Axial length measurement and detection rates using a swept-source optical coherence tomography-based biometer in the presence of a dense vitreous hemorrhage

Abstract

Purpose:To evaluate axial length (AL) measurements and detection rates using a swept-source optical coherence tomography (SS-OCT) biometer (OA-2000) compared with the IOLMaster, LENSTAR, and ultrasound (US) measurements in eyes with vitreous hemorrhage.Setting:Eye Hospital of Wenzhou Medical University, China.Design:Comparative evaluation of a diagnostic test or technology.Methods:Patients were adults with vitreous hemorrhage in one or both eyes. Lens Opacities Classification System III was used to determine they type and severity of cataract and the Forrester system was used to classify the grade of vitreous hemorrhage. The AL measurement was taken with the OA-2000, IOLMaster (v5.4), and LENSTAR (v2.1), and with US (Axis-II). The Bland-Altman test was used to assess the agreement between devices.Results:Forty eyes were scanned with each of the 4 biometers. For Forrester grade III, the detection rate was 100%, 41.7%, and 41.7% with the OA-2000, IOLMaster, and LENSTAR, respectively. For Forrester grades I and II, the detection rate was 46.4%, 3.57%, and 3.57% with the OA-2000, IOLMaster, and LENSTAR, respectively. The detection rate with US was 100% for all grades. The Bland-Altman limits of agreement between the 4 biometers indicated an excellent agreement: OA-2000 vs IOLMaster (95% limits of agreement -0.15 to 0.05 mm), OA-2000 vs LENSTAR (-0.04 to 0.12 mm), and OA-2000 vs US (-0.3 to 0.56 mm).Conclusions:In vitreous hemorrhage, the detection rate with the SS-OCT OA-2000 biometer was better than that with the IOLMaster and LENSTAR. In cases in which all biometers were able to acquire an AL measurement an excellent agreement was found.