Non-inferior comparative study comparing one or two day antimicrobial prophylaxis after clean orthopaedic surgery (NOCOTA study): a study protocol for a cluster pseudo-randomized controlled trial comparing duration of antibiotic prophylaxis

Antimicrobial prophylaxis (AMP) is one of the most important measures for preventing surgical site infections (SSIs) in various orthopaedic procedures [1, 2], (http://​www.​chemotherapy.​or.​jp/​guideline/​jyutsugo_​shiyou_​jissen.​pdf), [3]. Generally. AMP is not encouraged for long durations postoperatively, but adequate AMP duration remains controversial in clean orthopaedic surgery. The American Society of Health-System Pharmacists (ASHP) [4] and the Musculoskeletal Infection Society (MSIS) [5] permit postoperative AMP administration, but recommend discontinuation within 24 h after surgery. On the contrary, the recent guidelines from the World Health Organization (WHO) [2] and the Center for Disease Control and Prevention (CDC) [3] recommend AMP administration only for the intraoperative period, with no additional doses after the closure. Theoretically, shorter duration may be superior for not contributing to antibiotic resistance; however, no high-quality evidence in the field of orthopaedic surgery has demonstrated reduction in SSI rate or resistant pathogens by restricting AMP duration to the intraoperative period only [1], (http://​www.​chemotherapy.​or.​jp/​guideline/​jyutsugo_​shiyou_​jissen.​pdf).

There are several limitations of the WHO and CDC guidelines regarding the generalization of orthopaedic procedures. First, although the WHO developed guidelines following GRADE (Grading of Recommendations Assessment, Development and Evaluation) methods based on randomized control trials (RCTs) [2, 6], only a few studies on orthopaedic surgeries were cited and these studies mostly had small sample size [7‐9]. Second, although CDC guidelines cited 21 RCTs (N = 14,285) to conclude that there was no significant difference in SSI rates between intraoperative AMP and those discontinued within 24 h after surgery (P = 0.15) [3], there were only six orthopaedic-related studies (four on fracture surgery, and two on arthroplasty) [7, 8, 10‐13]. The pooled risk ratio of these six studies showed no difference in the SSI rate [7, 8, 11‐13], but there was a trend towards favoring postoperative AMP (pooled risk ratio 0.63 (95% CI 0.32–1.23, P = 0.18), with one study reporting higher SSI rate with intraoperative AMP (Fig. 1) [3]. Moreover, in WHO guideline, when the data were limited to clean orthopaedic, cardiovascular and thoracic surgeries, the pooled risk ratio of the quoted literatures was 0.51 (95%CI 0.36–0.73), significantly favoring postoperative AMP (Fig. 2). Among these studies, 2 studies showed significant SSI reduction with postoperative AMP, while no studies showed significant SSI reduction with intraoperative AMP [2]. From these available data of clean surgeries, the evidence for recommending intraoperative AMP seems weak and inadequate, and thus may not be appropriate for clean orthopaedic procedures.

The Japanese Society of Chemotherapy (JSC) and the Japan Society for Surgical Infection (JSSI) recently developed a practical guideline for AMP (http://​www.​chemotherapy.​or.​jp/​guideline/​jyutsugo_​shiyou_​jissen.​pdf), making recommendations for each common procedure. In total, 11 orthopaedic procedures were selected. According to this guideline, there was insufficient evidence to support shorter AMP duration to prevent SSIs in any orthopaedic procedure [14]. In contrast, they cite evidence indicating that shorter duration may actually increase the risk of SSIs in several orthopaedic procedures (http://​www.​chemotherapy.​or.​jp/​guideline/​jyutsugo_​shiyou_​jissen.​pdf), [15, 16]. From the subgroup analysis of the Cochrane review, a higher SSI risk was observed for a single-dose AMP compared with multiple-doses AMP in closed fracture surgery [16]. In another systematic review, postoperative AMP administration for > 24 h demonstrated lower SSI rate than shorter AMP duration for long-bone tumor surgery with endoprosthetic reconstruction [15]. For spinal surgery, only one RCT was conducted that compared intraoperative with prolonged AMP administration. In this study, the SSI rate was slightly higher in the intraoperative group, although there was no significant difference due to the small sample size [9].

