Does earlier bathing increase the risk of surgical site infection? A meta-analysis of 11 randomized controlled trials

Introduction

In outpatient or emergency department settings, patients recovering from sutured wound closure often ask whether and when the direct water contact with the skin wound is permitted. If regular bathing and body cleaning are not carried out during long-term treatment, body odor and negative emotions can result, especially in areas with high humidity and high temperature (1). Not bathing for a week or even several months significantly affects a patient’s comfort and reduces their quality of life.

In the 1980s, numerous authoritative studies addressed the question of whether wounds without stitches could be soaked in water and bathed. The traditional practice was to keep the wounds dry until the stitches were removed (2). However, guidance issued by an expert panel at the U.S. National Institutes of Health (NIH) states that it is relatively safe to bathe the epidermis and cover the wound within 48 h of wound closure (3). Furthermore, studies have found that earlier postoperative bathing does not increase the risk of suturing wound infection compared with bandaging without bathing (4, 5, 6).

Therefore, the possibility of earlier postoperative bathing has become the focus of international controversy (7), mainly concerning the potential for infection and its potential effects on healing the injured tissue if the wound surface becomes wet. The risk of infection depends on the type of wound (open wounds are more susceptible to infection than epidermal wounds), medical complications, type and quality of water used, and wound complications. However, because it is clinically difficult to determine the exact probability of infecting a sutured wound, it is difficult to fully determine whether bathing truly increases the probability of wound infection. Although as early as 1976, earlier postoperative bathing was reported not to increase the incidence of postoperative complications (8), over the years, surgeons have been reluctant to let patients bathe.

To further investigate the evidence or lack thereof regarding the timing of postoperative bathing, we conducted a systematic review of the literature to clarify whether and how direct water contact influences the healing of closed wounds, providing a reliable proof for the later formulation of guidelines and the promotion of concepts.

Methods

The study conforms to the principles outlined in the Handbook of the Cochrane Collaboration (9), along with the guidelines established by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (10). The protocol for this meta-analysis was registered on PROSPERO (Registration No.: CRD 42022303738).

Results

Eligible studies and study characteristics

We initially identified a total of 1801 unique published studies via the search methods detailed above and finally included 11 eligible trials in the meta-analysis, all of which reported surgical site infection (5, 8, 12, 13, 14, 15, 16, 17, 18, 19, 20). The reference flow is shown in Fig. 1. The details of included trials are shown in Table 1. These trials comprised 2964 participants, of whom 122 developed an infection at the surgical site. To ensure reliable study data, we evaluated the risk of bias and quality of evidence for each of the included studies. Figure 2 presents the risk of bias for each study.

View Full Size

Flow diagram for search and selection of included studies.

Citation: EFORT Open Reviews 9, 6; 10.1530/EOR-23-0062

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Risk of bias summary: review authors’ judgments about each risk of bias item for each included study (5, 8, 12, 13, 14, 15, 16, 17, 18, 19, 20).

Citation: EFORT Open Reviews 9, 6; 10.1530/EOR-23-0062

• Download Figure
• Download figure as PowerPoint slide

Table 1

Characteristics of included studies. Values are presented as the mean ± s.d., unless otherwise indicated.

Study	Country	Study design	Participants, n	Age, mean ± s.d.		Surgical site	Definition of SSI	Intervention
				Earlier bathing	Delayed bathing			Earlier bathing	Delayed bathing
Fraser et al. (8)	England	RCT	108	NA	NA	Miscellaneous	Own definition	Bath 48 h PO	Bath-forbidden
Ghandi et al. (18)	Iran	RCT	468	28 ± 8.1	31 ± 6.9	Dermatological surgeries	Defined by two dermatologists	Wet 24 h PO	Wet 48 h PO
Gök et al. (19)	Turkey	RCT	51	NA	NA	Coronary artery bypass graft	CDC criteria	Bath 48–72 h PO	Bath-forbidden
Goldberg et al. (5)	USA	RCT	200	NA	NA	Clean head and neck lacerations or incisions	Not defined	Bath 8–24 h PO	Bath-forbidden
Heal et al. (12)	Australia	RCT	857	56.5 ± 16.2	55.9 ± 16.6	Minor skin excision	CDC criteria	Bath 12 h PO	Bath 48 h PO
Hsieh et al. (13)	Taiwan	RCT	440	55.3 ± 14.2	55.6 ± 13.8	Miscellaneous	CDC definitions	Bath 48 h PO	Bath-forbidden
Neues & Haas (14)	Germany	RCT	572	39 (22–78)*	53 (23–81)*	Varicose surgery	Not defined	Bath 48 h PO	Bath-forbidden
Palese et al. (15)	Italy	RCT	53	51.4 ± 17.6	52.5 ± 15.6	Neurosurgical surgery	CDC criteria	Bath 72 h PO	Bath-forbidden
Riederer & Inderbitzi (16)	Germany	RCT	101	NA	NA	Inguinal hernia	Not defined	Bath 24 h PO	Bath-forbidden
Voorhees et al. (17)	USA	RCT	82	NA	NA	Miscellaneous	Own definition	Bath 48 h PO	Bath-forbidden
Yu et al. (20)	USA	RCT	32	NA	NA	Total knee arthroplasty	Not defined	Bath 48 h PO	Bath-forbidden

