ICD-10-CM coding uncovers the gap between serological and clinically identified coeliac disease prevalence: A population-based study.

Abstract

The International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) is a global health coding system that provides a standardised language for recording, reporting and monitoring diseases. However, epidemiological studies using real-world data on clinically diagnosed coeliac disease (CD) based on the ICD-10-CM remain limited. To evaluate in primary care: (1) the diagnostic adequacy of CD; (2) the accuracy of the ICD-10-CM code K90.0 for identifying CD; and (3) the prevalence of clinically diagnosed CD compared with the known CD seroprevalence in the same geographical area (5‰). This was a population-based, cross-sectional study in the city of L'Hospitalet de Llobregat (269,382 inhabitants), Catalonia, from 2005 to 2020. The data retrieved with the K90.0 code from electronic medical records were cross-checked against laboratory and pathology registries. The CD diagnostic criteria were validated on a case-by-case basis. To calculate the accuracy of the K90.0 code for identifying CD, patients were classified into confirmed versus uncertain/misdiagnosed CD. Overall, 536/737 patients (73 %) had confirmed CD, and 201/737 (27 %) were misdiagnosed. The accuracy of the K90.0 code for identifying CD was as follows: sensitivity, 91.63 %; specificity, 99.95 %; positive predictive value, 63.85 %; and negative predictive value, 99.99 %. The prevalence of clinically diagnosed CD was 1.99‰, with a decreasing age-related trend of -7.5 %. The ICD-10-CM code K90.0 is accurate for identifying clinically diagnosed CD and is thus a great tool for epidemiological disease surveillance. The gap between CD seroprevalence and the prevalence of clinically diagnosed CD (5‰ vs. 1.99‰) calls for the implementation of case-finding programmes to reduce underdiagnosis.