Applying the RE-AIM framework in a process evaluation of the introduction of the Non-Pneumatic Anti-Shock Garment in a rural district of Zimbabwe

The Non-Pneumatic Anti-Shock Garment (NASG) is a first aid tool that can halt and reverse hypovolemic shock secondary to obstetric hemorrhage. The World Health Organization recommended the NASG for use as a temporizing measure in 2012, but uptake of the recommendation has been slow, partially because operational experience is limited. The study is a process evaluation of the introduction of NASG in a public sector health facility network in rural Zimbabwe utilizing an adapted RE-AIM, categorizing observations into the domains of: reach, effectiveness, adoption, implementation and maintenance. The location of the study was Hurungwe district, where staff members of 34 health facilities at primary (31), secondary (2) and tertiary (1) levels of care participated. We found that all facilities became skilled in using the NASG, and that the NASG was used in 10 of 11 instances of severe hemorrhage. In the cases of hypovolemic shock where the NASG was used, there were no maternal deaths and no extreme adverse outcomes related to obstetric hemorrhage in the study period. Among the 10 NASG uses, the garment was used correctly in each case. Fidelity to processes was high, especially in regard to training and cascading skills, but revisions of the NASG rotation and replacement operating procedures were required to keep clean garments stocked. Clinical documentation was also a key challenge. NASG introduction dovetailed very well with pre-existing systems for obstetric emergency response, and improved clinical outcomes. Scale-up of the NASG in the Zimbabwean public health system can be undertaken with careful attention to mentorship, drills, documentation and logistics. The Institutional Review Board of The Medical Research Council of Zimbabwe approved this study in December 2018. The approval number is: MRCZ/A/2400. Written informed consent was obtained from participating health workers and oral informed consent was obtained from participating patients. This decision was made in collaboration with the Medical Research Council of Zimbabwe. Two key points considered were: 1) Hypovolemic Shock from obstetric hemorrhage is a medical emergency that requires fast action to save lives. In that context, obtaining written consent seemed like a potential risk to safety of patients; 2) the Medical Research Council of Zimbabwe had prior positive experience with the NASG randomized control trial in Zimbabwe, where verbal consent using a very similar script was obtained from enrolled patients. This was an observational study to evaluate the process of incorporating the NASG into the obstetric emergency response within the Zimbabwe public health system using an explicit implementation strategy. Study sites were facilities within a discrete geographical area of Zimbabwe, comprising the public health system of Hurungwe Province and its referral hospital in the adjacent district. The study was conducted prior to national introduction of NASG, with the intent of providing operational guidance for future policy and program planning. Data collection was aligned to the RE-AIM framework, looking at the process of the intervention in terms of its reach, effectiveness, adoption, implementation and maintenance [19]. Originally presented in 1999, the RE-AIM framework organizes evaluation into five domains, which, examined together, estimate an intervention’s public health impact [20]. While in research settings NASG has been shown to be of great clinical value, its uptake has been minimal, partially due to lack of evidence that it could work in a real-world environment, outside the controlled, well-resourced conditions of a clinical trial. Assessment utilizing the RE-AIM framework can aid the translation from clinical evidence to widespread implementation by virtue of including domains outside of efficacy and cost-effectiveness. The duration of this study was brief; although mortality was recorded, the process evaluation was not designed to capture statistically significant changes in maternal mortality, which is a relatively rare event requiring large populations and longer durations to assess. Additionally, the study duration was not long enough to utilize the framework to estimate public health impact reliably. However, interpretation through the RE-AIM framework does provide an organized, realistic evaluation of a spectrum of aspects of the process of NASG introduction to a rural public health system in a lower-income country. These organized observations can provide valuable guidance for practical action to take the NASG to national scale. The original RE-AIM domains and how they were adapted for use in this study are outlined in Table 1 below. In consultation with government stakeholders, Hurungwe District of Mashonaland West Province in north central Zimbabwe, bordering Zambia, was selected for this process evaluation. Reasons for this choice were multi-factorial, including non-urban or peri-urban setting, relative lack of implementation partners, burden of maternal morbidity and mortality, and feasibility of access by road. Hurungwe, with a population of 358,000, is a predominantly rural district in the province; 92% of people reside in rural settings where in general, only primary care is accessed. 