Development of Urinalysis Screening Criteria for Rhabdomyolysis in Acute Kidney Injury

KEY POINTS: Rhabdomyolysis accounts for approximately 7%-10% of AKI cases but is frequently missed because creatine kinase testing is rarely performed. Automated electronic screening of urinalysis parameters with clinical decision support to order creatine kinase may be a viable mechanism to increase rhabdomyolysis diagnosis. BACKGROUND: Rhabdomyolysis (RM) is the etiology in 7%-10% of cases of AKI and benefits from early, aggressive volume repletion. RM as an etiology of AKI is often missed in clinical setting as it requires specific diagnostic testing (creatine kinase [CK] testing). We evaluated various criteria that could be applied to urinalysis (UA) to identify individuals at elevated likelihood of RM. METHODS: A retrospective electronic chart review of adult inpatients hospitalized at the Yale New Haven Hospital was performed between January 2020 and December 2022. A total of 12 screening criteria were formulated using combinations of heme (≥1+, ≥2+, and ≥3+) and red blood cells (RBCs; ≤5, ≤10, ≤20, and ≤30) per high power field thresholds from the closest UA before AKI. The accuracy was assessed using post-AKI CK >1000 U/L as the definition of RM diagnosis. RESULTS: A total of 12,273 patients were included of which 20.5% patients had serum CK levels checked after AKI. Of those, 222 (8.8%) met the diagnostic criteria for RM. The median time from laboratory AKI diagnosis to CK measurement was 2.7 (0.9-6.5) days. UA criterion 1.4 (≥1 heme and ≤30 RBCs) met by 35.1% demonstrated the highest sensitivity for RM. UA criterion 3.1 (≥3+ heme and ≤5 RBCs) met by 2.0% demonstrated a specificity of 96.7% and sensitivity of 16.7% included the highest proportion of patients with CK measurement (45.3%), and more than a quarter of those tested had levels consistent with RM. CONCLUSIONS: RM-suspected patients with greater heme/RBC discrepancy on UA were more likely to be tested for serum CK levels and diagnosed with RM. CK testing was uncommon under all definitions, suggesting missed RM cases. Automated electronic screening of UA parameters with clinical decision support to order CK may be a viable mechanism to increase the diagnosis of RM among AKI patients.