Background

Maternal mortality is a significant concern in the developing world, where the majority of all maternal deaths occur [1]. Most maternal deaths are preventable and disproportionately affect the poorest and youngest women worldwide. The 2016–2030 Sustainable Development Goals’ target for maternal mortality is to reduce the global maternal mortality ratio to less than 70 per 100,000 live birth by 2030 [2, 3]. Although there have been global reductions in maternal mortality in recent years, Latin America still faces challenges. In particular, Guatemala has one of the highest maternal mortality rates in Latin America, with a rate of 88 deaths per 100,000 live births [4] in 2015. Causes for maternal mortality in Guatemala include postpartum hemorrhage (PPH), hypertension, sepsis, and unsafe abortion [5].

Given that the high maternal mortality rate may in part be due to limited training of health workers to adequately identify, prevent, or handle such emergencies [6], healthcare worker training programs have grown in popularity. Simulation-based training (SBT) for educating health professionals is one model that has consistently been found to improve knowledge, skills, behaviors, and patient-related outcomes [4]. SBT can be especially helpful in low and middle income countries (LMICs) given that it does not require expensive resources beyond human capital to facilitate teaching. By participating in simulations, trainees practice medical skills, effective communication, and teamwork under pressure, which can enable them to provide better and safer care to patients. Given the high acuity of obstetric emergencies and need for team-based care, SBT has been used to train healthcare workers in the management of obstetric emergencies in an effort to reduce maternal mortality [7,8,9,10,11].

We and others have published studies [12] evaluating SBT in LMICs with a focus on short and longer-term retention of learning, however there is still a need to better understand the process of implementation and uptake of training into local practice cultures [13]. In order to better elucidate the long-term impact of SBT on obstetric emergencies, this study sought to provide an in-depth look at one institution’s experience implementing SBT for the management of postpartum hemorrhage (PPH) in Guatemala. Using a mixed-methods approach, we aimed to assess if the introduction of SBT impacted provider attitudes, practice patterns, or patient management.

Methods

This study was approved by the Stanford University Institutional Review Board.

Study activities took place at the Instituto Guatemalteco de Seguridad Social (IGSS) Pamplona hospital in Guatemala City, Guatemala. Data collection was overseen by Stanford University and IGSS hospital faculty.

IGSS Pamplona is a large, tertiary academic hospital that manages approximately 8000 to 9000 deliveries per year. It is considered a high-risk hospital that receives referrals from neighboring hospitals for complicated pregnancies. IGSS Pamplona does not have an Adult Intensive Care Unit (ICU), however, and sends postpartum ICU patients to an affiliated, nearby sister site. IGSS Pamplona has PPH management protocols in place that are based on World Health Organization (WHO) and national guidelines.

The Global Outreach-Mobile Obstetrics and Medical Simulation (GO MOMS) program is a SBT program developed at Stanford University to provide a standardized obstetrical and gynecological training tool for faculty and resident learners within teaching hospitals in low-resource settings [14]. Topics covered in GO MOMS include B-Lynch, uterine balloon tamponade (UBT), management of pre-eclampsia/eclampsia, shoulder dystocia, placement of foley catheters for cervical ripening, and maternal cardiac arrest. A needs assessment was carried out prior to the first GO MOMS course in Guatemala. This revealed that PPH was a leading cause of morbidity, thus GO MOMS directed a large component of the SBT to addressing PPH management. Through SBT, GO MOMS teaches the techniques of B-Lynch suture and uterine balloon tamponade (UBT), recommended for the treatment of PPH by the WHO and others [15, 16]. For UBT, two techniques were taught using different devices: 1) Bakri balloons and 2) condom-catheter uterine balloon tamponades. Both devices were demonstrated given that access to Bakri balloons is limited in many places and condom-catheter UBTs may be more accessible but require additional steps for assembly.

Study participants

The intervention group consisted of Guatemalan OB/GYN residents and/or attending physicians who had participated in the GO MOMS simulation program at least once prior to 2019 (and who were participating in GO MOMS training again in 2019) (n = 24). The comparison group consisted of Guatemalan OB/GYN residents or attending physicians who had not yet participated in the GO MOMS simulation program (and who were participating in the program for the first time in 2019) (n = 15). The 2019 GO MOMS simulation program was held at a large conference, and individuals could opt in or out of participating in our study after completion of the GO MOMS simulation training session.

