The influence of the assessment method on the incidence of visual field progression in glaucoma: a network meta-analysis

Purpose: To study and quantify the difference in incidence of progression between methods for the assessment of glaucomatous visual field progression. Methods: We identified 2450 articles published up to April 2009 in the following data sources: PubMed, EMBASE, and Cochrane. Ten studies covering 30 methods were included. All studies aimed to compare different methods for the assessment of glaucomatous visual field progression in the same study population. A network meta-analysis using a mixed-effects model was performed to combine within-study between-method comparisons with indirect comparisons from other studies. The summarized incidence of progression was calculated for every method, and methods were ranked according to this incidence. Results: In total, methods were compared in 1040 eyes of 948 patients with glaucoma. On average, 21% of the eyes progressed. When all 30 methods were ranked, the incidence ranged from 2% to 62%. These incidences are corrected for a baseline mean deviation (MD) value of -7 decibels and a mean follow-up time of 6 years. Besides the assessment method, the incidence was only determined by the follow-up period and baseline MD value, leaving no unexplained variance in the incidence of progression. Conclusion: The incidence of progression varies considerably between different studies. This is mainly caused by the variety of methods used to assess progression but also by differences in follow-up time and baseline visual field loss.