Abstract
Purpose: :
Plus disease is critical for identification of severe retinopathy of prematurity (ROP). However, diagnosis is subjective and variable, and the reasoning process is not well-understood. This study investigates the diagnostic reasoning process of experts using a qualitative analysis protocol.
Methods: :
6 ROP experts were video recorded while being asked to independently review a set of 7 wide-angle retinal images from infants with ROP. Experts were asked to explain their diagnostic reasoning process in detail (think-aloud protocol), to mark key findings relevant to their diagnostic reasoning for each image, to diagnose each image (plus vs. pre-plus vs. neither), and to indicate their degree of certainty. Subsequently, each expert viewed the images again while being asked to examine arteries and veins in isolation, report whether there was sufficient arterial tortuosity and venous dilation for plus disease, and answer specific questions. Video recordings were transcribed and reviewed. The diagnostic process of experts was analyzed using a published cognitive model.
Results: :
Based on the think-aloud protocol, 5/6 experts agreed on the same diagnosis in 3 study images, 4/6 experts agreed on the same diagnosis in 1 study image, and 3/6 experts agreed on the same diagnosis in 3 study images. When experts were asked to rank images in order of severity, the correlation coefficient between pairs of experts ranged from -0.04 to 0.75 (mean 0.33). All experts considered arterial tortuosity and venous dilation while reviewing each image. Other factors considered by some experts included venous tortuosity, arterial dilation, vascular branching, and peripheral retinal features. When experts were asked to re-review images to diagnose plus disease based strictly on presence of sufficient arterial tortuosity and venous dilation, all but 1 expert made diagnostic differences compared to the think-aloud protocol. Responses by 3 experts who made different diagnoses for the same image were reviewed in detail to illustrate differences in the retinal features considered, differences in the diagnostic approach, and differences in interpretation of the same vascular findings among experts.
Conclusions: :
Agreement in plus disease diagnosis among experts is imperfect. This study demonstrates that experts may differ in their diagnostic reasoning process, the retinal features that they focus on, and their interpretations even after viewing the same retinal features. Understanding these factors may have benefits for diagnosis and education in ROP.
Keywords: retinopathy of prematurity