advertisement

---

Editors Selection IGR 13-4

Comments

Comment by James Morgan on:

22358 Disc hemorrhages and treatment in the early manifest glaucoma trial, Bengtsson B; Leske MC; Yang Z et al., Ophthalmology, 2008; 115: 2044-2048

See also comment(s) by Don Budenz • Kenji Kashiwagi • Chris Leung • Jody Piltz-Seymour • Hidenobu Tanihara • Boel Bengtsson •

Find related abstracts

---

The statistical analysis of the study is of particular concern.
1. The finding of no significant difference in disc hemorrhages between control and treatment groups should be interpreted with caution. EMGT was designed to compare glaucoma progression between treated and untreated patients with POAG and the sample size was estimated to detect differences in progression rates, not differences in the presence or frequency of disc hemorrhages.

The current study may not have the statistical power to address the effect of IOP reducing treatment on the development of disc hemorrhages

To detect a difference of 10% in disc hemorrhages between the groups (e.g., 50% in the treatment group; 40% in the control group), it is estimated that at least 385 eyes in each group would be required (α = 0.05, β = 0.2).The current study may not have the statistical power to address the effect of IOP reducing treatment on the development of disc hemorrhages. Likewise, the logistic regression analysis may not have the statistical power to evaluate the predictors of disc hemorrhages. An appropriate power calculation should be carried out.
2. As the number of patients with ever observed disc hemorrhages was similar by ophthalmoscopy and by review of photographs and there was good agreement between disc hemorrhage recordings in clinical forms and on photographs, it is unclear why refractive error, baseline/follow-up IOP and gender were associated with the presence of disc hemorrhages when the logistic regression analysis was based on clinical forms whereas only gender had significant association with disc hemorrhage when the analysis was based on photographs. This discrepancy deserves an explanation.
3. Follow-up duration is a major confounding factor in the comparison between the control and treatment groups. There would be a reduced chance to detect disc hemorrhages in the control group as 62% of patients (78/126) demonstrated progression at four years (no data were included from control patients after progression). Although the authors argued that both groups had similar percentages of disc hemorrhage frequency when all study visits were considered, such comparison would be biased if the follow-up duration is not matched.

---

• ← Previous
• Next →

---