Creating a quality measure for diagnosis of lung cancer

DESCRIPTION

Lung cancer is the most common cause of cancer-related death in the US. Many patients have symptoms for several months before they are diagnosed, resulting in late-stage disease where treatment options are more limited, and patient outcomes are far from ideal.

While we believe there are gaps and potential delays in the diagnostic pathway from first feeling unwell, to being diagnosed, there is not currently much research in the US which has pin-pointed where these gaps/delays in care are. Until we can measure the diagnostic pathway, and find these ‘pain points’, it is difficult to improve care.

This project aims to develop and test a clinical quality measure for lung cancer diagnosis using electronic medical record (EMR) data from UW Medicine. The quality measure will incorporate three metrics to lung cancer diagnosis. 1) Time intervals within the diagnostic pathway, including from patients first reporting symptoms, to their initial clinical presentation, to definitive histological diagnosis; 2) measurement of patient centered outcomes of testing, and 3) efficiency of the testing process.

We have identified a cohort of individuals who were diagnosed with lung cancer over a ten year period, and will use both coded data (e.g. diagnostic codes, procedures codes), as well as natural language processing of free text in clinical notes, to identify symptoms, tests, referrals, and populate a timeline of diagnosis and efficiencies of testing.