A mobile caesarean birth center as a solution to improve access to surgical birth in rural Ethiopia: a mixed methods research protocol

Margo S Harrison (  margo.harrison@cuanschutz.edu ) University of Colorado Anschutz Medical Campus: University of Colorado Anschutz Medical Campus https://orcid.org/0000-0003-2940-2241 Teklemariam Yarinbab Mizan-Tepi University Brooke Dorsey-Holliman University of Colorado Anschutz Medical Campus: University of Colorado Anschutz Medical Campus Gregory Aarons University of California San Diego Biomedical Library: University of California San Diego School of Medicine Ana Pilar Betran World Health Organisation: Organisation mondiale de la Sante Robert Goldenberg Columbia University Medical Center: Columbia University Irving Medical Center Margaret Muldrow Village Health Partnership


Abstract Background
As an evidence-based intervention to prevent maternal and neonatal morbidity and mortality, cesarean birth at rates of under 2%, which is the case in rural Southwest Ethiopia, is an unacceptable public health problem and represents an important disparity in use of this life-saving treatment compared to more developed regions. The objective of this study is to explore an innovative clinical solution (a mobile cesarean birth center) to low cesarean birth rates resulting from the Three Delays to emergency obstetric care in isolated and underserved regions of Ethiopia, and the world.

Methods
We will used mixed but primarily qualitative methods to explore and prepare the mobile cesarean birth center for subsequent implementation in communities in Bench Sheko and West Omo Zones. This will involve interviews and focus groups with key stakeholders and retreat settings for user-centered design activities. We will present stakeholders with a prototype surgical truck that will help them conceive of the cesarean birth center concept and discuss implementation issues related to sta ng, supplies, referral patterns, pre-and post-operative care, and relationship to locations for vaginal birth.

Discussion
Completion of our study aims will allow us to describe participants' perceptions about barriers and facilitators to cesarean birth and their attitudes regarding the appropriateness, acceptability, and feasibility of a mobile cesarean birth center as a solution. It will also result in a speci c, measurable, attainable, relevant, and timely (SMART) implementation blueprint(s), with implementation strategies de ned, as well as recruitment plans identi ed. This will include the development of a logic model and process map, a timeline for implementation with strategies selected that will guide implementation, and additional adaptation/adjustment of the mobile center to ensure t for the communities of interest.

Background
Background & Rationale: When medically indicated, cesarean birth saves maternal, fetal, and neonatal lives. [1][2][3] Historically, the World Health Organization (WHO) recommends that a 10 -15% population cesarean birth rate is appropriate among all global populations. 3 Although the exact ideal cesarean section (CS) rate is unknow, CS rates as population level between 9-19% have been associated with reductions in mortality. [3][4][5] As an evidence-based intervention to prevent maternal and neonatal morbidity and mortality, cesarean birth at rates of under 2%, which is the case in rural Southwest Ethiopia, is an unacceptable public health problem and represents an important disparity in use of this life-saving treatment compared to more developed regions.
Low cesarean birth rates plague many regions of sub-Saharan Africa, so the context of rural Southwest Ethiopia is generalizable to other settings. 6 Barriers to cesarean birth as an intervention to prevent morbidity and mortality in sub-Saharan Africa include those described by the Three Delays Model: 1) the delay in the decision to seek care, 2) the delay in reaching appropriate emergency obstetrical care, and 3) receiving adequate care when the facility is reached. 7 Preliminary data from our target community in Southwest Ethiopia found that the Three Delays is representative of barriers to accessing cesarean birth; delays made it "virtually impossible" for many women to reach the hospital. 8,9 The Three Delays Model describes delays to emergency obstetric care and was published 25 years ago, but little to no progress has been made in addressing these delays and it is still relevant as a barrier to progress in the eld. 7 In a survey of 389 women living in our study region, delays were experienced by 76% of respondents. 8 On average the women reported it took ve hours to reach a facility (the second delay), with 96% of respondents reporting long wait times (the third delay). 8 The Ethiopian pregnancyrelated mortality ratio is 412 deaths per 100,000 live births. 10 The leading cause of maternal and perinatal death is obstructed labor, which can be resolved with cesarean birth. 9,11-16 Therefore, there is a critical need for innovative solutions to low cesarean birth rates resulting from the Three Delays to emergency obstetric care in isolated and underserved regions of Ethiopia, and the world. 8,17 Mobile Cesarean Birth Center & Prior Research: Médecins Sans Frontières, a clinical aid organization, has been providing cesarean birth in tents in con ict zones in low-and middle-income countries. 18 In 2017, a quarter of the surgeries (35,000) they performed were cesarean births. 18 This recent prior experience suggests that mobile cesarean birth centers are a safe and effective method of delivering obstetric care.
