Executive summary: Cesarean delivery, or section (CS), is the single most common surgical procedure performed. Estimates indicate that in low resource settings, CS comprises up to 50% of more of the total volume of operations performed. The World Health Organization recommends national CS rates of between 10-15% to save lives and improve maternal and neonatal outcomes. Population-based work indicates that CS rates of up to 19% are demonstrably related to improved maternal and neonatal survival. However, complications are common, and gynecological and obstetric surgical interventions are associated with high rates of morbidity. In low resource settings, complication rates are particularly high.

The intervention being tested is based on a previously developed program called Clean Cut. Clean Cut is an adaptive, multimodal surgical infection prevention program that integrates perioperative process improvement and patient outcomes measurement using process mapping, training and improved management practices, and compliance with critical standards of surgical antisepsis. It was successfully piloted in five surgical departments in Ethiopia, and reduced the relative risk of infection by 35%. This has been adapted specifically for obstetric and gynecological operations and will be evaluated in a cluster randomized stepped wedge trial design in ten maternity hospitals/departments in Ethiopia in order to reduce infections and other complications for women undergoing cesarean delivery and other obstetric and gynecologic operations.

Background: Cesarean section (CS) is the single most common surgical procedure performed worldwide. In Ethiopia, estimates suggest that CS accounts for 30-50% of all operations performed. Population-based work indicates that CS rates of up to 19% are demonstrably related to improved maternal and neonatal survival. In Ethiopia, the national CS rate per live births was 1.9% in 2016, but rates are highly variable by region. In addition, complications are common, and gynecological and obstetric surgical interventions are associated with high rates of morbidity. Infections and complications following CS are estimated to cause 15% of maternal deaths in the country, and the overall SSI rate following CS is estimated at 9%. Failure to administer preoperative antibiotics has been highlighted as a particular improvement opportunity, and is one of the critical processes this program will focus on.

Lifebox, a charity devoted to improving surgical and anesthesia safety, developed a program to improve compliance with the WHO Surgical Safety Checklist and improve adherence to critical standards of perioperative infection prevention. This initiative, called Clean Cut, is an adaptive, multimodal surgical infection prevention program that integrates perioperative process improvement and patient outcomes measurement using process mapping, training and improved management practices, and compliance with critical standards of surgical antisepsis. The program was the result of a joint collaboration between the Ethiopian Federal Ministry of Health (FMOH), the Surgical Society of Ethiopia, and Lifebox and resulted in a 35% relative risk reduction in postoperative infections.

Aims -

Primary:

1. To reduce postoperative infections in patients undergoing CS Secondary
2. To reduce postoperative infections in patients undergoing other obstetrical and gynecological operations
3. To improve compliance with a core set of critical perioperative infection prevention and control practices that are essential to reducing infectious risks from surgical intervention
4. To reduce the need for reoperation in patients undergoing obstetrical and gynecological operations
5. To reduce the length of stay due to infectious and other complications for patients undergoing obstetrical and gynecological operations
6. To reduce mortality rates in mothers undergoing CS
7. To reduce mortality rates in women undergoing obstetrical and gynecological operations
8. To reduce mortality rates in neonates delivered by CS Ancillary
9. To assess facility readiness for and capacities to engage in quality improvement programs in surgery

The Intervention: Clean Cut focuses on improving compliance with six critical perioperative infection prevention standards:

• appropriate skin preparation of the surgeon's hands and the surgical site
• maintenance of the sterile field by ensuring the integrity and sterility of surgical gowns, drapes, and gloves
• confirmation of instrument sterility
• appropriate antibiotic administration
• complete swab counts
• routine use of the WHO Surgical Safety Checklist

Clean Cut is implemented in five phases:

1. Identification of a Clean Cut team to include members from all perioperative disciplines: Ob/Gyn, surgery, nursing, anesthesia, QI personnel and operating room (OR) management;
2. Establishment of a baseline by conducting context assessments and process mapping the six perioperative standards, and establishing a data collection system;
3. Modification and implementation of the WHO Surgical Safety Checklist to fit local practices;
4. Ongoing data feedback to connect baseline data with process maps, coupled with site-specific action plans for improvement;
5. Targeted training, workshops, and refresher courses using local providers, coupled with a transition to hospital management for sustaining the program.

Clean Cut has been adapted specifically for obstetric and gynecological operations and will implement it in ten maternity hospitals/departments in Ethiopia in order to reduce infections and other complications for women undergoing cesarean delivery and other obstetric and gynecologic operations. Clean Cut has been designed with sustainability at its core - the strategy emphasizes teamwork and collective leadership to identify and address critical gaps in perioperative safety processes.

Study Design: The impact of the CS-tailored Clean Cut program will be tested through a cluster randomized stepped-wedge study design at 10 hospitals over the course of 24 months. Each selected hospital will be distributed into two groups based on the type of facility: university teaching and referral hospitals in one group, and regional, district, and community hospitals in another. Five clusters and their order will be randomly established, paring a teaching/referral hospital with a regional/district/community hospital. These hospitals will start by collecting inpatient and outpatient outcomes on all patients undergoing CS prior to implementation. Over the course of 10 months, 2 hospitals at a time will begin implementing Clean Cut at 2-month intervals. Each starts by assembling a multidisciplinary improvement team, undergoing team training on WHO Surgical Safety Checklist use and implementation, and gathering compliance information about intraoperative safety practices. The initiation of Clean Cut will also involve creating facility-specific process maps of each critical perioperative practice. Once these process maps and compliance data are complete, usually after 2-4 weeks, the team establishes an adaptive, facility-driven improvement plan based on process gaps and barriers to best practice.

