As multicancer early detection tests move from laboratory promise toward population-scale deployment, a critical question has gone largely unasked: does rolling out mass screening strain the same diagnostic infrastructure that symptomatic patients depend on? New evidence from England's NHS-Galleri trial suggests the answer may be yes — and the implications deserve careful scrutiny before any national rollout.
This cross-sectional, difference-in-differences analysis examined all 21 cancer alliance regions in England across a roughly three-and-a-half-year window spanning April 2021 to September 2024. Eight regions participated in the NHS-Galleri trial of a cell-free DNA-based multicancer early detection (MCED) test; thirteen did not. The primary outcome was the proportion of patients referred for suspected cancer evaluation who waited longer than 28 days to reach diagnostic resolution — a surrogate for system-level capacity stress. The analysis focused on three cancer types not subject to routine screening: head and neck, lung, and upper gastrointestinal cancers, representing nearly 1.88 million referrals in total. The core finding is that regions actively running the MCED trial showed measurably higher diagnostic delay rates compared with control regions, suggesting that absorbing large-scale screening volumes displaced timely diagnostic capacity for symptomatic patients.
This finding sits at an underappreciated intersection between screening science and health systems research. Most clinical trials evaluating screening interventions are designed to detect cancer earlier in enrolled participants — not to monitor what happens to the patients who never enrolled. These "spillover effects" rarely appear in trial protocols, yet they can determine net population benefit. A test that catches early cancers in screened individuals while simultaneously delaying diagnosis in symptomatic patients could, in theory, produce no net mortality improvement or even harm. The study's cross-sectional, ecological design cannot confirm causality — confounding by regional differences in baseline capacity, staffing, or referral behavior remains plausible. Still, the magnitude and directionality of the signal across nearly two million referrals makes this more than a statistical artifact. For health systems contemplating MCED adoption, capacity modeling must become a prerequisite, not an afterthought.