Changing gloves during cesarean section for prevention of postoperative infections: a systematic review and meta-analysis

This systematic review and meta-analysis was conducted to assess associations between changing gloves during cesarean section (CS) and postoperative infection. A literature search was conducted using the major electronic databases MEDLINE, Scopus, ISI Web of Science, PubMed, CINAHL, and CENTRAL from their inception to September 2020. Randomized controlled trials (RCTs) comparing glove change during CS to no glove change were included. Outcomes of interest were endometritis, febrile morbidity, and incisional surgical site infection (SSI). GRADE approach was applied to assess the quality of evidence. Ten reports of six studies involving 1707 participants were included in the analyses. Glove change was associated with a reduction in the risk of incisional SSI following CS (pooled RR 0.49, 95% CI 0.30, 0.78; moderate quality of evidence). Compared to no glove change, glove change during CS did not reduce the risks of endometritis (pooled RR 1.00, 95% CI 0.80, 1.24; low quality of evidence) or febrile morbidity (pooled RR 0.85, 95% CI 0.43, 1.71; very low quality of evidence). Changing gloves during CS was associated with a decreased risk of incisional SSI. The risks of postoperative endometritis and febrile morbidity were not altered by changing gloves.


Study selection and data extraction.
Titles and abstracts of studies retrieved by electronic searching were screened independently by two review authors. Studies whose titles and abstracts did not meet the inclusion criteria were excluded. The full texts of potentially eligible studies were retrieved and independently assessed by two review authors. The risk of bias of the included studies was independently evaluated by two authors using the Cochrane Risk of Bias Tool for Randomized Controlled Trials 11 . Data were extracted onto a data abstraction form specifically designed for the review. Any disagreements were resolved through discussion with a third person. We collated multiple reports of the same study so that each study rather than each report was the unit of interest. This study included the relevant intervention groups in pairwise comparisons of intervention groups that met the criteria for including studies in the review. We combined groups to create a single pairwise comparison.
The outcomes of interest were endometritis, febrile morbidity, and incisional SSI. Endometritis was defined as a body temperature ≥ 38.0 °C (100.4 °F), fundal tenderness, and purulent discharge from the uterus. The criterion for the diagnosis of postoperative febrile morbidity was defined by the study authors, but most definitions were a temperature ≥ 38 °C following the procedure, excluding the first 24 h postoperatively. SSI was defined according to the Centers for Disease Control and Prevention's National Healthcare Safety Network Criteria for Surgical Site Infection 12 . Incisional SSI assessed in this review included either superficial or deep incisional SSI.

Data analysis.
The random-effects model with Mantel-Haenszel weighting was applied for all two-level meta-analyses to calculate the risk ratios (RRs) and their 95% confidence intervals (CIs) 13 . For studies where no events were observed in one arm, a fixed value of 0.5 was added to all cells of the results table. In the study that did not use one single outcome variable but instead reported the effect of the intervention on different variables, a three-level meta-analysis was applied to avoid an overestimate of confidence in the pooled results 14,15 . We analyzed three-level meta-analyses by using a restricted maximum likelihood estimation method (REML) for estimating the parameters in the model and executed without any covariates. We presented effect estimates of three-level meta-analyses with pooled RRs and their 95% CIs.
Statistical heterogeneity in the two-level meta-analysis was assessed using the I 2 statistic and chi-square test with a significance level of 0.10. For the three-level meta-analysis, three sources of variance were modeled: the sampling variance of the primary studies (level 1), the variance between effect sizes within studies (level 2), and the variance between studies (level 3). We applied a one-sided log-likelihood ratio test to examine heterogeneity within studies and between studies with a significance level of 0.10. We also presented the quantification of variance within-study heterogeneity ( σ 2 2 ) and between-study heterogeneity ( σ 2 3 ). The two-level meta-analysis was performed using RevMan software version 5.4 16 . The three-level metaanalysis was conducted using the metafor package 17 of the R statistical software 18 . Steps of a three-level metaanalysis were followed as previously recommended 19 . Sensitivity and subgroup analysis. Subgroup analysis was carried out according to the types of placental delivery and time of glove change. Sensitivity analysis was conducted by repeating the analysis excluding studies judged to be at a high or unknown risk of selection bias. www.nature.com/scientificreports/ Quality of evidence. We used the GRADE approach to assess the quality of evidence for each outcome.
The GRADE approach covers five domains: (1) risk of bias in the included studies, (2) inconsistency between studies, (3) imprecision in the effect estimate, (4) indirectness of evidence and (5) publication bias. The GRADE approach rates the overall certainty of evidence as high, moderate, low, or very low quality 20 .
Ethical approval. An institutional review board for ethical approval was not required for this study of deidentified information available in the public domain through prior publications. Figure 1 displays the PRISMA flowchart for study selection. A broad search yielded 156 reports from the electronic database searches. Fifty-nine reports were identified from other sources. After deduplication, 138 reports were screened, and 122 reports that did not meet the inclusion criteria were excluded. After reviewing the full texts of 16 reports that potentially met the review inclusion criteria, six reports were excluded because of unavailable full texts (n = 3) and ongoing studies (n = 3). Ten reports of six studies involving 1707 participants were included in the quantitative synthesis (Supplementary Tables S3-8).

