Clinical features of lung cancer prior to diagnosis: Case control study

Project summary

Some patients who are diagnosed with lung cancer have symptoms that take them to see their primary care clinician long before their time of eventual diagnosis. We want to see whether we can find out which symptoms are most likely to indicate that lung cancer is beginning to develop.

To do this, we are identifying patients who have been diagnosed with lung cancer, and control patients who do not have lung cancer from the UW Medicine health care system. We will then extract all the information that appears in their clinical electronic medical record notes, including using a new technique which can pick out words from the clinician’s detailed notes. We will compare which symptoms seem to differ between cases and controls, and how far back before someone’s diagnosis this can be seen.

If we are successful, we hope to be able to propose new and improved ways for patients, caregivers and primary care clinicians to identify who may have lung cancer. This could lead to earlier investigation and potentially more timely treatment.