Editors Selection (IGR: 9-4) | International Glaucoma Review #308

Editors Selection IGR 9-4

Medical treatment: Prospective

Comment by Jennifer Burr on:

15258 Relative risk of progressive glaucomatous visual field loss in patients enrolled and not enrolled in a prospective longitudinal study, Henson DB; Shambhu S, Archives of Ophthalmology, 2006; 124: 1405-1408

Henson et al. (1147) hypothesise that estimates of risk of glaucoma progression from prospective cohort data are biased in that only people at low risk of progressive glaucoma are included, or that the study conditions create an atmosphere where progression becomes less likely. The stated aim of the study was to quantify the extent of any selec-tion bias between different study populations. The risk of progressive visual field loss in a sample selected from glaucoma patients under follow up at a hospital eye department was estimated relative to the risk observed in a sample (66 eyes), selected from participants in a separate, prospective longitudinal study.¹ The clinic sample was selected to match the cohort sample on length of follow up, and extent and pattern of visual field loss. Linear regression analysis of the depth of visual field loss (Peridata, mean defect) and number of damaged locations (pattern deviation < 0.05) was used to estimate progression. Henson et al. report the relative risk of progressive visual field loss in the 'non-study' population to be 368% higher than the 'study' cohort.

Risk of progression may be higher in a group of random glaucoma patients than in those selected for long-term follow-up studies

Thus, an estimated relative risk of 4.68. There are significant numerical inconsistencies in the data presented and a lack of clarity in how the final result has been obtained which may have over-estimated the increased risk, but the effect is still apparent. The authors conclude that selection bias may explain differing risk estimates observed in cohort studies and clinical practice. However, the issues are more complicated than this.

It is hard, but of crucial importance to disentangle the effects of bias from the desired effects of close monitoring linked to intensive treatment on reducing the risk of progression in 'study' populations compared with observations from clinical practice; a limitation recognised by the authors in their discussion. Extended follow up of existing randomised controlled trials, and collaborative prospective cohort studies, with consecutive recruitment to represent all stages of severity would provide optimal data to evaluate risk of progression.

References

Kwartz AJ, Henson DB, Harper RA, Spencer AF, McLeod D. The effectiveness of the Heidelberg Retinal Tomograph and laser diagnostic glaucoma scanning system (GDX) in detecting and monitoring glaucoma. Health Technol Assess 2005; 9:

Comments

The comment section on the IGR website is restricted to WGA#One members only. Please log-in through your WGA#One account to continue.

Log-in through WGA#One

Issue 9-4

Table of Contents Editor's Selection