Results
Characteristics of included studies. Five studies were published in peer-reviewed journals [21][22][23][24][25] . One included study was available only in an academic search engine 26 . Table 1 displays the detailed characteristics of the included studies. Five included studies were undertaken in the United States, whereas the remaining included study was conducted in India. One included study recruited only women who underwent elective CS 23 .
Of the six included studies, glove change was performed by all surgical team members in three studies 23,25,26 . One included study did not state the person responsible for glove change 24 . In the two remaining included studies, glove change was performed by only the primary surgeon (one study) 22 or primary and assistant surgeons (one study) 21 .
Glove change was performed after delivery of the fetus and after delivery of the placenta in three and two included studies, respectively [21][22][23][24][25] . In a study by Devvoor et al. 26 , there were three comparison groups: the group without glove change, the group with glove change after delivery of the fetus, and the group with glove change after delivery of the placenta.
The effect of glove change assessed on multiple variables for the same outcome was noted in Devvoor et al. 26 . This study reported three outcomes related to incisional SSI including induration, gaping, and pus. We therefore synthesized incisional SSI using three-level meta-analysis.
Three included studies did not state their applied methods of placental delivery applied 23,25,26 . Prophylactic antibiotics were routinely administered in five included studies 21,[23][24][25][26] . In one included study, prophylactic antibiotics were selectively given based on the decision of the attending physician 22 (Table 1).  Figure 2 shows a summary of the risk of bias for each included study.

Risk of bias in included studies.
Two included studies were judged as having an unclear risk of selection bias 24,26 . Because blinding of personnel was technically impossible, all included studies were thus at high risk of performance bias. A high risk of attrition bias and reporting bias was noted in one 23 and two included studies 25,26 , respectively (Supplementary Table S5).

Effects of interventions.
Incidence of endometritis. Changing gloves during CS did not significantly reduce the risk of endometritis compared to no change in gloves (pooled RR 1.00, 95% CI 0.80, 1.24; six studies; 1707 participants; Fig. 3) [21][22][23][24][25][26] . Subgroup analyses showed no significant difference in the risk of endometritis between the two comparison groups when stratified by the type of placental delivery method and timing of intraoperative glove change (Fig. 4).
Incidence of febrile morbidity. There was no significant difference in the risk of postoperative febrile morbidity between women assigned to the group with intraoperative glove change and those who were assigned to the group without glove change (pooled RR 0.85, 95% CI 0.43, 1.71; three studies; 744 participants; Fig. 3) 22,23,26 . Subgroup analyses indicated no significant benefit of glove change during CS in reducing the risk of postoperative febrile morbidity across the types of placental delivery methods and timing of the glove change (Fig. 4).
Incidence of incisional SSI. Overall, intraoperative glove change was associated with a reduction in the risk of incisional SSI (pooled RR 0.49, 95% CI 0.30, 0.78; four studies; 836 participants; Fig. 3) 22,23,25,26 . However, when focused on the included studies that stated their placental delivery methods, the impact of changing gloves on the risk of incisional SSI was inconclusive, as there were only small, included studies with very low event rates contributing to the analysis (Fig. 4). Additional subgroup analysis according to the time of glove change indicated that intraoperative glove change after delivery of the placenta may be preferable to glove change after delivery of the fetus in reducing the risk of incisional SSI (Fig. 4).

Sensitivity analysis.
Sensitivity analyses by excluding studies 24,26 with a high or unknown risk of selection bias showed no marked difference in terms of the magnitude of associations for all outcomes ( Table 2).
Quality of evidence. The quality of evidence for endometritis was low due to a lack of information on selection of participants and lack of blinding in the included studies. We graded the quality of evidence for febrile morbidity as very low due to a lack of information on selection of participants and lack of blinding in the included studies and imprecision of estimation. For incisional SSI, we graded the quality of evidence as moderate due to lack of blinding in the included studies.

