Minimum Imaging Sets for Diagnosis, Activity Assessment and Complications in Non-Infectious Posterior Uveitis: MUV Report 9.

Purpose
To evaluate a minimum imaging set (MIS) to support ophthalmologists in diagnosing, monitoring disease activity and identifying disease-specific complications for six types of non-infectious posterior uveitides (NIPU).Design
Visual questionnaire answered live by a group of ophthalmologists.Participants
International group of ophthalmologists including uveitis experts, medical retina experts, general ophthalmologists and ophthalmologists-in-training.Methods
Five groups of uveitis experts were selected by the Multimodal Imaging in Uveitis (MUV) taskforce and asked to create a MIS for each of the following entities: multiple evanescent white dots syndrome (MEWDS), acute posterior multifocal placoid pigment epitheliopathy (APMPPE), multifocal choroiditis and panuveitis (MFCPU), punctate inner choroidopathy (PIC), serpiginous choroiditis (SC) and birdshot chorioretinopathy (BSCR). The MIS were designed to improve an ophthalmologists' ability to diagnose and manage these NIPU without the need for a complete multimodal imaging (MMI) approach. Five visual questionnaires including the color fundus photography (CFP), used as a proxy for funduscopic examination, and the MIS of a series of cases were constructed and proposed to the group of ophthalmologists. The number of correct answers achieved by participants reviewing the MIS was compared to those obtained by examining the CFP alone.Main Outcome Measures
Sensitivity, specificity and accuracy of MIS as compared to CFP alone in diagnosing, assessing activity and identifying complications in NIPU.Results
156 ophthalmologists participated in the visual questionnaire. For diagnostic purposes, the MIS outperformed CFP in MEWDS, APMPPE and BSCR. For assessing activity, MIS allowed participants to achieve significantly better results in all NIPU but APMPPE. Identification of complications (choroidal neovascularization) was tested in SC and MFCPU/PIC only, and in both cases MIS outperformed CFP alone. The MIS performance was not significantly influenced by ophthalmologists' level of training or geographical region of practice. The MIS outperformed CFP both in classic and challenging cases, but it exceeded 75% accuracy only in classic cases.Conclusions
The validation exercise confirmed the utility of most of the proposed MIS. These MIS increased the ability of ophthalmologists to diagnose and manage most cases of NIPU regardless of physicians' background and clinical setting. A complete MMI approach is still needed for more challenging phenotypes.