The persistent breach between evidence and practice in the prevention of surgical site infection. Qualitative study

Despite the dissemination of guidelines for surgical site infection (SSI) prevention, a gap between the theoretical measures and their compliance persists. Accurate estimates of the implementation of preventative measures is crucial before planning strategies. A web-based was of 11 of and surgeons. Questions determine their of evidence, personal and actual use of the main pre- ventative measures. of show a moderate skin antisepsis with solutions, Antibiotic prophylaxis is 24 of Screening for S. aureus is performed 27.6%. Hair by shaving is used of responders. The most common antiseptic solutions are alcoholic chlorhexidine (57.2%) and aqueous povidone (23.3%). 62.8% of surgeons allow the solution to air dry before applying surgical drapes. Adhesive drapes in the surgical field are used routinely in 33.4% of cases. Perioperative normothermia, glucose control and hyperoxia are used in 84.3%, 65.9% and 23.3% of cases. Antimicrobial sutures and negative pressure therapy are used by 20.2% and 43.5% of teams, respec-tively. Prior to closing the incision, 83.9% replace surgical instruments always or selectively. Wound irrigation before closing is used in 78.1% of cases, mostly with saline. Check-lists, standardized orders, surveillance, feedback and educational programs were rated most highly by respondents as a means to improve compliance with preventative guidelines, but few of these strategies were in place at their institutions. Conclusion: Gaps in the translation of evidence into practice remain in the prevention of SSI among different surgical specialities. Several areas for improvement have been identified, as some core prevention measures are not in common use.

antiseptic solutions are alcoholic chlorhexidine (57.2%) and aqueous povidone (23.3%). 62.8% of surgeons allow the solution to air dry before applying surgical drapes. Adhesive drapes in the surgical field are used routinely in 33.4% of cases. Perioperative normothermia, glucose control and hyperoxia are used in 84.3%, 65.9% and 23.3% of cases. Antimicrobial sutures and negative pressure therapy are used by 20.2% and 43.5% of teams, respectively. Prior to closing the incision, 83.9% replace surgical instruments always or selectively. Wound irrigation before closing is used in 78.1% of cases, mostly with saline. Check-lists, standardized orders, surveillance, feedback and educational programs were rated most highly by respondents as a means to improve compliance with preventative guidelines, but few of these strategies were in place at their institutions. Conclusion: Gaps in the translation of evidence into practice remain in the prevention of SSI among different surgical specialities. Several areas for improvement have been identified, as some core prevention measures are not in common use.

Introduction
Postoperative infections continue to be the most common complication among patients undergoing any type of surgery, and can represent up to 25% of hospital-acquired infections globally [1][2][3]. The development of a surgical site infection (SSI) causes a substantial increase in the economic burden of surgery, due to prolonged hospitalisation, associated morbidity, diagnostic tests, and treatment [4].
More than fifty specific measures with the potential to reduce SSI rate have been investigated. Periodically, national and international health organizations analyse the evidence of these proposed preventative measures and grade it in the form of clinical practice guidelines. These guidelines should be the translation distillation of all this knowledge into recommendations based on systematic reviews and meta-analysis, and their dissemination should improve outcomes and infection rates in a uniform way among hospitals and among surgeons. However, the level of awareness of and compliance with preventative protocols for SSI seems to be highly variable, as gaining acceptance and compliance requires substantial individual, cultural and organizational changes.
Surgeons are often identified as being key factors in non-compliance; some being unable to change personal and professional behaviour to comply with protocols and checklists [5]. In previous studies among general [6] and colorectal surgeons [7] our group found a low level of awareness about some preventative measures and many areas for improvement were detected to be addressed by scientific societies and health organizations.
The present study was designed by the Observatory of Infection in Surgery to understand the current level of compliance with these guidelines by a range of surgical specialists, prior to grouping the most important preventative measures into bundles and planning a dissemination strategy that could increase their level of implementation at a national level.

