|Home | About | Journals | Submit | Contact Us | Français|
Acute kidney injury (AKI) is common and portends mortality in several neonatal cohorts. Fluid overload is independently associated with poor outcomes in children and adults but has not been extensively studied in neonates.
Between February 2010 and May 2011, we followed 58 neonates who met the following criteria: birth weight >2,000 g, gestational age ≥34 weeks, 5-min Apgar ≤7, and parental consent. Serum creatinine (SCr) was measured daily for first 4 days of life. AKI was defined as a rise in SCr of >0.3 mg/dl or persistent SCr above 1.5 mg/dl.
AKI was present in 9/58 (15.6 %) neonates and was associated with higher birth weight, being male, lower 5-min Apgar scores, lower cord pH, delivery room intubation, and absence of maternal pre-eclampsia. Percent weight accumulation at day 3 of life was higher in those with AKI [median= 8.2, interquartile range (IQR)=4.4–21.6)] than without AKI (median= −4) (IQR=−6.5 to 0.0) (p<0.001). Infants with AKI had lower survival rates than those without AKI [7/9 (72 %) vs. 49/49 (100 %) (p<0.02)].
AKI incidence in this neonatal population is similar to other neonatal cohorts. Near-term/term infants with AKI have a higher mortality rate and a net positive fluid balance over the first few days of life.
Acute kidney injury (AKI) is common in the neonatal intensive care unit (NICU). Based on single-center prospective and retrospective data, the incidence of AKI is 19 % in premature infants , 39 % in newborns who receive hypothermia for severe perinatal asphyxia , 71 % in newborns with congenital diaphragmatic hernia who receive extracorporeal membrane oxygenation (ECMO) , and 52 % in newborns who undergo cardiopulmonary bypass for congenital heart disease . In all these neonatal cohorts, mortality rate is higher in those with AKI, whereas some studies suggest that this association is independent of other confounders that could also account for increased mortality rates [1, 5], In addition, AKI is an important risk factor for chronic kidney disease (CKD), whereby the rates of CKD are alarmingly high in children [6, 7] and adult  hospital survivors. A comprehensive description of the epidemiology of AKI has not been performed in sick late-term/ term infants admitted to level 2 or 3 neonatal intensive care unit.
The cohort described herewith were newborns who ≥34 weeks’ gestational age (GA), had a birth weight >2,000 g, were born with 5-min Apgar scores ≤7, and were admitted to our level 2 or 3 NICU. Although the relative risk of mortality for near-term infants is lower than premature infants, near-term newborns still comprise about 10 % of all neonatal deaths in the United States. Improving our understanding of the incidence, and outcomes ascribed to AKI could lead to preventive and therapeutic interventions that could improve outcomes . Over the last decade, it has become apparent that fluid overload is associated with mortality in pediatric and adult patients with AKI [10–18]. This potentially modifiable risk factor has not been thoroughly assessed in any neonatal population. Some neonatal data suggest that premature infants with fluid accumulation are at high risk for chronic lung disease [19, 20]. We describe the association between AKI and fluid overload in the stated cohort. We conducted a 15- month prospective cohort study of sick near-term/term infants to evaluate four distinct questions: (1) What is the incidence of AKI in this cohort? (2) What are the risk factors associated with AKI? (3) Do infants with AKI have higher fluid overload than those without AKI? (4) Is AKI associated with mortality? We hypothesized that infants with AKI will have high mortality rates, higher degree of fluid overload, and distinct perinatal risk factors compared with infants without AKI.
This prospective cohort study was conducted between January 2010 and March 2011 in the regional quaternary NICU located on the University of Alabama at Birmingham (UAB) campus. Criteria for inclusion included birth weight >2,000 g, GA ≥34 weeks, 5-min Apgar score ≤7, and parental informed consent. Infants were excluded if they did not survive to 24 h of life, had known congenital abnormality of the kidney, or were undergoing whole-body hypothermia for severe asphyxia. The University of Alabama at Birmingham’s Institutional Review Board approved the study.
