Difference-in-Differences (DID) analysis quantifies the difference in the change in outcomes over time between two groups (treatment and control). In certain situations, it may not be possible to have a control population such as when a nationwide policy is adopted.
We performed a DID analysis using SEER data to identify trends of the proportion of patients with metastatic disease at the time of diagnoses across 4 periods from 2007 to 2016. Period 1 (P1) and period 2 (P2) occurred before the publication of the National Lung Cancer Screening Trial (NLST) to establish baseline trends (2007-2009 and 2010-2011 respectively). Period 3 (P3) and period 4 (P4) were after the publication of NLST (2012-2014) and LDCT implementation (2015-2016) respectively.
The proposed DID design (DID design.pdf) can serve as a reference for DID analysis in cases when there is a single group available. Please note that, propensity score matching is recommended when using this approach. The SAS code is available in the repository, however the data could not be shared in accordance with the SEER Research Data User Agreement. Feel free to reach out at brihh.ansari@gmail.com for questions related to the code.