MotoMeds; Telemedicine and Medication Delivery Service

Our innovation addresses the problem of limited access to acute care for common childhood illnesses, including acute respiratory infection, diarrheal disease, and malaria. These three diseases are leading causes of pediatric death for children between 1 month to 5 years of age globally1. 

Early Access. Well-established low-cost treatments exist for acute respiratory infections, diarrheal disease, and malaria. However, these treatments are most effective when administered early, which is difficult when healthcare access is limited, especially at night. Delayed treatment can result in rapid progression from a pre-emergent to emergent state in children. In resource limited settings, emergency care is considerably more difficult to access and more expensive than pre-emergency care.  

Barriers. What are some of the barriers families face to access care? Our team conducted an NIH funded needs assessment in semi rural Haiti and identified accessibility, affordability, nighttime, infrastructure and distance as the key barriers to care. 

Nighttime. Families are most isolated from knowledge and care at nighttime. In many parts of the world once clinics have closed for the day the only options are to seek unqualified care (eg. medicine vendors, unlicensed pharmacists) or wait until the morning. However many pediatric conditions can quickly transition from pre-emergency to emergency without early access to care. 

1 United Nations Interagency Group for Child Mortality Estimation, Levels and  Trends  in Child Mortality, New York: United Nations Children’s Fund, 2020.

Our solution is the MotoMeds pediatric telemedicine and delivery service (TMDS) that is powered by scalable clinical-decision support tools and software for case management (currently in development) and dispatch. 

Workflow. When a child is experiencing an acute illness at nighttime, the parent places a call to the MotoMeds call center and is connected to a provider. The provider uses clinical decision-support tools that our team adapted from in-person WHO guidelines to first triage the child for danger signs (emergent patients are routed to the hospital). The provider then gathers basic exam findings and medical history from the parent over the phone to generate an assessment and treatment plan. The plan may include oral medications, oral fluids and/or clinic follow-up. If the child lives within a delivery zone, a MotoMeds driver is dispatched to transport items to the child’s home, otherwise the family receives advice only.  We request a flat fee of about $5US for delivery and accept any amount on a sliding scale down to 0.

Process. The core elements of MotoMeds are being designed, built, deployed and evaluated through a series of studies called the Improving Nighttime Access to Care and Treatment (INACT1-4) studies. The approach is being tested in Haiti but the model is intended to be country agnostic. INACT1 was a 2018 healthcare needs assessment, and revealed poverty, distance, and nighttime hours were barriers to care. In 2019 we responded with INACT2, a pre-pilot of the MotoMeds service. The study established clinical safety and feasibility in implementation of the concept. The ongoing INACT3 study, launched in 2020, has laid the groundwork to scale MotoMeds to transition from a single call-center hub with one delivery zone to two geographically distant delivery zones. INACT4 (pending 2022) will evaluate a digital clinical decision support tool that we expect will improve efficiencies of the workflow and allow MotoMeds to scale. We will simultaneously evaluate the portability of the solution in a different setting and conduct the INACT1-2 studies in Ghana. 

MotoMeds primarily helps families isolated by poverty, transportation, infrastructure, and other geographic factors who are experiencing pediatric illnesses at night. This intervention is intended to be a high-quality, low-cost solution that will avert progression to emergencies. Secondary benefits are financial and time savings. This population, families with sick children at nighttime, is underserved in that healthcare options for non emergencies are almost non existent at night. The service is available to all families with children 10 years and younger and those living within a delivery zone are eligible to receive medication/fluid delivery regardless of ability to pay. Additionally, the virtual and anonymous nature of telemedicine has the potential to bring marginalized families into the formal healthcare network.

Secondary beneficiaries are healthcare funding agencies like ministries of health or NGOs. We modeled the cost effectiveness of a TMDS against hospital level care (the only alternative in the study area) and found the TMDS to be highly cost-effective, and DALYs averted by the TMDS and hospital level care were 199.76 and 22.37, respectively. The corresponding incremental cost-effectiveness ratio is estimated at $1,288 signifying the service costs an additional $1,288 to avert one additional DALY. This value is under 23 the per capita GDP of Haiti, $1,177 in 2020,25 therefore, the telemedicine and medication delivery service is considered cost effective by WHO standards. In many low resource settings healthcare, including emergency services, is subsidized by the government or NGOs. These agencies realize cost savings with the reduction of emergency care expenditures. 