SSIs with resistant pathogens following clean orthopaedic procedures are generally a disaster. Considering the potential risk of antibiotic resistance with longer AMP duration, the JSC/JSSI recommend that AMP be discontinued within 48 h after closure for spinal instrumentation surgeries [1, 9, 17] and arthroplasties [18, 19]. In contrast, the ASHP and MSIS recommend discontinuation of AMP within 24 h [4, 5] and the CDC and WHO recommend discontinuation even earlier [2, 3]. However, with the lack of high-volume comparative studies in the field of orthopaedic surgery, it remains unclear whether shorter AMP duration really decreases antibiotic resistance, or provides better outcomes after orthopaedic surgeries.

Health care–associated infections (HAIs), including postoperative urinary tract infection (UTI) and respiratory tract infection (RTI), may also become disastrous complications after orthopaedic surgeries. They are known to increase medical costs, prolong hospital stays, and increase mortality rates [20‐22]. Theoretically, AMP may also be effective in preventing these HAIs [3, 6], but it is still unclear what clinical impact does AMP duration have on these complications following orthopaedic procedures. A large, well-designed comparative prospective study is needed to evaluate the association between AMP duration and HAIs after orthopaedic surgeries.

Considering the importance of all these infectious complications, we plan to conduct a comparative study evaluating the effectiveness of AMP duration on all HAIs, including SSIs, as a combined outcome. To the best of our knowledge, there has been no comparative trial in the field of orthopaedic surgery comparing the effectiveness of AMP duration for 24 h versus 24–48 h on these complications. In our previous study, AMP was continued for > 24 h in about half of our procedures [23]. Since none of the domestic SSI prevention guidelines recommend stopping AMP after closure [14], it is still not an established general practice to stop AMP without additional postoperative administration. Therefore, it was difficult to recruit participating hospitals to collaborate on a comparative study comparing intraoperative AMP alone with longer AMP duration. Moreover, although the JSC/JSSI recommend that AMP be discontinued within 48 h, there is not enough evidence to justify prolonged AMP of > 24 h following clean orthopaedic procedures. Accordingly, the objective of this study was to examine the effectiveness of AMP duration (24 h vs 24–48 h) on postoperative HAIs after clean orthopaedic surgeries.

Adequate use of AMP is encouraged worldwide, but is generally discontinued as quickly as possible in consideration of the potential risk of antibiotic resistance when used too long [6]. Prolonged use of postoperative antibiotics is discouraged because of possible antimicrobial toxicity and unnecessary expenses [32]. In the recent WHO report, > 50% of Staphylococcus aureus was methicillin-resistance in several countries, including Japan and the US [34]. The threat of antimicrobial resistance is now a global concern, and the Japanese Ministry of Health, Labor and Welfare has launched a national action plan to reduce the methicillin-resistant S. aureus rate from 51 to < 20% by 2020, while reducing the total amount of antibiotics used (https://​www.​mhlw.​go.​jp/​file/​06-Seisakujouhou-10900000-Kenkoukyoku/​0000120769.​pdf). A shorter AMP duration may theoretically reduce the incidence of resistant pathogens as well as the total amount of antibiotics used. Therefore, shorter AMP is an attractive method to combat antimicrobial resistance and should be encouraged if its effectiveness for preventing postoperative complications is comparable to that of longer AMP duration.

