Potential spillover effects on diagnostic delay for cancer during the NHS-Galleri trial: a quasi-experimental difference-in-differences study

Sean Mann; Pedro Nascimento de Lima; Joshua Eagan; Agne Ulyte; Beth Ann Griffin

Abstract

Objective To examine whether regions participating in the NHS-Galleri trial of a cell-free DNA-based multi-cancer early detection (MCED) test experienced changes in cancer diagnostic delay rates that could indicate the presence of spillover due to constraints on care delivery. Design Quasi-experimental difference-in-differences analysis. Setting England, April 2021-September 2024. Participants All 21 cancer alliance regions in England, 8 of which participated in the trial. Main outcomes Cancer diagnostic delay rates (primary outcome) and referral rates (secondary outcome). Primary analysis focused on the group of three cancer types (head and neck, lung, upper gastrointestinal) that are prespecified in the trial protocol, have higher signal detection for cell-free DNA-based testing, and were not subject to routine screening. A difference-in-differences analysis was used to compare outcomes in regions participating in NHS-Galleri to regions that did not participate, before and after the start of the trial. Results Diagnostic delay rates increased in the first six months of the trial for the primary group of three cancer types in participating regions compared to non-participating regions (adjusted difference-in-differences 3.4 percentage points, 95% CI: 1.9-5.0, p<0.001). These differences persisted from 6 to 12 months after trial start and receded over the following two years. No significant change in referral rates for this group of cancer types was observed. Conclusions Participation in the NHS-Galleri trial was associated with increased diagnostic delay rates for the primary group of high-detection cancer types. The trial may have led to higher demand for limited diagnostic resources (e.g. imaging, endoscopy) required for follow-up of suspected cancer arising from positive MCED test results or usual care processes. If patients in the control group faced delayed diagnosis as a result, this would represent spillover during the first year of the trial and could lead to overestimation of MCED benefit.