Background Admission blood lactate has previously guided therapy and is a valid estimator of prognosis in children with cerebral malaria (CM). Since bolus intravenous fluid administration to children with febrile coma and signs of hypovolemia (including high blood lactate) may be harmful, the use of blood lactate to guide therapy has waned. Its utility as a prognostic biomarker, either at admission or during hospitalization, remains unclear. Methods We performed a retrospective cohort study of 1674 Malawian children with CM hospitalized from 2000–2018 who had blood lactate measurements every 6 hours for the first 24 hours. We evaluated the strength of association between admission lactate or values measured at any time point in the first 24 hours post-admission, and outcomes (death or neurologic disability in survivors). We assessed the optimal duration of lactate monitoring of children with CM. Results When lactate is analyzed as a continuous variable, children with CM who have higher values at admission have a 1.05-fold higher odds (95% CI: 0.99–1.11) of death compared to those with lower lactate values. Children with higher blood lactate at 6 hours have 1.16-fold higher odds (95% CI: 1.09–1.23) of death, compared to those with lower values. If lactate levels are dichotomized into hyperlactatemic (lactate > 5.0 mmol/L) or not, the strength of association between admission lactate and death increases (OR = 2.49, 95% CI: 1.47–4.22). Neither blood lactate levels obtained after 18 hours post-admission nor trends in lactate concentrations during the first 24 hours of admission are associated with outcomes. Blood lactate during hospitalization is not associated with adverse neurologic outcomes in CM survivors. Conclusions In children with CM, blood lactate is associated with death but not neurologic morbidity in survivors. To comprehensively estimate prognosis, blood lactate in children with CM should be assessed at admission and for 18 hours afterwards.

Hypoglycemia, defined as a blood glucose < 2.2 mmol/L, is associated with death in pediatric cerebral malaria (CM). The optimal duration of glucose monitoring in CM is unknown. We collected data from 1,674 hospitalized Malawian children with CM to evaluate the association between hypoglycemia and death or neurologic disability in survivors. We assessed the optimal duration of routine periodic measurements of blood glucose. Children with hypoglycemia at admission had a 2.87-fold higher odds (95% CI: 1.35–6.09) of death and, if they survived, a 3.21-fold greater odds (95% CI: 1.51–6.86) of sequelae at hospital discharge. If hypoglycemia was detected at 6 hours but not at admission, there was a 7.27-fold higher odds of death (95% CI: 1.85–8.56). The presence of newly developed hypoglycemia after admission was not independently associated with neurological sequelae in CM survivors. Among all new episodes of blood sugar below a treatment threshold of 3.0 mmol/L, 94.7% occurred within 24 hours of admission. In those with blood sugar below 3.0 mmol/L in the first 24 hours, low blood sugar persisted or recurred for up to 42 hours. Hypoglycemia at admission or 6 hours afterward is strongly associated with mortality in CM. Children with CM should have 24 hours of post-admission blood glucose measurements. If a blood glucose less than the treatment threshold of 3.0 mmol/L is not detected, routine assessments may cease. Children who have blood sugar values below the treatment threshold detected within the first 24 hours should continue to have periodic glucose measurements for 48 hours post-admission.