13,674 births were reported in Hurungwe district in 2018. Of the 13,476 live births reported, 93.3% were institutional, the largest proportion of which (41.4%) were conducted at primary care facilities; 34.1% of facility births occurred at district and mission hospitals, 11.3% at rural health centers, and 13.1% at rural hospitals. Staff from the provincial health administration, Hurungwe District health management staff and those providing obstetric care at the primary, secondary and tertiary levels participated in the process evaluation. NASG was introduced for use in 34 health facilities: 1 tertiary level, 2 secondary, and 31 primary level facilities. The study area is depicted in the map displayed in Fig 1. Reprinted under a CC BY license, with permission from Brighton Gambinga, original copyright 2020. Over the course of three months, NASG was introduced into Hurungwe District through didactic and practical training in a ‘training of trainers’ style, where NASG trainers from the study team trained selected health workers from the participating facilities, and those trainees were expected to cascade NASG knowledge to co-workers at their facilities. The training demonstrated the use of the NASG, the theory behind it, and the necessary steps to introduce it, including how to fill in the study-specific OH case and transfer form, and data and logistics registers. Deployment of NASGs and data and logistics tools was done at the time of the training. For the remainder of the study period, participants were followed up at their facilities. Further, mentorship was delivered through a pre-existing provincial-district level supportive supervisory program. During those visits, NASG skills were observed in health workers through simulated obstetric hemorrhage response, data collection tools were reviewed, mentorship was provided, and informal qualitative feedback from health workers who were trained in NASG. Answers and comments from the health workers were recorded on supervisory visit forms. Non-medical janitorial staff were also selected to receive training in cleaning and storage of the garments; their role in successful incorporation of the NASG into obstetric emergency response was critical, as the logistics of cleaning, rotating, and tracking garments was paramount to its success in a clinical network. Standard operating procedures for emergency transport and NASG logistics were developed by District staff, supported by the research team. Availability of ambulances, informal arrangements for emergency transport, contact phone numbers and suggested troubleshooting were assessed and mapped in the referral tree of District facilities. All data collection was paper-based and done continuously throughout the three-month period, from 7 January 2019 to 31 March 2019. To capture as much detail about how the district system responded to OH after NASG introduction, each case of OH presenting to participating facilities was documented, regardless of whether hypovolemic shock developed from the OH, or whether the NASG was used as part of treatment. In addition to clinical details, communication and transportation details were documented: which facility was called at what time, and how long it took for transport to arrive once called. Other data was collected through NASG logs and clinical registers, and through documentation of supportive supervision that assessed retention and feedback of NASG skills among facility workers. Supportive supervisory visits were conducted by two District clinical supervisors, who were accompanied by two midwives on the study team who are NASG trainers. The pairings were not always the same, and the four debriefed on the findings of each supervisory visit, in order to minimize interobserver variation. In addition to assessing NASG skills, District supervisors also collected informal qualitative data, asking open-ended questions during visits, and recording some comments health workers made about the experience of learning how to use NASG and incorporating it into their pre-existing obstetric emergency response protocols. This einfomral qualitative data was collected from health workers who were trained, regardless of whether or not they had used NASG clinically yet. Clinical registers were augmented by study-specific forms to document clinical details of obstetric hemorrhage cases, such as estimated blood loss, signs of hypovolemic shock, treatments provided, if NASG was applied, if applied when and where removed, and clinical outcomes, such as shock recovery. Incremental costing was done through monitoring of expenses related to NASG introduction, including district sensitization, NASG and sundries procurements, service provider training, as well as post training follow up and supportive supervision visits.