Despite advances in resuscitation, mortality remains high after cardiac arrest (CA). While serum phosphate is linked to outcomes in critically ill patients, its prognostic value specifically in post-CA populations remains unclear. This study investigated the association between admission serum phosphate levels and 90-day mortality in CA patients. We conducted a retrospective cohort study using the MIMIC-IV database (2008–2019), including adult CA patients admitted to the ICU. The primary outcome was 90-day all-cause mortality. Multivariable Cox proportional hazards models adjusted for potential confounders were used to estimate hazard ratios (HRs). Kaplan–Meier curves and restricted cubic spline (RCS) models were employed to assess survival differences and nonlinear dose–response relationships. A total of 1,079 adult patients with CA were included (median age: 67 years; 61.9% male). Compared with survivors, non-survivors were older, had higher severity scores (SOFA, CCI), and exhibited significant disturbances in multi-organ function markers, including elevated serum phosphate, creatinine, liver enzymes, and coagulation parameters (all P < 0.05). Notably, non-survivors exhibited significantly shorter ICU and hospital lengths of stay, likely reflecting higher early mortality rates. Kaplan–Meier analysis showed distinct survival curves across phosphate quartiles (P < 0.001). In the fully adjusted Cox model, the highest phosphate quartile (Q4 ≥ 5.21 mg/dL) was independently associated with a 45% increased risk of 90-day mortality compared to the lowest quartile (HR = 1.450, 95% CI: 1.098–1.915, P = 0.009). Restricted cubic spline (RCS) analysis revealed that while the unadjusted association was J-shaped, multivariable adjustment uncovered a linear positive correlation, with mortality risk significantly increasing when admission phosphate levels exceeded 4.0 mg/dL. Elevated admission serum phosphate is an independent predictor of 90-day mortality in post-CA patients. Beyond traditional markers, it offers complementary prognostic value, supporting early risk stratification and potentially guiding individualized management strategies in this critically ill population.
Fu et al. (Fri,) studied this question.