The pilot of MotoMeds was in Haiti, where UF had an existing collaboration with the Haitian Ministry of Health and Université d'État d'Haiti- Faculté de Médecine et de Pharmacie (State College of Medicine) to operate a public health research laboratory. Our team's initial engagement in Haiti was through a stakeholder meeting with members of the local healthcare community. The results were encouraging so we formed a team composed of a Haitian nurse and Haitian enumerator to conduct a formal needs assessment to evaluate the healthcare seeking behaviors and existing healthcare provider network in a study area encompassing semi-urban and rural areas. We went to great lengths to ensure we captured a representative sampling of households, not just those that were easy to access. In total 568 household and 65 provider surveys were completed, analyzed and results published in BMJ Open. The data revealed families did in fact face barriers to seeking care from their intended providers and distance, nighttime, and financial barriers were key obstacles. We then held a workshop with local motorcycle taxi drivers as delivery is a key component to a functional TMDS.  The workshop focused on the feasibility of nighttime delivery in a setting without an address system, well paved roads or street lights and understanding safety of operating at night in a country experiencing volatility. This workshop led to a formal partnership with two Haitian physicians to provide local oversight and culturally relevant clinical guidance. We then recruited TMDS staff, conducted trainings and launched the MotoMeds pilot INACT2 study. Throughout the INACT2 and continuing on to the INACT3 study we collect qualitative feedback from users at a 10 day follow up call and we have regular staff meetings to gather their input and ideas. At multiple points we have modified clinical guidelines and workflows based on advice from our local partners. For example we have a relationship with a Haitian nursing school whose director expressed concern about whether families could properly execute the treatment plans generated by call center providers. We valued her input and turned her concern into the basis of a research study to formally evaluate treatment plan adherence. Preliminary data have already led to the implementation of a number of quality improvement measures.

We are about to repeat this entire process in Ghana where we have an existing relationship with the Ghana National Ambulance Service based on education, clinical training and collaborative research. Strong partnerships and consistent  with staff and MotoMeds users will remain the foundation of future work in additional locations.  

We are approaching the 5 year mark of developing our solution. To date our work has been housed in academia with a focus on establishing proof of concept through research studies and funding through grants and donations. We will soon reach the point where our approach will need to transition out of academia and  into a more business minded. Our team has limited experience and knowledge with developing business models to allow us to transition out of this space. developing partnerships outside of this space.  We are advisors to help us transition our solution outside. One particular obstacle we face is funding models

The design, implementation and evaluation of MotoMeds by our team is unique compared to conventional health care delivery methods. Utilization of telemedicine and the spectrum of telemedicine applications have both increased considerably since the onset of the COVID-19 pandemic1,2. However, provision of information without resources is of limited value. Coupling telemedicine with medication delivery closes the service loop towards reducing healthcare access barriers. Use cases of telemedicine and medication delivery are generally limited to high income countries3, chronic illnesses4 and daytime services5. Prioritizing nighttime pre-emergency care is also unique in that nighttime care is almost exclusively relegated to emergency level care. Additionally the idea of switching the directionality of patients traveling to centralized healthcare resources to services traveling to patients is unique. To our knowledge MotoMeds is the only service, framed within an academic research effort, providing nighttime pediatric care through both telemedicine and delivery in a resource limited setting. 

We expect to see an increase in telemedicine services outside of high income countries. We also expect increasing interest in the need for high quality pre-emergency after hours care. Ideally those working in this space will recognize the extensive amount of research that has gone into building the MotoMeds concept and collaborate with our team.  

1Temesgen ZM, DeSimone DC, Mahmood M, Libertin CR, Varatharaj Palraj BR, Berbari EF, 2020. Health Care After the COVID-19 Pandemic and the Influence of Telemedicine. Mayo Clinic Proceedings 95: S66-S68.

2Smith AC, Thomas E, Snoswell CL, Haydon H, Mehrotra A, Clemensen J, Caffery LJ, 2020. Telehealth for global emergencies: Implications for coronavirus disease 2019 (COVID-19). Journal of Telemedicine and Telecare 26: 309-313.

3NowRX Telehealth, Telehealth the way it should be. Available at: https://nowrx.com/telehealth/. Accessed December 15, 2021.

4National Health Security Office, Postal medicine delivery plays key role in telemedicine. Available at: http://eng.nhso.go.th/view/1/home/Postal-medicine-delivery-plays-key-role-in-telemedicine/230/EN-US. Accessed December 14, 2021.

