(182) Validation of Rhabdomyolysis in Claims Using Electronic Health Data in Taiwan

Background: Rhabdomyolysis, a critical clinical syndrome caused by muscle breakdown, can lead to serious complications such as electrolyte disorders and acute kidney injury. Given its low incidence, retrospective analysis using large-scale databases is an efficient approach. However, there have been no studies validating the diagnosis of rhabdomyolysis in these databases.

Objectives: To validate the diagnosis of rhabdomyolysis in Taiwan's health data.

Methods: A retrospective study was conducted using the Chang Gung Medical Research Database (CGRD), which represents 6% of Taiwan’s population. We analyzed patient laboratory data from 2016 to June 2023. Rhabdomyolysis was defined by three severity levels: 1) creatine kinase (CK) >1,000 IU/L within 72 hours of hospitalization, 2) CK >1,000 IU/L with elevated myoglobin, LDH, K+, Cr, AST within 72 hours, and 3) CK > 5,000 IU/L within 72 hours. As diagnosis fields from hospitals are required to be reported to the national health claims, we assessed the use of ICD-10-CM code M62.82 in diagnosis fields under two scenarios: from any inpatient diagnosis field, and from any inpatient diagnosis field with emergency department admission. The sensitivity and positive predictive value (PPV) of the M62.82 code were calculated for each scenario.

Results: A total of 8211, 5858, and 1929 medical visits were identified for each reference standard, respectively. The test cohorts included 1516 inpatient visits with a diagnosis of M62.82. Among these, 1437 visits involved patients admitted through the ER. Sensitivity ranged from 10% to 27% across the reference cohorts, while PPV varied from 33% to 62%. The stricter the reference cohort definition, the higher the sensitivity but lower the PPV for identifying rhabdomyolysis using M62.82. Sensitivity and PPV showed similar patterns in both test cohorts.

Conclusions: The use of ICD-10 code M62.82 to identify rhabdomyolysis patients shows generally low performance, suggesting a significant risk of misclassification when defining rhabdomyolysis in health claims using ICD-10 codes. Future research should focus on identifying relevant diagnosis combinations and developing more effective algorithms in claims databases to enhance the accuracy of related database studies.