Accuracy of ICD-10-PCS codes to identify Cardiac Implantable electronic devices in Portugal - A single center chart review

• Type: Another Publication in an International Scientific Journal
• Date: 2025

Authors

Couto, S (Author) Other; Lobo, M (Author) Other; Lopes, F (Author) Other; Cardoso, JCS (Author) Other; Moreira, E (Author) Other; Rocha, A (Author) Other; Freitas A (Author) FMUP

Journal

Title: International Journal of Medical Informatics

Vol. 199

ISSN: 1386-5056

Publisher: Elsevier

Indexing

ISI Web of Knowledge - 0 Citations

Scopus - 0 Citations

Other information

Authenticus ID: P-018-F73

DOI: 10.1016/j.ijmedinf.2025.105885

Abstract (EN): Background: Hospital use of Cardiac Implantable Electronic Devices (CIEDs) is documented in the NHMD (National Hospital Morbidity Database) using the International Classification of Diseases Procedure Coding System (ICD-10-PCS), providing a comprehensive real-world perspective on CIED usage. Objective: To evaluate the accuracy of ICD-10-PCS codes in the NHMD for identifying episodes of patients implanted with a Cardiac Resynchronization Therapy (CRT) device. Methods: This study analyzed hospital episodes from 2017 and 2021 at a single center, comparing ICD-10-PCS codes for CRT insertion or replacement in the NHMD against clinical records. Coding accuracy was first evaluated globally for CRT devices as a single category, and then by year, device type (CRT-P or CRT-D), and lead type. Positive predictive value (PPV) assessed coding accuracy. Results: We sampled 180 hospital episodes with an ICD-10-PCS code for CRT insertion or replacement (2017: n = 86; 2021: n = 94). The PPV remained above 90 % when the analysis was done by generator type (i.e. CRT-P or CRT-D). However, when analyzed by lead type, accuracy dropped significantly, with sensitivity and PPV occasionally reaching 0 %. Conclusions: ICD-10-PCS coding demonstrated a strong ability to identify CRT generator implantation procedures in the NHMD. However, concerns arose regarding its accuracy in identifying lead types. The discrepancy between coding and clinical records highlights that there is room for improvement in coding processes and data quality.

Language: English

Type (Professor's evaluation): Scientific

No. of pages: 6