Tulinde Vizazi Vijavyo

In 2017, approximately 295,000 preventable deaths related to childbirth occurred throughout the world. Low- and middle-income countries disproportionately carry the burden of maternal mortality, with 99% of deaths related to childbirth occurring in these countries annually. In response, United Nations Sustainable Development Goal Target 3.1 aims to reduce the global maternal mortality ratio to less than 70 deaths per 100,000 live births by 2030. 

In 2017, approximately 3 out of every 1,000 mothers died from childbirth-related causes in Kenya, a ratio almost 10-fold higher than that in the United States. Maternal morbidity and mortality can be reduced through early identification and management of pregnancy related complications or pre-existing complications. High quality antenatal care has shown to improve women’s health seeking behavior, promoting access to and use of skilled care at birth. 

As part of high-quality antenatal care, the World Health Organization recommends one obstetrical ultrasound prior to 24 weeks gestation. Though limited data exists to determine the proportion of pregnant people in Kenya receiving this recommended imaging, a study in Kenya’s Migori county found that only 16% of pregnant people underwent ultrasound imaging at any point in their pregnancy. Moreover, another study showed the median availability of ultrasound was 0-15% across all healthcare facilities in Kenya, with greater access available in urban areas relative to remote areas of the country. Access is likely even lower in fragile contexts like the northern Rift Valley of Kenya, the site of implementation for this solution. 

Point of Care Ultrasound (POCUS) is the use of portable ultrasound technology by trained medical professionals to answer specific clinical questions. It differs from traditional radiology in that ultrasound is used to answer specific clinical questions. In obstetrics, POCUS is used to determine multifetal gestation, fetal lie and position, fetal heart rate, placental location, and amniotic fluid depth. POCUS can thus aid in identifying potentially high-risk pregnancies early allowing time for adequate provision of care prior to and during delivery. Thus, application of obstetrical point of care ultrasound (OB POCUS) in resource limited settings, especially when paired with the capability to refer to higher levels of care, has the potential to improve maternal morbidity and mortality. 

In Kenya, task-shifting models whereby healthcare workers such as nurses, midwives, clinical officers and medical officers are trained in POCUS have been used to expand access to ultrasound. Prior literature suggests that longitudinal mentorship following intensive POCUS training is a significant barrier to continued POCUS use in clinical settings. While POCUS shows promise to address a health need and improve maternal morbidity and mortality, ensuring accuracy of image capture and interpretation and promoting long-term use of POCUS continue to be challenging.

Tulinde Vizazi Vijavyo (Let’s Protect Future Generations) is an innovative, technology based solution that seeks to improve antenatal care in the isolated and fragile context of the northern Rift Valley of Kenya and to decrease maternal morbidity and mortality in this region. It is a POCUS education model that expands evidence-based task-shifting strategies to promote the use of POCUS among healthcare workers in order to improve access to ultrasound for the general population. POCUS education is provided through in-person, hands-on training sessions. Mobile health and cloud based technology facilitate longitudinal POCUS mentorship by allowing trained health care workers to communicate and review de-identified images with POCUS educators and mentors when needed. 

Tulinde Vizazi Vijavyo will specifically apply a task shifting strategy to train and promote use of high-quality obstetrical POCUS (OB POCUS) among healthcare workers including medical officers, clinical officers, nurses and midwives in the northern Rift Valley of Kenya. It will continue to develop and refine mobile health and cloud based technology to facilitate longitudinal OB POCUS mentorship. Finally, Tulinde Vizazi Vijavyo will integrate artificial intelligence/machine learning (AI/ML) models for OB POCUS in this region, providing real-time clinical support for healthcare workers to promote early identification of certain high-risk obstetrical conditions. 

AI/ML models have been operationalized for other areas of POCUS, but the application of these clinical support models are limited for OB POCUS. AI/ML modeling is therefore an innovative addition to an evidence-based approach to POCUS education that has been successfully used in this region for the last decade. Tulinde Vizazi Vijavyo applies a technology driven approach to OB POCUS education designed for the unique challenges in the fragile context of the northern Rift Valley of Kenya, where poor physical infrastructure, limited healthcare networks, low population density and ongoing conflict makes training for healthcare workers challenging and access to diagnostic imaging for pregnant people limited.