*Median (range).

RCT, randomized controlled trials; CDC, Centers for Disease Control and Prevention; NA, not applicable; PO, post-operation.

Primary outcome: surgical site infection

The pooled infection rate was 3.57% (55/1542) in the earlier postoperative bathing group and 4.71% (67/1422) in the delayed postoperative bathing group. There was no statistically significant difference in the proportion of participants who developed surgical site infection between these two groups (OR: 0.84, 95% CI: 0.58–1.22, P = 0.368, I² = 0%; Fig. 3) The Harbord test showed no small-study effects (P = 0.117).

View Full Size

Forest plot of comparison: earlier bathing versus delayed bathing; outcome: incidence of wound infection. Insufficient data: studies that reported no events relevant to our outcome measure. Risk of bias legend: (A) random sequence generation (selection bias); (B) allocation concealment (selection bias); (C) blinding of participants and personnel (performance bias): objective outcomes; (D) blinding of outcome assessment (detection bias): objective outcomes; (E) incomplete outcome data (attrition bias): objective outcomes; (F) selective reporting (reporting bias); and (G) other bias (5, 8, 12, 13, 14, 15, 16, 17, 18, 19, 20).

Citation: EFORT Open Reviews 9, 6; 10.1530/EOR-23-0062

• Download Figure
• Download figure as PowerPoint slide

Cumulative meta-analysis

Cumulative meta-analyses were performed in order of years of publication. The results showed that the cumulative effect size stabilized, and the 95% CI narrowed gradually. The cumulative effect OR was 0.57, and the 95% CI was 0.27–1.19 (Fig. 4), indicating no difference between groups in the rate of infection.

View Full Size

Cumulative meta-analyses of the included studies for the incidence of surgical site infection. Risk of bias legend: (A) random sequence generation (selection bias); (B) allocation concealment (selection bias); (C) blinding of participants and personnel (performance bias): objective outcomes; (D) blinding of outcome assessment (detection bias): objective outcomes; (E) incomplete outcome data (attrition bias): objective outcomes; (F) selective reporting (reporting bias); and (G) other bias (5, 8, 12, 13, 14, 15, 16, 17, 18, 19, 20).

Citation: EFORT Open Reviews 9, 6; 10.1530/EOR-23-0062

• Download Figure
• Download figure as PowerPoint slide

Other wound complications

Rates of other wound complications were reported in three of the included studies. These mostly included redness and swelling, wound dehiscence, secretion, and hematoma. For the incidence of hematoma, there was a significant difference between the earlier and delayed postoperative bathing groups (OR: 0.25; 95% CI: 0.09–0.73, P = 0.01; Fig. 5); for the other wound complications, no significant difference was found.

View Full Size

Forest plot of comparison: earlier bathing versus delayed bathing; outcome: incidence of wound complication. Risk of bias legend: (A) random sequence generation (selection bias); (B) allocation concealment (selection bias); (C) blinding of participants and personnel (performance bias): objective outcomes; (D) blinding of outcome assessment (detection bias): objective outcomes; (E) incomplete outcome data (attrition bias): objective outcomes; (F) selective reporting (reporting bias); and (G) other bias (13, 14, 16).

Citation: EFORT Open Reviews 9, 6; 10.1530/EOR-23-0062

• Download Figure
• Download figure as PowerPoint slide

Satisfaction

Four of the included studies evaluated patient satisfaction (13, 14, 15, 16). Data from the study by Palate et al. were not pooled because no data were provided (15). In the remaining three studies, the earlier postoperative bathing group showed a significant increase in satisfaction compared with the control group (OR: 101.91, 95% CI: 36.92–281.29, P < 0.0001, I² = 62.3%; Fig. 6).