Mobile surgical trucks have been successfully used in Latin America to deliver surgery (including gynecologic procedures) under similar conditions, and Médecins Sans Frontières provided 35,000 cesarean births in mobile, temporary tents in 2017 in low-resource settings. 18-21 Prior to their success, surgical trucks were used for fteen years in Ecuador to deliver surgical services to remote areas of the country. [19][20][21] The Ecuadorian truck was out tted with an operating table, surgical light, anesthesia machine, cautery, laparoscopy, and a scrub sink and changing area. 19 It was air conditioned and had its own water supply. 19 Of 4,545 surgeries performed in the truck, 651 (14.3%) were gynecologic. 19 There were reportedly no deaths, four major complications (cardiac arrest, pulmonary embolism, gastrointestinal injury, transfusion), and three minor complications (two wound infections and one laparotomy conversion), which is lower than complication rates reported after general surgery. 19,22 A "Hospitainer" is a customizable shipping container that has been out tted to provide medical and surgical care; it includes an operating room, pre-and post-operation room, a storage/sterilization room, medical equipment, consumables, a generator, and medicines for surgical procedures. 23 Required speci cations of such units include detail on climate control, clean water supply, that all equipment should be battery powered but rechargeable, and ideally the vehicle should be parked for procedures. 23,24 Hospitainer reports that one 800 liter water and one 800 liter waste tank are included with the container with each requiring refreshing/emptying about every week. We will use this Hospitainer as our proposed prototype (picture below; provided with consent of Hospitainer); presenting a concrete example that will give stakeholders a model to consider as they explore, design (adapt), and prepare the unit to meet their needs. These Hospitainers have already been used in Sudan, Congo, Sierra Leone, Somalia, and Gabon with success (as well as in other regions, including the Middle East, Europe, and Americas).
In terms of sta ng such a mobile cesarean birth center, Ethiopia has increased surgical staff nationally through task-shifting emergency surgery to non-physician providers called Integrated Emergency and Surgical O cers (IESOs). 15,[25][26][27][28][29] The program was introduced in 2009; IESOs pursue a three-year course in emergency obstetrics and general surgery, with Helping Babies Breathe and newborn resuscitation as integral components of the training. 15,30 From 2012 -2014, 4,075 operations were performed by IESOs, 63% of which were cesarean births. 15 During this timeframe the cesarean birth rate was 12.5%. 3,4,15,31 However, IESOs work primarily at referral facilities, which highlights the lack of accessible, high-quality, facility-based surgical care and the need for mobile units that still exists in rural areas despite their training.

Methods/design
Scientific Premise: Despite extensive research to improve access to care in low-and middleincome countries (LMIC), access to caesarean section to all women in need is not universal, even though the Three Delays model was published 25 years ago. 7 As such, there is a critical implementation gap in determining how best to provide cesarean birth in vast, rural regions of the continent. Our overarching hypotheses are: 1) the cesarean birth surgical disparity in rural Ethiopia can be addressed by the implementation of a novel, mobile cesarean birth center staffed by advanced practice providers, and 2) the methods we will use to explore, prepare, and design the center for eventual adaptation, implementation, and dissemination will be generalizable to other underserved settings and/or surgical disparities. Preliminary data from 21 individuals from target communities and 10 physicians at the regional referral center found that the mobile cesarean birth center concept was highly (90-100%) acceptable, appropriate, feasible, and usable in their setting. Therefore, our overall objective is to adapt a mobile cesarean birth concept for use in geographically isolated and underserved areas of Ethiopia in order to prepare for subsequent implementation and potential dissemination.
