Purchase this article with an account.
Jonathan C Clarke, Renata Puertas, Paul J Foster, Keith Barton; : Inter-observer agreement and sensitivity - specificity analysis of a technology led, remote review glaucoma clinic. Invest. Ophthalmol. Vis. Sci. 2014;55(13):5563.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Patients with well controlled glaucoma and ocular hypertension are suitable for follow up assessment in “remote review clinics”. The clinical findings and investigations are collected at one site and the results analysed by a glaucoma physician at another site at a later time. The separation of data acquisition and clinical decision making can lead to an improved patient experience and workflow. Although this type of clinic set up exists, there is little information concerning the safety or physician agreement. We investigate the safety and level of clinical agreement of a remote review glaucoma clinic.
Low risk patients were selected by reviewing the clinical notes of a general adult glaucoma clinic. Patients were reviewed clinically after all the appropriate investigations to assess glaucoma were performed. A decision about stability was made. More than 1 year later, the same clinicians reviewed the data collected and decided whether the glaucoma appeared stable or unstable. These outcomes are compared between physicians as a measure of inter-observer agreement. A sensitivity and specificity assessment comparing clinical decision making and remote decision making is performed.
The records of 871 patients were assessed and 286 were invited to participate and 217 accepted. There were sufficient data collected by both clinical systems to allow accurate decision making in 200 patients. Inter-observer agreement was moderate with kappa scores of 0.50 (95% CI 0.374 to 0.619). For the more senior physician, 8 of 116 patients (6.9%) were progressing in the clinical examination and 6 of these patients were identified remotely with sensitivity of 75.0% and specificity 79.6%. The remaining 2 patients were slowly progressing and diagnosis is likely to have been made if imaging data were available in the remote review system.
Telemedicine and remote review medical assessments are increasingly common systems for treating patients. The safety of these systems has not been reliably assessed. We conclude that there is moderate agreement between two assessors of a group of stable glaucoma patients. Sensitivity of the system to detect progression is likely to improve if the data is presented systematically by an electronic patient record.
This PDF is available to Subscribers Only