Methods
A Web-based survey (SurveyMonkey; https://es.surveymonkey. com/r/BKSJG9D) was distributed to members of the association of perioperative nurses and 10 surgical associations of different specialities at a national level. A link to the site page containing the survey was distributed via email, newsletter and Twitter. The survey was open for 60 days. The questionnaire was designed by a core team with previous experience in conducting scientific surveys, and was submitted for evaluation to a panel of experts belonging to the 11 surgical societies participating in the study. The resulting survey contained 48 general questions aimed at evaluating the practices of perioperative nurses and surgeons in preventative measures of SSI prevention endorsed by international guidelines. The questions of the survey can be found in eAppendix 1. At the end of the questionnaire there were specific questions for each surgical specialty. Furthermore, the questions addressed the level of agreement between their beliefs and the protocols or the usual practice of their units. The agreement rate between the beliefs and the usual practice of all respondents was calculated on a scale from 0 to 100. Other questions were related to those policies already in place or that should be introduced to reduce SSI at respondents' hospitals.
The responses were compared with the recommendations of the most recent clinical practice guidelines: the WHO [8] and CDC [9] guidelines, plus The National Institute of Health and Clinical Excellence (NICE) Guideline (2008 [10] and 2019 update [11]); the Clinical Practice Guide for Surgical Patient Safety of the National Health System of Spain (2010) [12]; the Canadian Patient Safety Institute Guideline (2014) [13]; the 2014 update of the SHEA/IDSA Recommendation [14] the National For personal use only. No other uses without permission. Copyright ©2020. Elsevier Inc. All rights reserved. Health Service Scotland Guideline (2015) [15], and the American College of Surgeons and Surgical Infection Society: Surgical Site Infection Guidelines, 2016 Update [16].
The project was registered with the ClinicalTrials.gov Identifier: NCT04310878, and has been reported in line with the "Consolidated criteria for Reporting Qualitative Research (COREQ)" criteria. The results are expressed in percentages of the total answers obtained. Responses were entered into a computerized database that was analysed using the SPSS program (v.10.0, Chicago, IL, USA). To analyse the relationship between two categorical variables, the chi-square test has been used. Statistical significance was defined at p < 0.05.

Results
A total of 1105 nurses and surgeons answered the survey. Their distribution according to their Association is shown in Fig. 1. Half of respondents work in high-volume teaching centres, most of them pertaining to the National Health Service. Demographics of the respondents and the actual level of use of preventative measures are summarized in Table 1.
Up to 50% of respondents have no regular institutional feed-back of their SSI rate, with significant differences among surgical specialities.
When comparing the awareness of evidence and actual practice (Fig. 2), two patterns of responses were found. First, a pattern of a good level of agreement, with less than a 15% difference between the perception and the implementation rate. For instance, the respondents know there is not enough evidence for the use of an increased fraction of inspired oxygen, negative pressure therapy on the closed wound or antiseptic coated sutures, and consistent with this perception, they are not using them. By contrast, a different pattern shows a high level of disagreement between the perception of evidence and the practice of the respondents. Many of them reported removing hair from the surgical field, and 27% still do it by shaving, even though more than 40% of them are aware of the evidence against these preventative measures. Similarly, the evidence supporting maintenance of preoperative normothermia, plastic wound retractors, alcoholic solutions for skin preparation, surgical site lavage, glove changing and wound irrigation are considered low, but they are widely used practices.
When asking not only about their awareness of scientific evidence, but also their personal beliefs, the results are slightly different (Fig. 3). If evidence perception and beliefs are considered together, many measures recommended by guidelines are also supported by respondents (no hair removal, clipping, alcoholic solutions), and many other measures with a low level of evidence, but based on common surgical good sense and included in several preventative bundles, have a high rate of acceptance (change of surgical instruments before closing the wound or glove changing).
Regarding antibiotic prophylaxis, Fig. 4 shows the perceived barriers preventing staff from following protocol appropriately. Most respondents felt that forgetting prescription or administration due to concurrent tasks, lack of hospital protocol, of precision on who is responsible for prescribing the antimicrobials, and lack of verbal communication among team or computerized decision support were the most common reasons for failing to administer antimicrobials properly.     89.8% of them. Hair removal by shaving is still used by 16.6% of respondents, and is performed in 5.1% of cases by the patient and in 21.1% inside the operating theatre. The most common antiseptic solutions are alcoholic chlorhexidine (57.2%) and aqueous povidone (23.3%) (Fig. 7). Only 62.8% of surgeons allow solution to air dry before applying surgical drapes, 30.8% are in the habit of drying the antiseptic with gauze or absorbent paper, and 6.3% apply drapes before the solution is dry. At the same time, 32.6% of respondents knew about a safety problem or incident related to the use of alcoholic solution use in their theatres. Adhesive drapes in the surgical field are used routinely in 33.4% of cases, and 76.8% of responders operate with a single pair of gloves. A non-impervious device for wound edge protection is used in 30.1% of operations, while 16.2% of respondents do not protect edges (Fig. 8). Perioperative normothermia, glucose control and hyperoxia are used in 84.3%, 65.9% and 23.3% of cases, respectively. Most teams combine different methods to warm patients, such as forced-air blanket (90%), warming mattress (25%) and liquid warming sets (50.8%).
Antimicrobial sutures are rarely used to prevent infection (20.2%), probably because 31% of surgeons believe there is no evidence for its use, and 19% are not aware of them. At the end of a clean-contaminated procedure, 34.2% of teams report always replacing surgical instruments before incision closure, 47.6% change them selectively and 16.1% never do so. Wound irrigation before closing is used in 78.1% of cases, mostly with a saline solution.
Most respondents felt that there was a high discrepancy between published guidelines and actual clinical practice, which translates into an overall disagreement rate of 70%. National and international guidelines have been mostly rated as important or very important by respondents (Fig. 9). As methods of bringing practice closer to evidence, educational programs (79.6%), Enhanced Recovery After Surgery (ERAS) protocols (75.5%), a specialty SSI coordinator (73%), provider specific feed-back (71.1%), standardized orders (62.5%), and surveillance of infection (50.6%), are most highly rated by respondents, but few of these strategies are in place at their institutions (Fig. 10).

