Can Serum Lactate on Admission to Intensive Care Predict Length of Stay? A Retrospective Analysis of Critically Ill Paediatric Oncology Patients

Abstract

Introduction: Nearly 40% of hospitalised oncology patients are admitted to the paediatric intensive care unit (PICU). Clinicians utilise serum lactate as a biomarker to guide resuscitation and ongoing management of critically ill children. The significance of raised serum lactate remains unknown in oncology patients. Those who require an admission to the PICU have relatively poorer outcomes compared to those only managed on hospital wards, where there may be an increased tendency to apply additional interventions to normalise lactate. We assessed the association between serum lactate and the length of stay in PICU in paediatric oncology patients.

Patients and methods: We conducted a retrospective observational study of patients admitted to the oncology service, who had at least one PICU admission, at a tertiary children's hospital between 2015 and 2020. The outcome was PICU length of stay (LOS). Multilevel regression modelling assessed the association between lactate measured at PICU admission and PICU LOS, controlling for patient demographic and clinical covariates.

Results: There were 633 admissions to PICU from 426 children. Of these 633 admissions, 197 (31.1%) had hyperlactatemia (serum lactate ≥ 2 mmol/L) on admission, 229 (36.2%) had normal lactate levels (< 2 mmol/L) and the remaining 207 (32.7%) did not have lactate measured on admission. Of the 426 patients, 292 (68.5%) were diagnosed with solid tumour, 82 (19.2%) hematologic malignancy, and 52 (12.2%) had received a bone marrow transplant (BMT). The median PICU length of stay was 29 h (interquartile range 21, 68). Hyperlactatemia was associated with an average 1.25-fold (95% CI [1.04, 1.51]) increase in PICU LOS when we assumed that missing lactate indicated absence of hyperlactatemia. Interaction effects revealed that this association was only present in the solid tumour group. In sensitivity analyses including only admissions with measured lactate, this effect was reduced and no longer significant (β 1.20, 95% CI [0.99, 1.46]).

Conclusion: After adjusting for clinical and demographic covariates, increased lactate at PICU admission was associated with PICU LOS in paediatric solid tumour patients. While further larger prospective studies are required, serum lactate levels may serve as a predictive biomarker for PICU length of stay in children with solid tumours.

Keywords: lactate; length of stay; paediatric intensive care; paediatric oncology; sepsis.

FIGURE 1

Estimated adjusted marginal means and 95% CI of PICU LOS for children with and without hyperlactatemia within each of the three diagnostic groups. PICU LOS, paediatric intensive care unit length of stay.