Study information was provided to all participants. They were informed that we were conducting a study to help improve simulation training programs and evaluate experiences of participants. Verbal consent was obtained for all participants for both the multiple-choice knowledge survey and qualitative components of the study.

Multiple-choice knowledge survey

A multiple-choice knowledge survey was designed to gather demographic information, practice patterns, prior training, knowledge, and experience with B-Lynch, UBT, and medical abortion (for the management of spontaneous abortion). The data on medical abortion will be published separately.

For the multiple-choice knowledge survey, we utilized convenience sampling by allowing all residents and attending physicians participating in the GO MOMS program in 2019 (some of who had also participated in the GO MOMS program prior to 2019) to participate in the study. Study participants completed the multiple-choice questionnaires via REDCap in September 2019. The Stanford REDCap platform (http://redcap.stanford.edu) is developed and operated by the Stanford Medicine Research IT team. Participants completed the RedCap survey using a laptop computer, iPad, smartphone, or via pen and paper. Study participants who completed the REDCap surveys could “opt-in” to also participate in qualitative interviews.

Qualitative interview

A qualitative interview guide was designed to gather information concering challenges, positive experiences, and attitudes of providers around B-Lynch and UBT. The qualitative interview guide included question prompts and interviewer scripts.

A purposeful sampling strategy was utilized for the qualitative interviews. Our goal was not to generalize to a population with statistical confidence, but to select an information-rich population. This was deemed appropriate for our study as we sought to understand perspectives and in-depth points of view around provider experiences with implementation of B-Lynch and UBT for PPH beyond the multiple-choice knowledge survey alone [17]. We selected OB/GYN residents or attending physicians who had participated in GO MOMS training in the past (at least once prior to 2019) to gain more information about their experiences incorporating knowledge gained from GO MOMS into clinical practice.

All qualitative interviews were conducted using the qualitative interview guide via an interpreter fluent in Spanish with the support of a research assistant who speaks English. Interviews were audiotaped, translated into English, and transcribed by study personnel.

Multiple-choice knowledge survey and qualitative interview

For both the multiple-choice knowledge survey and the qualitative interview guide, all questions were written in English and validated via peer review by U.S. board-certified OB/GYNs for accuracy. The questionnaires were translated into Spanish and reviewed by IGSS faculty. A back translation was performed to ensure that the questions asked in Spanish represented the same content and meaning as those in English.

Data analysis

For the multiple-choice knowledge surveys, Fisher’s exact test was utilized with p < 0.05 considered statistically significant. Analyses were performed using SAS Enterprise Guide 7.1 (SAS Institute Inc., Cary, NC).

For the qualitative interviews, grounded theory methodology, as described by Charmaz [18], was utilized for data analysis. This method of qualitative analysis is constructivist, as it generates concepts and theories from data. Two independent investigators carried out the qualitative analysis by conducting line-by-line coding of interview transcripts. Preliminary themes were then developed by grouping similarly coded phrases. Emergent concepts were then derived by grouping preliminary themes into categories. One investigator was fluent in Spanish and English to ensure that any lack of clarity surrounding the translated content could be addressed. Investigators who participated in coding had prior experience in qualitative methods and conducting OB/GYN-related research.

Results

Multiple-choice survey

A total of 39 physicians participated in our study: 24 individuals in the intervention group and 15 individuals in the comparison group. Table 1 provides demographic information for all study participants who responded to the RedCap multiple-choice knowledge surveys. For the purposes of this study, only residents and attending physicians with at least 1–4 years of hospital practice were included in our analysis (n = 39). This meant that no first year residents were included, as they had less than one year of hospital practice.

