Abstract
Purpose :
Inherited retinal diseases (IRDs) are a diverse group of diseases that can possibly lead to blindness. Ongoing clinical trials are testing the potential for gene therapy as treatment for certain IRDs. Physicians treating IRD patients should know which patients qualify for these trials. However, one barrier for identifying these patients is the limited scope of ICD codes for IRDs. In this retrospective chart review, we use peripheral and macular IRDs to test the hypothesis that some retinal dystrophies are not well represented in current ICD nomenclature compared to more well-known IRDs like retinitis pigmentosa (RP) and Stargardt disease (STGD).
Methods :
This study was conducted with approval of the Vanderbilt University Medical Center Institutional Review Board/Ethics Committee. Manual chart review was performed for 250 patients who were isolated from the Vanderbilt Eye Institute electronic health record based on ICD10 H35.52 (peripheral retinal dystrophies) and for 281 patients who were isolated based on ICD10 H35.50 or H35.53 (macular dystrophies). Phenotypes of peripheral disease that were analyzed include RP, rod-cone dystrophy, and Usher syndrome. Phenotypes of macular disease that were analyzed include STGD, macular pattern dystrophy, vitelliform dystrophy, Sorsby fundus dystrophy (SFD), and cone/cone-rod dystrophy. The proportion of patients whose ICD designation aligned with the diagnosis made by an IRD specialist was then calculated for each phenotype.
Results :
In the peripheral disease group, the percentage of accurate diagnoses are as follows: 70.8% for RP, 4.8% for rod-cone dystrophy, and 7.6% for Usher syndrome. In the macular disease group, the percentage of accurate diagnoses are as follows: 14.9% for STGD, 31.3% for macular pattern dystrophy, 1.07% for vitelliform dystrophy, 1.4% for SFD, and 8.9% for cone/cone-rod dystrophy.
Conclusions :
Our results reveal low concordance between the clinician’s diagnosis and the ICD codes for peripheral and macular IRDs, even including well-known diseases such as RP, STGD, and macular pattern dystrophy. This suggests that current ICD nomenclature is unable to adequately encompass many IRD phenotypes, making big data analysis challenging for these diseases. More specific ICD10 codes are needed to phenotype these IRD patients to help physicians more readily identify those who qualify for IRD clinical trials.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.