Discussion
This study is an assessment of the associations between glove change during CS and postoperative complications, including endometritis, febrile morbidity, and incisional SSI. This systematic review highlights that changing gloves during CS reduced the risk of incisional SSI. Risks of postoperative endometritis and febrile morbidity, however, were not altered by changing gloves. Sensitivity analyses applying the risk of selection bias showed no clinically important changes in the direction of these associations. Interestingly, intraoperative glove change after delivery of the placenta may be preferable to glove change after delivery of the fetus. Almeida et al. 27 published their conference proceeding abstract, which was a systematic review investigating the impact of changing gloves during CS on the risks of postoperative complications. The authors concluded that women assigned to undergo changing gloves after delivery of the placenta had a lower incidence of wound infection than those assigned to the control group (RR: 0.34, 95% CI 0.18, 0.65).
Our review, which is based on ten reports of six RCTs involving 1707 unique participants, reaffirmed that changing gloves during CS reduced the risk of incisional SSI with an overall reduction in the rate of 51% (pooled RR 0.49; 95% CI 0.30, 0.78). Our review can provide additional insights into the results of the existing review in that the sensitivity analyses excluding the studies with a high or unclear risk of selection bias underlined the robustness of the results. The characteristics of the included studies, including surgical team members responsible for the glove change, administration of prophylactic antibiotics, method of placental delivery, timing of glove change, reported outcomes, and risk-of-bias assessment, were determined and presented in greater detail by our review. Subgroup analyses that attempted to assess whether methods of placental delivery and timing of the glove change alter the benefit of changing gloves were also performed in our review. To our knowledge, this is the first review to provide the full details of a meta-analysis on this issue. In addition, our review applied a three-level meta-analysis model for the outcomes that involved jointly analyzing several and related variables to minimize an overestimate of the pooled results 14,15 .
Although available evidence notes a reduction in the risk of incisional SSI following changing gloves during CS, the most appropriate timing of the glove change remains inconclusive. A single RCT involving 100 www.nature.com/scientificreports/ participants that was undertaken to assess the impact of the timing of glove change on postoperative infectious complications noted similar risks of endometritis, postoperative febrile morbidity, and wound complications between a glove change after delivery of the fetus and a glove change after delivery of the placenta 26 . However, a small number of studies with small sample sizes and very low event rates contributed to this finding, thus precluding any meaningful conclusion. It is worth noting that our subgroup analysis according to the timing of the glove change suggested that changing gloves after delivery of the placenta may be preferable to changing gloves after delivery of the fetus (Fig. 4). High-quality RCTs with sufficient sample sizes are required to confirm this interesting finding. Based on a previous systematic review conducted to provide evidence-based guidance for surgical decisions during CS, recommendations with high-certainty evidence according to the US Preventive Services Task Force included preskin incision prophylactic antibiotics, cephalad-caudad blunt uterine extension, spontaneous placental removal, surgeon preference on uterine exteriorization, single-layer uterine closure when future fertility is undesired, and suture closure of the thick subcutaneous tissue. Glove change during CS, however, was not recommended based on available evidence in this review 28 . In our updated review, changing gloves during CS significantly minimized the risk of incisional SSI with an overall reduction in the incidence of incisional SSI of approximately 50%. The quality of this evidence according to the GRADE approach is moderate. Intraoperative change in gloves, therefore, seems to be a promising intervention to lessen the risk of SSI following CS.
Nevertheless, the limitations of current evidence are worthy of consideration in the interpretation of review findings. First, ample evidence suggests that the methods of placental delivery and infectious complications following CS are related [29][30][31][32] . Routine manual removal of the placenta at CS increases postpartum maternal www.nature.com/scientificreports/ infectious morbidity [29][30][31][32] . Information regarding the method of placental delivery, however, was not provided in three of six studies included in this review, which did not enable us to evaluate the impact of the method of placental delivery on the benefits of changing gloves during CS. Second, six records retrieved from the comprehensive search were classified as potentially included studies, but their full texts could not be obtained (three records); additionally, there were ongoing studies awaiting an assessment (three records). Future updated metaanalyses are required when these results are made available. Third, given limited information, evidence regarding whether the type of surgical team member in the operating field who is assigned to change gloves during CS makes a difference in the risk of infectious complications remains unknown. Fourth, subgroup analyses were based on few included studies hence they may have been underpowered to detect differences in the effects of interventions. Finally, the limited number of studies included in this review precluded our ability to assess the potential of small-study effects in this meta-analysis. These limitations should be considered when integrating the policy of glove change during CS into clinical practice.
In conclusion, updated evidence implies that changing gloves during CS can significantly minimize the risk of incisional SSI. Risks of postoperative endometritis and febrile morbidity, however, were not altered by changing gloves. Glove changes after delivery of the placenta may be preferable to glove changes after delivery of the fetus. Any further studies should focus on an assessment of glove change among women undergoing spontaneous placental delivery, which is currently a standard practice 28 .

Data availability
The datasets generated and/or analyzed during the current study are not publicly available but could be made available on reasonable request.  www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.