Case Study: Carta Healthcare cuts clinical data processing time by 66% with Claude

Carta Healthcare cuts clinical data processing time by 66% with Claude

Carta Healthcare began with a challenge experienced at Stanford Children’s Hospital: the slow, manual process of clinical data abstraction was hindering progress. This painstaking work required trained abstractors to spend hours combing through complex medical records to extract answers for critical registries, a process that was both expensive and difficult to sustain at a large scale.

Using the Claude platform via Amazon Bedrock, Carta Healthcare implemented an AI engine for its Lighthouse system. Claude semantically interpreted structured and unstructured medical data to find answers and provide justifications from patient records. This solution from Claude reduced data abstraction time by up to 66%, achieved 99% Inter-rater Reliability scores, and transformed clinicians' roles from manual data hunters into high-value validators, saving one health system thousands of hours of labor annually.

Carta Healthcare

Andrew Crowder

Vice President of Engineering and CISO