Table 1 Demographics and practice patterns among survey respondents (n = 39)
Full size table

Table 2 compares where providers were first taught B-Lynch and UBT based on whether or not they had received previous GO MOMS training. Participants without GO MOMS training were significantly less likely to have ever been taught UBT and B-Lynch (p = 0.0353 and p = 0.0153, respectively). Specifically, participants with previous GO MOMS training were more likely to have been taught B-Lynch and UBT by the GO MOMS course or another resident or attending compared to those who had not received previous GO MOMS training (p = 0.0086 and 0.0068, respectively).

Table 2 Previous training experience with B-Lynch and UBT among survey respondents (n = 39)
Full size table

Table 3 compares provider comfort with B-Lynch and UBT based on whether or not they had received previous GO MOMS training. Participants with previous GO MOMS training reported feeling more comfortable doing B-Lynch with and without supervision and were also more likely to have taught B-Lynch to someone else compared to those who had not received previous GO MOMS training (p = 0.0030 and 0.0049, respectively). However, there was no difference between those with and without previous GO MOMS training in terms of comfort with UBT or teaching UBT to others.

Table 3 Previous teaching experience and comfort with B-Lynch and UBT among survey respondents (n = 39)
Full size table

Qualitative interviews

In total, 11 participants took part in the qualitative interviews. Eleven themes were generated from the qualitative analysis. Themes included hospital practice patterns and use of techniques (UBT and B-Lynch), hospital resource and personnel limitations, decision-making hierarchies, challenging nature of emergencies, and impact of SBT (Table 4).

Table 4 Themes and illustrative quotes pertaining to themesa
Full size table

The qualitative analysis demonstrated that unfamiliarity and time/resource limitations influenced the ability of providers, especially residents and trainees, to implement new procedures (or new ways of doing existing procedures). For example, UBT was a procedure many were familiar with, but due to limited Bakri balloons, lack of supplies on hand for condom-catheter UBTs, and lack of practice with UBT in general, UBT uptake was a significant challenge. Conversely, B-Lynch was more easily implemented given the comfort in the OR and readiness of supplies.

Not everyone knows how to place [UBT], the majority of us have never done it, and second because there aren’t any. In the labor area where postpartum hemorrhage happens, the condoms and everything to do it are not very available. It takes time to get all the materials. So it’s lost time.

Multiple interview respondents noted the importance of attending physicians in decision-making and that decision-making hierarchies impacted procedure implementation. If attending physicians were unfamiliar with a certain procedure, it seemed less likely that residents would be able to implement it.

Attendings in our hospital are not familiar with this suture (B-Lynch). So it’s not something we use because when we find ourselves dealing with an obstetric hemorrhage, usually we call the attending to make a decision together, so when an attending doesn’t have experience doing this type of suture, they don’t feel comfortable doing it with us, who are in training.

Discussion

Our study showed that providers with previous GO MOMS simulation training reported more comfort with B-Lynch for the management of PPH compared to those with no prior training. We also found that providers with previous GO MOMS simulation training reported more comfort with UBT for the management of PPH compared to those with no prior training, although this result was not statistically significant. Qualitative interviews provided insight into the challenges that hinder uptake of UBT with an emphasis on resource limitations and decision-making hierarchies. Utilization of a mixed-methods approach allowed us to evaluate provider attitudes, practice patterns, and patient management for PPH before and after the introduction of SBT in Guatemala. Analysis of multiple-choice knowledge surveys indicated that providers with previous GO MOMS training were significantly more comfortable doing B-Lynch with and without supervision and were also more likely to have taught B-Lynch to someone else compared to those who had not received previous GO MOMS training. There was no difference between groups in terms of comfort with UBT or teaching UBT to others. These findings suggest that SBT is successful for skills-based teaching and can also promote a “ripple effect” of knowledge acquisition and transfer among colleagues, specifically with procedures that are more commonly practiced in the local context, such as B-Lynch.

The qualitative interviews revealed challenges that hinder uptake and implementation of UBT compared to B-Lynch. Several respondents pointed to resource limitations (lack of commercially available uterine balloons), additional time required to assemble supplies for condom-catheter uterine balloon tamponades, and overall hospital culture as factors limiting UBT use. Implementation gaps due to health system barriers have similarly been identified in a mixed-methods study evaluating implementation of PPH guidelines in Uganda [19]. Data from our qualitative interviews also suggested that education of the majority of attending physicians is crucial for implementation and uptake of any new procedures since emergent decisions around patient management are often made at the attending level, and residents feel less comfortable suggesting a procedure that their attendings are unfamiliar with.

