Acute kidney injury, persistent kidney disease, and post-discharge morbidity and mortality in severe malaria in children: A prospective cohort study

Abstract

Background: Globally, 85% of acute kidney injury (AKI) cases occur in low-and-middle-income countries. There is limited information on persistent kidney disease (acute kidney disease [AKD]) following severe malaria-associated AKI.

Methods: Between March 28, 2014, and April 18, 2017, 598 children with severe malaria and 118 community children were enrolled in a two-site prospective cohort study in Uganda and followed up for 12 months. The Kidney Disease: Improving Global Outcomes (KDIGO) criteria were used to define AKI (primary exposure) and AKD at 1-month follow-up (primary outcome). Plasma neutrophil gelatinase-associated lipocalin (NGAL) was assessed as a structural biomarker of AKI.

Findings: The prevalence of AKI was 45·3% with 21·5% of children having unresolved AKI at 24 h. AKI was more common in Eastern Uganda. In-hospital mortality increased across AKI stages from 1·8% in children without AKI to 26·5% with Stage 3 AKI (p < 0·0001). Children with a high-risk plasma NGAL test were more likely to have unresolved AKI (OR, 7·00 95% CI 4·16 to 11·76) and die in hospital (OR, 6·02 95% CI 2·83 to 12·81). AKD prevalence was 15·6% at 1-month follow-up with most AKD occurring in Eastern Uganda. Risk factors for AKD included severe/unresolved AKI, blackwater fever, and a high-risk NGAL test (adjusted p < 0·05). Paracetamol use during hospitalization was associated with reduced AKD (p < 0·0001). Survivors with AKD post-AKI had higher post-discharge mortality (17·5%) compared with children without AKD (3·7%).

Interpretation: Children with severe malaria-associated AKI are at risk of AKD and post-discharge mortality.

Funding: This work was supported by the National Institutes of Health National Institute of Neurological Disorders and Stroke (R01NS055349 to CCJ) and the Fogarty International Center (D43 TW010928 to CCJ), and a Ralph W. and Grace M. Showalter Young Investigator Award to ALC.

Keywords: Acute kidney disease; Acute kidney injury; Blackwater fever; Cerebral malaria; Disability; Malaria; Mortality; NGAL; Neurologic deficit; Sub-Saharan Africa.

Figure 1

Study flow diagram. During hospitalization, 598 children with severe malaria and at least one creatinine measure available had acute kidney injury (AKI) defined using the Kidney Disease: Improving Global Outcomes (KDIGO) criteria based on a 1.5 fold increase in creatinine from estimated baseline or a 0.3 mg/dL change in creatinine over the first 24 h of hospitalization. AKI was stratified into stage 1 AKI or severe AKI (Stage 2 or Stage 3). Among survivors with a repeat creatinine value at 1-month follow-up we defined acute kidney disease (AKD) which reflects a 1.5-fold increase in creatinine from the estimated 1-month baseline creatinine or an estimated glomerular filtration rate < 90 mL/min per 1.73m² using the Bedside Schwartz equation. Kidney disease was defined as a 1.5-fold increase in creatinine from the estimated 1-month baseline creatinine or an estimated glomerular filtration rate < 90 mL/min per 1.73m² using the Bedside Schwartz equation in the community children. Children were recruited at Mulago National Referral Hospital in Kampala, Central Uganda and Jinja Regional Referral Hospital in Jinja, Eastern Uganda.

Figure 2

Proximal tubular injury is more common in children with persistent AKI or fatal outcomes and is elevated in children with AKD at 1-month follow-up. (Left) Bar graph depicting the frequency of a high-risk NGAL test result (NGAL ≥ 300 ng/mL) on admission in community children, children with severe malaria without AKI over hospitalization, children with AKI separated by AKI recovery status at 24 h (in gray shaded area), and children who died in-hospital. (Right) Scatter plot and median depicting NGAL levels at 1-month follow-up in survivors based on the presence of AKD. Analysis by non-parametric Wilcoxon rank sum test where **** adjusted p < 0·0001.

Figure 3

Forest plot evaluating predictors of AKD in severe malaria survivors. Forest plot evaluating risk factors for AKD at 1-month follow-up with the unadjusted odds ratio (OR) and 95% CI generated using logistic regression. Use of acetaminophen in hospital was associated with a reduced odds of AKD at 1 month follow-up while children from Jinja, with severe AKI on admission, unresolved AKI at 24 h, a history of blackwater fever and a high-risk NGAL test had an increased odds of AKD. Predictors significant following adjustment for 27 comparisons are indicated in pink.