Acute Kidney Injury in Pediatric Acute Decompensated Heart Failure Shivanand S. Medar, MD1; Daphne T. Hsu, MD2; Jacqueline M. Lamour, MD2; Scott I. Aydin, MD1,2
Objective: Acute kidney injury in adult patients with acute decompensated heart failure is associated with increased mortality. There is limited literature in pediatric patients with acute decompensated heart failure and acute kidney injury. We aim to study acute kidney injury in the pediatric acute decompensated heart failure population and its association with specific outcomes. Design: Retrospective, case-control study. Setting: Cardiac ICU in a children’s tertiary care hospital. Patients: Index admissions of patients younger than 21 years with acute decompensated heart failure between January 2008 and December 2012. Interventions: None. Measurements and Main Results: Index admissions of patients younger than 21 years with acute decompensated heart failure between January 2008 and December 2012 were reviewed, and the presence or absence of acute kidney injury at admission was determined based on the Pediatric Risk, Injury, Failure, Loss, End-Stage criteria. Descriptive statistics and multivariate analyses were performed to determine the association between acute kidney injury and a composite outcome of cardiac transplantation and/or mortality. Fifty-seven patients, with median age 12 years (interquartile range, 1.1, 16), were included for study. The median left ventricular ejection fraction was 27% (interquartile range, 18, 48). Twenty-one patients (36%) underwent cardiac transplantation and five patients (8.7%) died. Of the 57 patients, 44 (77%) had evidence of acute kidney injury (41% Risk; 39% Injury; 20% Failure). Of the 44 patients with acute kidney injury, 25 (57%) met the composite outcome, compared with 1 (7%) without acute kidney injury. Multivariate analyses demonstrated that a left ventricular ejection fraction up to 25% was significantly associated with the presence of acute kidney injury (adjusted odds ratio, 12.3; Division of Critical Care Medicine, The Children’s Hospital at Montefiore, Bronx, NY. 2 Division of Pediatric Cardiology, The Children’s Hospital at Montefiore, Bronx, NY. The authors have disclosed that they do not have any potential conflicts of interest. Address requests for reprints to: Scott I. Aydin, MD, The Children’s Hospital at Montefiore, 3415 Bainbridge Avenue, Rosenthal 1, Bronx, NY 10467. E-mail:
[email protected] Copyright © 2015 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies DOI: 10.1097/PCC.0000000000000412 1
Pediatric Critical Care Medicine
95% CI, 1.4–109; p = 0.03), and acute kidney injury was significantly associated with the composite outcome (adjusted odds ratio, 19.1; 95% CI, 2.3–160; p < 0.001). Conclusions: Acute kidney injury is common during the initial presentation of pediatric patients with acute decompensated heart failure. A left ventricular ejection fraction up to 25% is associated with acute kidney injury. The presence of acute kidney injury in this population is significantly associated with cardiac transplantation and/or death. (Pediatr Crit Care Med 2015; 16:535–541) Key Words: acute kidney injury; cardiomyopathy; cardiorenal syndrome; heart failure; outcomes; transplantation
A
cute decompensated heart failure (ADHF) is an increasing reason for hospitalization for children in the United States. Acute heart failure admissions in children are estimated to total 11,000–14,000 per year with mortality rates as high as 7% (1). Despite the recent introduction of new medical and surgical therapies, readmission rates and mortality remain unchanged (2–4). Renal dysfunction may play an important role in the prognosis of ADHF. In the adult population, approximately 30% of patients with ADHF exhibit evidence of acute kidney injury (AKI) (5). Adult patients with AKI experience increased length stay, higher costs, higher readmission, and in-hospital mortality rates (6–8). To date, AKI and renal function, in general, have not been as well characterized in pediatric ADHF population compared with the adults. Price et al (9) examined a population of children with ADHF and concluded that worsening renal function, based on an increase in serum creatinine (SCr) of 0.3 mg/dL or more, was associated with increased length of stay, need for mechanical circulatory support, and in-hospital mortality. The aim of this study was to assess the prevalence of AKI in a population of children admitted with their index episode of ADHF. We sought to describe risk factors for the development of AKI, as well as the association of AKI to specific patient-related short-term clinical outcomes.