Enrolled infants were followed prospectively from the time of birth until 36 weeks postmenstrual age (PMA) or hospital discharge, whichever occurred first. Serum creatinine (SCr) was obtained daily during the first 4 days of life and measured using the Jaffe reaction. AKI was defined as an acute rise in SCr of at least 0.3 mg/dl within 48 h [stage 1 of the Acute Kidney Injury Network (AKIN) definition]. Each day’s SCr was compared with the lowest previous SCr measured. Those with a persistent SCr ≥1.5 mg/dl for at least 3 days after birth were also considered to have AKI. The degree of fluid accumulation was assessed by determining the percentage weight change at 3 days of life. Infants who were discharge home or were in stable good/stable condition at 36 weeks PMA were classified as survivors.
Descriptive statistics were performed to determine differences between neonates with and without AKI. Shapiro–Wilk test and normal probability plot were used to test for normality of data. Normally distributed continuous variables were compared using Student’s t test. Those without a normal distribution were compared using the Mann–Whitney U test. Categorical variables were analyzed using Fisher’s exact test. Categorical variables with multiple variables were compared using Mantel Haenszel chi-square test. Correlation analysis using linear regression compared percentage weight change at 3 days of life with cord pH. SAS version 9.2 (SAS Institute Inc., Cary, NC, USA) was used for all statistical analysis.
Of the 129 infants screened for enrollment who met inclusion criteria, parents of 58 infants consented and parents of 71 infants did not consent to participate in the study (35 were not available, 30 were not interested, four were transferred to other units, two unknown) (Fig. 1). AKI occurred in 9/58 (15.6 %) consented patients (Fig. 1). There were no significant differences in race, gender, birth weight, GA, and 1-min or 5-min Apgar scores between screened infants with parental consent compared with those who did not consent to participate (Table 1), ensuring that consented infants represented the entire cohort.
Differences in maternal and infant demographics between patients with and without AKI are shown in Table 2. Infants whose mothers had pre-eclampsia were less likely to have AKI. Infants with AKI had lower birth weight, were more likely to be male, had lower Apgar score at 5 min, were born with a lower cord pH, and were more likely to require mechanical ventilation in the delivery room. Survival was lower in those with than those without AKI [7/9 (77 %) vs. 49/49 (100 %); p<0.02] (Table 3). Those with AKI had higher percent weight change at day 3 compared with those without AKI [median=8.2, interquartile range (IQR)=4.4– 21.6] vs. (median=−4, IQR=−6.5 to 0.0) (p<0.001) (Fig. 2).
In order to explore the association between severity of perinatal depression and fluid overload, we used cord blood pH as a surrogate for perinatal depression and percent weight change at day 3 as a surrogate for fluid overload. For this analysis, data was available for 24/49 infants without AKI (two were not weighed at 3 days of life; 23 did not have perinatal pH available) and from 7/9 infants with AKI [one excluded because he died on day 2 of life and no weight data at day 3 of life; the other was an extreme outlier, as he had >40 % fluid overload (twice that of any other infant) and died at day 5 of life]. For those with AKI, there was a tight correlation between percent fluid weight gain and cord blood pH (F value 23.9; df=1; adjusted R2=0.79; p values<0.005). For those without AKI, there was a no statistical significance (F value 1.1; df=1; adjusted R2= 0.003; p values<0.31) (Fig. 3).
In this prospective cohort study, we show that sick near-term/ term infants admitted to the NICU have a 15 % incidence of AKI. Maternal pre-eclampsia was a protective risk factor, whereas low birth weight, lower GA, male gender, lower 5-min Apgar score, lower cord pH, and intubation at delivery were associated with AKI. Those with AKI had higher mortality rates and higher weight accumulation (fluid overload) over the first 3 days of life.
We recently performed a similar prospective cohort study at the same institution for very-low-birth-weight (VLBW) infants (≤1,500 g)  and found similar AKI rates as in this cohort (18 % and 15 %; respectively). In both cohorts, mortality rates of those with AKI were higher than in those without; however, VLBW infants with AKI had higher mortality rates than infants in the cohort reported here (42 % vs. 22 %; respectively). Unlike the VLBW cohort, in which we were able to show that AKI was an independent predictor of mortality, we were unable to ascertain whether AKI was an independent mortality predictor in this cohort.