Specific Objectives: Our multidisciplinary team of cesarean birth and obstetrics, implementation science, and public health experts as well as representative stakeholders from Ethiopia are poised to successfully achieve our objectives, guided by the Exploration-Preparation-Implementation-Sustainment (EPIS) Framework (implementation process framework), as follows: 32,33 Objective 1: EXPLORE (first phase of the EPIS framework) the outer and inner contexts of the communities in rural Ethiopia where we will study pre-implementation of the mobile cesarean birth center 33 Implementation Strategies: Identify barriers and facilitators to the delivery of cesarean and test the appropriateness, acceptability, and feasibility of a mobile cesarean birth center as one potential solution 34 Objective 2: PREPARE (second phase of the EPIS framework) to address the outer and inner contextual components of the communities where we will study pre-implementation of the mobile cesarean birth center 33 Implementation Strategies: Develop a formal implementation blueprint (with implementation strategies) for the center that addresses the barriers and facilitators (including all Three Delays) to the delivery of cesarean birth 34 Research Frameworks: In order to adapt and prepare the mobile cesarean birth center to address the public health problem of low cesarean birth rates in the region, our research is guided by the EPIS implementation research process and determinant framework; the author of the framework is part of our research team and has applied the framework in sub-Saharan Africa, previously. 14,32,33,35-37 EPIS provides guidance on understanding barriers, facilitators, and adapting our intervention (Exploration Phase, AIM 1); taking what is learned in exploration and preparing to implement it (Preparation Phase, AIM 1); putting structures, processes, and action into place (Implementation Phase, AIM 2), and beginning with the end goal in mind (Sustainment Phase) so that implementation gains are realized and have the greatest public health impact. 14 The framework examines both the "Outer Context", which refers to system-level factors in the country and community and "Inner Context", which refers to organizational level factors, and "Bridging Factors" that link outer and inner contexts (e.g., policies, collaborations), interorganizational relationships, and innovation factors (e.g., cesarean procedures in Hospitainers). 32,33,38 Using a stakeholder framework, we will include patients and the public, providers, purchasers, payers, policy makers, product makers, and principal investigators. 39 In our target region, this translates to women, husbands, and community leaders (patients and the public); IESOs/physicians (providers); Ministry of Health and Local, Zonal, Regional, and Federal Government representatives (purchasers, payers, policy makers); and our research collaborators (principal investigators). These will be the stakeholders involved in our research activities, guided by the EPIS framework, to ensure rigorous execution of our objectives. and West Omo Zones. Women account for 51% of the population, and 48% are of reproductive age; this suggests that about 612,000 women of reproductive age live in these two Zones. 10 Based on the overall population size, the WHO would recommend five emergency obstetric facilities in the region. 40 40 We intend for these four containers to initially serve the needs of at least four tribes in five communities: the Bench, Me'en , Dizy, and Suri tribes. Because each tribal group is unique, this population includes significant heterogeneity.
We visited all of these communities with our partners in April 2019 to discuss the project and begin the necessary preparations for recruitment; the concept for the mobile cesarean birth center was enthusiastically received at that time (preliminary unpublished data, unable to be cited).
Sample Size: We will recruit volunteers in the communities with the help of community leadership until we have at least 10 participants in each morning and afternoon session, though we expect numbers closer to 20 participants per session. At MTUTH we will invite all physicians and administrators to participate, as well as Zonal health leadership. 41 Recruitment: Women of Reproductive Age: Women 18 years and older seeking antenatal or postnatal care at the community health clinics will be offered enrollment, in person, by study staff. We will offer enrollment until 10 women agree to participate.
Men: Community (religious or tribal) leaders will offer men enrollment in person at community gatherings that they host, until we have at least 10 participants. partnerships. These are the most crucial stakeholders to engage, in our experience and according to the physicians at MTUTH as presented in the preliminary data section.
Clinicians/Administrators/Health Officials: We have already conducted preliminary assessments of the health centers in our target communities and as such have met with the aforementioned stakeholders, including some of the zonal and woreda (regional) health officials.
