Abstract
USC) in an outpatient context in NHS Greater Glasgow and Clyde.This is a stepped-wedge, cluster-randomised study consisting of a retrospective technical evaluation and prospective clinical effectiveness study alongside the assessment of acceptability via qualitative work and a cost-effectiveness evaluation.The primary objective is to assess the clinical effectiveness of qXR software to prioritise patients with USC on CXR for follow-up computed tomography (CT) across three hospital clusters in NHS Greater Glasgow and Clyde.Secondary objectives report against time to treatment, mortality, stage at diagnosis, health economics and the acceptability of the intervention.The study protocol can be viewed at https://bmjopen.bmj.com/content/14/9/e081062#ref-2.87,893 prospectively collected CXRs were included in this evaluation between October 2023 and October 2024.qXR flagged 9-14% of CXRs across the three clusters with 6-8% of flagged CXRs subsequently deemed to be positive for USC by a radiologist.Median time to CXR report for USC images increased over all sectors (+2.10hrs, p = 0.01) with significant variation between sectors.Reporting time for all CXRs increased over the evaluation period reflecting challenges to radiology capacity.Time to cancer treatment was reduced across all sectors (-20 days, p = 0.01).This evaluation demonstrates mixed results following the implementation of an AI-enabled pathway for USC detection, with increased time to X-ray report but decreased time to cancer treatment.X-ray reporting times were impacted by a service-wide deterioration in reporting capacity.Further analysis will be conducted to adjust for X-ray reporting and treatment trends in each sector.