Assessing the Uniformity of Uveitis Clinical Concepts and Associated ICD-10 Codes Across Health Care Systems Sharing the Same Electronic Health Records System

Abstract

IMPORTANCE Big data studies may allow for the aggregation of patients with rare diseases such as uveitis to answer important clinical questions. Standardization of uveitis-related variables will be necessary, including the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) codes used to identify patients of interest. There are currently limited data on the uniformity of diagnosis mapping to ICD-10 codes for uveitis diagnoses among different health systems.OBJECTIVE To assess the degree of uniformity in mapping of uveitis clinical concepts to ICD-10 codes across health care systems using the same electronic health record (EHR) system. DESIGN, SETTING, AND PARTICIPANTSThis multicenter survey study was conducted between September 14 and October 9, 2020, at 5 academic health care systems that use the Epic EHR.

Colombian Ocular Infectious Epidemiology Study (COIES): Ocular Toxoplasmosis Incidence and Sociodemographic Characterization, 2015-2019

“…Also, Toxoplasmosis diagnosis was determined by ICD-10 codes. Since we do not have access to clinical charts and previous studies based in SISPRO had determined that the concordance rate of the database using ICD-10 with medical records up to 83.4%, we assumed that Toxoplasmosis diagnosis was made by clinical and laboratory findings, and the diagnosis of OT was made by ophthalmic detection of retinochoroiditis and laboratory confirmation (Informe Quincenal Epidemiológico Nacional, 2013;McKay et al, 2021;Singh and Woodward, 2021). This bias control confers confidence in the epidemiologic description of our data.…”

Exploring the association between precipitation and population cases of ocular toxoplasmosis in Colombia

“…To control bias, we only included patients with a new diagnosis of OT ("new confirmed"). Additionally, it is essential to note that ophthalmologists most commonly use the B58.0 code, SISPRO-based studies using ICD-10 have shown 83.4% of concordance with the medical record, and ICD-10 has demonstrated acceptable accuracy in studies on uveitis using big data [27][28][29].…”

The Rigorous Work of Evaluating Consistency and Accuracy in Electronic Health Record Data

“…Measurement error can be evaluated by expert medical record review for a subset of patients, by sensitivity analyses, by using a priori thresholds of cohort accuracy, by natural language processing methods to probe clinical notes, and by using more easily verifiable findings, such as laboratory values to define cohorts, whenever relevant. Publication of error estimates, such as that performed in the study by McKay et al, 2 can help other researchers determine a suitable sample size to test their hypotheses before performing their own validation checks.…”

“…Health systems differ in observed coding patterns for several reasons: (1) text mapping of the diagnosis phrase to the billing code (eg, posterior uveitis), (2) differences in coding practices, and (3) difference in the actual case mix of patients. In this issue of JAMA Ophthalmology , McKay and colleagues report on this issue of coding uniformity; different EHRs may map text to different ICD-10 codes. In other words, when 2 clinicians type in the same words to describe a clinical condition into 2 EHRs, is that term mapped to the same ICD-10 code?…”