Clinical Epidemiology (Oct 2024)
Validating ICD-10 Diagnosis Codes for Guillain-Barré Syndrome in Taiwan’s National Health Insurance Claims Database
Abstract
Cheng-Yang Hsieh,1,2,* Po-Ting Chen,3,* Shih-Chieh Shao,2,4,5 Swu-Jane Lin,6 Shu-Chen Liao,3,7 Edward Chia-Cheng Lai2,5 1Department of Neurology, Tainan Sin Lau Hospital, Tainan, Taiwan; 2School of Pharmacy, Institute of Clinical Pharmacy and Pharmaceutical Sciences, College of Medicine, National Cheng Kung University, Tainan, Taiwan; 3Department of Emergency Medicine, Keelung Chang Gung Memorial Hospital, Keelung, Taiwan; 4Department of Pharmacy, Keelung Chang Gung Memorial Hospital, Keelung, Taiwan; 5Population Health Data Center, National Cheng Kung University, Tainan, Taiwan; 6Department of Pharmacy Systems, Outcomes and Policy, College of Pharmacy, University of Illinois at Chicago, Chicago, IL, USA; 7Chang Gung University College of Medicine, Taoyuan, Taiwan*These authors contributed equally to this workCorrespondence: Shu-Chen Liao, Department of Emergency Medicine, Keelung Chang Gung Memorial Hospital, 222, Maijin Road, Anle District, Keelung City, 204, Taiwan, Tel +886-2-24313131, Email [email protected]: To validate the International Classification of Diseases, 10th Revision (ICD-10) codes for Guillain-Barré syndrome (GBS) in Taiwan’s insurance claims database.Methods: We identified adult patients hospitalized at any Chang Gung Memorial Foundation branch hospital between January 1st, 2017, and December 31st, 2022, with ICD-10 code G61.0 in any of the five discharge diagnosis positions, indicating possible Guillain-Barré syndrome. We then validated the possible GBS diagnosis using data from electronic medical records of the identified patients, based on the diagnostic criteria established by the National Institute of Neurological Disorders and Stroke. We determined the positive predictive values (PPV) of various operational definitions, including the position (primary or other) where the code was recorded in the discharge diagnosis, nerve conduction study (NCS) claims, and / or specific GBS treatments.Results: The final validation cohort of 484 patients with ICD-10 code for GBS in the discharge diagnosis was found to include 368 true GBS patients. Identifying inpatients using only the ICD-10 code for GBS in any of the five positions for discharge diagnosis yielded a PPV of 76.0%. With more restrictive definitions (primary diagnosis only, or requiring additional claims for NCS and/or treatments), the PPV tended to increase, but with fewer true GBS patients identified. Using ICD-10 GBS code in the primary diagnosis plus NCS and treatment claims yielded the highest PPV (98.3%); however, 140 (38.0%) of the true GBS patients were missed using this definition. In contrast, using the ICD-10 GBS code in any position, plus claims for NCS, achieved a relatively good PPV (85.8%) with minimal loss of true GBS patients (13, ie, 3.5%).Conclusion: In Taiwan’s NHI claims data, identifying true GBS patients using only the ICD-10 code yielded a PPV of 76.0%; however, adding claims for diagnostic procedure and GBS treatment increased the PPV to 98.3%.Keywords: validation, positive predictive value, diagnostic codes, ICD-10, Guillain-Barré syndrome
