Authors: AC Viswanathan, FW Fitzke, RA Hitchings.

Institute of Ophthalmology, Bath Street, London, UK and Moorfields Eye Hospital, City Road, London, UK.

Purpose: To compare the ability of various methods of analyzing visual fields to detect conversion from ocular hypertension to glaucoma.

Methods: From a cohort of 350 ocular hypertensive patients 21 were chosen whose visual fields showed some clinical evidence of change (27 ‘suspicious’ eyes). Fields from each eye of these patients were analyzed by Statpac 2 (compares each point with a baseline from initial fields), PROGRESSOR (performs linear regression of sensitivity on time for each point), and AGIS (compares each point with threshold values based on population means).

Results: Statpac 2 classified 23, PROGRESSOR 24 and AGIS 7 eyes as converting. The latter 7 were all from the ‘suspicious’ group. Statpac 2 and PROGRESSOR agreed on 18 eyes as converting: 13 were from the 27 ‘suspicious’ eyes used as entry criteria. For these 18 eyes Statpac 2 detected conversion earlier than PROGRESSOR in 1 case, the two programs detected conversion simultaneously in 5 cases and PROGRESSOR detected conversion earlier than Statpac 2 in 12 cases (p < 0.05, Wilcoxon Matched-Pairs Signed-Ranks Test).

Conclusions: Combined analysis of serial visual fields is helpful in attempting to identify ocular hypertensives at risk of conversion to glaucoma. PROGRESSOR detects change earlier than Statpac 2 and would be more helpful in glaucoma management.

(Presented at the Royal College of Ophthalmologists Annual Congress, Edinburgh 1996)