Sprague et al. Trials (2020) 21:712 https://doi.org/10.1186/s13063-020-04611-9 METHODOLOGY Open Access Cluster identification, selection, and description in cluster randomized crossover trials: the PREP-IT trials Sheila Sprague1,2* , Taryn Scott1, Shannon Dodds1, David Pogorzelski1, Paula McKay1, Anthony D. Harris3, Amber Wood4, Lehana Thabane2, Mohit Bhandari1,2, Samir Mehta5, Greg Gaski6, Christina Boulton7, Francesc Marcano-Fernández8, Ernesto Guerra-Farfán9, Joan Hebden10, Lyndsay M. O’Hara3, Gerard P. Slobogean11, on behalf of the PREP-IT Investigators Abstract Background: In cluster randomized crossover (CRXO) trials, groups of participants (i.e., clusters) are randomly allocated to receive a sequence of interventions over time (i.e., cluster periods). CRXO trials are becoming more comment when they are feasible, as they require fewer clusters than parallel group cluster randomized trials. However, CRXO trials have not been frequently used in orthopedic fracture trials and represent a novel methodological application within the field. To disseminate the early knowledge gained from our experience initiating two cluster randomized crossover trials, we describe our process for the identification and selection of the orthopedic practices (i.e., clusters) participating in the PREP-IT program and present data to describe their key characteristics. Methods: The PREP-IT program comprises two ongoing pragmatic cluster randomized crossover trials (Aqueous- PREP and PREPARE) which compare the effect of iodophor versus chlorhexidine solutions on surgical site infection and unplanned fracture-related reoperations in patients undergoing operative fracture management. We describe the process we used to identify and select orthopedic practices (clusters) for the PREP-IT trials, along with their characteristics. (Continued on next page) * Correspondence: sprags@mcmaster.ca 1 Department of Surgery, Division of Orthopaedic Surgery, McMaster University, 293 Wellington Street North, Suite 110, Hamilton, ON L8L 8E7, Canada 2 Department of Health Research Methods, Evidence, and Impact, McMaster University, 293 Wellington St. N., Suite 110, Hamilton, ON L8L 8E7, Canada Full list of author information is available at the end of the article © The Author(s). 2020 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://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Sprague et al. Trials (2020) 21:712 Page 2 of 12 (Continued from previous page) Results: We identified 58 potential orthopedic practices for inclusion in the PREP-IT trials. After screening each practice for eligibility, we selected 30 practices for participation and randomized each to a sequence of interventions (15 for Aqueous-PREP and 20 for PREPARE). The majority of orthopedic practices included in the Aqueous-PREP and PREPARE trials were situated in level I trauma centers (100% and 87%, respectively). Orthopedic practices in the Aqueous-PREP trial operatively treated a median of 149 open fracture patients per year, included a median of 11 orthopedic surgeons, and had access to a median of 5 infection preventionists. Orthopedic practices in the PREPARE trial treated a median of 142 open fracture and 1090 closed fracture patients per year, included a median of 7.5 orthopedic surgeons, and had access to a median of 6 infection preventionists. Conclusions: The PREP-IT trials provide an example of how to follow the reporting standards for cluster randomized crossover trials by providing a clear definition of the cluster unit, a thorough description of the cluster identification and selection process, and sufficient description of key cluster characteristics. Trial registration: Both trials are registered at ClinicalTrials.gov (A-PREP: NCT03385304 December 28, 2017, and PREPARE: NCT03523962 May 14, 2018). Keywords: Cluster, Randomized crossover, Pragmatic, Cluster characteristics, Orthopedic, Surgical site infection Background our cluster identification and selection process. Add- In clinical research, large high-quality randomized con- itionally, we present data on the key characteristics of trolled trials (RCTs) are considered the highest level of the clusters to illustrate the level of data collection that evidence to determine the effectiveness of an interven- is required for appropriate reporting in CRXO trials. tion [1]. While most RCTs follow a parallel group design and randomize individual participants to one or more Methods intervention and control groups, there are certain situa- The PREP-IT program tions in which randomization at this level is not feasible The Program of Randomized trials to Evaluate Pre- or practical [2]. Cluster randomized trials overcome operative antiseptic skin solutions in orthopaedic Trauma these challenges by randomizing predetermined groups (PREP-IT) aims to determine the effectiveness