Combatting Antimicrobial Resistance through AI and Analytics in LMICs

This solution employs advanced analytics and AI to identify and address priority antimicrobial misuse and resistance hotspots, clusters, and profiles in Kenyan LMICs and beyond using mobile and digital platforms. This will enable the targeted allocation of advanced analytics and control resources for enhanced and precision healthcare decision-making and interventions.

Dr. Patrick Njage is highly experienced in the incorporation of complex and unstructured data using predictive and prescriptive analytics, network analysis and other advanced modeling approaches to combat antimicrobial resistance. 

The specific problem being addressed is the crisis of antimicrobial resistance (AMR) resulting from the rampant misuse and overuse of antimicrobials, particularly in Low- and Middle-Income Countries (LMICs) including Kenya which we estimated to have among the highest AMR globally[1]. AMR threatens to render common infections untreatable, leading to increased mortality, morbidity, and healthcare costs. Globally, AMR is responsible for 700,000 deaths annually, a number projected to rise to 10 million by 2050 if unchecked. The causes of AMR include over-prescription of antibiotics, lack of diagnostic tools, poor adherence to treatment guidelines, and inadequate patient education. This proposal focuses on leveraging mobile and digital platforms for dynamic data collection and analysis to leverage antimicrobial data for real-time identification of AMR hotspots, understanding antimicrobial usage patterns, and swift detection of anomalies. By applying advanced analytics and artificial intelligence, the solution aims to enable healthcare systems to take informed actions to promote rational use of antimicrobials and combat AMR. This approach addresses the cause of inadequate surveillance and data analysis capabilities, aiming to reduce the incidence of AMR in the targeted communities and, by extension, globally.

1. Hendriksen, .. Njage, P. et al.  Nat Commun 10, 1124 (2019). 

This solution primarily serves healthcare professionals, policymakers, and regulatory agencies in Low- and Middle-Income Countries (LMICs). It addresses their critical need for comprehensive data and insights into antimicrobial usage (AMU) and resistance (AMR) patterns to combat the global health threat of AMR effectively. By leveraging advanced analytics, mobile and digital platforms, and artificial intelligence, the solution provides these stakeholders with the tools to identify AMU anomalies, predict AMR hotspots, and plan targeted in-depth monitoring and interventions. Engagement with these groups is achieved through existing collaborations in ongoing and previous projects on antimicrobial resistance surveillance, ensuring the solution aligns with their needs and respects ethical considerations and patient confidentiality. We are working with the Nairobi County Antimicrobial Stewardship Interagency Committee (CASIC) as advisors in a scoping project to roll out AMR mitigation. This involves engagement with prescribers, laboratories,  pharmacists and infection control from whom we have gained a deep understanding of the problem scope and potential mitigation strategies. Our approach will not only aid in promoting rational antimicrobial use but also support the development of policies and practices to mitigate the rise of AMR, directly impacting public health outcomes in these communities.

This solution contributes a significant public good by addressing the health challenge of AMR through a novel integration of technology, analytics, and collaborative data-sharing. The solution offers a comprehensive understanding of AMR dynamics by leveraging mobile and digital platforms for dynamic data collection and employing advanced analytics, including artificial intelligence, to analyze antimicrobial usage (AMU) patterns. This approach not only enhances the ability of healthcare systems, especially in LMICs, to identify and respond to AMR hotspots but also facilitates the rational use of antimicrobials, thereby preserving their effectiveness for future generations.

Creating open-access, real-time dashboards and reporting tools as a direct output of this work provides a tangible public good by making critical information on AMR trends and anomalies accessible. These tools empower healthcare professionals and policymakers with actionable insights, enabling informed decision-making to mitigate AMR risks. Furthermore, the ethical framework for data sharing established by this solution ensures that the benefits of this innovation are shared widely, under fair, reasonable, and non-discriminatory terms, contributing to global health security. By providing knowledge, data, and analytical tools as public goods, this solution fosters a collaborative and informed approach to combating AMR, exemplifying a commitment to improving public health outcomes globally.

This solution aims to significantly improve healthcare systems in Low- and Middle-Income Countries (LMICs), particularly in Kenya, in combating antimicrobial resistance (AMR) through the use of advanced analytics and mobile technology for comprehensive data collection. By identifying antimicrobial usage (AMU) patterns and anomalies, it enables the implementation of informed interventions to promote rational AMU and combat AMR spread. Primarily benefiting vulnerable populations in LMICs, who suffer from limited healthcare access and a high prevalence of infectious diseases, this initiative seeks to enhance the effectiveness of antimicrobial treatments and reduce the morbidity and mortality associated with resistant infections, thereby improving public health outcomes. Drawing on successful experiences from Ethiopia, Tanzania, Nigeria, and Mozambique, where mobile phone surveys were effectively used for data collection, and our experience employing artificial intelligence for predictive and prescriptive modeling of AMR, the solution demonstrates potential in forecasting trends and pinpointing AMR development hotspots. By integrating data collection, analytical insights, and targeted interventions, this approach is poised to make a significant impact on AMR management in LMICs, improving healthcare outcomes and access to effective treatments for vulnerable communities.