Ultimately, this solution will equip healthcare workers in the northern Rift Valley of Kenya to use OB POCUS to identify potentially high risk pregnancy conditions. Patients can then be referred to higher levels of care earlier in pregnancy, reducing their risk during delivery. Thus, Tulinde Vizazi Vijavyo may ultimately decrease maternal morbidity and mortality in this fragile region of Kenya. 

The northern region of Kenya comprises approximately 60% of the geographic land mass of Kenya, but only 18% of Kenya’s approximately 53 million people. This expansive region has historically been politically and economically isolated from the rest of the country, exacerbated by limited infrastructure in the region and ongoing conflict throughout it. As of 2021, less than 50% of people in the region were within a one-hour travel time of any healthcare facility – public, private or combined. In three counties of northern Kenya, the average travel time to the nearest public health facility was greater than three hours. Referral systems in these regions are also weak due to poor physical infrastructure, particularly unsafe roadways, and limited ability for people to finance higher level medical care.  

Tulinde Vizazi Vijavyo focuses on the counties of West Pokot, Elgeyo- Marakwet, and Trans Nzoia in northern Kenya -- the northern Rift Valley of Kenya. Based on the 2019 Kenya Population and Housing Census, West Pokot has a population of 621,241 people; Elgeyo-Marakwet has a population of 454,480 people and Trans Nzoia has a population of 990,341 people. The gross domestic product per capita in these counties range from 1,810USD to 2,582USD. Collectively, these counties have a population of just under one million people of childbearing age. West Pokot has the fifth highest fertility rate in Kenya with approximately 5.6 births per 1,000 women. Maternal mortality in the fragile context of the northern Rift Valley of Kenya remains high with an estimated maternal mortality ratio of 70 deaths per 100,000 live births in West Pokot, 230 deaths per 100,000 live births in Elgeyo-Marakwet and 137 deaths per 100,000 live births in Trans Nzoia.

Tulinde Vizazi Vijavyo, through a collaborative educational model for POCUS education, would enhance access to OB POCUS for pregnant people in the northern Rift Valley of Kenya. As a result, high risk pregnancy conditions may be identified early allowing for coordination of the appropriate level of care for patients at the time of delivery. While there is a paucity of data specifically related to uptake and access to antenatal care and obstetrical ultrasound in this region, it is likely that limited access to these services contribute to high maternal mortality ratios in this area. Tulinde Vizazi Vijavyo directly addresses this limitation and thus could significantly decrease maternal morbidity and mortality in this fragile context of Kenya.

Tulinde Vizazi Vijavyo is built on the collaboration of Moi University College of Health Sciences (MUCHS), Moi Teaching and Referral Hospital (MTRH), the Global Ultrasound Institute (GUSI) and AMPATH (Academic Model Providing Access to Healthcare).   

Moi University College of Health Sciences (MUCHS) and Moi Teaching and Referral Hospital (MTRH), located in Eldoret, Kenya serve as one of the largest academic medical  centers in Kenya with work ranging from HIV/AIDS to chronic disease management to maternal and child health. 

The Global Ultrasound Insitute (GUSI) is a community of expert POCUS educators from a variety of clinical backgrounds with a mission to advance POCUS training and education by equipping medical professionals throughout the world with comprehensive POCUS skills using evidence-based educational models. 

The AMPATH global network builds cross-cultural partnerships between universities and academic health centers to strengthen health systems and tackle health disparities, train future global health leaders and foster healthcare innovations to improve health worldwide. AMPATH Kenya is a partnership between MUCHS, MTRH and the AMPATH Consortium, a global network of universities and academic health centers led by Indiana University. It partners with the Kenyan government to support care for a population of more than 24 million people throughout Kenya and leads foundational training and research programs. Community-based outreach activities of AMPATH Kenya include livelihood enhancement interventions directly targeting food and income insecurity. 

MUCHS has partnered with GUSI to promote POCUS training among healthcare workers throughout Kenya. Since 2017, this collaboration has successfully provided POCUS training to over 90% of family physicians in Kenya and POCUS training continues in family medicine training programs at MUCHS. GUSI has also provided training in OB POCUS for over 500 healthcare workers in Kenya. GUSI recently began developing AI/ML models as a clinical and educational support tool to improve POCUS image interpretation. 