5Rockethealth Telemedicine Laboratory Pharmacy,. Available at: https://www.rockethealth.shop/. Accessed December 14, 2021.

1 year impact goals

• Build and evaluate a digital clinical decision support tool that results in 97.5% or greater provider adherence to MotoMeds' clinical guidelines in the Haiti deployment. 
• Pilot MotoMeds in Ghana to establish clinical safety and feasibility of the MotoMeds model in a setting outside of Haiti. 

5 year impact goals

•  MotoMeds coverage of delivery zones in Haiti will reach 50% of the country.
• Inpatient hospitalizations for acute respiratory infection, and dehydration from diarrheal disease in Haiti's MotoMeds coverage areas will be reduced by 25%.
• MotoMeds will be operational in 5 countries

We are currently using the following indicators to establish proof of concept and refine the solution.

Clinical

• Proportions of case severity categorizations
• Enumeration of case disease types
• Rates of clinical guideline adherence and enumeration of guideline deviations
• Rates of severe adverse events (per 1,000)
• Disposition at 10 day follow up

Operational

• Median time of phone consultation
• Median time to delivery
• Proportion of failed deliveries
• Enumeration of qualitative feedback by end users
• Percentage of repeat users (parents and/or children receiving service from MotoMeds on ≥ occasion)
• Percentage of paying users
• # hours of training provided to MotoMeds staff

Once we transition out of the development phase and into a growth phase we will also use the following indicator;

• Enumeration of cases consulted per night, per week, per month, per year.

Our theory of change is based on the idea that children need early access to simple and cost-effective treatments for common childhood illnesses to prevent illness progression. Traditional methods of providing families with basic healthcare are limited especially during the nighttime hours, forcing families to wait until morning to seek qualified care. A pediatric telemedicine and delivery service will fill this nighttime gap in pre-emergency healthcare access. MotoMeds is designed with the intent to bridge the timing and distance barriers to pre-emergency care, and thus avert patient decline. In Haiti, we have seen our theory of change manifest in reality as a positive feedback loop where high-quality professionalism, reasonable payment structure and word-of-mouth all perpetuate MotoMeds with minimal advertising. After 2 years of operations we have consulted and provided care to over 1000 children. At 10 day follow up 94% reported their condition as recovered/improved, % same, and <1% worse. 

The INACT2-3 studies exposed a critical need for digital clinical decision-support (dCDS) tools in order to improve guideline adherence and efficiency of the MotoMeds model at scale. MotoMeds providers currently use a paper case report form for data capture of clinical symptoms and medical history and a set of paper clinical guidelines to classify severity and form treatment plans. The forms are cumbersome and complex, contributing to long call durations (median 20mins; range 3-60mins) and avoidable guideline deviations (8% of INACT2 cases included a guideline deviation). The extended call length hinders the ability to respond to subsequent calls, burdens the caller who pays by the minute, increases the likelihood that the caller’s phone battery will run out and delays time to delivery. Training new staff has been time and resource intensive, particularly when launching new delivery zones. A digitized version of the form that incorporates the workflow, clinical guidelines and medication dosing calculator will streamline the overall process, increase guideline adherence, facilitate onboarding of new staff and improve operational outcomes (e.g. reduce call duration, time to delivery). 

We are currently in the design stages of building a digital clinical decision-support tool that can be used on and offline on a mobile device or web interface to guide the user to an evidence-based treatment plan in less than ten minutes, while still allowing the provider autonomy to deviate from the guideline for unique situations. The user interface will draw on 7 years of technology and clinical algorithm development experience, and three years of formative iteration of the MotoMeds paper clinical decision support tools. The interface will adhere to an input/output format. The first step is entry of the clinical complaint(s). The selection of complaints is based on INACT1 and INACT2 data and include: “fever”, “cough/breathing”, “vomiting”, “diarrhea”, “ear pain”, “skin”, “pain with urine” and “other”.  The second step is triage using WHO Integrated Management for Childhood Illness danger signs. The third step is gathering additional information specific to each complaint and allergies. The input section will have expandable ‘input page’ that will expand and contract based on the data entered on the ‘input homepage’. A progress bar will provide a visualization of where in the user is at in the data input phase. After data entry on the input page(s), the output page will provide a patient summary with the age, estimated weight, severity level (green, yellow, red), and recommended disposition. The weight estimation will be based on the WHO 50th percentile weight for age. All patients will be screened for dehydration status. Medication recommendations will align with WHO guidelines and dosages and duration will be calculated. Anticipatory guidance will be provided for clinic follow-up. The user-interface will conclude with messaging on the scope of the decision-support tool, including text that the provider should deviate from the recommendation if the case is outside the intended scope. Our timeline to begin evaluating a prototype of the digital tool is fall 2022.

