Out-of-the-box large language models (LLMs) like GPT-4 and Gemini Pro have sparked enormous interest in automating healthcare tasks, including medical coding. But real-world performance tells a different story.
In a study by Mount Sinai, published in NEJM AI, even the most advanced LLMs struggled with accuracy, hitting just 34% exact matches for ICD-10-CM codes and often generating imprecise or even fabricated results. This is a stark reminder that scale alone doesn’t equal precision, especially when dealing with the clinical and financial consequences of coding in healthcare.
At IMO Health, we’ve spent decades building a deep clinical knowledge layer that addresses this exact challenge. By combining LLMs with rich terminology, curated mappings, and proven AI techniques, we transform raw model output into explainable, trustworthy, and high-performing medical coding.
Want to know how it works in practice?
Download our latest guide, authored by IMO Health’s Data Science and Chief AI Architect, Jingqi Wang, PhD: Optimizing LLMs for precise analytical output: The IMO Health approach.
The following excerpt highlights one of the most powerful differentiators in our approach: transparency.
Explainable and trustworthy code selections
One of the key advantages of IMO Health’s knowledge layer is its ability to explain why certain medical codes are chosen and why they are more suitable compared to other similar codes. By prompting the LLM with our mapping and terminology resources, the generated explanations are more clinically logical, with fewer hallucinations and false statements. This makes the results more acceptable and trustworthy to medical coders when they review the output.
The explainable nature of code selections is particularly valuable when there is ambiguity or multiple potential codes for a given medical condition or procedure. By providing clear and clinically sound reasoning for the chosen codes, the system instills confidence in medical coders and facilitates a more efficient review process.
This transparency also enables coders to quickly identify and address any potential discrepancies or uncommon cases, further improving the overall accuracy and reliability of the coding output (below). 
Explore the full guide to see how IMO Health enhances AI performance through domain-specific data, structured terminology, and advanced techniques like retrieval-augmented generation (RAG) and fine-tuning.