To scale the impact on combating antimicrobial resistance (AMR) over the coming years, a strategic expansion and enhancement plan will be deployed, initially solidifying the project's foundation in Kenya and then extending into more Low- and Middle-Income Countries (LMICs). The plan leverages mobile technology for data collection and analysis, refining analytics and AI algorithms for adaptability across diverse regions, emphasizing customization to local healthcare systems. Over three years, the aim is to significantly broaden the project's geographical reach into multiple regions in Kenya and other LMICs (in the list below) at high risk of AMR, developing strong partnerships with health organizations, government bodies, and the private sector to access larger datasets, improve predictive accuracy for AMR hotspots, and ensure ethical data sharing. Investment in technology infrastructure is key, including sophisticated AI models and user-friendly digital platforms for real-time analysis and reporting. The project also focuses on integrating solutions into national healthcare policies, promoting rational antimicrobial use, and building capacity among healthcare professionals for effective AMR management. Through expanding reach, enhancing technology, and cultivating partnerships, the initiative seeks to transform the AMR response in the LMICs, aiming for a lasting public health impact in LMICs and beyond.

Measuring success against our impact goals for combating AMR will involve a structured approach to monitoring and evaluation, focused on specific, quantifiable indicators and metrics that reflect the effectiveness and reach of our solution:

 1. Reduction in AMR Incidence: Measuring the decrease in reported AMR cases within the target populations before and after implementing our solution. Success will be indicated by statistically significant reduction in AMR rates.

2. Changes in Inappropriate Antimicrobial Prescription Rates: Tracking the rate of antimicrobial prescriptions and target infections to monitor changes towards more rational use, aiming to reduce unnecessary prescriptions.

 3. Accuracy and Predictive Value of Analytics: Evaluating the accuracy of our AI-driven predictions against observed AMR and AMU outcomes, aiming for high precision and recall in identifying AMR hotspots and trends.

4. User Engagement with Digital Platforms: Monitoring the adoption and active use of our mobile and digital platforms by healthcare professionals, measured by user login frequency, engagement rates, and feedback surveys.

5. Compliance with Antimicrobial Stewardship Practices: Assessing the implementation of recommended interventions and adherence to antimicrobial stewardship guidelines among healthcare providers, through audits and feedback loops.

Metrics from a pilot project will form a baseline for these to guide and monitor our impact goals.

The project faces several barriers over the next one to three years, including financial constraints, technical challenges, legal and regulatory hurdles, cultural and educational gaps, and policy and market barriers. Overcoming financial constraints necessitates partnerships with the private sector and government support to fund scaling technology, research, and interventions. Technical challenges, such as infrastructure limitations in LMICs and the need for local customization, will require collaboration with local tech companies, investment in infrastructure, and development of adaptable platforms. Legal and regulatory hurdles, including data privacy laws and data sharing agreements, will be navigated through building relationships with regulatory bodies and aligning the project with national health priorities. Cultural and educational resistance from healthcare providers or communities will be mitigated through comprehensive training, community engagement leveraging our existing projects, and leveraging local influencers. Lastly, aligning project goals with existing healthcare policies and market dynamics poses challenges that will be addressed through advocacy and showcasing pilot study success stories. The strategy to navigate these barriers includes leveraging partnerships, utilizing pilot data to attract further investment, and continuously adapting the solution based on field feedback, aiming to scale the impact and contribute significantly to the global fight against AMR.

We are applying to The Trinity Challenge because it aligns perfectly with our mission to use advanced analytics and digital technology to combat antimicrobial resistance (AMR), a growing global health threat. The Challenge's focus on harnessing data and analytics to improve public health outcomes mirrors our project's objectives, making it an ideal platform for support and collaboration. 

  Our primary barriers include the need for additional funding to scale our solution across LMICs, access to broader networks of global health experts, and the integration of our data analytics platform with existing healthcare systems. The Trinity Challenge can help us overcome these obstacles by providing not only the necessary financial resources but also by facilitating partnerships with a global network of experts and organizations. This support would enable us to refine our solution, expand our reach, and enhance our impact on public health systems. Moreover, participation in The Trinity Challenge would provide validation and visibility, crucial for attracting further investment and collaboration opportunities

To initiate, accelerate, or scale our solution, we seek to collaborate with a variety of organizations including:

Global Health Organizations: Partnerships with entities like the World Health Organization (WHO) and the African Center for Disease Control and Prevention could provide access to a wealth of AMR and public health expertise, along with global networks for dissemination and advocacy.

Local Health Departments and Regulatory Agencies: Collaborating with local health authorities, such as the Ministry of Health in Kenya, will facilitate access to national health databases and ensure our solution aligns with local healthcare policies and needs.

Technology and Mobile Companies: Teaming up with technology firms, especially those with a strong presence like Safaricom in Kenya, could enhance our mobile data collection platform's reach and effectiveness.

Academic Institutions: Partnerships with universities and research institutions, particularly those with AMR and public health research programs, for leveraging ongoing efforts and partnerships.

Non-Governmental Organizations (NGOs) in Healthcare: Collaboration with NGOs working on the ground can provide insights into local challenges and community needs, ensuring our solution is community-driven and sustainable. These collaborations will offer critical resources, local insights, and technological support to refine and scale our solution, making it more effective in combating AMR.