INTRA-OBSERVER VARIABILITY OF QUANTITATIVE CORONARY ANGIOGRAPHY MEASUREMENTS USING THE CARDIOVASCULAR MEASUREMENT SYSTEM AT GREEN LANE HOSPITAL ANGIOGRAPHIC CORE LABORATORY B. Webber,* T. West, J. Ormiston Cardiac Investigation Rooms, Green Lane Hospital,
Auckland, NZ. Quantitative coronary angiography (QCA) has become an important tool in
angiographic trials. Because of its objectivity, accuracy and reproducibility
in assessing coronary anatomy, QCA results are frequently used as primary or
secondary end points. QCA has a degree of variability, both inherent and
operator introduced which impacts upon the accuracy of the results and ability
to detect change. Core angiographic laboratories need to assess their own
contribution to this variability. Forty randomly selected projections from 27 Green Lane Hospital patients
who were entered into a randomised trial had QCA analysis of the same lesion on
three occasions; baseline, at one week and four weeks from baseline.
Projections were not pre-selected for suitability for factors that may
influence variability. Each film had a "zero" frame identified and each
projection had the calibration frame and the frame selected for QCA identified
and recorded. Identical frames per lesion were then used for calibration and
measurement over three periods. Strict QCA rules were adhered to, including
defining the beginning and end points for analysis as at major branch points
and within the centre of the coronary lumen. For each of the three analysis
periods the forty projections were randomised into a different order to
eliminate prior knowledge affecting the measurements. Results: (mm) Baseline One
week Four
weeks P
value MLD 1.036 1.011 0.979 0.161 Ref.
Diam 3.261 3.213 3.269 0.165 Diam
% 68.05 68.01 69.73 0.096 Obs.
Length 12.41 12.02 12.27 0.409 Accuracy and precision (mm) One
week Accuracy One
week Precision Four
weeks Accuracy Four
weeks Precision MLD -0.005 0.145 -0.034 0.158 Ref.
Diam -0.048 0.227 0.008 0.181 Diam
% -0.640 4.235 -1.023 4.905 Obs.
Length -0.390 1.969 -0.141 1.870 MLD = Minimum luminal diameter, Obs. = Obstruction,
Ref. = Reference, Diam = Diameter, In conclusion,
no significant difference was found between baseline, one week and four week
measurements of the same frame. The range of accuracy and precision values
compares favourably with other published data.