MATERIALS AND METHODS Design We performed a retrospective single-center review of records from the Children’s Hospital at Montefiore and identified all pediatric patients with newly diagnosed primary ventricular www.pccmjournal.org
535
Medar et al
dysfunction who were admitted with ADHF between January 2008 and December 2012. Data were extracted from the patient’s medical record of their index admission for ADHF. The study was approved by the Institutional Review Board of the Albert Einstein College of Medicine with a waiver of parental or patient consent. Patient Population Eligible patients were 21 years old or younger admitted with ADHF. Only the index admission of patients with ADHF was considered for inclusion into the study. We excluded patients older than 21 years, patients with structural heart disease, patients previously or currently treated with renal replacement therapy (RRT), patients with preexisting chronic renal insufficiency, and patients with history of renal and/or heart transplantation. Definition of ADHF ADHF was determined to be new-onset signs and symptoms of ventricular dysfunction that required in-patient treatment. Ventricular dysfunction was defined qualitatively as mildly depressed or worse and quantitatively as a left ventricular ejection fraction (LVEF) up to 55%. In addition, the patients met criteria for ADHF with the presence of fluid overload requiring IV diuretics and/or low cardiac output requiring IV inotropic support or afterload reduction. Definition of AKI Estimated creatinine clearance (eCCL in mL/min/1.73 m2) was calculated by using the Schwartz formula (patients < 1 yr) and modified Schwartz formula (patients > 1 yr) as follows: term infants younger than 1 year: 0.45 (length, cm)/SCr and children: 0.413 (height, cm)/SCr (10–12). For determination of baseline eCCL, the lowest value of SCr in the 6 months preceding study enrollment was obtained from the hospital records. If no previous SCr was available, the patient was assumed to have normal renal function and assigned a baseline eCCL of 100 mL/min/1.73 m2. Pediatric Risk, Injury, Failure, Loss, EndStage (pRIFLE) criteria were assessed at two time points: at admission and at the lowest eCCL. The maximal pRIFLE score was calculated using the lowest eCCL, and patients were classified as having no AKI, class R (Risk; eCCL decrease by 25%), class I (Injury; eCCL decrease by 50%), and class F (Failure; eCCL decrease by 75% or eCCL < 35 mL/min/1.73 m2) (13). If patients received RRT during their admission, the maximal pRIFLE class was listed as F. Data Collected Data were retrospectively collected from hospital medical records. The etiology of ventricular dysfunction, demographic data, SCr values, blood urea nitrogen (BUN) values, and N-terminal pro-brain natriuretic peptide (NT-proBNP) values were obtained at admission and at the time of the maximum level of SCr measured during the hospitalization. LVEF and qualitative assessment of left ventricular function at admission were recorded. The age-based Ross classification (14) 536
www.pccmjournal.org
for heart failure in children 5 years or less or the New York Heart Association Functional Classification (15) for heart failure in patients 5–21 years at admission was determined. The administration of diuretics, vasoactive, and/or nephrotoxic medications and the need for RRT were recorded; in addition, the highest inotrope score (16) was calculated. Outcome data collected included the institution of mechanical circulatory support (i.e., extracorporeal membrane oxygenation [ECMO] or ventricular assist device [VAD]) and heart transplantation during the index admission. Duration of mechanical ventilation, ICU and hospital length of stays, and in-hospital mortality were obtained. Statistical Analysis The data were analyzed using SAS software (Version 9.2, SAS Institute, Cary, NC). Descriptive statistics were performed. Continuous variables were expressed as mean ± sds or as median with interquartile ranges for parametric and nonparametric variables. Student t test was used for normally distributed continuous variables, Mann-Whitney U test was performed for nonnormally distributed continuous variables, and chi-square test or Fisher exact analysis was performed on all categorical variables. A lambda value was calculated to determine the strength of the relationship between the severity of AKI and the composite outcome (0–0.10, no relationship; 0.10–0.20, weak relationship; 0.2–0.3, moderate relationship; 0.3–0.4, moderate-to-strong relationship; and 0.4 and above, strong relationship). Univariate logistic analysis was performed to examine the risk factors for any evidence of AKI (pRIFLE = R, I, F) with any variable with a p value up to 0.1 entered into a multivariate step-wise logistic regression analysis. Odds ratio (OR), CI, and p value were calculated for each variable. Univariate logistic analysis was performed to examine the relationship between multiple clinical variables including the presence of AKI and clinical outcomes (ICU length of stay, hospital length of stay, duration of mechanical ventilation, and the composite outcome of cardiac transplantation and/or in-hospital mortality). Variables with a p value up to 0.1 were then cast into a multivariate step-wise logistic regression analysis. OR, CI, and p values were calculated. Statistical significance was defined as a p value up to 0.05.
