Inaccuracies Abound in OCT-Angiography Studies

By David Douglas

December 11, 2019

NEW YORK (Reuters Health) - More than 90% of studies involving optical-coherence-tomography angiography (OCTA) fail to allow for the possible difference in axial length between the model eye of the device and that of the patient, according to a new literature review.

"Our study demonstrated that the vast majority of studies involving OCTA are using inaccurate data," Dr. Joseph Carroll of the Medical College of Wisconsin, in Milwaukee, told Reuters Health by email. "This is important because OCTA is being used more and more to detect and diagnose a number of retinal and systemic diseases."

Dr. Carroll and colleagues found 989 articles on OCTA retinal imaging that met their inclusion criteria. All were published between 2015 and 2018.

They reviewed each in detail to assess whether the OCTA measurements needed correct lateral scaling, and if so, whether axial length was reported or used to scale the images. In all, 509 studies required such scaling, but only 41 (8.0%) reported measuring and using axial length to correct the lateral scale.

Moreover, of the 468 articles that did not correctly scale their images, only 18 (3.8%) mentioned this as a limitation to the study.

The team notes, "A common argument made by researchers and clinicians against acquiring axial length is the inconvenience of having to use a different device, so a solution may be to integrate axial length and other ocular biometry functionalities within the OCTA devices. A compromise may involve simply reporting scan dimensions in visual angle units (eg, arcmin and degrees)."

"Without knowing if or how data may be reused by future studies," they add, "knowingly publishing inaccurate data should be avoided (or at the very least should be disclosed)."

Dr. Carroll said the impact of publishing such data is akin to "a clinician (knowingly) relying on an uncalibrated EKG or blood pressure cuff to try and diagnose heart disease."

Ophthalmologist Dr. Andrew Rabinowitz, chief medical officer at American Vision Partners in Phoenix, Arizona, told Reuters Health by email that the article "brings to light the observation that a large number of peer-reviewed articles on OCTA measurements fail to factor in differences in the axial length. Axial length of the globe directly impacts the lateral magnification of in vivo retinal images."

"Thus, in order to gain the most insight from studies of OCTA images," concluded Dr. Rabinowitz, who was not involved in the study, "a universal scale must be created and deployed. This scale will allow for more clinically valid conclusions."

SOURCE: JAMA Ophthalmology, online November 27, 2019.