Importance: The incidence and associated outcomes of recurrent acute kidney injury (rAKI) in neonates remain largely unknown.
Objective: To determine the incidence, risk factors, and clinical outcomes associated with rAKI in critically ill neonates.
Design, setting, and participants: This cohort study was a secondary analysis of the multicenter, international Assessment of Worldwide Acute Kidney Injury Epidemiology in Neonates retrospective study. Comparisons were made among neonates with no AKI, a single AKI episode (sAKI), and rAKI. All neonates younger than 14 days who were admitted between January 1 and March 31, 2014, to 24 participating level II to IV neonatal intensive care units and received intravenous fluids for at least 48 hours were considered for inclusion. Neonates with congenital heart disease requiring surgery within the first week of life, lethal chromosomal anomalies, death within 48 hours of admission, or severe congenital kidney abnormalities were excluded. Data were analyzed from May 23, 2022, to December 8, 2023.
Exposure: Recurrent AKI using the neonatal Kidney Disease: Improving Global Outcomes criteria. Determination of each rAKI required a complete return to the baseline serum creatinine level that defined the prior AKI episode.
Main outcomes and measures: Incidence and risk factors of rAKI and associations of rAKI with length of stay (LOS; ie, birth to hospital discharge) and mortality.
Results: The study cohort (n = 2162) included 1233 male neonates (57.0%). Gestational age distribution was less than 29 weeks for 276 neonates (12.8%), 29 to less than 36 weeks for 958 (44.3%), and 36 weeks or older for 928 (42.9%). Of 605 neonates with AKI, 133 (22.0%) developed rAKI with risk factors including younger gestational age, lower birthweight, and higher stage of initial AKI. Infants with rAKI experienced longer median LOS (no AKI, 17 [IQR, 8-34] days; sAKI, 18 [IQR, 9-45] days; rAKI, 60 [IQR, 25-109] days; P < .001). Time-varying Cox proportional hazards regression models suggest rAKI is independently associated with a lower hazard of discharge (adjusted hazard ratio, 0.7 [95% CI, 0.6-0.9]; P = .01) when compared with sAKI, but mortality did not differ between groups (adjusted hazard ratio, 1.4 [95% CI, 0.6-3.0]; P = .44).
Conclusions and relevance: In this cohort study, neonatal rAKI was independently associated with longer LOS when compared with sAKI, suggesting that rAKI in neonates may be an important clinical distinction warranting further study and careful monitoring after an initial AKI episode.