Discussion
Surgical infections have become one of the most important health-   For personal use only. No other uses without permission. Copyright ©2020. Elsevier Inc. All rights reserved.
care associated infections (HAIs) and pose increased cost, morbidity and mortality. If SSI have such a big impact on patient outcomes, it would seem logical to apply any available measure to prevent them. Anonymous surveys can be a very useful tool for identifying problems regarding compliance, or why staff fail to follow protocols in specific areas of interest. Despite the publication of several clinical guidelines for the prevention of SSI during the last decade, compliance with passively disseminated policies is poor [6,7,17]. This study was aimed at analysing not only the level of adherence to preventative measures by different surgical specialists, but also at knowing their beliefs and level of awareness of the published evidence.

Contributions of the study
It is quite astonishing that half of the respondent surgeons do not routinely receive feedback about their SSI rate. Significant differences were found among specialities, with cardiothoracic and colorectal being the specialities with the highest rate of feedback, and unfortunately some other specialities, such as head and neck, obesity and vascular surgery providing a very low level of feedback to their surgical teams.
Although guidelines have been considered important, in some topics a low rate of awareness and compliance of the recommendations of international guidelines has been shown. When not only the perceived level of evidence, but also the personal beliefs of respondents are added, the results are different, as many measures well considered by guidelines are better rated by respondents. The comparison of Fig. 2 with 3 illustrates these differences. Although this may be true, their level of implementation of the main recommendations is still low.
Particularly worrisome are the low rate of preoperative nutritional assessment before major surgery, the policies on hair removal (including a rate of 21% of hair removal inside the operating theatre), the high level of use of plastic adhesive drapes on the surgical field, and the products and methods for skin antisepsis (66% aqueous solutions, 87% of brushstroke application, 76% of multiple-use antiseptic bottles, 30.8% of manual drying after antiseptic application). It should be remembered that, regardless of the antiseptic used, allowing time for the preparation solutions to air dry is imperative to maximize its efficacy and prevent a fire hazard [9]. In addition, the habit of drying the antiseptic with absorbent paper can lead to a break of antisepsis if areas not treated with antiseptic are inadvertently touched. Likewise, fire safety policies  should be reinforced, as only two thirds of respondents allow the solution to air dry and 6% apply drapes even when the solution is not dry. An excessive prolongation of antibiotic prophylaxis has been also detected, as some specialists extend prophylaxis for more than 24 h, in special Cardiac, Aesthetic and Head & Neck surgeons. Not all prophylactic measures can be or have been sufficiently evaluated. Nonetheless, despite a lack of evidence some measures are universally accepted by the surgical community resulting from common sense and clinical practice. Among them, organ/space surgical site lavage and wound irrigation at the end of the procedure can be listed. Surgical site lavage at the end of operation is an unresolved issue and the recommendations of the guidelines are disparate [9][10][11][12][13][14][15][16][17], but our study shows a widespread custom of performing cavities and wound lavage with saline, and a less extended use of antiseptic or antibiotic solutions.
Few surgeons use double gloving in this study (23.2%). There is a high rate of perforation of the gloves during surgery, and it has been shown that the addition of a second pair of gloves reduces perforations of the innermost gloves [18]. However, there is no evidence to correlate glove perforation rate and incidence of SSI. Despite this, some institutions have included double gloving in their recommendations, including the Royal College of Surgeons of England (2005) and NICE (2008). Although there is no solid evidence in this regard, there seems to be room for improvement in the glove changing policy, given that only 30.4% of surgeons change them at the end of a digestive anastomosis, 34.8% do not do so before closing the incision and 23.2% never change them.
Replacement of surgical instruments and ancillary devices before closing the surgical wound in clean-contaminated surgery, although recommended by surgical common sense, has a low level of scientific evidence, with some bundles which include the measure having negative results [19,20]. The high rate of instrument replacement in the survey is noteworthy (83.9%).
Altogether, the results support our hypothesis that there is an actual breach between the recommendations of the main clinical guidelines and daily clinical practice in SSI prevention. Moreover, the comparison between the beliefs of the respondents and their usual practice also showed remarkable differences.
The literature on knowledge translation warns of the difficulties related to uptake and compliance with guidelines [21]. Studies have previously shown a significant mismatch between the best scientific evidence and clinical practice as far as SSI prevention is concerned. These results demonstrate that many senior surgeons fail to follow the best surgical practices despite being aware of the evidence underlying them. Most respondents of the survey are aware of the problems associated with converting knowledge into to surgical practice, and suggest different strategies to solve it. The Normalization Process Theory [22], is a descriptive model that may help researchers and clinicians to improve practice, and perhaps may facilitate the introduction of multifaceted processes and new technologies in health systems [23].
Some surveys with similar goals have been published, but most have been done in narrower fields [17,24,25], very specific procedures (such as coronary artery by-pass [26], caesarean sections [27], or arthroplasty [28]. A few surveys have been addressed to operating room nurses [29], and others to members of surgical societies, such as general [6], colorectal [7] or paediatric surgery [30]. The present study collects the opinions of perioperative nurses and surgeons from a range of specialities at a national level, and is also the one that with the highest number of responses.

