Mar 5, 2024
Matt Hollingsworth, Co-Founder of Carta Healthcare, has identified the need to change how hospitals complete clinical registries, which require information from clinical notes. Currently, nurses manually fill out the forms primarily based on their understanding of the notes from the doctors. Using AI to read through notes, extract relevant information, and enter the data into the clinical registry form saves nurses 95% of their time, improves accuracy, and highlights actionable data. This approach allows the abstractors to spend more time on complex cases and uncover overlooked medical conditions.
Matt explains, "What we do is collect data for these things called clinical registries, which are required quality submissions that hospitals have to do. And we bring AI to the table so that we can make it a not-so-manual process. Historically, the way these forms have been completed has been with just nurse time. Also, I don't think there's a word in the English language to quite correctly describe form in this particular scenario. Just to give a concrete example, if a PCI, so a stent installation via cath is performed, it takes a nurse about an hour and a half to fill out the documentation required to do the submission to the registry. It takes a really long time."
"The problem is that most of it is in the notes. So one of the questions in a registry was the patient hypertensive upon admission. And the way you figure that out is you look, in most cases, in the H&P note and read through it. It'll say somewhere in there that there's either a history of hypertension or they took a blood pressure, and now they have it, etc. And so, the nurse, to paint a picture about what this looks like here, what the nurse is doing is they have the EHR open on one screen, a form to fill out on the other, and they're reading through notes, reaching conclusions, and then filling out the form. So, most of the data needed for this whole process is in the clinical notes, not in some structured form."
"Yes, so this is how we define quality in the whole shift to value-based care. A necessary component is that you measure quality in it. And this is just how the sausage is made for doing that. The way you measure something is you collect data for it. These registries are there to collect data, and in many cases, they are procedure-based. Sometimes, it's condition-based, like there's a COVID registry or what have you. Sometimes, it's a location-based thing like this patient was admitted to this unit like a CVICU or an acute care floor."
#CartaHealth #GenAI #HealthSystems #2024MedicalPredictions #HealthTech #MedTech