of iodophor of participants (i.e., clusters) to interventions [3]. In clus- compared to chlorhexidine solutions at reducing surgical ter randomized crossover (CRXO) trials, each cluster site infection (SSI) and unplanned fracture-related reopera- serves as their own control group by participating in tions in fracture surgery patients. The PREP-IT program in- both the treatment and control arm at least once at vari- cludes two ongoing multi-center pragmatic CRXO trials ous periods of time throughout the trial. An advantage that study the effects of four antiseptic solutions in three in- to this design [CRXO] over the parallel group cluster dependent populations of surgically treated fracture pa- randomized trial is that fewer clusters are required to tients: the Aqueous-PREP (A Pragmatic Randomized trial achieve statistical power. Evaluating Pre-operative aqueous antiseptic skin solutions Researchers undertaking CRXO trials must ensure that in open fractures) trial and the PREPARE (A Pragmatic they take into consideration multiple differences in the con- Randomized trial Evaluating Pre-operative Alcohol skin so- duct and reporting between RCTs and CRXO trials [4, 5]. lutions in Fractured Extremities) trial (Fig. 1). The One of these fundamental differences is the requirement to Aqueous-PREP trial compares 4% aqueous chlorhexidine provide information on the characteristics and flow of clus- versus 10% povidone-iodine in open extremity fracture pa- ters throughout the conduct of a CRXO trial (i.e., enrol- tients. The PREPARE trial compares 2% chlorhexidine in ment, allocation, follow-up, analysis), as described in the 70% isopropyl alcohol (ChloraPrep™) versus 0.7% iodine CONSORT extension for cluster randomized trials [4]. This povacrylex in 74% isopropyl alcohol (DuraPrep™) in both is accomplished by including a priori cluster selection cri- open extremity fracture patients and patients with closed teria (similar to participant eligibility criteria) and a cluster lower extremity or pelvic fractures. Both trials follow a flow diagram that shows the cluster selection and flow over single master protocol which has been described in a the duration of the trial (similar to the patient flow dia- previous manuscript [6]. gram). This information enables knowledge users to assess Briefly, the order of treatment allocation for each the generalizability of the trial and determine if the results orthopedic practice (i.e., cluster) was randomly assigned are applicable to their setting and practice. using a computer-generated randomization table. Sim- To disseminate the early knowledge gained from our ple, non-stratified randomization was used. Cluster experience initiating two large infection prevention randomization is performed centrally at the Methods CRXO trials in orthopedic fracture patients, we describe Centre by personnel who are unaware of the cluster Sprague et al. Trials (2020) 21:712 Page 3 of 12 Fig. 1 The PREP-IT research program characteristics. Randomization is performed when a infection, wound healing problems, and fracture healing cluster has received local ethics approval and their clin- problems. The Aqueous-PREP trial will enroll a minimum ical trial agreement is in place. of 1540 open fracture patients, and PREPARE will enroll a Each orthopedic practice starts with the surgical prep- minimum of 1540 open fracture patients and 6280 closed aration solution to which they were initially allocated fracture patients. and subsequently crosses over to the other solution for their second cluster period. This process of alternating Cluster definition treatments repeats approximately every 2 months as dic- We defined clusters as orthopedic practices within par- tated by the initial randomization. This process of alter- ticipating hospitals. Each participating hospital has only nating treatment periods (crossovers) will continue until one participating orthopedic practice. Clustering occurs the minimum sample size is achieved for each fracture at the level of the orthopedic practice as only orthopedic population and the study’s planned recruitment period is surgeons were required to follow the study protocol and completed. Clinical sites must complete at least two use the randomized surgical preparation solution. Dur- treatment periods (one crossover), and the majority of ing the design phase, we considered defining clusters as clinical sites will complete 12 treatment periods (11 either the hospitals with which orthopedic practices crossovers), and 24 months of enrollment. were affiliated or the operating rooms used by ortho- For both trials, the primary outcome is SSI as defined by pedic practices; however, we decided against these defi- the Centers for Disease Control and Prevention (CDC) [7] nitions as other specialities associated with the hospital which