CodMed is an AI-assisted medical coding tool that helps hospital physicians pre-code medical procedures and diagnoses by suggesting relevant CCAM and CIM-10 codes from dictated or written reports. The platform uses semantic search powered by OpenAI's API, and includes an integrated explorer to navigate both coding nomenclatures and refine selections. It is currently in development, with data hosted on French servers (OVHcloud) and processed in compliance with GDPR.
Funding
Funding not disclosed
Founders
Product
Problem
In hospitals, physicians must pre-code medical procedures and diagnoses so that the Department of Medical Information (DIM) can perform final coding. This added administrative burden is time-consuming and eats into time spent with patients, often leads to imprecise codes due to the complexity of the nomenclatures, and requires the DIM to send follow-up requests that delay the final coding process.
Solution
CodMed is an AI-assisted search tool for CCAM and CIM-10 codes designed to make pre-coding as fluid as possible. From a dictated or written medical report, CodMed extracts relevant medical procedures and diagnoses, then uses AI-powered semantic search to suggest the most pertinent codes. The tool includes an integrated explorer that lets users navigate the CCAM and CIM-10 tree structures to refine their choices. By centralizing the process in a single interface, CodMed accelerates pre-coding, improves accuracy, and reduces the need for corrections from the DIM.
Target Audience
The primary users are hospital physicians who perform pre-coding of medical acts and diagnoses, as well as hospital Departments of Medical Information (DIM) that rely on accurate pre-coded data to expedite final coding.
Features
- AI-powered semantic search that identifies medical procedures and diagnoses from dictated or written reports and suggests relevant CCAM and CIM-10 codes
- Integrated explorer for navigating the CCAM and CIM-10 hierarchical nomenclatures to fine-tune code selection
- Speech-to-text transcription of audio content via OpenAI's API
- Data filtering to remove personally identifiable information before storage
- End-to-end encryption, with data in transit secured via HTTPS and data at rest protected by storage volume encryption
- Hosting on OVHcloud servers located in France, ensuring data sovereignty and compliance with GDPR