, knowing the effect of nonsterile vs sterile gloves on postoperative SSI rates would be important.3,7,9,18,23- 25 We performed a systematic review and meta-analysis to examine this topic.
Question What is the difference in rates of surgical site infection with the use of sterile vs nonsterile gloves during outpatient cutaneous surgical procedures?
Findings In this systematic review and meta-analysis that included 11 071 patients, no difference was found in the rate of postoperative surgical site infection between outpatient surgical procedures performed with sterile vs nonsterile gloves.
Meaning Considering the difference in cost and the increasing number of minor surgical procedures performed worldwide, use of nonsterile gloves is an acceptable option.
This study was performed in accordance with a protocol that prespecified important criteria such as study selection, exclusion criteria, data extraction, and statistical analysis, as well as heterogeneity and measurements of inconsistencies. The methods used in this manuscript were in accordance with the Cochrane collaboration guidelines (http://www.cochrane.org), the Newcastle-Ottawa Scale for quality assessment (http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp), and the QUOROM statement.26
The search strategy was initially designed in the expanded Ovid MEDLINE file, which includes in-process and nonindexed citations from 1946 to the present. Medical subject heading (MeSH) terms and text words were used to ensure that all relevant citations were included. MeSH terms are a controlled vocabulary used to minimize the number of synonyms necessary to describe each concept.
MeSH search terms used for this study included gloves, surgical; dermatologic surgical procedures; and surgical infections. Text words used for this study included: gloves, sterile, clean, and nonsterile. We used publication types and text words to filter for randomized clinical trials, cohort studies, follow-up studies, or comparative studies. The strategy was then translated into database-specific terms for the Ovid Cochrane Central Register of Controlled Trials (1991 to present), Ovid EMBASE (1988 to present), and EBSCO Cumulative Index to Nursing and Allied Health Literature (1980 to present). The databases Scopus (1996 to present) and Web of Science (1975 to present) were primarily searched with text words.
Studies included in this systematic review and meta-analysis were randomized clinical trials or prospective or retrospective observational comparative studies. The patients in the studies underwent an outpatient cutaneous or mucosal surgical procedure, including MMS, laceration repair, standard excisions, and tooth extractions. The procedures used sterile or nonsterile gloves, and documentation of postoperative SSI was present. Study eligibility criteria were assessed by 2 independent investigators (J.D.B. and A.B.G.J.).
All of the included trials and observational studies were reviewed in detail for methodologic features most pertinent to potential bias. The 2 independent reviewers (J.D.B. and A.B.G.J.) assessed features according to the Cochrane collaboration guidelines for the randomized clinical trials, which included randomization, intention-to-treat analysis, completeness of follow-up, and outcome assessment and attrition rates. Allocation of concealment, blinding, and masking of allocation were not evaluated as part of the assessment of quality given that, by the nature of these studies, all parties could tell whether the gloves used for the procedure were sterile or nonsterile. The Newcastle-Ottawa Scale was used to assess the quality of observational studies. All discrepancies in quality assessment were resolved by consensus-forming discussions between the 2 investigators (eTables 1 and 2 in the Supplement).
The primary outcome variable in this study was postoperative wound SSI. Data from studies that included other postoperative complications such as dry socket or inflamed socket in the setting of dental procedures were not included. Assessment of the primary outcome was performed independently by the 2 investigators (J.D.B. and A.B.G.J.), and no discrepancies or disagreements existed.
After ensuring that each study included the correct primary outcome, we abstracted data regarding the use of sterile vs nonsterile gloves, in addition to the number of procedures in each arm of the study. Because this comparative study used dichotomous data measures for the primary outcome variable, the relative risk (RR) was the unit of analysis. No missing data were identified in any of the included studies. All data were abstracted and analyzed using Review Manager software (version 5.3; Cochrane). In the case of the study that included only data for 1 study arm—nonsterile gloves only and SSI—data regarding the primary outcome and number of procedures were abstracted.
As a secondary meta-analysis, all studies pertaining to only cutaneous surgery (removing all dental procedure studies) and all studies pertaining to only dental procedures were analyzed separately. Finally, a subgroup meta-analysis was performed on studies pertaining to only MMS.
A total of 512 publications were initially considered for inclusion after a comprehensive systematic review. From these, 463 were deemed irrelevant based on the abstract and content. The remaining 49 publications were reviewed in full. Of these, 14 met the inclusion criteria (Table, Figure 1, and eTable 3 in the Supplement).4- 7,9- 17 Among the 35 articles excluded after full-text review, 4 articles were in discordance between the 2 independent reviewers. Consensus was reached to ultimately exclude the 4 articles because of differences in study design and outcome.3,28- 30 One of these studies28 also had no clear evidence of the necessary data needed for inclusion. The authors of the study were contacted, and after multiple failed attempts, the study was excluded.