View Full Size

Forest plot of comparison: earlier bathing versus delayed bathing; outcome: satisfaction. Risk of bias legend: (A) random sequence generation (selection bias); (B) allocation concealment (selection bias); (C) blinding of participants and personnel (performance bias): objective outcomes; (D) blinding of outcome assessment (detection bias): objective outcomes; (E) incomplete outcome data (attrition bias): objective outcomes; (F) selective reporting (reporting bias); and (G) other bias (13, 14, 16).

Citation: EFORT Open Reviews 9, 6; 10.1530/EOR-23-0062

• Download Figure
• Download figure as PowerPoint slide

Discussion

Earlier bathing improves patient satisfaction, reduces the incidence of respiratory infections and venous thrombosis (7), and does not increase the incidence of wound complications. To provide stronger evidence supporting this opinion, we conducted a meta-analysis to assess the effect of earlier postoperative bathing on wound healing. In this meta-analysis of 11 randomized controlled trials with a total of 2964 participants, all included patients who underwent elective surgery and underwent strict sterilization before surgery. As a result, patients were well prepared for surgery. Except for two articles comparing earlier water exposure (after 12 or 24 h post procedure) and late (after 48 h post procedure) (12, 18), all other comparisons were made between contact post-surgery and no contact with water before suture removal. The pooled results show that earlier bathing is not significantly associated with surgical site infections. Furthermore, the findings suggest that earlier bathing improves patient satisfaction but does not increase the rate of other wound complications such as redness and swelling, wound dehiscence, and secretion. Although a statistically significant increase was found for hematoma, only one study reported this rate, and the sample size was small, with the potential for bias being large, and the evidence level being low. To speculate on potential reasons for the observed lower incidence of hematoma in the early bathing group, we can consider several hypotheses based on existing knowledge and clinical experience such as improved wound healing due to better blood flow and oxygenation, enhanced removal of inflammatory mediators or debris, and reduced bacterial colonization through early cleansing and subsequent wound care. As shown in Fig. 2, the risk of bias assessment revealed a high-risk level. This classification is primarily attributed to performance bias. It is imperative to acknowledge that high risk of bias may have introduced limitations to the reliability of our findings, potentially leading to overestimation or underestimation of results. We have conducted sensitivity analyses to assess the actual impact of high-bias risk on our results. While highlighting this limitation, we also emphasize the need for cautious interpretation. It is essential to consider other aspects such as the number of studies, consistency, and effect size when evaluating the overall evidence. Additionally, we suggest future research directions, including stricter methodology, comprehensive reporting, and enhanced quality control measures, to mitigate the influence of high bias risk in similar studies.

Besides, the included study by Yu et al. provides valuable insights into patient bathing practices specifically related to total knee arthroplasty and its implications for orthopedic practice. The findings from the study by Yu et al. suggest that early showering, starting at 2 days post operation, does not increase the risk of bacterial recolonization at the surgical site in total knee arthroplasty (TKA) patients. There was no significant difference in the rate of colonization or bacterial type between the early showering group and the delayed showering group. This finding supports the consideration of early showering as a safe practice for TKA patients, enhancing their comfort, facilitating early mobilization, and aligning with patient preferences. Early showering can contribute to improved patient satisfaction, psychological well-being, and compliance with infection control measures. These findings, along with the conclusions drawn from other studies in abdominal surgery, neurosurgery, and various surgical subspecialties, suggest that patient bathing practices may not significantly impact surgical site outcomes in terms of bacterial recolonization. However, it is important to acknowledge the heterogeneity in the available literature and the need for further research specifically focused on patient bathing after orthopedic surgical procedures. While the conclusions for orthopedic patients based on the current evidence may be limited, the collective evidence suggests that patient bathing practices can be approached cautiously and individualized based on the specific surgical context and patient preferences.

Studies have shown that bathing earlier postoperatively can effectively remove sweat, dirt, microorganisms, and necrotic tissue, thereby accelerating wound healing and reducing the risk of infection. Not bathing for 2 or 3 days after surgery has a negative impact on patient comfort and can lead to the risk of infection (2). The traditional recommendation is to cover the wound with a dressing for at least 48 h, as this is the period when wound epithelialization occurs (21). Post-bathing, there is no evidence of increased growth of bacteria on the skin around the wound (20). Considering all the above evidence, it seems illogical to prohibit patients from taking an earlier postoperative bath. However, more than half of doctors nonetheless reject the evidence and say they will not apply it to the clinic, especially in cardiovascular surgery, orthopedics, and neurosurgery. In clinics, the wound was either taped over, or the patient was advised to wait until the wound had healed before bathing. In addition, most patients do not realize that the wound can be exposed to water, and even after receiving the doctor’s explanation, they may be nonetheless worried about the occurrence of complications such as wound infection.

