Hypophosphatemia and adverse composite outcome in mechanically ventilated critically Ill children: A single-center retrospective cohort study

Aim: Hypophosphatemia may impair diaphragmatic contractility and prolong weaning from mechanical ventilation, but pediatric prognostic data are scarce.

Materials and Methods: This single-center retrospective cohort included 206 children (1 month – 18 years) invasively ventilated for at least 48 hours (April 2024 – January 2026). Conditions distorting phosphate homeostasis were excluded. Serum phosphate was measured at admission, day 3 (both required for inclusion), and day 7 in patients still in the pediatric intensive care unit (PICU); hypophosphatemia was defined as < 2.5 mg/dL. The primary outcome was a composite of PICU mortality or prolonged ventilation (≥ 14 days); secondary outcomes were mortality and ventilation duration. Multivariable models were adjusted for age, PRISM III, sepsis, albumin, furosemide, and vasoactive/inotropic exposure. A pre-specified sensitivity analysis used admission hypophosphatemia as the exposure.

Results: Hypophosphatemia occurred at any timepoint in 107 patients (51.9%) and the composite outcome in 71 (34.5%; 46.7% vs 21.2%, p < .001). After adjustment, any-timepoint hypophosphatemia remained associated with the composite outcome (adjusted OR 2.52, 95% CI 1.25 – 5.09, p = .010), driven by the ventilation component (adjusted OR 3.77, 95% CI 1.61 – 8.84, p = .002) rather than mortality (adjusted OR 1.85, p = .158), and with longer ventilation (adjusted ratio 1.57, 95% CI 1.30 – 1.91, p < .001). In the sensitivity analysis the composite association was no longer significant (adjusted OR 1.56, 95% CI 0.79 – 3.09, p = .197), and discrimination was limited (all AUCs < 0.70).

Conclusion: Hypophosphatemia is common in mechanically ventilated children and, when assessed at any timepoint, is associated with mortality or prolonged ventilation; the association is driven by the ventilation component and is lost when restricted to admission values. Hypophosphatemia therefore appears to be predominantly a marker of illness severity rather than an independent, admission-identifiable driver of outcome. Whether serial monitoring and correction improves outcomes requires prospective evaluation.