Pulmonary embolism (PE) is a common and potentially life-threatening condition encountered in acute medicine. Clinical prediction rules (CPRs), including PERC, Wells, Revised Geneva, YEARS and LEGEND, have been developed to support rational use of computed tomography pulmonary angiography (CTPA). This study evaluated whether retrospective co-application of Revised Geneva, YEARS or LEGEND alongside an existing PERC/Wells-based diagnostic pathway could reduce CTPA utilisation in a Same Day Emergency Care (SDEC) population.
We conducted a retrospective cohort study of all patients who had CTPA for suspected PE in a UK SDEC unit between January and June 2024. CPRs were applied retrospectively using documented clinical data. The primary outcomes were the potential reduction in CTPA utilisation and the proportion of scans that would have contained PE.
Of 433 CTPAs identified, 362 patients met inclusion criteria. The prevalence of PE among those scanned was 16.6%. Retrospective co-application of Revised Geneva, YEARS and LEGEND would have reduced CTPA utilisation by 5.5%, 29.0% and 39.2% respectively. No PEs were identified among scans potentially avoided using Revised Geneva, whereas PE was present in 4.8% and 5.6% of avoided scans using YEARS and LEGEND. Missed events occurred in a clinically stable ambulatory cohort and were not associated with radiological evidence of right ventricular strain; however, the clinical significance of these findings remains uncertain.
In this selected SDEC population, layering additional CPRs onto an existing diagnostic pathway may have the potential to reduce CTPA utilisation, although strategies achieving larger reductions were associated with missed diagnoses. These findings should be regarded as hypothesis-generating and require prospective validation before informing changes to clinical practice.