Tom Sanderson
About Tom Sanderson
Tom Sanderson is a Senior Contributor at Houston Methodist in Texas, managing a team of 49 people and overseeing coding for 1,600 clinicians. He tested an AI-powered medical coding tool from Epic Systems and believes AI can assist with routine procedures while complex cases will still need human expertise.
Tom Sanderson Senior Contributor Experience at Houston Methodist
Tom Sanderson holds the position of Senior Contributor at Houston Methodist in Texas. With a dedicated team of 49 individuals under his leadership, Sanderson manages the coding process for 1,600 clinicians. His responsibilities include processing between 70 to 250 claims per day, depending on their complexity. His role is pivotal in ensuring efficient and accurate coding in a high-demand healthcare environment.
Tom Sanderson's Involvement in AI-Powered Medical Coding Tools
Tom Sanderson has played a significant role in testing an AI-powered medical coding tool developed by Epic Systems, based on GPT-4. As one of the first dozen individuals to evaluate the coding assistant prototype, he has been instrumental in providing feedback. The tool ingests and summarizes clinician notes to suggest diagnosis and procedure codes. This system is designed to autosuggest codes, particularly for minor procedures and simple surgeries, while linking back to the information that led to each recommendation to address potential inaccuracies.
Feedback on AI Medical Coding Tool by Tom Sanderson
During the testing phase, Tom Sanderson provided valuable insights on the AI tool's performance. He noted that the AI correctly coded a simple hernia repair surgery but failed to accurately code a general office visit. Furthermore, he pointed out that the AI did not register missing information that human coders would typically notice. Sanderson's detailed feedback highlights areas for improvement in the tool's development, ensuring it can better assist human coders in the future.
Tom Sanderson's Vision on AI in Medical Coding
Tom Sanderson is enthusiastic about the potential of artificial intelligence to enhance the medical coding process. He believes that AI can autosuggest codes for routine procedures such as X-rays, EKGs, knee replacements, hernia repairs, and tonsillectomies. However, he emphasizes that complex cases, such as coding for transplant patients, genetic abnormalities, and advanced surgeries, will continue to require human expertise. His vision involves integrating AI tools with human oversight to maintain high coding accuracy and efficiency.