There have been relevant meta-analysis studies in the past to explore this scientific problem (7, 22), but each has had their own shortcomings. Copeland-Halperin et al. searched only within PubMed and restricted the publication language to English, thereby increasing the potential bias of the report. Additionally, they included only seven articles (one retrospective study and six randomized clinical trials (RCTs)), concluding that earlier postoperative bathing did not increase the risk of infection (22). In 2018, Yu et al. also conducted a meta-analysis, which included a total of 12 articles (two retrospective and ten RCTs), and came to the same conclusion (7). However, this article included a study on the use of saunas; we investigated whether bathing affects wound infection, mainly because the wound increases the risk of infection when exposed to water. Because sauna use mainly implies contact with sweat, which is relatively sterile, it is not the focus of our discussion, and our study excluded it (23). Unlike previously published studies, our meta-analysis only included RCT studies, did not restrict language, and searched multiple databases. Additionally, as years have passed and published evidence has accumulated, our analysis was able to include more published RCT studies with larger sample sizes are larger and higher evidence levels.

Although we conclude that earlier postoperative exposure to water has no significant effect on wounds, most of these studies were conducted in developed countries. Sanitation conditions vary from place to place, and some developing areas or island communities may still be living with rainwater, well water, and other sources when groundwater is scarce or polluted. Rainwater is the source of drinking water most likely to be contaminated by feces, and during collection, it can be contaminated by roofs and other collection surfaces, and often has no mobility, which can easily lead to bacterial growth when stored (24). In developing countries, tap water may not be disinfected, and the number of bacteria in tap water may greatly exceed the standard. Even if tap water is disinfected, water can become susceptible to infection during storage and pipeline transportation (25). Further, the bacterial requirements for domestic water are developed based on levels of common pathogenic bacteria that are pathogenic to the human digestive system after drinking. For water used to clean wounds, the requirements for bacteria should be higher. It is generally accepted in the medical profession that wounds can become infected when the bacterial colonies per gram of tissue or per milliliter of liquid exceed 10⁵ colony-forming units (CFUs) (26). In summary, if the patient is given permission for earlier postoperative bathing, we recommend the following: First, for these patients, it is best to choose intradermal sutures or use smooth silk threads or leather nails for sutures, because if silk thread is used, the water will be immersed in the deep tissue along the silk thread. Secondly, the bathroom used for bathing should be disinfected. Thirdly, domestic water should be boiled and cooled before use, which further eliminates most of the bacteria. Fourthly, rubbing the wound should be avoided during bathing. Fifthly, it is recommended to carry out local disinfection of the wound after bathing. Sixthly, the patient’s pain should be managed because wound pain will reduce the likelihood of bathing (27). RCTs involving more countries are needed to further support these recommendations.

Based on the available studies, our findings provide a high level of evidence that earlier bathing does not increase the incidence of infection and improves patient comfort. However, a gap between evidence-based knowledge and decision-making exists at all levels of health care. Some patients may still refuse earlier postoperative access to water, mainly because of a lack of awareness, the treating team’s hesitance to allow an earlier bath, and the lack of advocacy from a trusted institution. Therefore, the medical community and government health authorities should develop and disseminate, including via social media, clinical practice guidelines that allow clinicians to agree and adhere to recommendations, and guide patients to make correct, beneficial treatments to promote proper care.

Conclusion

This meta-analysis shows that earlier postoperative bathing does not increase the risk of wound infection and other complications, and can improve patient satisfaction with medical care. Given this medical evidence, the medical community, health authorities, and governments should develop and disseminate clinical practice guidelines that allow clinicians to adhere to recommendations that guide patients to evidence-based and beneficial treatment.

ICMJE Conflict of Interest Statement

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the study reported.

Funding Statement

This work did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector.

Author contribution statement

YR and HL collected data, performed data analysis, interpreted results, and wrote the first draft of the manuscript. WJP and HY reviewed the protocol, screened articles, extracted data, and reviewed the results and manuscript. LC and HL contributed to the systematic review protocol and critically reviewed the results and manuscript. HL contributed to the protocol development and reviewed the manuscript. ZW provided critical advice during the revision process. All authors contributed to the conception and design of the study and reviewed all documents and materials.