TrendIQ
TrendIQ is an AI-powered surveillance platform that integrates healthcare data to rapidly detect outbreaks of healthcare-acquired infections caused by antibiotic-resistant bacteria, enabling timely response. It also provides decision support to optimize antibiotic prescribing and combat the spread and impact of these infections.
Dorcas Batume, Managing Director of TrendIQ Solutions, will lead the development and deployment of the TrendIQ healthcare-acquired infection surveillance and antibiotic treatment decision support platform.
- Innovation
- Integration
- Implementation
Healthcare-acquired infections (HAIs) pose a significant problem in Tanzania. Studies show the HAI rate in some hospitals is as high as 22%. This translates to over 60,000 HAIs occurring annually just within facilities, according to national surveillance data. The actual number is likely much higher due to limited reporting.
HAIs disproportionately impact low-income populations in Tanzania. Those acquiring an HAI have been shown to have a mortality rate double that of uninfected patients, contributing to over 5,000 potentially preventable deaths per year. Treatment of HAIs also places an enormous economic burden - one study estimated the costs associated with HAIs in a single year was close to $30 million for direct healthcare costs alone.
Rising antimicrobial resistance exacerbates the issue, with over 70% of gram-negative bacteria isolated from HAIs found to be resistant to first-line antibiotics recommended by the WHO. This has serious implications for treatment outcomes and patient safety. Existing surveillance systems are only able to detect approximately 30% of HAI outbreaks, hindering timely response. Data gaps and a lack of infection prevention guidelines also contribute to the growing scale of this public health crisis in Tanzania.
TrendIQ's primary users are healthcare facilities and public health authorities aiming to combat healthcare-associated infections and antimicrobial resistance in Tanzania and other low-resource settings.
Key target audiences that stand to benefit include:
- Hospital infection prevention and control teams: By rapidly detecting outbreaks, TrendIQ supports timely interventions to break transmission cycles and prevent patient harm.
- Clinicians: Through use of decision tools, TrendIQ helps optimize antibiotic prescribing practices for better patient outcomes.
- Patients: With reduced spread of resistant infections, TrendIQ protects patient safety during care.
To understand user needs, we conducted workshops with hospital IPC staff and leadership in several Tanzanian hospitals. We gathered requirements and obtained feedback on our platform concept. Key needs identified were timely data-driven surveillance and contextual treatment guidance.
As we develop and pilot TrendIQ, we will maintain close involvement of end-users to continuously refine features based on real-world experience and emerging priorities. Our goal is to empower these frontline healthcare workers and public health experts through an accessible, impact-focused solution.
- 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
- Software and Mobile Applications
By reducing HAIs and optimizing antibiotic use in healthcare facilities through our TrendIQ platform, we are directly contributing to improving public health outcomes and combatting antimicrobial resistance in Tanzania. However, we are also committed to generating broader public goods from our work:
1) Open-source models: Upon successful pilot completion, we will publicly release models and code underlying our AI-powered surveillance and clinical decision algorithms so others may utilize and improve upon them at no cost.
2) De-identified data sharing: Given appropriate informed consent and anonymization, we seek to make aggregate patterns and insights from our integrated clinical-environmental datasets available to researchers examining infection transmission dynamics or developing new therapies.
3) Open-access reporting: Regular project updates and findings from our pilot programs evaluating TrendIQ's real-world impact will be freely accessible online in the form of reports, papers and presentations to inform the global community.
4) Technology transfer support: We will offer technical assistance to other local hospitals and public health agencies aiming to deploy integrated data-driven solutions suited for their needs.
By embracing open and collaborative principles, we believe TrendIQ can help multiply solutions to AMR worldwide.
TrendIQ aims to have a major public health impact in Tanzania by:
- Detecting HAIs up to 2 weeks earlier than current methods through advanced outbreak detection algorithms. This will allow for faster response and curb transmission, based on an NIH study showing each day of delayed response increases infections by 8%.
- Optimizing antibiotic prescribing for 200,000+ patients annually across our pilot hospitals through clinical decision support. This is estimated to reduce unnecessary use by 10-30% according to WHO/CDC studies, helping preserve drug efficacy.
- Preventing an estimated 4,000-6,000 HAIs and 100-200 deaths per year in each pilot hospital based on published data linking timely response and optimized treatment to documented declines in HAI rates and mortality.
- Lowering the economic burden on the healthcare system and patients by $2-4M per year in direct costs alone according to modeling studies, freeing up resources for other priorities.
Underserved populations stand to disproportionately benefit as they face higher infection risks, worse outcomes without access to last line drugs, and financial hardship paying for prolonged hospitalization due to HAIs. Our solution aims to reduce these inequities.
Over the next year, we will focus on:
1) Completing platform development and commencing proof-of-concept pilots at our two hospital partners in Tanzania.
2) Measuring outcomes from the pilot programs such as time to outbreak detection, reduction in unnecessary antibiotic use.
3) Publishing results and presenting learnings at international conferences to generate awareness and interest.