Study Design: For objective 1, we will conduct semi-structured interviews and focus groups with representative stakeholders (community members and leaders, clinicians, administrators); participants will discuss the center and alternative solutions and ways to ensure the solution(s) respond to the Three Delays and meet community needs. 42 To achieve objective 2, representative stakeholders in modeling and simulating the proposed cesarean birth center [or other potential solution(s)] we will conduct cyclic consensus discussions to optimize the cesarean birth center into a clinically implementable innovation that will be adaptable for future dissemination 43 To achieve our objectives, we will host five, two-day retreats, one with each of the Bench, Me'en, Dizzy, and Suri communities at a location chosen by community leadership, and a sixth retreat at MTUTH with the purchasers/payers/policy makers/providers contingent of outer and inner context stakeholders. The exploration phase of the EPIS framework begins when stakeholders are aware of a public health need and are considering ways to address it. 8 Providers at MTUTH were surveyed and reported that the three delays are relevant to care for their patients, and preliminary data we collected from patients at the facility who hailed from all areas of the zone (we did not collect their self-identified tribal group so this data may not reflect our study population) suggested that women and their husbands are also aware of the need for improved access to emergency obstetric care, and consider the mobile cesarean birth center as an acceptable, appropriate, and feasible solution. 17 Study Activities, Objective 1, Exploration Phase: To achieve the first objective, the Exploration Phase, will focus on exploring the Hospitainer for pilot testing in the region. To do this, on the first day of the visit, we will have a morning session with women and an afternoon session with men and community leaders as most community leaders are male (female leaders will participate with the female focus groups); each iteration noted in the table below (Table 1) will take about 45 minutes to complete. The goal of these retreats is to ensure that the Hospitainer is explored to address specific barriers and facilitators to emergency obstetric care in each community, using Human-Centered Design methods and considering EPIS constructs. [43][44][45] Human-Centered Design "offers problem solvers…a chance to design with communities, to deeply understand the people they're looking to serve…and to create innovative new solutions rooted in people's actual needs". [43][44][45] These methods have been used successfully in rural Ethiopia with teff (grain) farmers to co-create interventions that have been designed through academic-community collaboration, but it is innovative to apply them to optimizing a surgical intervention for cesarean birth, as outlined in Study Activities, Objective 2, Preparation Phase: To achieve our second objective, in the preparation phase we will produce a detailed implementation plan to capitalize on implementation facilitators and address potential barriers, and further assess needs for adaptation. 27,47 Critical to this phase is planning implementation strategies and developing an positive implementation climate in which the adapted Hospitainer is valued and supported; in order to achieve this goal, during the exploration phase (Table 1) and preparation activities ( Table 2) the service and policy environment, and the characteristics of the individuals (women who are patients and consumers) who will use the Hospitainer must be clarified. 47 In order to consider the inter-organizational relationships between entities such as governments, funders, professional societies, and consumers, the second day of the retreat will focus on taking the explored and preliminarily adapted Hospitainer approach that stakeholders have determined is Page 9/18 appropriate, acceptable, and feasible, and develop a specific, measurable, attainable, relevant, and timely implementation blueprint(s), with implementation strategies defined, through activities listed in Table 2, below. 32,33,47  Data Collection Methods: At the beginning of the retreats, verbal consent will be obtained from participants and sociodemographic information will be collected. The sessions will be recorded and transcribed in Amharic. Translation into English will be performed by the facilitators (analysts).
Data Management: No identifiable information will be collected in the proposed research project. There will be no biospecimens or other records obtained. Deidentified transcribed data will then be transmitted securely in Amharic and English and stored on password protected computers at the University of Colorado. This data will not be linked to any other previously collected data.
Outcomes: Completion of objective 1, or the Exploration Phase, will allow us to describe participants' perceptions about barriers and facilitators to cesarean birth and their attitudes regarding the appropriateness, acceptability, and feasibility of the Hospitainer as a solution.
Each community may have variants in the prototype; given the eventual plan for four Hospitainers, tribes will be able to adapt the unit per their preferences (external color, decorations). This approach is consistent with work in adaptation of evidence-based practices while preserving core elements (e.g., surgical setting and procedures) while making adaptations to fit local culture and preferences. 48 Observing, addressing, and documenting these adaptations may assist with subsequent dissemination to other local or global regions.
Completion of objective 2, or the Preparation Phase, should result in a specific, measurable, attainable, relevant, and timely (SMART) implementation blueprint(s), with implementation strategies defined, as well as recruitment plans identified. 32,33,47 This will include the development of a logic model and process map, a timeline for implementation with strategies selected that will guide implementation, and additional adaptation/adjustment of the Hospitainer to ensure fit for the communities of interest. 34 The implementation will be more successful if there is a high degree of fit between the values and needs of the stakeholders and the characteristics of the innovation to be implemented. 27 A summary of the outcomes are presented in Table 3, below. Analytic Methods: We will utilize qualitative content analysis to analyze the data. 49 Using an inductive approach, the team will develop a set of codes from multiple readings of the transcripts using Atlas.ti qualitative data management software. 50 The senior professional research assistant will code the transcripts with the PI and qualitative expert heavily involved in codebook development (e.g., coding the first few iterations), with feedback and participation of the facilitators. All discrepancies in the code definitions and applications will be reconciled through consensus. Codes will be clustered into related categories, which will then guide theme development. 42 We will utilize a quantitative assessment tool to have participants rate the appropriateness, acceptability, and feasibility of the intervention during iteration 5 (Table   1). 46 The tool uses four questions to ask about each concept with a grading system to quantify the results. 46 It will be translated, back-translated, and piloted to ensure applicability to the study populations. Qualitative research with community members has been successful in Ethiopia, previously, including in this region and regarding barriers to surgical care, specifically. 8, [51][52][53][54][55][56][57] The two types of data will be triangulated to produce a joint display of our qualitative and quantitative findings. The same analytic methods that were used during the exploration phase will be used during the preparation phase.
