The Problem
Lower back pain (LBP) is one of the most prevalent health complaints amongst adults, with up to 84% of individuals reporting significant back pain at some point in their life. One of the challenges in triaging patients with low back pain is to make sure that patients go the ‘right’ provider, whether that be general practitioners, spine health practitioners such as chiropractors, invasive proceduralists or spine surgeons.
Our Solution
We hope to implement a machine learning predictive model that will allow patients to be triaged to the most appropriate practitioner. This will enhance patient and practitioner satisfaction and improve health efficiencies – two important missions of the Duke Health System.
Anticipated Impact
The impact on the health system will be profound. There are approximately 70,000 unique patient encounters in the Duke system for low back pain in one given year. For a given surgeon, only 10-20% of the patients that they see ultimately receiving surgery. This leaves a large population that would be best served if they went initially to another provider. Similarly, there are patients that go to non-operative practitioners but would be best served if they went to a surgeon to start.
