136 The use and efficacy of ffr-ct – a real-world multi-centre audit of clinical data and cost-analysis

Introduction

UK and US Guidelines now recommend HeartFlow FFR-CT (Fractional Flow Reserve-Computed Tomography) in stable chest pain (CP), but its clinical effectiveness and costs in real-world practice are not known. Our objectives were to audit HeartFlow FFR-CT clinical practice against England’s National Institute for Health and Care Excellence (NICE) guidelines, and to assess its diagnostic accuracy and cost.

Methods

A multi-centre audit was undertaken covering the three-year period when FFR-CT was centrally funded in England. For coronary computed tomography angiograms (CCTAs) submitted for FFR-CT, centres provided data on symptoms, CCTA and FFR-CT findings, downstream testing, and further management. Audit standards included the use of FFR-CT only in patients with stable CP and stenosis of uncertain functional significance (50–69%). Diagnostic accuracy was evaluated against invasive coronary angiography and FFR, where performed. Follow-up for non-fatal myocardial infarction and all-cause mortality was performed. The real-world cost of an FFR-CT strategy was assessed.

Results

Twelve centres submitted 2298 CCTAs for FFR-CT analysis. Stable chest pain was the main symptom in 77% of patients, with CCTA stenosis of 50–69% in 41%. FFR-CT ≤0.80 was found in 996 (47%), resulting in revascularisation in 348 (35%). Compared to invasive FFR, the patient-level sensitivity, specificity, positive and negative predictive values of FFR-CT were 90%, 24%, 49%, and 76%, respectively. A total of 46 events occurred over a mean follow-up of 17 months, and FFR-CT with a cut-off of 0.80 was not predictive. Economic modelling demonstrated that the FFR-CT strategy cost £3913 per patient, compared with an average cost of £2148 for non-FFR-CT pathways.

Conclusion

In clinical practice, NICE recommendations for the use of FFR-CT were met in only 77% of cases for symptoms and 41% for stenosis. FFR-CT showed low specificity, making its use more expensive than strategies using other functional tests. FFR-CT was not predictive of cardiac events.

Conflict of Interest

None