Abstract
Purpose :
The VIBES (Vienna Imaging Biobank Eye) Study is a retrospective real-world analysis of all patient data from the Department of Ophthalmology, Medical University of Vienna, from 2007-2017 to better understand the situation in Austria’s largest tertiary-care hospital.
Methods :
We analyzed data from 2 electronic patient records (2007-2015, 2014-2017), the hospital controlling information on treatments and 3 major OCT imaging systems. We present our primary analysis from 2007-2015 electronic records of the macula clinic (n=7831 patients) matched with the controlling database (n=42965 procedures on 6928 patients), which had an overlap of 3657 treated eyes. Matched patients were filtered for diagnosis entries, defined by a retina specialist team, to identify patients with AMD. Included patients had received at least 1 injection of Ranibizumab/Aflibercept during the relevant time to classify them as active neovascular AMD. Their visual acuity (VA) also had to have been measured 60-0 days before the first injection, remaining 1925 eyes eligible for analysis. For comparison of results to pivotal clinical trials (e.g. VIEW), eyes having VA <20/320 were excluded from analysis. If available, OCT scans were analyzed in addition, to define morphologic subgroups.
Results :
1925 eyes had a baseline (BL) VA of 0.33±0.27, of which 19% were below 20/320. Remaining 1560 eyes had a BL VA of 0.4±0.25. 794 eyes were monitored over 1, 272 over 2 and 78 over 3 years (Y). The mean number of injections was 3.02±1.89 in Y1, 2.72±2.25 in Y2 and 2.96±2.59 in Y3. The BL VA/number of injections in each year for eyes still in follow-up was at no time different from these patient cohorts’ at preceding timepoints. Mean VA at follow-up was 0.41±0.29 (n=794) showing visual stability after Y1, but 0.33±0.26 after Y2 and 0.29±0.24 after Y3, consecutively. The VA of patients receiving more than 4, 5 or 6 injections in Y1 had not improved more than that of their matched group's BL VA (with an identical BL VA). Eyes gaining at least 0.1 of VA in Y1 had had slightly more injections (3.47±2.21). Morphologic features have been analyzed, too.
Conclusions :
The VIBES study shows the state-of-the-art in a large real-world setting in Europe including OCT image analysis. The high BL VA indicates an efficient referral system. VA was largely maintained despite a low retreatment frequency. Nevertheless, under treatment is recognized as an overall real-world deficit.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.