includes superficial incisional SSI within 30 days or operating rooms were not required to follow the and deep incisional or organ/space SSI within 90 days of study protocol. definitive fracture management surgery. We continue to follow participants for SSI for 12 months to inform our Cluster eligibility criteria sensitivity analyses. The secondary outcome is unplanned To be eligible for inclusion in the PREP-IT trials, ortho- fracture-related reoperations within 12 months to manage pedic practices were required to meet the predetermined Sprague et al. Trials (2020) 21:712 Page 4 of 12 eligibility criteria as specified in the trial protocols. Spe- questionnaire, which is comprised of 44 questions, cap- cifically, the inclusion criteria were as follows: (1) ad- tures important data on hospital characteristics, current equate research personnel infrastructure to manage the surgeon preferences and practices for pre-operative surgi- study, (2) adequate open fracture volume and closed cal preparation techniques, and infection co-interventions lower extremity and pelvic fracture volume to complete known to reduce the incidence of SSIs. Each orthopedic enrollment within the study timeline (i.e., a minimum of practice’s questionnaire data is updated every 4 months 77 open fractures (for Aqueous-PREP and PREPARE) over the enrollment and follow-up periods. and 314 closed lower extremity fractures per year (for PREPARE only)), (3) commitment from all or most Statistical analyses orthopedic surgeons to participate in the trial, and (4) Our statistical analysis plan was determined a priori. We ability to use the two skin preparation solutions. The ex- included data from the orthopedic practices currently clusion criteria were as follows: (1) lack of interest in the enrolling participants into the Aqueous-PREP trial and trial; (2) anticipated challenges with complying with the the PREPARE trial. We used descriptive statistics to protocol; (3) conflicting studies, in the judgment of the summarize all characteristics of the orthopedic practices Principal Investigators, that would inhibit patient partici- (frequencies and percentages for categorical variables, pation; (4) budgeting or contract constraints; and (5) and medians and interquartile ranges (IQRs) for con- conflicting roles within the trials (e.g., member of the tinuous variables). We used Microsoft Excel 2016 to Data and Safety Monitoring Board). conduct all statistical analyses. Cluster identification and selection Results We identified orthopedic practices through professional Cluster selection contacts with the PREP-IT Principal Investigators. Each Of the 58 potentially eligible orthopedic practices that orthopedic practice received an email containing infor- we identified, 47 (81%) accepted our invitation to mation about the PREP-IT trials, as well as an invitation complete a feasibility questionnaire and 11 (19%) de- to apply to participate in either one or both trials. For clined our invitation and were excluded due to a lack of orthopedic practices who declined this invitation, we interest in the trial (Figs. 2 and 3). Of the 47 practices identified and documented the reason for exclusion that completed the feasibility questionnaire, 12 were de- using our predetermined eligibility criteria. Orthopedic termined to be ineligible and 35 were invited to partici- practices who expressed a desire to participate were re- pate in a teleconference with the Principal Investigators quired to complete a feasibility questionnaire. This ques- and study team to further assess eligibility. Of the 35 tionnaire consisted of 17 questions and assessed the orthopedic practices that participated in the teleconfer- orthopedic practice’s research experience and infrastruc- ences, 5 were determined to be ineligible, 5 were se- ture, fracture volume, current practice patterns, and lected to participate in both the Aqueous-PREP and interest in participating in one or both trials. We used PREPARE trials, 10 in the Aqueous-PREP trial only, and the responses to the feasibility questionnaire to prelimin- 15 in the PREPARE trial only. Specific reasons for ortho- arily screen orthopedic practices for eligibility. Ortho- pedic practice exclusions for each trial are documented in pedic practices that were found to be ineligible through Figs. 2 and 3. One of the practices participating in the this preliminary screen were notified, and the reason for Aqueous-PREP trial was discontinued post-randomization exclusion was documented. Orthopedic practices that due to an inability to follow the trial protocol. Specifically, were eligible to participate following the preliminary this cluster did not meet the a priori threshold of treat- screen were invited to participate in a series of telecon- ment compliance (e.g., 90% of patients receive the