The 14 articles that met the inclusion criteria for systematic review included a total of 12 275 unique patients who had undergone 12 275 unique outpatient procedures with sterile or nonsterile gloves and had follow-up regarding SSI. Incidence of the primary outcome variable for these studies is shown in eFigure 1 in the Supplement. One study16 included 1204 patients who had undergone MMS with nonsterile gloves only. Thus, because the study was noncomparative, these patients could not be included in the meta-analysis. A total of 11 071 patients were included in the final meta-analysis, of whom 2741 were randomly assigned to sterile gloves (n = 1360) or nonsterile gloves (n = 1381) as the intervention in a clinical trial. The remaining 8330 patients were part of prospective and retrospective observational studies; these patients had undergone an outpatient cutaneous surgical procedure with sterile (n = 4680) or nonsterile (n = 3650) gloves.
Low heterogeneity was identified among the studies included in the main meta-analysis as an I2 of 0% (P = .43) for the randomized clinical trials and an I2 of 0% (P = .94) for the observational studies. Between the 2 different study designs, the I2 was 0% (P = .43) with an overall I2 of 0% (P = .73). These statistics suggest that the 2 different study designs were not heterogeneous and that heterogeneity is not a problem across the included studies overall. All subgroup and sensitivity analyses also had an I2 of 0% except for the subgroup analysis for cutaneous reconstruction (I2 = 21.9%; P = .26) and the sensitivity analysis for Martin et al6 (excluding Lilly and Schmults18), which had an I2 of 7.9% (P = .30) for heterogeneity among the different study groups.
Of the 11 071 patients included in the final meta-analysis, 228 were documented as having postoperative SSI (2.1%), including 107 of 5031 patients in the nonsterile glove group (2.1%) and 121 of 6040 patients in the sterile glove group (2.0%). In addition, the single-arm prospective study with only nonsterile gloves documented 11 of 1204 patients with a postoperative SSI after MMS (0.9%).16 When we included these patients in the nonsterile glove cohort, a total of 118 postoperative SSIs occurred in 6235 total patients (1.9%).
The combined analysis of all studies gave an overall RR of 1.06 (95% CI, 0.81-1.39). The 8 clinical trials had an RR of 0.95 (95% CI, 0.65-1.40), and the 5 observational studies had an RR of 1.19 (95% CI, 0.81-1.73). None of these RRs were significant (Figure 2).
On contacting the authors of 1 study,6 we were made aware of a follow-up publication from the same institution.18 To maintain the assumption of independence, instead of including both studies separately, the data from the 2 studies were combined (Martin et al6 and Lilly and Schmults18) for the final analysis: a total of 1035 surgical procedures performed with nonsterile gloves and 585 with sterile gloves. A sensitivity analysis to determine whether the outcome was different if either one of the studies was excluded showed no difference: the RR including only Martin et al6 was 1.10 (95% CI, 0.84-1.44) and the RR including only Lilly and Schmults18 was 1.03 (95% CI, 0.78-1.34).
Quality assessment was concordant between the 2 reviewers, with no discrepancies. All studies were deemed to be high quality with minimal risk for bias (Table).27 The risk for publication bias is more of a factor with retrospective observational studies. Because all studies were believed to be of high quality and had no major discordances or outliers, especially among the observational studies, the risk for publication bias was believed to be minimal.
For the subgroup analyses, low heterogeneity was identified among the included studies in the MMS and dental procedures cohorts, with an I2 of 0% for both and P = .82 and P = .70, respectively. For the dental procedures, the overall RR for these studies was 3.36 (95% CI, 0.81-13.99) (eFigure 2A in the Supplement). The subgroup analysis for cutaneous reconstruction in the setting of MMS yielded an overall RR of 1.15 (95% CI, 0.68-1.97) for these studies (eFigure 2B in the Supplement). Finally, studies in the cutaneous surgery group included 4 randomized clinical trials and 5 observational studies. The subgroup analysis for these cutaneous reconstruction studies yielded an overall RR of 1.02 (95% CI, 0.78-1.34) (eFigure 2C in the Supplement).
Our meta-analysis found no difference in the rates of postoperative SSIs when comparing sterile vs nonsterile gloves in the setting of outpatient minor surgery. This finding that simple cutaneous surgical procedures can be safely performed with clean, nonsterile gloves was similar in separate subgroup analyses of dental procedures, cutaneous surgery, and MMS (grade of recommendation, 1; strength of evidence, A to B).
