Overview
Real-world evidence offers promising opportunities for comparative effectiveness research, however challenges arise when patient characteristics and treatment choices are influenced by factors not always captured in the electronic health record. This study explored a new statistical approach, called regression discontinuity in time (RDiT), to better estimate the real-world benefits of immunotherapy (pembrolizumab) compared to chemotherapy (docetaxel) for patients with advanced non-small cell lung cancer (aNSCLC) who had already received prior treatment.
Using data from nearly 2,000 patients in the Flatiron Health Research Database, researchers compared survival outcomes for those treated with pembrolizumab versus docetaxel between 2011 and 2023. The study found that patients receiving pembrolizumab had better survival than those on docetaxel, with results from the RDiT method closely matching those seen in clinical trials. Importantly, the RDiT approach provided more conservative—and potentially more accurate—estimates of treatment benefit than traditional statistical methods.
Why this matters
Understanding how new cancer treatments perform outside of clinical trials is essential for patients, doctors, and policymakers. This research demonstrates that advanced statistical methods like RDiT can help overcome some of the challenges of real-world data, providing more reliable real-world evidence on which treatments work best for patients. By validating these methods against clinical trial results, the study supports the use of real-world evidence to inform treatment decisions and improve care for people with advanced lung cancer.