The identification of abnormal narrowing of a blood vessel from non-invasive coronary angiography remains an exciting goal. While Computed Tomography (CT) has established a clear role in patients it retains some limitations. Primarily, alterations in blood flow leading to a restriction in blood supply to tissues cannot be accurately predicted from the anatomical assessment of an abnormal narrowing in a blood vessel - a limitation of both invasive, and computed tomographic, angiography. This is important, as there is now a raft of evidence demonstrating that revascularisation should only be considered in the presence of a restriction in blood supply to tissues.

The gold standard method of identifying a restriction in blood supply to tissues is measurement of the fractional flow reserve (FFR), a dimensionless value generated as a ratio between the circulatory pressure before and after a narrowing of a blood vessel.

At present, patients with indeterminately significant abnormal narrowing in a blood vessel identified using CT must go on to have a further test for a restriction in blood supply to tissues, prior to any decision about revascularisation being made.

As part of our governance commitment to maintaining our reporting standards, we undertake ongoing audit of our cardiac CT practice. The investigators recently identified a group of patients who have undergone FFR measurement as part of routine clinical care, and compared these to our clinical reports, to ensure that appropriate recommendations were being made. It appeared that the ability to predict important abnormal narrowing in a blood vessel was reasonable (~70%) and hence we propose the formulation of a research question to explore this further.

The primary objective of this study is to determine the accuracy of a visual assessment of the likely significance on blood flow of a narrowing of a blood vessel identified on cardiac CT, in comparison to invasive, fractional flow reserve.

The secondary objective is to evaluate features of an abnormal narrowing in a blood vessel that is likely to have a significant impact on blood flow.

The investigators' working hypothesis is that cardiac CT FFR is a useful predictor of abnormal narrowing in a blood vessel.

This is a retrospective study of diagnostic accuracy. We do not intend to undertake any intervention or require patient involvement at any stage of this research.

The study would utilise anonymised patient data, identified as part of the previous audit. FFR values are already known for each patient. The image dataset would be retrieved and matched to an FFR. At this stage all data, including the images, would be anonymised. The images would be reviewed independently by two expert readers who would record their impression of the likely haemodynamic significance of a particular stenosis. Two weeks later, to minimise bias, the readers would re-review the images to identify features of the stenoses known, or likely, to cause haemodynamic significance, including stenosis length, territory of the vessel, calcification.

The results would be directly compared to the FFR readings and the features would be subjected to multivariate analysis to identify any salient anatomical feature which may predict haemodynamic significance.

Inclusion Criteria:

• Patients with coronary stenoses which have been analysed with both CT and FFR

Exclusion Criteria:

• Under 18 years

Patients with coronary stenoses which have been analysed with both CT and FFR