This systematic review and meta-analysis provide important insight into an area of medical practice that is large in scope and far reaching. Many health care specialties are involved in outpatient cutaneous minor surgery on a regular basis. Given the current interest in a more mindful approach to the use of medical resources and cost containment, understanding whether the use of sterile vs nonsterile gloves affects SSIs in outpatient minor surgery is important because of the large cost difference between these different glove types.7,28
The strengths of this systematic review and meta-analysis include the pooling of a large number of patients (n = 11 071) and the quality of the studies and trials that were pooled, all being deemed as high-quality research with minimal heterogeneity and risk for bias. Thus, the evidence from the analyses performed as part of this systematic review and meta-analysis would also be considered of high quality. Subtle differences may have existed among the studies analyzed that were not reported, including details of surgical site antisepsis measures, draping, use of topical or systemic antibiotics, or details of hand scrubbing or washing. However, given the large number of patients summarized and the low heterogeneity found in the included studies, we do not believe that these possible differences affect the outcomes of this review or the power of the overall assessment.
Because several of the studies used for this meta-analysis were observational studies, the possibility of publication bias remains. This bias seems unlikely, however, given the number of randomized clinical trials that demonstrate similar results and the relatively similar outcomes noted across all observational studies.
Given the thoroughness of the original search strategy and systematic review, the data in this study would be considered complete. Some studies that compared sterile vs nonsterile gloves were excluded because they had a different study design and outcome measure. One of these excluded studies evaluated the effects on open wounds after surgery3 and another determined the chances of urinary tract infection in the setting of cystoscopy.29 Although cystoscopy is an outpatient procedure, for the purposes of inclusion, it was not considered similar enough to an outpatient surgery to meet inclusion criteria. Specifically, no cutaneous or mucous membrane was cut or breached as part of the cystoscopy procedure.3,29
We should consider some previous findings that are in disagreement with the findings of our study. In 1 study6 that included heightened infection control—multiple sterile precautions, including gloves—the postoperative SSI rates were 2.5% vs 0.9% in the nonsterile and sterile glove categories, respectively. A follow-up study from this same institution18 demonstrated that decreasing the costs of infection control protocols, such as eliminating sterile gloves during the tumor removal stage of MMS and removing sterile gowns and half-sheet drapes during reconstruction, did not affect infection risk in an office-based MMS practice. In contrast, in a study separating outpatient surgical procedures by complexity, procedures that were more complex than simple excision had a postoperative SSI rate of 14.7% using nonsterile gloves, compared with 3.4% when sterile gloves were used.17 Thus, more extensive surgery or complicated repairs may have a higher risk for SSI, which should be taken into consideration. This information should be considered carefully when deciding surgical practice guidelines, because this systematic review may not be generalizable to more complex surgical procedures. Of interest, however, some evidence shows that having an open wound for a long duration in the setting of a complex surgery or repair may not affect the risk for postoperative SSI as much as once was thought.31 In contrast to these disagreements, ample evidence suggests that, in general, nonsterile gloves do not influence the chances of postoperative SSI in outpatient surgical procedures.
When considering surgical practices and guidelines, multiple factors should be considered, including the potential consequences of deviating from accepted sterile glove use and the potential challenges this could cause from a medicolegal standpoint. Patient perception of the sterile technique used should also be considered, in addition to the dexterity that comes from wearing a surgical glove that fits snugly, as opposed to a clean glove that gives the surgeon a different feel. Although the broad use of nonsterile clean gloves may be justified, caution is advised in generalizing this justification to more advanced outpatient surgical procedures that may not pertain to the information summarized in this review and meta-analysis. Future study could include whether duration of surgery and complexity of the repair influence postoperative SSI development in the setting of sterile vs nonsterile gloves.
This systematic review and meta-analysis found no difference in the rates of postoperative SSI in outpatient surgical procedures performed with nonsterile vs sterile gloves. Given the cost difference between these gloves and the multitude of outpatient surgical procedures performed worldwide, these findings could have a significant effect on and implications for current practice standards.
Corresponding Author: Jerry D. Brewer, MD, Division of Dermatologic Surgery, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (email@example.com).
Accepted for Publication: May 5, 2016.
Published Online: August 3, 2016. doi:10.1001/jamadermatol.2016.1965.
Author Contributions: Drs Brewer and Gonzalez had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Brewer, Roenigk, Otley, Erwin.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Brewer, Erwin.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Gonzalez.
Administrative, technical, or material support: Gonzalez, Erwin.
Study supervision: Brewer, Arpey, Roenigk, Otley.
Conflict of Interest Disclosures: None reported.