advertisement
Abstract #19694 Published in IGR 9-4
Use of a Bayesian network to predict the nighttime intraocular pressure peak from daytime measurements
Nordmann J-P; Berdeaux GClinical Therapeutics 2007; 29: 1751-1760
See also comment(s) by Christopher Bowd •
BACKGROUND: Intraocular pressure (IOP) fluctuation over 24 hours is an independent risk factor for glaucoma progression. Nighttime lOP measurement is not routine practice. OBJECTIVE: The aim of this study was to predict, using a Bayesian network (BN), the likelihood of a nighttime IOP peak >18 mmHg based on daytime measurements. METHODS: A pooled analysis was conducted using a BN. Data from three clinical trials of adult patients with glaucoma or ocular hypertension were used. IOP values at 0800, 1200, 1600, and 2000 hours were dichotomized according to the 18-mmHg threshold. Patients' lOPs were assessed from the pretreatment washout periods and during latanoprost or travoprost treatments. A BN was constructed to study associations between daytime and nighttime IOP values, and prostaglandin analogue IOP control adjusted for trial. The nighttime IOP peak was defined as the maximum IOP value between 2400 and 0400 hours. RESULTS: The study identified 382 daily IOP vectors (6 measures per day, every 4 hours for 24 hours; pretreatment, 208; latanoprost, 73; travoprost, 101). Based on the BN, IOP at 0800 hours was associated with IOP at 1200 hours, which was also associated with the IOP at 1600 hours. IOP at 2000 hours was predicted by the IOPs at 1200 and 1600 hours. The nighttime IOP peak was associated with IOP >18 mmHg at 1200 and 2000 hours. The percentage of patients with controlled IOP at 1200 and 2000 hours was higher in those receiving travoprost than in those receiving latanoprost. Travoprost was also associated with an increased probability of controlling nighttime IOP values >18 mmHg (travoprost, 76.9%-77.5% vs latanoprost, 66.7%-67.9%). Daily IOP fluctuations were not found to be associated with nighttime IOP peak. CONCLUSIONS: Daytime IOP measurements are highly intercorrelated. According to this BN, IOP at 1200 and 2000 hours are more strongly associated with the nighttime IOP peak than other IOP measurements. BN can. estimate the risk of a nighttime IOP peak >18 mmHg. Daytime IOP control was important for nighttime. IOP control. These findings require validation in a clinical setting.
Dr. G. Berdeaux, Alcon France, RueiI-Malmaison, France. Gilles.berdeaux@alconlabs.com
Classification:
6.1.2 Fluctuation, circadian rhythms (Part of: 6 Clinical examination methods > 6.1 Intraocular pressure measurement; factors affecting IOP)
