Purchase this article with an account.
S. Vandenbroeck, F. Dobbels, S. De Geest, I. Stalmans, T. G. Zeyen; Ophthalmologists Poorly Predict Patient Non-Adherence. Invest. Ophthalmol. Vis. Sci. 2008;49(13):1579. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Due to its suggested association with unfavorable outcomes, identification of non-adherence (NA) with eye drop treatment in patients with glaucoma is crucial. The diagnostic value of collateral report as measurement method for NA based on evaluation of the treating ophthalmologists needs to be scrutinized. The purpose of the study was therefore to evaluate the diagnostic value of ophthalmologists' judgement using patient self-report as gold standard.
A multi-center cross-sectional survey was performed in a convenience sample of 827 glaucoma patients treated in 81 centers/practices in Belgium. NA was assessed by written self-report asking patients how many times they had forgotten to administer eye drops during the past 2 weeks (never, seldom, once a week or daily). Because of underreporting of NA using self-report strategy, the most stringent definition of NA "missing 1 or more doses" was used. The ophthalmologists on the other hand were asked by means of 1 single question if they thought their patient was adherent or not (yes, no, don’t know). If "don’t know" was checked, the patient was excluded from analysis. To determine the diagnostic value of the ophthalmologist’s ratings, both sensitivity and specificity were calculated and a Roc-analysis was performed.
In total, 663 surveys were completed by both patient and ophthalmologist (80% participation rate), of which 635 (96%) were available for analysis. Self- and collateral report of NA were 38% and 2.2% respectively. Sensitivity (i.e. probability of a NA-rating among patients who really are NA) of collateral report was 3.3%, while the specificity (i.e. probability of adherence-rating among patients who are really adherent) was 98.5%. The positive and negative predictive values were 57% and 62.4% respectively. The Roc-curve showed an Area Under the Curve of 0.491 (95% C.I: 0.445-0.538), indicating a very poor diagnostic accuracy.
A strong discrepancy was demonstrated between NA reported by patients and ophthalmologists. Compared to the patient’s reports, ophthalmologists tended to consider more patients as adherent. These results demonstrate that a better tool than the ophthalmologist’s rating should be used as gold standard for screening of NA (e.g. electronic monitoring).
This PDF is available to Subscribers Only