With MUCHS as the leader, Tulinde Vizazi Vijavyo will expand a collaborative POCUS education model that has successfully trained healthcare workers throughout Kenya. Working within the AMPATH care networks in the northern Rift Valley of Kenya, it  will draw on existing community partnerships and mobilize care networks to train healthcare workers in OB POCUS in this fragile region. Furthermore, Tulinde Vizazi Vijavyo will draw on mobile health and cloud based technology for longitudinal support in POCUS education--technology that has been successfully used by GUSI and MUCHS for POCUS education in Kenya. Tulinde Vizazi Vijavyo will also further refine artificial intelligence/machine learning (AI/ML) models for OB POCUS currently being developed by GUSI to provide real time clinical support in remote and fragile regions of Kenya.

Finally, Tulinde Vizazi Vijavyo will leverage the expansive care network of AMPATH Kenya to develop clinical protocols that will guide management and referral of patients with high risk obstetrical conditions identified through OB POCUS. Identifying high risk pregnancy complications early and connecting patients to appropriate level of care for delivery can reduce maternal morbidity and mortality in this region.

The collaborative POCUS education model used for training by MUCHS and GUSI has been responsible for training over 90% of family physicians in Kenya in POCUS and continues to be used to train family medicine registrars at MUCHS in POCUS with more than 75 physicians trained through this program. As a result, these trainees are able to bring POCUS to their clinical work throughout the country. More recently, GUSI was involved in the training of over 500 healthcare workers in Kenya on OB POCUS.

AMPATH supports public sector health service delivery in much of western Kenya where MTRH serves as the referral facility for a catchment population of more than 24 million people. AMPATH care networks currently exist in the northern regions of West Pokot, Elgeyo-Marakwet and Trans Nzoia and beyond. Based on the 2019 Kenya Population and Housing Census, West Pokot has a population of 621,241 people; Elgeyo-Marakwet has a population of 454,480 people and Trans Nzoia has a population of 990,341 people. Based on the same dataset, approximately 24,511 births occurred in West Pokot in 2019; 13,212 births occurred in Elgeyo-Marakwet in 2019 and 29,005 births occurred in Trans Nzoia in 2019. Currently, AMPATH supports care for 25,000 people living with HIV across these three counties and for 120,000 people across six counties in western Kenya. The activities of AMPATH in this region extend beyond HIV care and include the care of thousands of people with non-communicable diseases, which include pregnant people, infants and children. In addition, the activities of AMPATH extend to the community through livelihood enhancement interventions directly targeting food and income insecurity.

Tulinde Vizazi Vijavyo proposes to expand an educational model for POCUS to improve access to high quality OB POCUS to the fragile region of the northern Rift Valley of Kenya with the goal of identifying high risk conditions of pregnancy early and connecting patients to the appropriate level of care. This solution depends on the presence of care networks that include care settings able to manage high risk pregnancy conditions, the provision of ultrasound equipment, an established OB POCUS training curriculum and POCUS educators, mobile and cloud technology based technology to provide longitudinal mentorship for trained healthcare workers, and the development of artificial intelligence/machine learning (AI/ML) clinical support models for real-time use in OB POCUS. 

The existing collaboration between MUCHS, GUSI, and AMPATH provides the care networks throughout Kenya that include the tertiary care setting of MTRH. Moreover, training curriculum for OB POCUS has been developed and deployed through GUSI in many regions in Kenya with demonstrated success. To reach the most remote and fragile regions in Kenya with this imaging modality, we request additional support in the following:  

1. Funding support to run previously successful and tested hands-on training workshops and longitudinal support technological infrastructure for healthcare workers in the northern Rift Valley of Kenya. 

2. Provision of ultrasound equipment/devices for healthcare facilities in these regions through industry partnerships and collaboration. 

3. Enhancement of mobile health and cloud based technology for longitudinal support for healthcare workers trained in OB POCUS designed for use in geographically isolated, fragile contexts like the northern Rift Valley of Kenya.  

4. Technical support in the further development of AI/ML clinical support models for real-time OB POCUS image acquisition and interpretation that is appropriate for this particular clinical context. Specifically, there is a need for a diverse library of de-identified obstetrical ultrasound images that consulting radiologists can label and categorize to train the AI/ML model specifically for deployment in this region of Kenya.