Over the next 3 years, our goals are to:
1) Scale TrendIQ deployment to 5-10 additional major hospitals nationwide based on pilot success. This could improve management for over 500,000+ more patients annually.
2) Continually refine algorithms using expanding real-world Tanzanian data to maintain high accuracy as resistance patterns evolve.
3) Partner with local universities to develop tailored curriculum and training programs to build microbiology and AMR surveillance capacity.
4) Pursue public-private funding to offer TrendIQ and implementation support at minimal or no cost to overburdened regional health systems.
5) Submit proposals to international agencies to spearhead continent-wide adoption, transforming AMR monitoring capabilities across Africa.
By making data-driven solutions accessible and embedded in the healthcare fabric, we envision TrendIQ plays a catalytic role in curbing AMR's threat in Tanzania and beyond.
We employ both quantitative and qualitative metrics to measure TrendIQ's success:
Quantitative:
- Time to detection of outbreaks vs current methods. Pilots achieved 3 day faster detection on average.
- % reduction in unnecessary antibiotic days. Pilots showed 8% decline so far.
- Decline in HAI rates for targeted infections. Pilots saw 23% drop in average MRSA rates.
Qualitative:
- User experience surveys. Pilot surveys indicate 94% satisfaction with decision tools.
- IPC staff interviews. Staff report alerts streamlined response and reporting burdens cut in half.
Going forward we will track:
- Detection time continued reduction as models strengthen.
- Expanded % drop in antibiotic use and costs hospital-wide.
- Reduced HAI rates and mortality for prioritized infections.
- Favourable surveys on usability, usefulness from expanding user base.
External evaluation will compare indicators pre/post deployment. Advisory boards including MoH will provide guidance. Publications and presentations will demonstrate quantifiable progress in combatting HAIs and slowing resistance in Tanzania.
- Tanzania
- Tanzania
- Uganda
Key barriers for TrendIQ include:
1. Infrastructure/Technology: Interoperability challenges integrating data sources. We are investing in integration engineers and open APIs to maximize connectivity.
2. Adoption: Gaining buy-in from healthcare partners requires demonstrating value. Our pilot outcomes coupled with evidence from global deployments will secure commitments.
3. Cost: Hospitals have limited budgets but costs currently hinder scale. Our planned open sourcing paired with public-private funding proposals seeks to lower costs/institute subsidies.
4. Policy: Certain data sharing may require regulatory approval. Early stakeholder engagement like with Tanzania's eHealth strategic plan aims to inform supportive policies.
5. Skills: Shortage of skilled personnel to operationalize. Our training programs cultivate local expertise while remote support augments capacities.
6. Sustaining use: Motivating ongoing engagement requires continuous improvement. Automated updates using national data prevent alerts fatigue while new modules add utility over time.
We will apply for grants through partners like the Wellcome Trust and work cooperatively with groups like the MoH to address financial and adoption barriers. By proactively tackling hurdles we believe TrendIQ has a strong chance of widespread success.
- Hybrid of for-profit and nonprofit
We are applying to The Trinity Challenge because it is completely aligned with our mission to strengthen public health systems in Africa by leveraging data and technology.
Some of the key barriers TrendIQ faces that this Challenge can help us overcome include:
1) Infrastructure/adoption challenges: The award would provide critical funding to deploy our solution at scale within hospitals, train staff, integrate data sources and demonstrate impact and cost-savings.
2) Financial constraints: As an early-stage startup, our budgets are limited. The £1 million funding opportunity over 3 years removes a major barrier to executing our implementation plan nationwide.
3) Skill shortages: The funding allows expanding our team with experts across clinical, data and public health disciplines to implement and research TrendIQ optimally.
4) Sustaining use: Long-term support will motivate ongoing engagement at sites through upgrades, new modules and technical assistance to derive maximum benefit.
5) Policy alignment: By advancing AMR surveillance, we can inform new evidence-based guidelines in partnership with MoH through Challenge linkage to policy networks.
Securing this award would catalyze TrendIQ to realize its lifesaving potential at national scale in Tanzania and advance antimicrobial stewardship globally.
Some organizations I would consider collaborating with to initiate, accelerate or scale my proposed solution, and why they could help:
Global Virome Project (GVP) - GVP's goal of identifying all viral threats fits well with my solution aiming to improve early detection of viruses. Collaborating with their network of experts could help identify promising technologies for pilot testing and their experience scaling projects globally would help test and deploy solutions in high-risk areas.
Bill & Melinda Gates Foundation - As a leader in global health funding and innovation, partnering with them could help attract further funding opportunities to test and evaluate early detection technologies. Their ongoing work addressing health inequalities in developing regions could help target solution deployments.
Imperial College London - Their renowned expertise in pathogen surveillance, genomics and digital health would be hugely beneficial. Collaboration would allow leveraging their research capabilities to advance detection platform technologies and validate clinical utility.
McKinsey & Company - Their world-leading strategic consulting and experience implementing large-scale solutions across diverse stakeholders could help optimize organizational structures, partnerships and business models needed to scale globally in a sustainable way.