allo- ferences with the PREP-IT Principal Investigators and cated antiseptic solution) and the clinical site was unable study personnel. The purpose of these calls was to re- to complete the case report forms, which led to incom- view trial and clinical practice logistics in detail and to plete data submission. The clinical site enrolled 14 partici- confirm whether each orthopedic practice met the eligi- pants and was withdrawn within 2 months of initiation. bility criteria for participation. For any orthopedic prac- Therefore, there are 14 orthopedic practices in the tice that was determined to be ineligible through these Aqueous-PREP trial and 20 orthopedic practices in the calls, we documented the reason for exclusion. Ortho- PREPARE trial. pedic practices that met all eligibility criteria were se- lected to participate. Characteristics of participating clusters Table 1 details the name and location of the hospital as- Collection of cluster characteristics sociated with each orthopedic practice participating in Upon selection, orthopedic practices were required to the Aqueous-PREP and PREPARE trials, as well as the complete a cluster definition questionnaire. This intervention sequence to which they were randomized. Sprague et al. Trials (2020) 21:712 Page 5 of 12 Fig. 2 PREPARE CONSORT flowchart for clusters The majority of orthopedic practices are affiliated with Orthopedic practices participating in the PREPARE trial hospitals that are level I trauma centers and owned by a treat a median of 142 open fractures annually and 1090 non-profit, non-religious organization (Table 2). The closed fractures annually. These orthopedic practices in- median population served is 1,625,000 for orthopedic clude a median of 7.5 orthopedic surgeons who treat practices participating in the Aqueous-PREP trial and 2, fracture patients and 9.5 who take trauma calls. 400,000 for the PREPARE trial. The characteristics of Participating orthopedic practices employ a variety of the withdrawn cluster are presented in the Appendix. surgical infection prevention measures (Table 4). Specif- Orthopedic practices participating in the Aqueous- ically, practices participating in the Aqueous-PREP trial PREP trial treat a median of 149 open fractures annually have a median of 2.2 infection preventionists per 250 and include a median of 11 orthopedic surgeons who hospital beds. Sixty-four percent of hospitals (n = 9) per- treat fractures and 8.5 who take trauma calls (Table 3). form active surveillance for methicillin-resistant Sprague et al. Trials (2020) 21:712 Page 6 of 12 Fig. 3 Aqueous-PREP CONSORT flowchart for clusters Staphylococcus aureus (MRSA) in open fracture patients. Orthopedic practices participating in the PREPARE None of the clinical sites perform active surveillance for trial employ a median of 2.6 infection preventionists per vancomycin-resistant Enterococcus (VRE) in open frac- 250 hospital beds. Twenty percent of orthopedic prac- ture patients. Sixty-four percent of practices (n = 9) have tices (n = 4) perform active surveillance for MRSA in a decolonization protocol for MRSA. All practices have open fracture patients, and 15% of orthopedic practices some infection control parameters in place in their oper- (n = 3) perform active surveillance in closed fracture pa- ating rooms (e.g., high-efficiency particulate air (HEPA) tients. No clusters perform active surveillance for VRE filters, positive pressure operating rooms, air changes in either open or closed fracture patients. Sixty percent per hour, temperature maintained between 20 and 24 °C, (n = 12) of orthopedic practices have a decolonization and humidity between 20 and 60%). protocol for MRSA. Additionally, all practices had some Sprague et al. Trials (2020) 21:712 Page 7 of 12 Table 1 Clusters and location Hospital (cluster) name Location Randomization sequence Trial (1/2 vs. 2/1) University of Maryland School of Medicine, R Adams Cowley Shock Baltimore, MD 2/1 (Aqueous-PREP); 1/2 Aqueous-PREP and PREP Trauma Center (PREPARE) ARE Greenville Health System Greenville, SC 2/1 (Aqueous-PREP); 2/1 Aqueous-PREP and PREP (PREPARE) ARE Hamilton Health Sciences – General Site Hamilton, ON 1/2 (Aqueous-PREP); 2/1 Aqueous-PREP and PREP (PREPARE) ARE IU Health Methodist Hospital Indianapolis, IN 1/2 (Aqueous-PREP); 1/2 Aqueous-PREP and PREP (PREPARE) ARE San Antonio Military Medical Center San Antonio, TX 1/2 Aqueous-PREP; 1/2 Aqueous-PREP and PREP (PREPARE) ARE Banner – University Medical Center Tucson Tucson, AZ 1/2 Aqueous-PREP Hospital Universitari Parc Tauli Barcelona, Spain To be randomized Aqueous-PREP McGovern Medical School at UTHealth Houston Houston, TX 1/2 Aqueous-PREP The CORE Institute Phoenix, AZ 2/1 Aqueous-PREP University of California, San Francisco San Francisco, CA 1/2 Aqueous-PREP University of Florida Gainesville, FL 