Participant Timeline: All our study activities are planned for October 2021 and will not require any ongoing follow-up of study participants.

Discussion
Harms: Risks to our study subjects includes loss of privacy and con dentiality related to participating in any of the focus groups or retreats. For participants who choose to participate with other members of their community, there may be psychological, social, and potentially cultural risks associated with engaging in study discussions. We will inform participants that they may leave the focus group at any time if they feel uncomfortable and will begin the sessions with setting ground rules in an effort to create an environment of safety and trust. It is also possible that there could be a breach of study data during secure transmission or storage of our transcribed and translated data, however these transcripts will not include any identifying data. If any identifying data was accidentally collected during the interview it will not be included in the transcript. We will inform participants of these risks prior to verbal consent, and if they do not wish to participate, they can decline involvement in the study; however, these risks are highly unlikely. No alternative treatments or procedures are relevant in the context of this study; patients can decline to participate, which is their only alternative to consenting to participate in study activities. We plan to collect and report any adverse events or unintended effects that arise during data collection as observed by study staff.
Protection Again Harms: The planned strategy for protecting against and minimizing all potential risks identi ed is to properly explain the social nature of focus groups to participants to help them understand they will be providing information publicly, in front of other community members and peers. Helping them to fully understand what focus groups are may assist them in deciding whether or not they want to be part of it and be exposed to potential psychological, social, and cultural risks that may exist. Additionally, in terms of the transfer of de-identi ed data, this will occur securely over an encrypted connection directly to CU whose password protected servers and networks can only be accessed by employees with access to the system. Even if the data were to be breached, given that it is encrypted and transmitted in a deidenti ed format, there is less of a chance of that data being directly attributed to any given individual. As we do have physicians performing the consent and the study activities, if they determine that a patient has incurred an adverse outcome related to the study activities, they will be able to manage the logistics of getting that patient to appropriate care in such an event.
Auditing: No audits are planned at this time as the Principal Investigator will be present to oversee all study activities as data is being collected.
Dissemination Policy: The investigators plan to communicate study results to participating professionals and the communities visited by presentation of a report and accompanying PowerPoint to be emailed to community and facility leadership. There is no intention to use professional writers to convey study results, and authorship eligibility guidelines will re ect those required by journal submission. There are no plans for granting public access to the participant-level dataset. MTUTH: This study has been submitted to Mizan-Tepi University Teaching Hospital for review and study initiation will not occur until the study has been approved.
Consent & Assent to Participate: We intend to seek verbal consent at the start of our interviews and focus groups. Verbal consents, with a discussion of risks, alternatives, and bene ts, will be performed before any focus group, interview, or retreat. We will consent in the presence of a non-study staff witness to ensure participants have support and can voice understanding regarding the intent of the study, the risks, bene ts, and alternatives were properly explained, and that the participants had an opportunity to ask questions about the study. We will offer the participants a postcard consent, in their native language, to review in case they have speci c questions regarding the consent process, and we will review the postcard consent with them if they are not literate. The consent will be sought by our local study staff, who are native Amharic speakers, which is the language in which both the consent will be sought, and study activities executed, although translators will be present to interpret if participants prefer to speak in the language of their tribe. As the study staff are also physicians, they will be able to make an informed determination of whether or not the adult subject being offered participation in the study has capacity to provide informed consent. Any subject who is inebriated, has a mental disability, or who cannot give informed consent will not be offered enrollment in the study. Legally, adolescents are unable to consent for their own care in Ethiopia but will be included if their parent/husband is able to consent and they are able to assent to participate. Consent for Publication: No individualized data is being presented that required consent for publication.
Availability of Data and Material: The data has not yet been collected but a data use agreement will be formally signed allowing all principal investigators (Harrison, Muldrow, Yarinbab) complete access to the data with provisions that all investigators need to review any analyses prior to submission for publication.
Competing Interests: The protocol contributors have no relationships to disclose that may be deemed to in uence the objectivity of this paper and its review. The authors report no commercial associations, either directly or through immediate family, in areas such as expert testimony, consulting, honoraria, stock holdings, equity interest, ownership, patent-licensing situations or employment that might pose a con ict of interest to this analysis. Additionally, the authors have no con icts such as personal relationships or academic competition to disclose. The ndings presented in this paper represent the views of the named authors only, and not the views of their institutions or organizations. Appendix.docx