Pandemic road safety: how COVID-19 policies shaped crash severity in England

Background COVID-19 pandemic restrictions changed travel volume, road-user composition and the traffic environment, but evidence on crash severity across distinct policy phases remains limited. We examined fatal or serious outcomes across 14 England policy periods. Methods Official police-reported collision, vehicle and casualty records for 2019–2022 were linked at crash level. The binary outcome was fatal or serious versus slight crash severity. We fitted a complete-sample, unweighted logistic regression adjusted for vehicle, road, environmental, temporal and police-force factors, using calendar-week cluster-robust CIs. Marginal standardisation estimated adjusted probabilities and risk differences; non-linear weekly trends were assessed in sensitivity analysis. Results Among 384 472 crashes, 81 931 (21.3%) were fatal or serious. Adjusted probability of a fatal or serious crash was 20.5% (95% CI 20.2% to 20.7%) before the pandemic, 22.0% (21.0% to 23.0%) during the first national lockdown and 22.6% (22.3% to 22.9%) during postrestriction follow-up. Absolute differences were +1.52 and +2.08 percentage points, respectively. Both associations remained positive after nonlinear weekly adjustment, but only the first-lockdown association remained statistically significant with police force by year controls. Motorcycle involvement (OR 3.17, 95% CI 3.07 to 3.28) and single-vehicle crashes (OR 2.45, 95% CI 2.39 to 2.51) had larger associations. Discussion and conclusion Policy-period associations were heterogeneous and modest in absolute magnitude. Vehicle involvement and crash context had stronger associations. Findings are observational and do not establish causal policy effects.