2/1 Aqueous-PREP Vall d’Hebron Hospital Barcelona, Spain To be randomized Aqueous-PREP Vanderbilt Medical Center Nashville, TN 2/1 Aqueous-PREP Wright State University Dayton, OH 2/1 Aqueous-PREP Brigham and Women’s Hospital Boston, MA 1/2 PREPARE Carolina’s Medical Center Charlotte, NC 2/1 PREPARE Dartmouth-Hitchcock Medical Center Lebanon, NH 2/1 PREPARE Duke University Hospital Durham, NC 2/1 PREPARE Inova Fairfax Medical Campus Fairfax, VA 2/1 PREPARE Massachusetts General Hospital Boston, MA 1/2 PREPARE MetroHealth Medical Center Cleveland, OH 1/2 PREPARE Royal Columbian Hospital New Westminster, 2/1 PREPARE BC Regional Medical Center of San Jose San Jose, CA 1/2 PREPARE Sanford Health Sioux Falls, SD 1/2 PREPARE University of Maryland Capital Regional Health Cheverly, MD 2/1 PREPARE University of Mississippi Medical Center Jackson, MS 2/1 PREPARE University of Pennsylvania Philadelphia, PA 1/2 PREPARE University of Utah Salt Lake City, UT 2/1 PREPARE Wake Forest Baptist Hospital Winston-Salem, 1/2 PREPARE NC infection control parameters in place in their operating transparency and upholding rigorous reporting stan- rooms. dards, this paper clearly defines the cluster unit, provides detailed information about the cluster identification and Discussion selection process, and describes key characteristics of The Aqueous-PREP and PREPARE trials represent two the included clusters (Fig. 4). These are some of the key ongoing CRXO trials within the field of orthopedics. reporting requirements described in the CONSORT ex- The successes we have experienced with cluster identifi- tension for CRTs [4]. cation and selection provide support for the feasibility of Given that the Aqueous-PREP and PREPARE trials use using this trial design in orthopedic trauma trials. As a pragmatic design, when selecting orthopedic practices part of the PREP-IT program’s commitment to (i.e., clusters), we aimed to include a sample that was Sprague et al. Trials (2020) 21:712 Page 8 of 12 Table 2 Cluster characteristics Characteristic Aqueous-PREP, N = 14* PREPARE, N = 20 Level of trauma care provided, n (%) Level I 14 (100) 17 (85) Level II 0 (0) 3 (15) Ownership of hospital, n (%) Non-profit, not religious order affiliated 8 (57) 15 (75) Government 4 (29) 4 (20) Private 1 (7) 1 (5) Private administration with public funding 1 (7) Hospital affiliation, n (%) Independent, free-standing 2 (14) 1 (5) Multi-facility organization (chain) 5 (36) 5 (25) Hospital system, attached 4 (29) 7 (35) Hospital system, free-standing 3 (21) 7 (35) Population size served, median (IQR) 1,625,000 (3,735,699) 2,400,000 (2,875,319) Number of inpatient beds, median (IQR) 679 (287.3) 616 (388.5) Number of operating rooms, median (IQR) 21 (18) 31 (22.5) Number of beds in intensive care unit, median (IQR) 48.5 (44.8) 74 (81.5) Hospital is a primary teaching hospital for a medical school, n (%) 14 (100) 16 (80) IQR interquartile range *The cluster that was withdrawn is not included representative of the orthopedic practices that treat the still ensuring that participating orthopedic practices patient population in “real world” usual care settings. would have the infrastructure and patient volume to This is an essential component of pragmatic trials which successfully follow the protocol and meet study time- aim to determine how well interventions work in the lines. Through our cluster selection process, we identi- “real world” as opposed to explanatory trials which aim fied 58 potentially eligible orthopedic practices and to determine how well interventions can work in ideal ultimately selected 5 sites to participate in both the settings [8]. To accomplish this goal, we designed our Aqueous-PREP and PREPARE trials, 10 in the Aqueous- cluster eligibility criteria to be as broad as possible while PREP trial only, and 15 in the PREPARE trial only. This Table 3 Orthopedic characteristics Characteristic Aqueous-PREP, N = 14* PREPARE, N = 20 Number of orthopedic surgeons who treat fracture patients, median (IQR) 11 (11.3) 7.5 (7) Number of orthopedic surgeons who take trauma calls, median (IQR) 8.5 (13) 9.5 (9.5) Orthopedic training programs, n (%)** Orthopedic fellowship 13 (93) 17 (85) Orthopedic residency 14 (100) 18 (90) Orthopedic clerkship 14 (100) 17 (85) None of above 0 (0) 1 (5) Annual number of operatively managed open fractures, median (IQR) 149 (125) 142 (185) Annual number of operatively managed closed fractures, median (IQR) 1200 (828.8) 1090 (630.5) Annual number of operatively managed closed lower extremity and pelvic fractures, median (IQR) – 725 (435) Number of operating rooms used to treat fractures, median (IQR) 4 (2) 4 (5.5) IQR interquartile range *The cluster that was withdrawn is not included **Percentages do not sum to 100 as categories are not mutually exclusive Sprague et al. Trials (2020) 