I’ve spent more than 30 years in nursing, and in recent years my work has shifted to something most people never think about: clinical data abstraction. My job is to review the medical record and extract the pieces of information that clinical registries rely on for research, quality reporting, regulatory requirements, and everyday decision-making. It’s specialized work that depends on clinical judgment, a deep understanding of clinical documentation in the EHR, and a commitment to getting every detail right.
Clinical data abstraction-the meticulous process of reviewing medical records to extract key information for research, quality reporting, and regulatory compliance-is a complex undertaking. It demands clinical judgment, a deep understanding of electronic health records (ehrs), and unwavering attention to detail. Information is often scattered across multiple systems and documented inconsistently by different clinicians, making the task both challenging and time-consuming.
As the initial reactions to the OpenAI and Anthropic healthcare announcements of two weeks ago have started to settle down, a consensus seems to be emerging: While both products target consumers through integrations into medical records, OpenAI seems to be doubling down more on the consumer market (partially evidenced through their acquisition of Torch) while Anthropic's seems to have a more pointed focus on enterprise use-cases. Following their showcase, we saw announcements from multiple vendors (Elation and Carta Healthcare) highlighting that they’re being "powered" by Claude. It’s likely both companies will continue to add more support for healthcare use-cases through additional integrations and partnerships. While OpenAI may be more consumer-focused on launch, we wouldn’t be surprised if they take on enterprise use-cases more directly in the coming months.
Health systems using Carta Healthcare reduce abstraction time by up to 66%, lower abstraction costs by 50% or more, and consistently achieve 98%–99% Inter-Rater Reliability (IRR), the highest standard for data quality. By enhancing the identification of clinically relevant information across complex medical records, Claude models help Carta Healthcare deliver consistent, high-quality outputs across large case volumes while keeping clinical judgment at the center of every workflow.
“The launch of Claude for Healthcare demonstrates how quickly clinical-language models are advancing, and what becomes possible when paired with deep domain expertise,” said Brent Dover, CEO at Carta Healthcare. “Anthropic’s investment in healthcare-specific reliability helps us iterate faster and deliver more consistent outputs for hospitals. This is the promise of hybrid intelligence: technology amplifying clinical insight to improve performance and reduce burden across health systems.”
Carta Healthcare, the award-winning leader in enterprise clinical data management, underscored the role of hybrid intelligence in advancing responsible AI for healthcare as Anthropic unveiled Claude for Healthcare during the 44th Annual J.P. Morgan Healthcare Conference.
Hybrid Intelligence Delivers Consistent Value in Clinical Workflows.
Carta Healthcare uses Claude models—securely deployed through Amazon Bedrock—to support its hybrid intelligence approach, which combines advanced AI with expert clinicians in a human-in-the-loop model. This combination reduces the time, cost, and burden of manual abstraction and improves the quality of the data that underpins patient safety, quality measurement, and performance improvement efforts.
Earlier this month, OpenAI launched ChatGPT Health, described as “a dedicated experience that securely brings your health information and ChatGPT’s intelligence together, to help you feel more informed, prepared, and confident navigating your health.”
Needless to say, this has prompted plenty of discussion. Here’s what several health industry experts had to say on the subject.
The company achieved 100% customer retention, signed 25 new contracts, and welcomed 10 hospitals that switched to Carta Healthcare from another clinical data abstraction solution provider, all while delivering 2X revenue growth and expanding its team with 144 new hires.
PSQH reached out to experts throughout healthcare to get their predictions for what will happen in patient safety and healthcare quality in 2026. We received so many predictions this time around that we’re breaking it up into two parts. Here’s Part 2 of what they had to say.
"Health information management [HIM] is evolving quickly as organizations push for greater accuracy and stronger safeguards around patient data. New approaches to data exchange and abstraction are helping HIM teams deliver timelier information, while advances in privacy and security are keeping pace with regulatory demands. The trend is toward practical innovations that make the work more reliable without adding extra complexity for professionals on the front lines." - Greg Miller
