Serum Calcium-Phosphate Product and the Risk of Kidney Replacement Therapy or In-Hospital Death in Critically Ill Patients with Rhabdomyolysis: a Multicohort Database Study

Abstract Background. Rhabdomyolysis may be complicated by kidney replacement therapy (KRT) or death. We evaluated whether the early serum calcium-phosphate product (Ca×P) was associated with these outcomes in critically ill adults with rhabdomyolysis. Methods. We conducted a retrospective multicohort study using MIMIC-IV, eICU, and NWICU. Adults with maximum creatine kinase (CK) ≥1000 U/L during ICU hours 0–24 were eligible. Total calcium and phosphate were paired when measured within 6 h; the earliest valid pair completion time was selected, with the smallest gap used for ties. A 24-h landmark excluded KRT by 24 h, end-stage kidney disease/chronic dialysis, and ICU stay ≤24 h. The primary outcome was incident KRT after 24 h or in-hospital death after 24 h. Logistic models adjusted for age, sex, and ln(CK). Results. Final cohorts included 2,744 MIMIC-IV, 1,644 eICU, and 109 NWICU patients, with 541, 370, and 18 composite events. Adjusted ORs per 10 mg²/dL² higher Ca×P were 1.416 (95% CI 1.324–1.514), 1.337 (1.246–1.434), and 1.515 (1.080–2.124), respectively. The pooled fixed-effect OR was 1.380 (1.316–1.448; I²=0%). In MIMIC-IV, restricted cubic spline analysis showed an overall association and evidence of nonlinearity (both P<0.001). Conclusions. Higher early Ca×P was consistently associated with the composite of incident KRT or in-hospital death across three ICU cohorts. The nonlinear MIMIC-IV pattern supports further evaluation of Ca×P as a risk marker but does not establish causality or a treatment threshold.