Purchase this article with an account.
P.A. Rai, I.E. Murdoch, K. Barton; Risk factors for Bleb Related Infection–Interobserver variation of clinical findings . Invest. Ophthalmol. Vis. Sci. 2004;45(13):5575.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose:The purpose of this study is to determine interobserver variation in the proposed clinical outcomes of a large case control study designed to identify bleb morphology and ocular surface risk factors for bleb–related infection.Methods: Twenty five eyes (19 patients) with trabeculectomy blebs at least 3 months following surgery, were examined independently by two experienced clinicians (PAR & IEM). Assessments were performed on blepharitis (acute and chronic), upper lid position in relation to superior limbus, bleb classification, bleb description, fluorescein and rose bengal staining of the cornea and conjunctiva. Bleb description included presence of bleb, bleb size (horizontal and vertical), bleb overhanging the cornea, area of thin conjunctiva, bleb leak and bubble formation.Results:kappa (Κ ) values for agreement between examiners were good to very good (Κ =0.61– 1.00) for area of conjunctival avascularity, corneal Rose–Bengal staining, presence of sutures and the presence of a bleb overhanging the cornea. Moderate agreement (Κ =0.41– 0.60) was obtained for fluorescein and rose bengal staining of the bleb. Fair agreement (Κ =0.21– 0.4) was found for bleb size, bleb leak, rose Bengal staining over the conjunctiva. Bubble formation, dellen and bleb presence elicited only poor agreement (Κ =< 0.2). Measurement of horizontal and vertical size of bleb and was within ± 1.0 mm and height within ± 2.5 mm., however there was some systematic bias between observers. There was good agreement with acute blepharitis (results within ± 1 on a 5–point scale). Similar agreement was demonstrated for chronic blepharitis (± 1.0 on a 5–point scale) with some systematic bias. Measurement of lid position in relation to the superior limbus showed agreement within ± 1.0mm, again with some systematic bias. Revalidation is in progress on the variables with poor agreement and for those with systematic bias. Conclusion: This study showed the confidence that could be placed in a large range of separate assessments and will inform the choice of outcomes and study power required for the main investigation. The findings have implications for the interpretation of previous studies relating to this subject
This PDF is available to Subscribers Only