(144) Validation of Diagnostic Codes to Identify Carbon Monoxide Poisoning in Taiwan's Claim Data

Publication Number: 475

Background: Previous studies have been published to better understand the epidemiological features, risk factors, clinical outcomes and treatment effectiveness of carbon monoxide poisoning (COP) from secondary data sources. However, possible impacts of misclassification bias may threaten the study validity if the accuracy of disease diagnosis codes is unclear in these secondary database research.

Objectives: To construct and validate the case definitions for carbon monoxide poisoning (COP) based on the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) and Tenth Revision, Clinical Modification (ICD-10-CM) diagnosis codes in the Taiwan’s claims data.

Methods: We included patients newly receiving the carboxyhemoglobin (COHb) test for any clinical diagnoses from four Chang Gung Memorial Hospitals in Taiwan during 2011-2020. We performed the 10% random sampling from the original cohort, and two independent clinical doctors ascertained the diagnosis of COP by reviewing medical records as the reference standard. We estimated the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of various case definitions (e.g., ICD-9-CM code, ICD-10-CM code and the uses of hyperbaric oxygen therapy) in the claim data.

Results: We randomly selected the 557 cases from the original cohort of a total of 5,571 cases newly receiving COHb test in the study hospitals. For 278 cases with ICD-9-CM codes of 986, 90, 35, and 9 cases were determined to be true-positive, false-positive, and false-negative episodes of COP, respectively. For 279 cases with ICD-10-CM codes of T58, 111, 34, and 6 cases were determined to be true-positive, false-positive, and false-negative episodes of COP, respectively. The sensitivity, specificity, PPV and NPV for COP using ICD-9-CM codes were 90.9%, 80.4%, 72% and 94.1%, respectively; the sensitivity, specificity, PPV and NPV for COP using ICD-10-CM codes and 94.9%, 79%, 76.6% and 95.5%, respectively. Among a total of 69 false-positive cases, the most frequent reason for miscoding was burn (n=21), followed by other substance poisoning (n=9), and peripheral vertigo (n=6). The PPV generally increased when COP was defined as a primary diagnosis or as ICD-CM codes of 986 or T58 plus oxygen therapy use. The PPV could reach 100% if patients with ICD-CM code of 986 or T58 received hyperbaric oxygen therapy during hospitalizations.

Conclusions: The COP-related ICD-CM codes may not accurately identify patients with COP, but the accuracy could be improved by adding the oxygen therapy to the ICD-CM codes in Taiwan’s claim data.