BactWatch: Community AMR prediction and surveillance through AI
East Africa is facing an AMR surge, which is worsened by sub-optimal case-identification, surveillance and data management for AMR at community-level. BactWatch is going to be a comprehensive AMR data management system, complimented by a mobile app used to collect, predict and identify AMR cases in communities, using artificial intelligence.
Dr. Godfrey Kigozi, MBChB, MPH, PhD
- Innovation
- Integration
- Implementation
Many innocent lives were lost to infectious diseases until 1928 when Sir Alexander Fleming discovered penicillin, which initiated the antibiotic revolution. Despite the revolunization of healthcare by the introduction of antibiotics, the world is currently faced with AMR and WHO has declared it as a global public health threat.
In 2019, 1.27 million deaths were directly caused by AMR globally. In Uganda in 2019, there were 7,100 deaths attributable to AMR and 30,700 deaths associated with AMR. Recent studies have showed that resistance to the most commonly-used antimicrobials (e.g. penicillins, tetracyclines, cotrimoxazole) was in some cases above 80 percent in Uganda.
Despite the increasing AMR rates, there is still sub-optimal case identification, surveillance and data management for AMR in Uganda at community-level. There is currently no standardized comprehensive data management system for collecting, analyzing and reporting AMR data, both at community-level and at health facility level. Furthermore, there is limited capacity in communities to detect, predict and identify AMR cases, thereby leading to sub-optimal patient management. There is also lack of AMR data collection electronic tools like a mobile app, to ease the collection, prediction and identification of AMR cases.
BactWatch is targeting to help;
- Community Health Workers (CHWs),
- Healthcare workers in different health facilities in Uganda, and
- Ministry of Health and District Health Teams.
BackWatch is going to support CHWs to collect, predict and identify AMR cases within communities, using its mobile application supported by artificial intelligence. BactWatch will also help healthcare workers in different facilities to also collect, predict and identify AMR cases, and thereby effectively manage the AMR cases.
BactWatch's web-based dashboards will enable Ministry of Health, District Health Teams and other partners to monitor the prevalence of AMR across the country in real time, and this will inform key policy decisions in the control of AMR in Uganda.
The initial steps of the project will involve aims 1 and 2, and these will involve engagements and interviews (focus group discussions and key informant interviews) with the target audience (CHWs, healthcare workers, Ministry of Health and District Health Teams) to get their perceptions, and experiences about AMR data management, and case identification at community and health facility levels. We shall also acquire their insights on what they want BactWatch to be like, and the functionalities that it should be having.
- Proof of Concept: A venture or organisation building and testing its prototype, research, product, service, or business/policy model, and has built preliminary evidence or data
- Artificial Intelligence / Machine Learning
- Big Data
- GIS and Geospatial Technology
- Internet of Things
- Software and Mobile Applications
BactWatch will make several publications about the innovation, so that other countries in Africa with similar settings like Uganda can also establish similar systems.
Furthermore, BactWatch is going to be freely accessible to Ministry of Health and its partners for use in AMR case identification and data management and surveillance, during its scale up.
BactWatch will also have web-based dashboards that will be used freely by key stakeholders at national and district levels to monitor the progress of the AMR program, and make key policy decisions regarding AMR control in Uganda.
BactWatch will also have a freely available mobile application, which will be used by community health workers in the different communities and healthcare workers in the different health facilities to predict and identify AMR cases, to improve patient management.
Our solution prototype for the AMR data system will be freely available to the other Ministries of Health with similar settings like Uganda, if in case they need to implement a similar solution to help in AMR control, like ours in Uganda.
BactWatch is going to involve a comprehensive AMR data management system with a real-time web-based dashboard. This dashboard will be used by Ministry of Health and District Health Teams to monitor the AMR program, and make key policy decisions regarding AMR control in Uganda. Hence this is a tangible impact, that will benefit the people of Uganda if right policy decisions are made to prevent AMR in the country.
BactWatch will also have a mobile application, installed on mobile phones to help community health workers and healthcare workers to predict and identify AMR cases in communities and health facilities, to assist in proper patients management. Hence this will be a tangible impact, that will benefit patients facing AMR in Uganda.
About five publications will also be made in peer-reviewed journals about BactWatch, and these will guide other countries to establish similar systems, which will help in the control of AMR in the Sub-Sahara Africa region.
From the start, our team is working with Ministry of Health and selected district health teams, so that they are aware of the solution.
After a successful pilot study in aim 3, our team will work with Ministry of Health and its partners to implement and scale up BactWatch in a phased manner, starting with about 5 selected health regions in the first year, and scaling it up to all the 16 health regions by the third year.
Our team will also share the results about BactWatch at international conferences and in peer-reviewed journals to disseminate the findings, but also stimulate stakeholder buy-in about the solution.
Our team has established a set of key performance indicators for monitoring and evaluation of our solution and these include;
- Number of AMR tests performed and entered into the data management system
- Number of mobile application downloads
- Number of login accounts opened into the system
- Number of AMR case predictions and identifications made
- Number of District Health Teams implementing the solution in their districts
- Number of health regions implementing the solution
Our team will have regular monitoring and evaluation meetings to discuss the progress of the solution implementation and identify the opportunities for improvement of the solution.
- Uganda
- Uganda
The major barrier is lack of a policy to guide AMR data management in Uganda. However we are working with Ministry of Health to advocate for the establishment of a policy to guide AMR data management in Uganda
Furthermore, the other challenge is limited funds to build the solution, however we are trying to establish partnerships with several donors to raise funds to establish the solution.
Currently, we have an established multi-disciplinary team of experts and we are currently starting on the process of conducting aims 1 and 2.
- Academic or Research Institution
Our team is applying to The Trinity Challenge because this challenge rightly aligns with our objective of contributing towards AMR control in Uganda, and beyond. We desire to establish a comprehensive AMR data management and surveillance system, however we have limited funds, and that is why we are applying to The Trinity Challenge.
Furthermore, we also believe that The Trinity Challenge will connect us to other AMR Experts, for peer-to-peer learning, and south-to-south collaborations.
We would mainly wish to collaborate with Ministry of Health Uganda and its partners including WHO, The Global Fund, PEPFAR, and CHAI among others. These would offer additional guidance to our team, to initiate the solution.