21:712 Page 9 of 12 Table 4 Surgical infection prevention information Characteristic Aqueous-PREP, N = 14* PREPARE, N = 20 Number of infection preventionists per 250 hospital beds, median (IQR) 2.2 (1.7) 2.6 (1.9) Infection preventionists certified in infection control, median (IQR) 3 (2.8) 4 (3) Perform active surveillance cultures for MRSA in open fracture patients, n (%) 9 (64) 4 (20) Perform active surveillance cultures for VRE in open fracture patients, n (%) 0 (0) 4 (20) Perform active surveillance cultures for other organisms in patients with open fractures, n (%) 2 (14) 3 (15) Perform active surveillance cultures for MRSA in closed fracture patients, n (%) – 3 (15) Perform active surveillance cultures for VRE in closed fracture patients, n (%) – 3 (15) Perform active surveillance cultures for other organisms in patients with closed fractures, n (%) – 1 (5) Decolonization protocol for MRSA, n (%) 9 (64) 12 (60) Policies and/or guidelines on maintaining normothermia during perioperative period, n (%) 14 (100) 16 (80) Airflow system(s) in operating room, n (%)** None 0 (0) 1 (5) Vertical laminar flow 9 (64) 14 (70) Conventional ventilation 6 (43) 7 (35) Other 0 (0) Parameters in place in hospital operating rooms, n (%)** High-efficiency particulate air (HEPA) filters 12 (86) 15 (75) Operating rooms are positive pressure 14 (100) 16 (80) Air changes per hour 13 (93) 17 (85) Temperature maintained between 20 and 24 °C and humidity between 20 and 60% 13 (93) 18 (90) *The cluster that was withdrawn is not included **Percentages do not sum to 100 as categories are not mutually exclusive represents an inclusion rate of 26% and 34%, respect- the result of changes in research infrastructure from ively. Using the Pragmatic-Explanatory Continuum Indi- the time of site selection to site initiation. The Ex- cator Summary (PRECIS-2) toolkit, a tool designed to ecutive Committee, with support from the Method- measure where a trial falls on the pragmatic versus ex- ology Core, made the decision to withdraw this site. planatory continuum on several domains, the setting of As the criteria for cluster withdrawal were estab- the Aqueous-PREP and PREPARE trials received a score lished prior to the initiation of the trial and a small of 4 out of a possible 5 indicating that the trials are proportion of data are affected, bias should not be “mostly pragmatic.” This score is reflective of the fact that introduced as a result of the decision to withdraw most of our included orthopedic practices are affiliated this cluster. Additionally, withdrawing this cluster with teaching hospitals that are level I trauma centers and will improve the overall integrity of the trial, as regional referral centers versus a mix of trauma centers allowing them to continue in Aqueous-PREP would and local community hospitals. This is unsurprising as have led to increased protocol violations, treatment large academic hospitals tend to have both the infrastruc- contaminations, and missing data. ture and patient volume to make participation in the As the primary outcome for both the Aqueous- PREP-IT trials feasible, whereas smaller community hospi- PREP and PREPARE trials is SSI, we documented in- tals may have insufficient research infrastructure. fection control processes. To remain pragmatic with Despite our rigorous cluster selection process, we data collection, we documented the infection control withdrew one cluster as they were unable to comply processes at three different levels: cluster level, sur- with the Aqueous-PREP protocol [6]. Specifically, geon level, and patient level. The most important SSI this cluster was unable to meet the a priori thresh- prevention strategies and infection strategies that are old 90% of patients receiving the allocated antiseptic likely to vary from patient-to-patient, including anti- preparation solution. Additionally, the clinical site biotic prophylaxis and blood sugar control, are docu- was unable to complete the case report forms, which mented at the patient level on the case report forms. led to incomplete data submission. These issues were Other infection control details (e.g., hair removal Sprague et al. Trials (2020) 21:712 Page 10 of 12 Fig. 4 Overview of the PREP-IT cluster identification, selection, and randomization processes methods, type of surgical gown) are collected at the use at 64% of Aqueous-PREP practices and 60% of surgeon level, as these typically vary by surgeon but PREPARE practices). remain consistent across patients. Infection control Providing the knowledge user with a high level of de- measures that are implemented at the hospital level, tail regarding the hospital characteristics, orthopedic including normothermia and decolonization protocols, surgery details, and infection prevention protocols and are documented at the cluster