Appropriateness of CT Pulmonary Angiography Requests at Croydon University Hospital: An Audit

Background CT pulmonary angiography (CTPA) is widely used in the evaluation of patients with suspected pulmonary embolism (PE); however, inappropriate requests can expose patients to unnecessary radiation and iodinated contrast, delay imaging for those with confirmed PE, and increase demand on radiology services. National guidance (National Institute for Health and Care Excellence (NICE) NG158) recommends applying the two-level Wells score and D-dimer testing to determine pretest probability before requesting CTPA. This two-cycle audit evaluated adherence to these standards at Croydon University Hospital and examined whether documentation and diagnostic pathway compliance improved over time. Methods A retrospective two-cycle audit was conducted. Cycle 1 reviewed all CTPAs requested for suspected PE between October 1 and 31, 2024. Cycle 2 repeated the audit in September 2025 using the same methodology and comparison standards. Data were extracted from the electronic health record and radiology system, including Wells score documentation, D-dimer testing, chest X-ray (CXR) findings, anticoagulation status, and CTPA outcomes. Audit standards were based on NICE NG158 and Royal College of Radiologists expectations for CTPA diagnostic yield (15-35%). Following Cycle 1, the Trust introduced mandatory Wells score and D-dimer fields in the Cerner CTPA request form and increased educational messaging regarding appropriate imaging. Results Cycle 1 included 58 patients, with 11 positive CTPAs (diagnostic yield, 19%). Wells score documentation was low, with only four documented cases (6.9%), and D-dimer testing was recorded in 45 requests (77%). In Cycle 2, 96 patients were included, with 16 positive CTPAs (diagnostic yield, 16.7%). Documentation improved markedly: Wells score recording increased to 84 cases (87.5%), and D-dimer documentation was present in 86 cases (89.6%). Among Cycle 2 patients, PE was detected in 14 of 61 patients (23%) with Wells scores greater than 4 and in two of 23 patients (8.7%) with Wells scores of 4 or less. Alternative diagnoses were identified on CXR in 25 cases (26%). Patterns of anticoagulation were similar between cycles. Conclusions Across two audit cycles, CTPA diagnostic yield remained stable and within the nationally expected range. However, documentation of the Wells score and D-dimer improved substantially by September 2025, reflecting better adherence to the recommended diagnostic pathway for suspected PE. These findings demonstrate the effectiveness of system-level changes, including mandatory request-form fields and improved clinician awareness. Continued monitoring is recommended to ensure sustained compliance and to support appropriate imaging utilization.