Evaluating clinical signs in trabeculectomized eyes

Aim To evaluate interobserver agreement for clinical signs in trabeculectomized eyes when examined face-to-face with slit-lamp biomicroscopy (SL) or by remote examination using telemedicine (real-time remote video imaging; TM). Method A system for examining trabeculectomized eyes was devised and validated. A prospective randomized interobserver agreement study was then undertaken to compare standard SL biomicroscopy and TM. Remote examination was performed using a 384 kbps Sony 5100 videoconferencing system. Three ophthalmologists each examined 40 eyes of 40 patients, who had previously undergone trabeculectomy. In rotation, two examiners used SL biomicroscopy. The third examined the eye remotely by TM. Analysis was performed to determine the variability in clinical signs and the presence or absence of systematic bias between ophthalmologists and examination methods. Results High levels of agreement were observed for paired examinations by SL biomicroscopy (SL/SL) for bleb vascularity ( score range 0 - 10) with no systematic bias. Paired examination by SL and TM (SL/TM) also showed good levels of agreement for bleb vascularity, although the spread of disagreement was wider (95% limits of agreement 2.57 vs 2.98 ( P = 0.054)). For anterior chamber depth, observers agreed within +/-10% of anterior chamber depth for 68% of eyes (SL/SL) and 51% of eyes (SL/TM) (P = 0.68). Agreement was 'good' for wall thickness (kappa 0.63 +/- 0.08), bleb height (kappa = 0.67 +/- 0.1), and the existence of bleb leak (kappa = 0.63 +/- 0.19), but poor for bleb morphology (kappa = 0.26 +/- 0.12). For the SL/TM comparison, agreement was fair for wall thickness (kappa = 0.39 +/- 0.13), poor for bleb height (kappa = 0.17 +/- 0.12), good for bleb leak (kappa = 0.56 +/- 0.19), and fair for bleb morphology (kappa = 0.31 +/- 0.12). Microcysts were not reliably detected using either technique. Conclusion SL biomicroscopy and TM telemedicine examination may permit reliable clinical assessment of trabeculectomized eyes. However, remote examination with TM is more limited with respect to assessing bleb height and bleb wall thickness. The assessment of bleb morphology and microcysts was unreliable with both instruments. We propose that TM examination of trabeculectomized eyes appears safe and appropriate in situations where face-to-face examination by an ophthalmologist is not practical.