Truveta has launched a care insights solution that uses daily updated, de-identified electronic health record data and AI-powered natural-language analysis to help pharmaceutical and medical device companies track how treatments are adopted and used in real-world clinical care.
Understanding whether a treatment works in a clinical trial is only one part of modern healthcare. Once a therapy reaches the market, healthcare providers, researchers and life sciences companies face a new set of questions: Which patients receive it? Who benefits? Why do some patients stop treatment or switch therapies? And how quickly does clinical practice change when new evidence emerges?
Truveta is attempting to address that information gap with a newly launched care insights solution built around de-identified electronic health record data from more than 140 million patients in the United States.
The platform combines longitudinal clinical information with de-identified clinical notes and Truveta Intelligence, an AI-powered natural-language interface designed to help users investigate emerging treatment patterns. The company says the data is updated daily, potentially providing pharmaceutical and medical device companies with earlier visibility into changes in clinical practice than traditional billing and claims datasets.
The launch reflects a broader shift in enterprise AI applications within healthcare: moving beyond predictive models toward AI systems that can help organizations interrogate large, complex collections of real-world clinical information.
Moving beyond billing data
Claims and billing-derived datasets have long been used to study treatment utilization and healthcare trends. Their limitations, however, can make it difficult to understand the circumstances behind clinical decisions.
A claims record may show that a prescription was filled or a procedure occurred, but it often provides limited information about why a patient discontinued therapy, experienced side effects or faced difficulties accessing treatment.
Those details can instead be found in electronic health records and clinical notes.
Truveta’s care insights platform combines structured clinical data with de-identified contextual information to examine treatment initiation, adherence, switching and discontinuation. Users can also investigate factors that may influence treatment decisions, including documented side effects, access barriers and provider decisions.
The company is using Truveta Intelligence as the AI layer for investigating those datasets through natural-language questions.
For enterprise healthcare teams, this approach could reduce the technical barrier to analyzing complex clinical datasets. Rather than requiring every question to begin with a traditional data query or a lengthy analytics workflow, users can ask questions about emerging treatment patterns through a conversational interface.
That represents a growing use case for generative AI technologies in enterprise healthcare: connecting large language model-style interfaces with specialized, governed data environments.
Early treatment adoption is only part of the story
Truveta highlighted research involving the rollout of an oral Wegovy formulation as an example of why timely clinical data may matter.
According to the company, researchers found that more than one-third of early users were new to GLP-1 medications rather than switching from an existing treatment. A later analysis found that the pattern persisted.
The larger challenge, however, is understanding what happens after treatment begins.
Truveta cited a study published in JAMA Network Open involving more than 125,000 adults with overweight or obesity. Researchers and collaborators found that approximately 65% of patients without type 2 diabetes discontinued GLP-1 therapy within one year, compared with 47% of patients with type 2 diabetes.
By combining longitudinal clinical data with de-identified clinical notes, researchers examined documented reasons associated with discontinuation, including side effects, cost, medication availability and provider decisions.
The example illustrates a key difference between raw utilization data and AI-enabled clinical analysis. Knowing that a patient stopped taking a medication is useful; understanding the context behind that decision can be more valuable for researchers and medical affairs teams attempting to identify gaps in care.
AI enters the real-world evidence workflow
Real-world evidence has become an increasingly important part of pharmaceutical and medical device research, particularly as companies seek to understand how therapies perform outside controlled clinical trials.
AI platforms could expand the speed and accessibility of that work by helping researchers analyze larger volumes of structured and unstructured information.
Major technology companies, including Microsoft, Google and Amazon, are investing in healthcare AI infrastructure, cloud platforms and generative AI capabilities. At the same time, specialized healthcare technology companies are developing systems designed around clinical workflows and real-world evidence.
Truveta’s approach focuses on the combination of large-scale EHR data, clinical context and AI-powered investigation.
The company also pointed to research on changes in clinical practice following updated medical evidence.
According to Truveta researchers, 95% of patients undergoing IUD insertion in 2025 had no documented pain management care, despite updated CDC recommendations. The finding highlights the potential gap between clinical guidance and its implementation in everyday practice.
In contrast, another peer-reviewed study using Truveta Data found that clinical behavior can sometimes shift quickly. Following new clinical trial findings and an FDA labeling update involving the HeartMate 3 cardiac device, aspirin use reportedly declined from approximately 90% of patients in 2023 to 53% in 2025.
Researchers found no increase in adjusted risk of major complications, stroke or blood clots among patients who did not receive aspirin, according to the study cited by the company.
From AI analysis to healthcare action
The strategic value of care insights will depend on whether faster access to treatment patterns leads to measurable improvements in healthcare decisions.
For medical affairs teams, earlier visibility could help identify unmet evidence needs, gaps in patient access and areas requiring scientific education. The platform could also support research collaborations with healthcare systems investigating whether changes in clinical practice are improving outcomes.
The broader technology trend is clear: enterprise AI is increasingly being deployed not simply to automate administrative tasks, but to help organizations make sense of complex, rapidly changing data environments.
In healthcare, that transition comes with significant requirements around privacy, governance and responsible use. AI-generated insights must be validated, and observational real-world data cannot automatically establish causation.
Still, Truveta’s launch highlights how AI automation platforms and natural-language interfaces are beginning to reshape access to real-world clinical evidence.
For pharmaceutical and medical device companies, the ability to see treatment patterns while they are still emerging could change how quickly organizations identify problems—and how quickly they respond.
Market Landscape
AI-powered healthcare analytics is expanding as pharmaceutical companies, health systems and medical device manufacturers seek faster ways to generate real-world evidence.
The market increasingly combines electronic health record data, cloud infrastructure, machine learning and generative AI interfaces. A major technology challenge is connecting structured clinical records with unstructured information contained in clinical notes while maintaining privacy, governance and appropriate controls.
Truveta’s care insights platform represents the growing use of enterprise AI applications to shorten the time between changes in real-world clinical care and the analysis of those changes.
Top Insights
- Daily clinical data could improve visibility: Truveta’s platform is designed to identify treatment patterns as they emerge rather than relying exclusively on delayed claims datasets.
- Clinical notes add critical context: De-identified unstructured information can help explain treatment discontinuation, access barriers, side effects and provider decisions.
- Natural-language AI simplifies analysis: Truveta Intelligence enables users to investigate complex clinical datasets by asking questions rather than relying solely on traditional queries.
- Real-world evidence can reveal care gaps: Timely data may help medical affairs teams identify differences between clinical guidance and actual patient experiences.
- AI insights still require validation: Observational healthcare data can identify important patterns, but further research is needed before drawing causal conclusions.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
