Development of a Decision Analytic Model of Changes in Mean Deviation in People With Glaucoma---A Merger of Data From the CIGTS, OHTS, and AGIS Trials

Abstract

Purpose: : To obtain a better understanding of the rate of change in visual field loss in people with glaucoma, we pooled patient level data from three large randomized trials examining the range from ocular hypertension to advanced glaucoma. Here we report the preliminary results describing the change in mean deviation (MD) over seven years predicted by the model.

Methods: : We constructed a Markov model using MD data from study participants in CIGTS (n=607), OHTS (n=1,636), and AGIS (n=560). The model was constructed with transition probabilities for each visit/MD combination over 7 years. We assessed the difference in MD from baseline to the 7 year visit for four simulated participants with starting MDs of -1 dB, -4 dB, -10 dB, and -20 dB. We report the result for the right eye only. For the model MD was characterized in integers. "Change" is a difference of 1 dB between visits. "Improvement" is a "less negative" MD score. "Worse" is a "more negative" MD score.

Results: : There was a difference in the rate of change over 7 years between those with -1 dB and -4 dB compared to those with -10 dB and -20 dB. At year 7, the participants with baseline MD scores of -1 dB or -4 dB were more likely to have the same or improved MD scores than the participants with baseline MD of -10 dB or -20 dB (65% vs. 35%) Those with baseline MD of -10 dB or -20 dB were more likely to have a worse MD in year 7 than an improved score. Almost 1/3 of those with a baseline MD score of -10 dB or -20 dB had a worsening of MD of 3 dB or more.

Conclusions: : Further validation of the model, including calibration, is necessary, but initial results seem to replicate those seen in other studies. Our model for changes in MD could serve as a versatile tool for evaluation of the effectiveness of interventions to prevent glaucoma progression.

Keywords: clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology • visual fields • clinical (human) or epidemiologic studies: treatment/prevention assessment/controlled clinical trials