We have not yet built the technology. Once built, prototypes will be tested with MotoMeds providers in an iterative process by running standardized case simulations. The case scenarios will cover mild, moderate, and severe cases. Each case simulation will be paired such that the case can be run twice with and without the dCDS to establish baseline behavior. Quantitative data will be gathered by enumerating duration of call/exam with stratification for time to determine severity, treatment plan, medication dosing, and disposition. Qualitative data will be obtained from both focus group and in-depth interviews.

Using the beta prototype we will evaluate the dCDS tool for use at at a TMDS in a pre-post clinical trial (INACT-4). The dCDS tool will be deployed at the TMDS call centers in Haiti and Ghana to determine if the dCDS tool is associated with increased provider guideline adherence and improved operational metrics (shorter phone consultation and time to delivery).

Our core team members are housed within The Emerging Pathogens Institute at The University of Florida. Dr Eric Nelson, is an Assistant Professor of Pediatrics at UF and is the Principal Investigator of the INACT studies. Additional team members are associated with Université d'État d'Haiti- Faculté de Médecine et de Pharmacie, and the Ghanaian National Ambulance Service. 

Our leadership team is comprised of physicians, public health professionals and researchers native to the United States, Haiti and Ghana. Prior to engaging in any country we feel that it is essential that we recruit members in leadership roles representative of that country. A culturally and ethnically diverse team is essential to what we do. 

Our team is small with mostly part-time contractual workers. However, we promote equity amongst our team members by providing training and professional development opportunities. An example of an equity minded decision was when first hiring nurses to work at the call center in Haiti our primary focus was not on years of experience or name recognition of the nursing schools they attended but on critical thinking skills and enthusiasm for the mission. We hired nurses with varying backgrounds and the more junior nurses spent time shadowing more experienced nurses before working autonomously.

The MotoMeds work environment is inclusive to all team members. The healthcare field can be hierarchal with unequal power dynamics between physicians, nurses, etc. and our team actively promotes the idea that everyone should be valued and has something to offer. An example of acting on this ideal was when we held a stakeholder workshop with Haitian healthcare professionals to learn more about access to healthcare in Haiti we intentionally invited non-conventional providers to learn about their  perspectives as well. 

In theory MotoMeds could be an effective healthcare delivery method during the day or night but we chose a nighttime model for two reasons. In many low income settings there are no alternative option to pre-emergency care so there is virtually no competition. But we also did not want to disrupt the existing daytime healthcare provider network by providing potentially free services.  

Our plans for longterm financial sustainability are still in the idea stages and the exploration of revenue models is one of the main areas where we are looking for additional support. For the next 2-3 years we will continue to fund our work through research grants and private donations. During this period we would like to begin attracting donations from corporate sponsors who will in turn receive social marketing opportunities. An example would be Digicel, a Caribbean based telecommunications company that is a vital component to MotoMeds operations. In terms of a longer term outlook we envision two models, both involve providing partners with an ‘off the shelf package’ of tools that involves the transfer of clinical and logistical expertise and capacity. Model 1 is 'hub and spoke' concept. For example, in Haiti that would involve a centralized call-center operated by the Ministry of Health or other large-scale healthcare provider with delivery zones (spokes) spread throughout the country and operated by a combination of the ministry of health and/or different healthcare NGO partners. This model leverages healthcare providers with existing localized infrastructure. Model 2 involves collaboration with existing emergency response service. MotoMeds would incorporate pre-emergency services into their workflow at a relatively low cost. This would provide the benefit of expanded services and reserve the emergency medicine infrastructure for emergencies. These models are not mutually exclusive.  

To date we have been successful in funding our research with the following sources of income.

• The formative research in Bangaldesh, the INACT1- Haiti needs assessment and the INACT2- Haiti pilot were funded by an NIH Early Independence Award to Dr Eric Nelson.
• The INACT3- Haiti study testing for scalability is being funded by private donations from over 50 individuals.
• Fees collected from MotoMeds users totaled around $2000 US and were used to purchase medications for the MotoMeds pharmacy.
• We received $200,000 from the US government (unable to share additional details) to complete the INACT1- Ghana needs assessment and INACT2- Ghana pilot study.