Although there is a trend toward short AMP duration worldwide, several concerns require further investigations. First, shortening the duration of AMP may not prevent SSIs. The supportive evidence for preventing SSIs in clean surgery is still limited, and several studies indicated the possibility of a higher SSI risk when the AMP duration is too short [10, 15, 16]. Second, the short duration may consequently increase other HAIs postoperatively, but these relationships are not well established. Third, there is no high-level evidence study supports the notion that shorter AMP duration decreases antimicrobial resistance in the field of orthopaedic surgery. The evidence of higher resistance risk with longer AMP duration is mainly based on observational studies [35, 36], which may not be appropriate to generalize in various clean orthopaedic procedures.

Many Japanese orthopaedic surgeons still prefer AMP of longer durations longer than 24 h. From our previous study, about half the procedures performed in our institutions were continued for > 24 h. Surgeons in participating hospitals were not aware of the CDC/WHO recommendations at the time we planned this trial. Therefore, it was very difficult to conduct a trial comparing those cases without an additional dose after closure, as this was too short for most participating surgeons and was not part of our general practice. Although the JSC/JSSI guideline recommends AMP discontinuation within 48 h (http://​www.​chemotherapy.​or.​jp/​guideline/​jyutsugo_​shiyou_​jissen.​pdf), [14], there is not enough evidence to justify prolonged AMP use for > 24 h in clean orthopaedic procedures. Unnecessary doses of antibiotics may only increase the chances of antimicrobial resistance as well as medical costs. Therefore, we have designed this comparative study to evaluate the efficacy of AMP duration, by comparing AMP duration of 24 h with that of 24–48 h after clean orthopaedic surgery.

Approximately 4% of patients in the US acute care hospitals have at least one HAI [22]. Postoperative HAIs are associated with greater deterioration in patients’ general health status and social/economic burden [6]. In this study, we have hypothesized that the risk of HAIs for 24 h of AMP will be equivalent to 24–48 h of AMP. If Group 24 reveals an equivalent or lower complication rate, this would encourage surgeons to shorten their AMP duration, which may minimize the chance of antibiotic resistance and reduce the antibiotic load. In contrast, if Group 24 reveals a higher HAI risk, we might have to reconsider longer AMP duration in regard to patient safety and push for future research to validate our results. Either result will be informative and has the potential to change our future practice. In Japan many surgeons still prefer AMP longer than 24 h. From the viewpoint of economics and potential resistance, shorter AMP duration should be encouraged. Therefore, we have chosen a non-inferiority design.

The limitation of this study design should be acknowledged. First, the allocation schedule which will be scheduled based on previous data to achieve similar background, will be announced to the participating hospitals before the study. This was to provide a reasonable time for preparing the study intervention, to avoid crossovers and decrease human errors regarding AMP administration. Thus, bias regarding allocation concealment may be present. To minimize the bias, participating hospitals were asked not to select patients based on our allocation schedules. We will also monitor the recruited number of patients, and explore patient backgrounds of the study populations thoroughly to assure comparability. Though patients may recognize which group they belong after 24 h after surgery, they do not know which group they belong at the time when they receive surgery. Bias regarding patients’ preference is less likely to be present, and it is hardly imaginable that the outcome may be influenced by unblinded patients. Second, the variation of doses repeated within and after 24 h in both groups may influence the results. Though, we have strictly controlled not to extend the planned duration for Group 24, and not to forget additional dose after 24 h for Group 48, to avoid cross overs and to improve protocol adherence, the number and amount of doses repeated in both groups will not be strictly controlled considering wide variety of procedures that are included in this study, and to maintain feasibility of our participating hospitals.

The NOCOTA study has been designed to provide evidence to determine adequate AMP duration, not just to prevent SSIs, but to prevent overall infectious postoperative complications. We believe that the results of this trial will have a substantial impact on the future management of clean orthopaedic surgery and provide deeper insights into patient safety.

The trial was registered in the UMIN000030929 and opened on January 22, 2018. The first patient was assigned on May 1, 2018. Patient recruitment was planned to be open between May 1, 2018 and December 31, 2018.