Study sites and facility eligibility: Study sites will consist of ten hospitals in Ethiopia that provide maternal surgical services. Five of these hospitals will be university teaching or referral hospitals (aka specialized or referral hospitals), and five will be regional, district or smaller community hospitals (aka general or regional hospitals).

Site randomization: Each selected hospital will be distributed into two groups based on the type of facility: university teaching and referral hospitals in one group, and regional, district, and community hospitals in another. One hospital in each group will be paired to create five clusters; these pairings will be purposive as district and referral hospitals in Ethiopia typically have long-standing relationships which will facilitate implementation at the cluster level and prevent inadvertent crossover of the intervention prior to randomization. The sequence of implementation for each of the five clusters will be established using computer-based randomization (https://www.randomizer.org/).

Participants: As obstetric and gynecologic operations are typically undertaken in separate, dedicated operating theatres, we will focus our prospective observations on patients admitted to these theatres. Any patient undergoing surgery at any time in one of the targeted operating theatres is eligible for inclusion; there will be no exclusion criteria. Enrollment will occur at the time of observation and will include various times (day and night) and days of the week (weekdays and weekends). As the standards being implemented are not in dispute and are considered critical for ensuring antisepsis and sterility, patient informed consent will not be obtained. While our focus will be on cesarean delivery, any obstetric or gynecological operation will be included, with the inclusion of other operations in this populations (such as appendectomy for appendicitis that is found incidentally or misdiagnosed as ovarian torsion, for example). There will be no age range limit.

We will also interview key hospital personnel to understand the context of each facility, its experience with quality improvement initiatives, and the perceived importance of this work to patient safety, patient care, and the work routine. These interviewees will be recruited from the implementation teams involved in Clean Cut. We will also administer surveys in conjunction with Ariadne Labs, a partner in this work, to understand the context, perceptions, and priorities of the various institutions, and how the Clean-CS team can support implementation at the time of intervention (phases 3, 4, and 5)

Sample size and power calculations: Our prior work and review of the literature indicate a baseline SSI incidence of 12% following CS. Given our past work has reduced SSI by 35%, we assume an effect size of Clean Cut resulting in a 25% absolute reduction in SSI (from 12% to 9%). We expect that the preintervention sample will match the postintervention sample in size and general demographic characteristics. As we do not have specific information on the interhospital characteristics or outcomes, we assume an intracluster correlation (ICC) of 0.1 With recruitment of 80-90 patients per cluster per month over 18 months, we expect the final sample to include 7,200 to 8,100 patients, which will be sufficient to detect our expected difference within a reasonable range of ICCs.

Details of our power calculations, data handling, and data forms will be available in our online protocol: https://www.lifebox.org/cleancs/

Analysis: We will conduct pre/post analyses with our planned assessment controlling for risk factors and other demographic and procedure variables listed below that are known to affect infection rates. We will compare patient demographics pre and post intervention as well to evaluate overall matching of patients in each part of the study. We will evaluate compliance individually and in an all-or-none manner as previously described in our Clean Cut pilot work. We will undertake a planned subanalysis of patients observed early during baseline (first two months) and compare them to patients undergoing surgery during the final stage of the study (last two months) after implementation of the program has had time to take effect to assess primary and secondary outcomes.

We will not conduct any interim analyses as we believe that neither benefit nor futility would be readily observable at an interim analysis given the implementation strategy and approach; furthermore, even if we were able to assess benefit or futility we would want to continue the trial to fully study the mechanisms of implementation. Regarding an assessment of harm, if such an outcome were noted we believe this would most likely be due to detection bias with improving ability to capture complications as the study progressed, and we would want to let the study play out to mitigate this detection bias over time.

Risks, benefits, and IRB review: The protocol was approved by the Armauer Hansen Research Institute (AHRI/ ALERT) Ethics Review Committee and the Ministry of Science and Higher Education. As the standards being implemented are not in dispute and are considered critical for ensuring antisepsis and sterility, patient informed consent will not be obtained and waiver for informed consent has been requested. The Investigator will ensure that this trial is conducted in accordance with relevant regulations and with Good Clinical Practice.

The benefits of this study are improvements in compliance with best practices in perioperative infection prevention. There are no known risks other than those typically associated with surgery that are not specific to this trial.

Institutional oversight: Lifebox and the Ethiopian Society of Obstetricians and Gynecologists are jointly organizing and running this study. Lifebox will be responsible for data security and analysis. The ten hospitals are listed in this protocol.

Funding: UBS Optimus Foundation through a Gates Grand Challenge Grant

Study Contact:

Thomas Weiser MD MPH Associate Professor of Surgery, Stanford University Consulting Medical Officer, Lifebox

Inclusion Criteria:

• Any patient undergoing obstetric and gynecologic surgery at any time in one of the targeted operating theatres is eligible for inclusion

Exclusion Criteria:

• there will be no exclusion criteria

women undergoing obstetric or gynecological operations

Data sharing will likely occur at the request of other researchers in accordance with Ethiopian policies governing these data.

patient data will be available upon completion of the trial

All reasonable requests