advertisement
Editors Selection IGR 13-2
Progression: Expert versus non-experts in assessing changes in OD photographs
Comment by Kouros Nouri-Mahdavi on:
45773 Agreement and accuracy of non-expert ophthalmologists in assessing glaucomatous changes in serial stereo optic disc photographs, Breusegem C; Fieuws S; Stalmans I et al., Ophthalmology, 2011; 118: 742-746
Breusegem et al. (1073) compared the performance of 37 general ophthalmologists for detection of glaucomatous optic disc progression to a reference standard consisting of the agreement of a panel of three experienced glaucoma specialists. The agreement among general ophthalmologists was poor overall (average Cohen's κ= 0.2) and increased only slightly (κ= 0.27) after a 30-minute training session. The interobserver agreement between the non-experts and the experts was fair (κ= 0.33) but was not reported after the training session. The percentage agreement between the non-experts and glaucoma specialists improved from 68.5% before training to 71.4% after the training session.
Review of optic disc photographs for progression by general ophthalmologists is suboptimal... Additional training would likely be helpful
Qualitative evaluation of optic disc progression is a complex subjective task and requires adequate initial training and ongoing practice. Additionally, performance of any given individual in setting of investigations such as this, be it an experienced glaucoma specialist or a general ophthalmologist, also depends on how obvious the progression is on pairs of slides demonstrating such change and the proportion of progressing eyes. It also depends on the initial training and experience of the general ophthalmologists reviewing the disc photographs. The number of the general ophthalmologists enrolled in this study was large enough to allay any concerns regarding the quality of their training and experience. Also, a reasonable proportion of the eyes progressed in the chosen study sample (15/40 eyes or 38%) to make the agreement results meaningful. Overall, the investigators confirmed that experience and expertise matters. In addition, although a short 30-minute training session somewhat improved the performance of the general ophthalmologists, the improvement was of questionable clinical significance. This research has immediate clinical implications for detection of glaucoma progression with optic disc photographs, which still remains the mainstay of detection of glaucoma progression. One can conclude, as a rule of thumb, that review of optic disc photographs for progression by general ophthalmologists is suboptimal at best. Additional training would likely be helpful and to be of benefit, the training needs to span over a longer period of time with possibly continuous feedback from experienced clinicians. From a research point of view, it would mean that progression rates based on optic disc review by glaucoma researchers may not necessarily apply to the practice of general ophthalmologist involved with care of glaucoma patients.