Limitations of the study
This survey has several limitations. First, it is difficult to calculate accurately its response rate, given the uncertainty about the number of members of the different Societies who actually received the survey invitation. Online surveys are probably able to get a large number of responses at the expense of obtaining a low percentage of response.
Nevertheless, the absolute number of respondents is high and seems sufficiently representative. Furthermore, there seems to be a balanced representation of different types of hospitals (size, teaching and ownership) and surgical specialities, which suggests that the results can be generalized to the reality of surgical practice in the area of interest. These results, although drawn from surgical specialists in a single country, may well represent the actual use of preventative measures in European countries. Secondly, the study may also be limited by selfreport bias; as self-reporting has been shown to overrate performance [31].
In summary, it seems that preoperative shower, surgical hand scrub of the surgical staff, use of impermeable surgical drapes and perioperative normothermia are the measures where current practice guidelines are most closely followed by operative nurses and surgeons. Other measures, such as organ/space surgical site lavage and wound irrigation with saline are frequently used, probably mainly because of surgical tradition. On the other hand, other measures which are highly recommended by the main guidelines are only loosely followed. Practices that can be beneficially modified include: the common practice of routine elimination of hair and razor shaving; the infrequent use of alcoholicbased solutions for cutaneous antisepsis; not allowing antiseptic solution to air-dry; the policy of intraoperatively changing gloves; and the use of liquid and bacterial permeable wound edge protectors.

Conclusions
Knowing the level of implementation of preventative measures and the level of awareness of the providers of the available scientific evidence is essential. Our results suggest that gaps in the translation of best evidence into actual practice in the prevention of SSI in surgery are persistent, even within academic environments. A national platform, such as the proposed Observatory of Infection, could be a way of unifying all these findings (such as surgeons' negative attitudes or noncompliance of measures) and could provide a comprehensive forum to disseminate solutions and improve education and compliance of SSI prevention practices. The active diffusion of homogenous SSI prophylactic recommendations supported by strong scientific evidence ought to reduce instances of SSI consistently. Implementation policies must concentrate not only on the professionals, but also on the context in which they perform.

Financial support
None reported.

Provenance and peer review
Not commissioned, externally peer-reviewed.

Data statement
The research data is deposited on the website which was used for the survey (https://es.surveymonkey.com), and has a specific identifier (SurveyMonkey; https://es.surveymonkey.com/r/BKSJG9D).
All data will be made available on request.

Declaration of competing interest
None.