Acute kidney injury e-alerts in pregnancy: rates, recognition and recovery
Rouvick Gama, Katherine Clark, Mahua Bhaduri, Amanda Clery, Kelly Wright, Priscilla Smith, Hayley Martin, Royce P Vincent, Satish Jayawardene, Kate Bramham
Abstract
BACKGROUND: Acute kidney injury (AKI) in pregnancy (Pr-AKI) is associated with substantial maternal morbidity and mortality. E-alerts are routinely used for detection of AKI in non-pregnant patients but their role in maternity care has not been explored. METHODS: All pregnant or postpartum women with AKI e-alerts for AKI Stages 1-3 (Kidney Disease Improving Global Outcomes (KDIGO) criteria) were identified at a tertiary centre >2 years. Two women matched by delivery date for each case were selected as controls. AKI stage, recognition of AKI, pregnancy outcomes, renal recovery, AKI aetiology and risk factors were extracted from electronic patient records. RESULTS: 288 of 11 922 (2.4%) women had AKI e-alerts, of which only 118 (41%) were recognized by the obstetric team. Common Pr-AKI causes included infection (48%), pre-eclampsia (26%) and haemorrhage (25%), but no cause was identified in 15% of women. Renal function recovered in 213 (74%) women, but in 47 (17%) repeat testing was not undertaken and 28 (10%) did not recover function. Hypertensive disorders of pregnancy and Caesarean section were associated with increased incidence of Pr-AKI compared with controls. CONCLUSIONS: Pr-AKI e-alerts were identified in ∼1 in 40 pregnancies. However, a cause for Pr-AKI was not identified in many cases and e-alerts may have been triggered by gestational change in serum creatinine. Pregnancy-specific e-alert algorithms may be required. However, 1 in 10 women with Pr-AKI had not recovered kidney function on repeat testing. Better understanding of long-term impacts of Pr-AKI on pregnancy and renal outcomes is needed to inform relevant Pr-AKI e-alert thresholds.