level. procedures allows the knowledge user to determine the We aimed to select orthopedic practices that follow generalizability of the trial and the applicability to their infection prevention protocols that are representative hospital and surgical practice. of the usual care that fracture patients would receive in the real world. We have found some variation in Conclusions the infection control processes across the clusters in When undertaking CRXO trials, it is imperative to fol- both the Aqueous-PREP trial and the PREPARE trial. low rigorous reporting standards, as outlined in the Specifically, we found that 64% of all orthopedic prac- CONSORT extensions for CRTs and CRXO trials, in- tices in the Aqueous-PPREP trial perform active sur- cluding providing knowledge users with a clear defin- veillance cultures for MRSA (64%). In the PREPARE ition of the cluster unit, a thorough description of the trial, we found that only 20% of orthopedic practices cluster identification and selection process, and sufficient perform active surveillance for MRSA in open frac- description of key cluster characteristics. The PREP-IT ture patients and 15% for closed fracture patients. In trials illustrate how this can be accomplished using two both trials, there was substantial variation in the use large infection prevention CRXO trials in orthopedic of decolonization protocols for MRSA (protocols in fracture patients as an example. Sprague et al. Trials (2020) 21:712 Page 11 of 12 Appendix Characteristics of the withdrawn cluster Table 5 Cluster characteristics Characteristic Withdrawn cluster Level of trauma care provided Level I Ownership of hospital Non-profit, not religious order affiliated Hospital affiliation Hospital system, free-standing Population size served 6,000,000 Number of inpatient beds 640 Number of operating rooms 17 Number of beds in intensive care unit 102 Hospital is a primary teaching hospital for Yes a medical school Table 6 Surgical infection prevention information Characteristic Withdrawn cluster Number of infection preventionists per 250 hospital beds 5 Infection preventionists certified in infection control 1 Perform active surveillance cultures for MRSA in open fracture patients No Perform active surveillance cultures for VRE in open fracture patients No Perform active surveillance cultures for other organisms in patients with No open fractures Decolonization protocol for MRSA Yes Policies and/or guidelines on maintaining normothermia during Yes perioperative period Airflow system(s) in operating room Conventional ventilation Parameters in place in hospital operating rooms High-efficiency particulate air (HEPA) filters Operating rooms are positive pressure Air changes per hour Temperature maintained between 20 and 24 °C and humidity between 2 and 60% Sprague et al. Trials (2020) 21:712 Page 12 of 12 Supplementary information Author details 1 Supplementary information accompanies this paper at https://doi.org/10. Department of Surgery, Division of Orthopaedic Surgery, McMaster 1186/s13063-020-04611-9. University, 293 Wellington Street North, Suite 110, Hamilton, ON L8L 8E7, Canada. 2Department of Health Research Methods, Evidence, and Impact, McMaster University, 293 Wellington St. N., Suite 110, Hamilton, ON L8L 8E7, Additional file 1. THE PREP-IT Investigators. Canada. 3Department of Epidemiology and Public Health, University of Maryland School of Medicine, Baltimore, MD, USA. 4Association of periOperative Registered Nurses, Denver, CO, USA. 5Department of Abbreviations Orthopaedic Surgery, University of Pennsylvania Perelman School of PREP-IT Program: The Program of Randomized trials to Evaluate Pre-operative Medicine, Philadelphia, PA, USA. 6Department of Orthopaedic Surgery, antiseptic skin solutions in orthopaedic Trauma; RCTs: Randomized controlled Indiana University School of Medicine, Indianapolis, IN, USA. 7Department of trials; CRXO: Cluster randomized crossover; SSI: Surgical site infection; Orthopaedics, Banner Health and the University of Arizona-Tucson, Tucson, Aqueous-PREP: A Pragmatic Randomized trial Evaluating Pre-operative AZ, USA. 8Orthopedic Department, Parc Taulí Hospital Universitari, Institut aqueous antiseptic skin solutions in open fractures; PREPARE: A Pragmatic d’Investigació i Innovació Parc Taulí I3PT, Universitat Autònoma de Barcelona, Randomized trial Evaluating Pre-operative Alcohol skin solutions in Fractured Sabadell, Spain. 9Department of Orthopaedic Surgery and Traumatology, Extremities; CDC: Center for Disease Control and Prevention; University Hospital Vall d’Hebron, Barcelona, Spain. 