Glaucoma risk estimates influenced by life expectancy

March 7, 2008

The predicted 5-year risk of developing glaucoma may be significantly altered when life expectancy is incorporated in the risk calculator, said Anne Coleman, MD, PhD.

The predicted 5-year risk of developing glaucoma may be significantly alteredwhen life expectancy is incorporated in the risk calculator, said Anne Coleman, MD, PhD.

Dr. Coleman, Jules Stein Eye Institute, University of California Los Angeles, presented thefindings from a study calculating glaucoma risk for a group of 1,000 hypothetical patientsusing 5-year mortality risk predicted by the Charlson Index and the unadjusted 5-year risk ofglaucoma using the OHTS/EGPS risk calculator. The results showed that when the glaucoma riskcalculator was used by itself, the greatest 5-year glaucoma risk was in the oldest age group.After adjusting for life expectancy using the Charlson Index score, however, youngerindividuals had the greatest risk for glaucoma.

"The AAO guidelines and OHTS/EGPS investigators suggest the decision to treat [suspected]glaucoma should be based on other factors in addition to risk development, including lifeexpectancy, comorbidity, and patient preference," Dr. Coleman said.

"Available literature shows clinicians are inaccurate and unreliable in estimating lifeexpectancy of patients they see in the office. The Charlson Index is a popular, validatedexplicit measure that has greater interrater reliability than implicit judgment of theclinician," said Dr. Coleman, explaining the motivation for the study.

She acknowledged, however, this research represents a beginning. Further studies are needed toinvestigate how clinicians prefer to use life expectancy data and also to consider lifeexpectancy over a longer term of decades rather than just 5 years.

"There are other limitations as well," Dr. Coleman said. "Current calculators estimate time toglaucoma development, not time to visual disability, the threshold for initiating treatment isnot firmly established, and how clinicians use life expectancy may depend on the adverseeffects of treatment alternatives. Of course, the one thing that we do know as clinicians isthat we really don't know how long someone will live."