Background Nailfold videocapillaroscopy (NVC) has been proposed as an outcome measure for systemic sclerosis [SSc] related microvascular disease.
Objectives To assess the reliability of both measure availability and measured value of 4 parameters extracted from NVC images: capillary density, capillary (apex) width, overall image grade (normal, early, active, late), and the presence of giant capillaries.
Methods 173 subjects were recruited in the study (101 with SSc, 22 with primary Raynaud's phenomenon and 50 controls). Panoramic NVC imaging was performed on all 10 digits using a 300x magnification microscope with green LED illumination for best capillary contrast.
Ten capillaroscopy experts (“raters”) from 7 centres assessed images using custom software. Raters were asked to: (1) assess overall image grade (“Normal”, “Early”, “Active”, “Late”, “Non-specific”, or “Ungradeable”), (2) mark the location of capillary apices (capillary density calculated from distal vessel locations), (3) categorically grade the distal vessels in terms of size and shape, and (4) measure the apical width of distal vessels using a click-and-drag tool.
Results The 10 raters analysed a median (range) of 129 (98–1641) images, returning 3401 evaluations of 1650 images, including a sub-sample analysed twice by the same rater to measure intra-rater reliability. Of these, 2731 evaluations (80%) satisfied the criteria to allow vessel density to be measured (at least 2 distal vessels identified).
The reliability (intra- and inter-rater) of both measure availability and measured value for the 4 parameters are shown in the table. Overall image grade was available in only around 45% of observations due to the exclusion of “non-specific” gradings from the analysis.
Conclusions Capillary density and (especially) apical width, parameters likely to be helpful in quantifying temporal change, were reliable measures in evaluable nailbeds. The reliability of measured values is generally high, but is conditional on measure availability. This must be accounted for when considering the applicability of outcome measures to clinical trial situations. With standardised training (and potentially increased automation of analysis), high magnification NVC has strong potential as an outcome measure of SSc-related microangiopathy.
Acknowledgement Work funded by the Raynaud's & Scleroderma Association.
Disclosure of Interest None declared