Our findings agree with those in the literature that B-Lynch suture and UBT are amenable to simulation training in resource-limited settings [20, 21]. Prior work by us (GO MOMS) [12] demonstrated that SBT was efficacious for use in training in low-resource settings by increasing short and long-term (> 6 month) clinical knowledge and self-efficacy regarding the management of obstetric emergences. In addition, others (PRONTO) have found that SBT can have an impact on local practice patterns, such as reducing the local incidence of cesarean sections [22]. We are able to add to this emerging body of literature by incorporating a mixed-methods approach, examining provider attitudes, local practice patterns, and patient outcomes. This wider view allows a deeper understanding of the challenges that arise when trainees attempt to implement new procedures after participation in a SBT session. Many evaluation studies assessing training programs examine pre- and post-test outcomes alone[13], but without robust evaluation of provider experiences after training, it is impossible to identify the human and structural barriers that may impede the translation of knowledge into implementation.

Potential learning opportunites of SBT that were not analyzed with this study but should be considered for future research in LMICs include the introduction of simulation “drills” to ensure a unifed approach by the entire labor and delivery team during obstetrical emergencies. Checklists, which have emerged as a powerful tool to unify information and task coordination [23] in urgent situations, are best introduced and practiced in a simulated setting. Additionally, the introduction and practice of communciation techniques such as closed loop or call outs are effectively taught in a simulated environment [24].

The strengths of this study are several: first, this study evaluated an educational platform that is easily reproducible in a low-resource setting; second, we incorporated local resources in our didactics and simulation activities which led to a strong collaboration from the local community and medical team; third, this study revealed important barriers to implementation of knowledge. However, this study is not without limitations. With regards to the multiple-choice knowledge survey, participants differed between groups, with more attending physicians in the group that had not previously received GO MOMS training. As a result of this, we are unable to determine if the increased comfort with B-Lynch observed among providers who had received previous GO MOMS training (more resident participants) versus those who did not (more attending physician participants) is attributable to SBT, level of training, or a combination of both. For instance, residents may attend more deliveries than attendings, and therefore may report more comfort with certain techniques. The presence of more residents in the group that had previously received GO MOMS training suggests that the GO MOMS trainings were more targeted towards residents, and residents may have also been more available to attend SBT sessions (compared to attendings). Future recruitment efforts should focus more resources on also involving attendings in SBT, given that they are at the “top” of decision-making hierarchies and their “buy-in” is needed for the successful implementation of new practices. In addition, given that a goal of our SBT program was to provide training that is relevant to providers in their local contexts, our SBT program, and others, can benefit from conducting frequent needs analyses to better understand what resources are (or are not) available among trainees, which would allow for training efforts to be better tailored to fit the local context. For example, our curriculum had already incorporated training around condom-catheter UBTs, given that Bakri balloons are limited in LMICs, and findings from this study suggest that future training efforts may benefit from focusing even more on condom-catheter UBTs. For the qualitative analysis, one possible limitation is the maintenance of objectivity; we attempted to minimize this by involving multiple investigators who participated in various levels of data coding and analysis.

Conclusions

This study utilized a mixed-methods design to better understand the impact of simulation-based training at several levels that would not adequately be captured with one type of evauation alone. Capturing data across several sources allowed us to gain insight into the mechanisms hindering uptake of procedures taught in simulation, allowing for future adjustments in program development. In our experience, there were also sociocultural factors that enhanced and deepened our results, which would not have been captured through one modality alone. Although this study is specific to procedures used to manage the obstetric emergency of PPH, our findings can be used by the global simulation learning community as a whole. Study designs or program evaluations that utilize a “one-pronged” approach may not be feasible in certain contexts based on numeous barriers that challenge data collection. A mixed-methods study could serve as a useful framework for others who are planning to carry out program evaluations in LMICs.