Integrated, Intelligent Healthcare to Improve Pediatric Health Outcomes in Tanzania

Dr. Elsa is a health assistant tool for care providers that offers symptom assessment, decision support, and next steps recommendations to improve health outcomes.
Tanzania

Background

The delivery of, and access to, high-quality healthcare is crucial to achieving national health objectives and the overall development of a country and its people. In a developing nation as diverse as Tanzania, health challenges create an imperative to act and an opportunity for innovation. Key among these challenges are the shortages of trained healthcare providers at all levels of delivery and high under-five morbidity and mortality rates. Emerging technologies such as Artificial Intelligence and Machine Learning are demonstrating that, with enough information, they can perform as well as highly- trained clinicians to improve the healthcare delivery system. It is with these tools that we can better support providers at the first points of care, more accurately predict patient outcomes and ensure that care services are reaching the people who need them most.

Problem

Tanzania faces significant health challenges, particularly for children and adolescents, and ranks one of the worst in the world in healthcare delivery/ outcomes. Healthcare workers, especially in rural areas, have limited training, resources, and tools to diagnose and treat complex conditions, resulting in high rates of misdiagnoses, prolonged care, and unnecessary services. Additionally, Tanzania has one of the worst doctor-to-patient ratios in the world (1:20,000), with over 75% of doctors living in urban areas. There is a severe lack of pediatric specialists in the country, causing 20 million children to experience poor health outcomes and low-quality care.

Additionally, youth and adolescents are particularly at risk for illness and malnutrition. The lack of quality healthcare services and access to adolescent-friendly care leads to high morbidity and mortality in young people. Children who acquire HIV infection vertically are likely to pass away before the age of 5 as a result of malnourishment and lack of access to care for basic opportunistic infections. As youth get older and become sexually active, their risk of acquiring HIV and other sexually transmitted diseases increases, proving a need for patient-centered, predictive care services.

Solution

The Dr. Elsa Health Assistant tool is powered by data and artificial intelligence and supports healthcare workers (HCWs) in rural areas through symptom assessment, diagnostic decision support, next step recommendations, and predicting disease outbreaks. Our tool is delivered to healthcare providers through a mobile application on a phone or tablet, and acts like an expert

pediatrician in a tablet – the tool takes in patient demographics, vitals, symptoms and test results (when available), and outputs a differential diagnoses and evidence-based recommendations for next steps. It supports low-cadre healthcare providers in making decisions about their patients through powerful symptom assessment, and is intended to augment the healthcare providers’ knowledge to produce better health outcomes.

With the support of dLab, our tool is built for pediatric care (ages 0-14), and has been tested in Bagamoyo health facilities in partnership with Ifakara Health Institute, as well as Kimundo Dispensary in Arusha. The tool supports 120 pediatric illnesses and has a unique question flow that supports accurate and timely data collection.

Process

This year, we have collected over 25,000 pediatric health records to build and train our models, and we have worked with expert pediatricians to ensure high quality decision making. The tool currently supports common pediatric illnesses (rash, gastroenteritis, etc), nutrition-related illnesses, high- mortality illnesses such as malaria, and sexual and reproductive health related illnesses (including HIV/AIDS). All of these are of high priority to the Tanzania government and other large health organizations.

The Dr. Elsa tool lives in the cloud and can be delivered across multiple devices to a large number of healthcare workers in any location. Through tablets or mobile phones, healthcare providers can access the tool, which is built into an easy-to-use interface. Healthcare workers input patient demographic information (gender, age, location, etc.), patient symptoms, and relevant history, and the tool outputs insights and next step recommendations to the healthcare worker. As per the national requirements, we also recommend nearest health facilities and forward relevant data to those facilities.

We partnered with Ifakara Health Institute to test the friendliness and usability of the tool in the clinical setting across three levels of institutions. In addition, we tested the tool with a Clinical Officer at the Kimundo Dispensary in Arusha. To date, we have received positive feedback about the use of the tool, and continue to iterate on the design and models based on new information we learn. We are also working with a consortium of organizations to test this tool with community health workers and private pharmacies in Tanzania. In 2020, we will work on scaling up the use of Dr. Elsa and building tools to support the HIV/AIDS epidemic.

Outcomes & Impact

Our smart tools provide value to healthcare providers, patients, health institutions/ organizations, and government agencies. The Health Assistant integrates into the current healthcare workflow so that health providers better diagnose and treat more patients with less

fatigue. With Dr. Elsa, pediatric patients are directly impacted through access to intelligent decisions and predictive symptom assessment. The social impact of our work is improved health outcomes for youth in East Africa (in support of SDG 3) and decreased rates of misdiagnoses. Additionally, we will see less crowded care facilities due to better triaging/ referral of patients, better diagnostics at the first point of care, better linkage to care, and more appropriate allocation of resources as a result of accurate infectious disease outbreak predictions.

We are tracking a number of other outcomes:

• Improved lives, measured by the health quality indicator “decreased number of misdiagnoses in the target population”
• Quality & Timely Care: Number of complex diagnoses managed by healthcare worker without referral, number of patients seen per day, and percent increase in time spent with patients using Elsa
• Reduced child and adolescent illness as a result of more evidence-based diagnostics
• Better understanding of disease/illness through pattern recognition algorithms
• Reduced self-medicating and decreased antibiotics resistance as a result of more evidence-based diagnostics
• Ability for Ministry of Health to relay vital health decision information to healthcare providers in a timely and efficient manner.

The world is currently undergoing a massive shift in healthcare – Dr. Elsa will be able to support that shift on the African continent and build tools that are effective and make a difference in people’s lives.

Key Collaborators

Ally Salim Junior is the Founder and CEO of Inspired Ideas, the company behind the creation of Dr. Elsa. Ally is a software engineer who is passionate about locally-built solutions that utilize technology, and about empowering others to ensure the health and sustainability of their communities. He works with the rest of his team to build and grow Dr. Elsa: Megan Allen, Happiness Watimanywa, and Jonathan Lukumay.

The team benefited from a $20,000 grant from the Data for Local Impact (DLI) Innovation Challenge in 2019, which has allowed them to obtain necessary technology, collect health data for training the tool, test and validate the models, and implement the tool in Bagamoyo.

Dr. Elsa has an established partnership with Ifakara Health Institute to test the tool. We are thankful for the support from DLIIC, which aims to engage, support, and connect Tanzanian innovators, developers, and solution providers to each other and to opportunities to make a difference in people’s lives. To learn more, visit http://www.dlab.or.tz.