Picture a world where healthcare research is an interconnected masterpiece of discovery and innovation. Imagine researchers from every corner of the globe, equipped with the finest tools and support, collaborating effortlessly across borders. The synergy ignites a cascade of breakthroughs, treatments, therapies, and drugs, all converging to address the most pressing medical challenges. In this world, the data flows with unparalleled clarity and precision, transforming into actionable insights that propel evidence-based decisions and drive patient care to new heights.
Instead, the world of healthcare research is marred by significant challenges: fragmented capacities, limited funding, and inequitable access to resources. Despite the growing volume of research, systemic inefficiencies hinder progress and innovation. Careframe.ai is redefining the siloed, friction-filled research model by seamlessly integrating disparate strands of fundamental approaches with its state-of-the-art platform.
Current state and challenges
Large studies often lack resources for analytical and statistical support, or they are inadequate for the size of the project. This creates a bottleneck in research productivity and quality. Careframe.ai addresses this gap by offering scalable technical support for quantitative research. By providing consistent and comprehensive assistance, Careframe.ai promotes equity in research capabilities that enable researchers and institutions across the globe to contribute effectively to the scientific community.
Careframe.ai CEO and founder Sanjeev Kumar explains, “Another problem affecting research and innovation in healthcare is that funds dedicated to ‘health services research and innovation’ account for only 0.04-1% of healthcare spending. This underinvestment stifles the development of new interventions and healthcare delivery models.”
In 2023, the Canadian Institutes of Health Research (CIHR) budget of $1.2 billion represented about 0.35% of Canada’s total health spending of $344 billion. Only about 0.041% of this spending is specifically allocated to health systems/services research. In the U.S., the National Institutes of Health (NIH) budget of $47.7 billion in 2023 accounts for approximately 1.06% of the country’s $4.5 trillion in health spending, with a focus primarily on biomedical research rather than health services research and innovation.
Careframe.ai addresses this challenge of inadequate funding and support for health services research by providing scalable support and improving data accessibility. Its platform is designed to facilitate the testing and implementation of novel approaches, thereby driving innovation in healthcare systems research.
Kumar adds, “In a study that reviewed submitted grant proposals, 24% were labeled as Interdisciplinary. The proportion of investigators who submitted only interdisciplinary proposals has increased by 9% over a 6 year period leading up to 2021. At the same time, the number of investigators submitting no interdisciplinary proposals decreased by the same percent value over the same period. This highlights the synergistic potential to contribute to momentum in solving complex healthcare challenges.”
Careframe.ai addresses this issue by enabling seamless collaboration across different research disciplines. Whether bridging basic biomedical sciences with social determinants of health or integrating diverse fields, Careframe.ai promotes a more comprehensive approach to tackling problems.
“Another reality is that only 50% to 60% of clinical decisions are based on high-quality evidence. This disconnect between research and practice can lead to suboptimal patient outcomes,” asserts Kumar. “Our platform accelerates the knowledge mobilization process and allows for careful synthesis and dissemination of research findings to bridge this gap.” By ensuring that these evidence-based practices are more readily integrated into clinical decision-making, Careframe.ai supports consistency and standardization.
The volume of published research is growing exponentially. Compounding the issue, Kumar notes, “Researchers often spend 42% of their time on administrative tasks rather than actual research.” This greatly impedes their ability to keep up and perform high-throughput research.
In addition, Kumar shares, “Cognitive biases were associated with diagnostic inaccuracies in 36.5%–77% of case scenarios per a systematic review.” One of the ways Careframe.ai tackles bias is by grounding responses in the objective of study protocols during the planning phase. In addition, it plans statistical tests for various subanalyses as part of the protocol planning process itself. Synthetic datasets grounded to the study protocol further enhance planning without waiting for real-world data, allowing for study protocol enhancements at various steps and ensuring standardized research outputs.
At a high level, the platform discretizes study protocols into fundamental building blocks that can be co-edited in an operator-in-loop manner with the latest frontier AI model. It is also working towards scaling the autonomous study protocols to adapt to the medical setting, it does this by using knowledge graphs to generate hypotheses which then can be independently executed. Users can organize the standardized results into organizational knowledge graphs that allow a point-in-time look into the significance of dependent and independent variables with respect to the study protocol. The platform simplifies study protocol planning, design, and execution, thereby enhancing access and reliability of studies on existing and real-time data.
Evolution of Careframe.ai
The genesis of Careframe.ai is deeply rooted in the personal experiences of its founder. Kumar’s tenure at Henry Ford Health System and the University of Michigan provided him with firsthand insights into the limitations of research capacity, the need for innovation, and the challenges of data accessibility. His experiences brought to light the urgent need for a solution like Careframe.ai that integrates sophisticated data processing pipelines with a nuanced understanding of healthcare research needs.
While there are competitors like large EMR providers and technology providers with healthcare integrations, Careframe.ai stands out for its niche focus on quantitative research on existing data using synthetic data pipelines coupled with deterministic exit points. By leveraging optimized agentic pathways to generate discrete sections of protocol, suggest data variables and generate synthetic datasets all grounded into the study objectives, it ensures that its pipelines scale with increasing aptitude and capacities of frontier large language models.
As LLMs increase in capacity, CareFrame’s study protocol designs improve as well. While there are risks associated with hallucination, it takes a layered approach to ensure that the outputs of pipelines are deterministic and handled by traditional programming approaches. While conducting study protocols one at a time is meaningful from an operator-in-loop perspective, it is increasingly becoming clear that the gap between the potential of quantitative research on existing data and our actions to harness that potential need to be better aligned.
As better network architectures and high quality pre-train datasets improvements in GenAI push the reasoning horizon further forward, it is becoming clear that quantitative research can benefit from these improvements. CareFrame benefits both from reliability gains in GenAI research and the increasing capacity of data centers to scale research with computers. Careframe.ai’s approach emphasizes a sophisticated, AI-driven, generative research model tailored to the medical space.
Careframe.ai builds upon existing practices from a new fundamental point of view. It intends to shine the light on GenAI’s vast potential in healthcare research, in particular on locked potential in existing data. By addressing the core challenges of fragmented research capacity, limited knowledge mobilization, and data accessibility issues in planning, Careframe.ai is paving the way for a more connected and innovative future in healthcare research.
