A comparative study of computerised visual field testing and optic disc morphometric parameters in the follow-up of primary open angle glaucoma

Purpose To evaluate the correlation between computerised visual field testi ng and optic disc morphometric parameters(rim area, rim/disc area ratio, cu p/disc area ratio and optic disc surface smoothness (ODSS)) in the follow-u p of a group of patients affected by primary open angle glaucoma (POAG). Methods Reliable automated perimetry (Humphrey 640 VFA, central 30-2 thresh old program) was performed at T-0 (the enrolment time), T-1 (after 6 +/- 1 months; range 5-7 months), T-2 (12 +/- 1 months), T-3 (18 +/- 1 months), an d T-4 (the end of the follow-up period: 24 +/- 2 months) in one eye randoml y chosen from each of 30 POAG patients. Computerised optic disc analysis (I MAGEnet X Rev-3-51b, Topcon Europe, The Netherlands) was performed at T-0 a nd T-4. To evaluate the correlation between morphometric parameters and com puterised visual field testing, the stability or worsening of visual field lest was evaluated by means of 'Mean Deviation linear regression analysis' (STATPAC 2 software); that of morphometric parameters was studied using the ir coefficients of variation. A rank of '0' was assigned to stability and a rank of '1' to worsening. The Spearman rank correlation coefficient was us ed to evaluate statistically the correlations between visual field analysis and morphometric parameters. Furthermore kappa statistic was used to evalu ate the agreement of morphometric parameter changes with visual field progr ession analysis. Results According to the MD slope regression analysis, in 18 patients the v isual fields were stable while in 12 they were worsening during the follow- up period. In 86.65% of patients (n = 26) the morphometric parameter ODSS a nd visual field analysis were concordant (p < 0.0001), In 80% of patients ( n = 24) the cup/disc area ratio and visual field analysis were concordant ( p < 0.001). The other morphometric parameters trim area, rim/disc area rati o) were less correlated with visual field analysis than ODSS (p < 0.05). Th e agreement of visual field analysis with all the morphometric parameters w as good (kappa = 0.690, 95% confidence interval (CI) of kappa = 0.589-0.790 ). The agreement of visual field analysis with ODSS and cup/disc area (kapp a = 0.688; 95% CI = 0.511-0.869 was better than the agreement of visual fie ld analysis with rim area and rim/disc area ratio (kappa = 0.438; 95% CI = 0.260-0.655), Conclusion Analysis of the progression of visual field damage and optic ner ve head morphometric parameters should both be taken into account in glauco ma follow-up. Among the morphometric parameters evaluated by means of Topco n IMAGEnet, ODSS and (to a lesser extent) the cup/disc ratio should have th e greatest weight.