In recent years, point of care ultrasound (POCUS) has gained popularity in a variety of clinical settings in both high income and low and middle income settings as a means to expand access to ultrasound imaging and improve clinical evaluation of patients. The majority of efforts thus far have focused on portable equipment distribution, widespread POCUS training for healthcare workers, and development of POCUS curriculum for medical education. 

Literature suggests that longitudinal mentorship following intensive POCUS training is a significant barrier to continued POCUS use in clinical settings. Moreover, such mentorship is critical to ensuring quality of images captured by healthcare workers and accurate interpretation of those images. This barrier is exacerbated in geographically, politically, and economically isolated areas. Tulinde Vizazi Vijavyo expands our POCUS education model and directly addresses these important barriers for obstetrical point of care ultrasound (OB POCUS) by not only including a clinical and educational support network for trained healthcare workers built on mobile health and cloud based technology, but also through the development of artificial intelligence/machine learning (AI/ML) real-time clinical support models for OB POCUS to promote accurate image acquisition and interpretation. 

In regions of the world that are geographically isolated and burdened by ongoing conflict, provision of medical care and continuing education of healthcare workers is limited. The continued refinement of mobile health and cloud-based technology coupled with the development of AI/ML models for real time clinical support is of particular importance in isolated regions in Kenya, like the northern Rift Valley, given the limited ability for OB POCUS educators to be on-site in these regions. 

While these technology based clinical support tools are being developed primarily with the goal of bringing OB POCUS to the isolated regions of Kenya, they have the potential to be scaled up for use in other isolated regions throughout the world, including rural regions in the United States. In addition, these technology based clinical support tools, when incorporated into OB POCUS training curricula for healthcare workers, may enhance training and skill development for learners thus promoting continued use of OB POCUS in clinical settings following training, further expanding access to this imaging modality throughout the world.

One Year Impact Goals: 

1. Train and provide longitudinal mentorship in OB POCUS for at least 25 healthcare workers in the northern Rift Valley of Kenya: This will be achieved through the proposed solution (Tulinde Vizazi Vijavyo).

2. Further develop and implement AI/ML models for OB POCUS to provide real-time clinical support for trained healthcare workers in the northern Rift Valley of Kenya: This will be achieved in collaboration with industry partners who may provide technical support.

3. Develop new protocols for patient management and referral incorporating OB POCUS: Family Medicine team members from Moi University College of Health Sciences (MUCHS), Moi Teaching and Referral Hospital (MTRH) and AMPATH (Academic Model Providing Access to Healthcare) have expertise in protocol development and will be able to create these protocols with input from healthcare workers trained in OB POCUS. 

Five Year Impact Goals:

1. Expand training program across AMPATH to train > 200 healthcare workers in OB POCUS: AMPATH has robust clinical training infrastructure and expansive healthcare networks throughout Kenya, so we will be able to replicate this model in other regions easily.

2. Incorporate AI/ML models for OB POCUS for real-time clinical support of trained healthcare workers across AMPATH: We will engage our radiology partners at MUCHS, MTRH and AMPATH to operationalize this.

3. Share training and AI/ML models for OB POCUS with the Kenyan Ministry of Health to promote replication and expansion: MUCHS, MTRH and AMPATH have strong relationships with the Kenyan Ministry of Health and county governments, through which robust collaboration occurs across all AMPATH sites. We will use these relationships to disseminate our progress and discuss replication.

Evaluation of Tulinde Vizazi Vijavyo will incorporate quantitative evaluation using the Moore’s framework for medical educational activities as a guide. First, the number of healthcare workers trained through Tulinde Vizazi Vijavyo will be measured. Second, we will administer structured questionnaires to participants immediately after OB POCUS training and six months following the training to assess self-reported confidence in conducting OB POCUS. Third, we will measure clinical impact by tracking the number of OB POCUS scans performed by trained healthcare workers. Finally, we will evaluate healthcare workers’ OB POCUS accuracy by comparing to AI/ML models as well as feedback from mentors via refined mobile health and cloud-based support network.