10Department of IQR: Interquartile range; MRSA: Methicillin-resistant Staphylococcus aureus; Medicine, University of Maryland School of Medicine, Baltimore, MD, USA. VRE: Vancomycin-resistant Enterococcus; HEPA: High-efficiency particulate air; 11 Department of Orthopaedics, University of Maryland, R Adams Cowley PRECIS-2: Pragmatic-Explanatory Continuum Indicator Summary Shock Trauma Center, Baltimore, MD, USA. Acknowledgements Received: 6 April 2020 Accepted: 15 July 2020 We acknowledge and appreciate the many patients, nurses, national organizations, and others who have contributed to the success of this trial. Additional file 1 details the names, affiliations, and roles of the PREP-IT team; References the Executive Committee is responsible for the overall conduct of the trial 1. Sackett D. Rules of evidence and clinical recommendations on the use of and is comprised of the Principal Investigators and a patient partner. The Ex- antithrombotic agents. Chest. 1989;95(2 Suppl):2S–4S. ecutive Committee is advised by a Steering Committee; multiple clinical, re- 2. Arnup SJ, Forbes AB, Kahan BC, Morgan KE, McDonald S, McKenzie JE. The search, and stakeholder specialty cores; and experts in patient engagement use of the cluster randomized crossover design in clinical trials: protocol for (University of Maryland PATIENTS Program). An Adjudication Committee re- a systematic review. Syst Rev. 2014. https://doi.org/10.1186/2046-4053-3-86. views participant eligibility and reported study events. The Methods Centre 3. Arnup SJ, McKenzie JE, Hemming K, Pilcher D, Forbes AB. Understanding is responsible for the day-to-day management of the PREP-IT trials, which in- the cluster randomised crossover design: a graphical illustraton of the cludes clinical site management, data management, and data analysis. The components of variation and a sample size tutorial. Trials. 2017. https://doi. Administrative Centre is responsible for piloting each trial protocol, contract- org/10.1186/s13063-017-2113-2. ing with each clinical site, and overseeing the Central Institutional Review 4. Campbell MK, Piaggio G, Elbourne DR, Altman DG. CONSORT 2010 Board activities. statement: extension to cluster randomised trials. BMJ. 2012;345(sep04 1): e5661. https://doi.org/10.1136/bmj.e5661. Authors’ contributions 5. Arnup SJ, Forbes AB, Kahan BC, Morgan KE, McKenzie JE. The quality of SS, MB, and GPS conceived the study. SS, TS, and LT analyzed and reporting in cluster randomised crossover trials: proposal for reporting items interpreted the data. SS, TS, SD, DP, and PM drafted the manuscript. All and an assessment of reporting quality. Trials. 2016. https://doi.org/10.1186/ authors critically revised the manuscript and gave final approval of the s13063-016-1685-6. version to be published. 6. Program of Randomized Trials to Evaluate Pre-operative Antiseptic Skin Solutions in Orthopaedic Trauma (PREP-IT) Investigators, Slobogean GP, Sprague S, Wells J, Bhandari M, Rojas A, Garibaldi A, Wood A, Howe A, Harris Funding AD, Petrisor BA, Mullins DC, Pogorzelski D, Marvel D, Heels-Ansdell D, The PREPARE trial is funded by the Patient-Centered Outcomes Research In- Mossuto F, Grissom F, Del Fabbro G, Guyatt GH, Della Rocca GJ, Nguyen U, stitute (PCORI) (PCS-1609-36512) and the Canadian Institutes of Health Re- et al. Effectiveness of iodophor vs chlorhexidine solutions for surgical site search (CIHR) (Foundation Grant); the Aqueous-PREP trial is funded by the US infections and unplanned reoperations for patients who underwent fracture Department of Defense (W81XWH-17-1-070) and the CIHR (Foundation repair: the PREP-IT master protocol. JAMA Network Open. 2020;3(4):e202215. Grant). McMaster University Surgical Associates funded start-up activities at https://doi.org/10.1001/jamanetworkopen.2020.2215. the Methods Centre and The Physician Services Incorporated provided fund- 7. Centers for Disease Control and Prevention (CDC). Surgical site infection ing to the Methods Centre and Hamilton Health Sciences for the Aqueous- (SSI) event.; 2017. PREP trial. The views in this publication are solely the responsibility of the au- 8. Ford I, Norrie J. Pragmatic trials. N Engl J Med. 2016. https://doi.org/10.1056/ thors and do not necessarily represent the views of the PCORI, its board of NEJMra1510059. governors, or methodology committee. Publisher’s Note Availability of data and materials Springer Nature remains neutral with regard to jurisdictional claims in The datasets used and/or analyzed during the current study are available published maps and institutional affiliations. from the corresponding author on reasonable request. Ethics approval and consent to participate Ethics approval for PREPARE and A-PREP was obtained through the Hamilton Integrated Research Ethics Board (Project #4913 and #4336) and Advarra (Pro00028360 and Pro00023709). Informed consent was obtained for all study participants enrolled in the PREP-IT trials. Consent for publication Not applicable. Competing interests The authors declare that they have no competing interests.