Attention to the degree of fluid accumulation has gained considerable attention as a potentially modifiable independent mortality predictor in pediatric and adult studies [10–18]. Although this study was not powered sufficiently to determine the association between fluid overload and mortality, we show the correlation between the degree of acidosis at birth with percent weight change over the first 3 days. This suggests a relationship between perinatal distress and cumulative fluid overload. This is likely because perinatal distress predisposes to AKI, which then leads to fluid overload; although other hypotheses, such as extensive fluid provision, the syndrome of inappropriate antidiuretic hormone (SIADH), may explain this correlation. Further studies on the etiology and role that fluid provision and fluid overload have on survival in neonates with AKI are greatly needed. Interestingly, we again report the negative association between maternal pre-eclampsia and neonatal AKI. Previously we reported that only 17% of premature infants with AKI had mothers with pre-eclampsia, whereas those without AKI had maternal pre-eclampsia rates of 35 % . There are several possible explanations for these findings. First, pre-eclampsia may represent a response to a problem of maternal origin rather than a primary fetal/neonatal problem; thus, the cause for prematurity or initial illness may be an issue with the mother, not the infant. A second possibility is that preeclampsia has a protective effect on AKI. This could arise from ischemic preconditioning whereby small, repeated ischemic events can prevent AKI [21–23]. A third possibility may be that a biochemical feature that induces pre-eclampsia protects against neonatal AKI. The reasons for this association cannot be deduced from data available in this study; however, studies to elucidate its cause may provide important insights.
The strengths of this study include the robust evaluation of serum for SCr (as 62 % of infants had at least two samples drawn in the first 3 days of life), the prospective nature of data collection, and the evaluation of a neonatal population that has not been previously described in the literature. Despite these strengths, this study is limited by being a single-center study with a relatively small sample size. In addition, we acknowledge the possibility of misclassification bias given the known problems with SCr to define AKI , including the use of the Jaffe reaction in neonates, which may have elevated bilirubin levels. In addition, because our hospital does not routinely document strict output in neonates, we did not include urine output as part of our AKI definition. Nonetheless, we used a contemporary definition of AKI that has been reported in several neonatal cohorts . Future work to test definitions that incorporate novel AKI biomarkers in large, prospective multicenter studies are greatly needed. We acknowledge that weight percent change may not be the most accurate method to document the cumulative percent of fluid overload, as most reports used daily intake and output to describe fluid overload.
We chose to use percent weight change as a surrogate for cumulative fluid overload because documentation of the exact fluid intake and output are not always accurate and, as in many of our infants, only the number of wet diapers are recorded in the medical record. We justify the use of percent weight change because Selewski et al. showed that using weight-based estimates of fluid overload is extremely similar to intake methods in pediatric cohorts . Similarly, we acknowledge that cord blood pH may not be the best surrogate for perinatal depression. Nonetheless, the association between perinatal depression and cumulative fluid overload in those with AKI is present using these surrogates and deserves further evaluation. Future neonatal AKI studies need to incorporate accurate input and output data to best describe the cohort and determine the etiology of fluid overload that could be due to excessive fluid provision, decreased urine output, SIADH, or other causes.
In conclusion, sick near-term/term infants admitted to the level 2 and 3 NICU are at risk for AKI. Although the exact predictability of survival in those with AKI cannot be ascertained by this small, single-center study, we found that neonates with AKI have a higher mortality rate. Fluid accumulation occurs in neonates with AKI and is associated with cord blood pH. Hypothesis-driven trials to prevent AKI, treat existing AKI, and test management strategies to improve outcomes in neonates with AKI are needed.
Funding source: This study was supported by grants provided by the American Society of Nephrology Career Development Grant, the Kaul Pediatric Research Institute and a pilot and feasibility grant from the NIH-sponsored O’Brien Center for Acute Kidney Injury Research (www.obrienaki.org). Dr. Askenazi is a consultant and serves as a speaker for Gambro Renal Products.