The long term outcome goal of Tulinde Vizazi Vijavyo is to decrease maternal morbidity and mortality in the northern Rift Valley, a fragile region within the country of Kenya. The theory of change underlying Tulinde Vizazi Vijavyo depends on five key determinants:

1.  Portable ultrasound technology availability and provision for facilities in this region

2. An established OB POCUS training curriculum and OB POCUS educators provided by the Global Ultrasound Institute (GUSI) and Moi University College of Health Sciences (MUCHS) which will inform initial hands-on training of healthcare workers (HCWs) in this region  

3. The integration of mobile health and cloud-based technology to provide remote longitudinal clinical and educational mentorship in OB POCUS for trained healthcare workers (HCWs). 

4. The development of artificial intelligence/machine learning (AI/ML) models for OB POCUS to provide real-time clinical support for healthcare workers 

5. The development of new protocols for patient management and referral to higher levels of care. 

Please refer to the attached logic model for Tulinde Vizazi Vijavyo where activities, outputs and short, intermediate and long term outcomes are further defined and underlying theory of change is described. 

Portable ultrasound technology, mobile health and cloud based technology, and artificial intelligence/machine learning (AI/ML) models power the innovative solution that is Tulinde Vizazi Vijavyo. 

1. Portable ultrasound technology has previously been used successfully in Kenya given its portable nature and ease of use. Tulinde Vizazi Vijavyo will also use portable ultrasound technology. 

2. Mobile health and cloud based technology has been previously used for longitudinal mentorship of healthcare workers trained in POCUS. Tulinde Vizazi Vijavyo aims to further refine this approach to facilitate longitudinal support in isolated regions where POCUS educators cannot always be physically present. This POCUS educational model has relied on cloud based technology whereby de-identified scans can be uploaded for feedback that can be provided remotely by POCUS educators. In addition, use of encrypted, mobile text messaging software allows for ongoing peer-to–peer discussion about cases that can be facilitated by POCUS educators remotely. There is great potential to expand these technology based approaches to longitudinal mentorship with additional industry support and innovation, particularly for remote contexts that may have limited connectivity. 

3. Artificial intelligence/machine learning (AI/ML) models will further power Tulinde Vizazi Vijavyo. While both mobile health and cloud based technology provides longitudinal support for healthcare workers trained in POCUS, robust support tools are needed in isolated regions because challenges in access limits the ability for educators to be on site to provide support. In addition, mobile health and cloud based technology, while important for continued learning, cannot guarantee real-time clinical support. Artificial intelligence/machine learning models can fill this gap. This technology will minimally entail the training of 3 ML models analyzing single-frame sonographic images – a) detection of multiple gestation (classification), b) non-cephalic presentation (classification) and c) gestational age (regression). It is expected that all can be implemented using  standard 2D-CNN foundational models available with the PyTorch framework and using off-the-shelf tools for image annotation.  Provided acceptable accuracy can be achieved, it is highly desirable that a small footprint model be developed where inference can be performed on high-end smartphones with GPU assist, since internet connectivity to cloud-based services may not always be available in these settings. 

Global Ultrasound Institute (GUSI) is a for- profit organization of expert POCUS educators whose founders have collaborated with Moi University College of Health Sciences (MUCHS) and Moi Teaching and Referral Hospital (MTRH) for close to a decade. AMPATH Kenya (Academic Model Providing Access to Healthcare) is a partnership between MUCHS, MTRH and the AMPATH Consortium, a global network of universities and academic health centers led by Indiana University. It partners with the Kenyan government to support care for a population of more than 24 million people throughout Kenya and leads foundational training and research programs.

Tulinde Vizazi Vijavyo aims to increase access to high quality, recommended obstetrical ultrasound to a marginalized population in the northern Rift Valley. The Pokot and Turkana people in this region have been in conflict since pre-colonial times. Colonialism introduced criminalization of Pokot-Turkana cattle raiding which further complicated the conflict. As in other conflict zones, access the healthcare is negatively impacted - fewer specialists such as radiologists and high-risk obstetrics practitioners will choose to practice in these areas and pregnant people are less likely to travel long distances to access diagnostics such as ultrasound. These regions have high fertility rates and similarly high maternal morbidity and mortality, the result of persistent inequities in healthcare access. Tulinde Vizazi Vijavyo democratizes OB POCUS skills and tools to healthcare workers in this fragile region to promote equitable access to high-quality recommended obstetrical care. Tulinde Vizazi Vijavyo will train local healthcare workers in these regions, diversifying the healthcare workforce who have OB POCUS skills, including individuals who are often unable to access trainings because they reside in an isolated region with ongoing conflict.