Field test reliability of a new lens opacity rating system utilizing slit-lamp examination

A field test of a lens opacity rating system was run. Findings were graded in 7 categories: Cortex: (1) Waterclefts, Vacuoles, & Flakes (WVF, 4 grades), (2) Wedges & Spokes graded by number of Quadrants affected (WS-Q), and (3) Inward extent of Wedges & Spokes (WS-I, 3 grades); Nucleus: (4) Opacification (NUC-O) and (5) Coloration (NUC-C), each using 4 grades; and Subcapsular areas: (6) Anterior (SUB-A) and (7) Posterior (SUB-P) opacification, each using 5 grades. An instruction booklet with corresponding guidelines, sketches, and photographs was developed. Reproducibility of grading was tested in 3 clinics using 16 ophthalmologist-examiners. At each clinic, examiners were randomly paired. They independently performed slit-lamp exams of assigned patients (9 on average) on the same day (inter-examiner) and 1-2 weeks later (intra-examiner). To test intra-examiner reliability, examiners were divided into two groups (one from each pair), with gradings analyzed by group. For each group, 53 to 58 patients had evaluable examinations. Intra-examiner concordance for each lens category (defined as at most one grade difference in one lens) was as follows (for examiner groups 1 and 2, resp.): WVF: 98% & 97%; WS-Q: 96% & 100%; WS-I: 100% & 100%; NUC-O: 98% & 100%; NUC-C: 100% & 98%; SUB-A: 90% & 100%; and, SUB-P: 92% & 100%. Inter-examiner concordances were slightly lower than those found for intra-examiner. Grading lens findings in the manner described above appears to have potential utility in large scale studies.