RESULTS The study population consisted of 57 patients determined to have ADHF. Etiologies of ADHF were 31 patients (54%) with dilated cardiomyopathy (DCM), 18 patients (32%) with myocarditis, and eight patients (14%) with restrictive or ischemic cardiomyopathy. The study cohort had a median age of 12 years (interquartile range [IQR], 1.1–16) and median weight of 40.5 kg (IQR, 8.3–68). The majority of patients (96%) in the cohort were treated with diuretic therapies as well more than half the patients treated with inotropes, with the mean inotrope score (16) of 6.26 ± 6.37. ICU and hospital lengths of stay were 10 days (IQR, 4–25) and 13 days (IQR, 7–27), respectively. Of the 57 patients, 25 patients (44%) were mechanically ventilated July 2015 • Volume 16 • Number 6
Cardiac Intensive Care
during their index admission. The composite outcome was met by 28 patients (49%) with 24 patients (86%) undergoing cardiac transplantation and four patients (14%) dying or expiring. Some degree of AKI was present in 77% (44 patients) during their ICU stay. The maximal pRIFLE score was Risk for 43.2% (19 patients), Injury for 38.6% (17 patients), and Failure for 18.2% (eight patients). Of the pRIFLE score, 53% of patients under “Risk” met the composite outcome, 59% of patients under “Injury” met the composite outcomes, and 88% of patients under “Failure” met the composite outcomes. Of the 44 patients who exhibited any evidence of AKI, 28 (64%) did so within 24 hours of admission, seven (16%) did so after 24 hours but less than 72 hours of admission, nine (20%) did so after 72 hours of admission to the ICU, and 26 (59%) continued to have evidence of AKI at the time of discharge or composite outcome. The median duration of AKI was 7.5 days (IQR, 4–14). Patient characteristics of non-AKI and AKI patient cohorts are demonstrated in Table 1. The median SCr levels were not different between the patients with and without AKI. Nephrotoxic mediations were limited to primarily vancomycin, angiotensin-converting enzyme inhibitors, and diuretics. Six patients (11%) did not receive nephrotoxic medications. Of the six patients, only two patients (33%) had AKI. The use of RRT was limited to only three patients (5%) all within the cohort of patients with AKI. All three patients had a diagnosis of DCM and presented with evidence of AKI at admission. Two patients received ultrafiltration only, and one patient received hemodialysis. Indication for RRT in the two patients who underwent ultrafiltration was refractory fluid overload in the context of ADHF. The patient who underwent hemodialysis was found to have HIV-associated DCM and subsequently developed a nephropathy. One patient who received ultrafiltration was ultimately transplanted after mechanical circulatory bridging. All three patients ultimately met the composite outcome. Mechanical circulatory support was used in a total of 16 patients (28%), all of which had AKI. AKI was present in all patients prior to cannulation for ECMO or insertion of VADs. Two patients (13%) did not meet the composite outcome; these patients were supported with ECMO and successfully decannulated. Diagnoses were myocarditis and restrictive cardiomyopathy, respectively. Table 2 demonstrates severity of AKI with the need for mechanical circulatory support. Risk for AKI After adjusting for variables found to be associated with AKI in the univariate analysis, multivariate step-wise logistic regression analysis identified LVEF less than 25% at the time of diagnosis (OR, 12.13; 95% CI, 1.35–109.44; p = 0.03) as the only significant factor associated with the development of AKI. Of note, age (per month increase) trended toward significance (OR, 0.99; 95% CI, 0.98–1.0; p = 0.09). Univariate variables entered into the multivariate model included age, weight, diagnosis of DCM, heart failure classification, LVEF up to 25%, inotrope score at least 10, and NT-proBNP more than 5,000 pg/mL. Pediatric Critical Care Medicine
Clinical Outcomes Comparison of clinical outcomes between the patients without AKI and patients with AKI are presented in Table 2. Of the 44 patients with AKI, 23 (52%) were mechanically ventilated (OR, 6.0; 95% CI, 1.2–30.4; p = 0.03) with 18 (40%) mechanically ventilated for at least 3 days (OR, 8.3; 95% CI, 0.99–69.7; p = 0.04). Multivariate step-wise logistic regression analysis demonstrated that only the presence of AKI was significantly associated with the need for mechanical ventilation (adjusted OR, 4.3; 95% CI, 0.78–23.7; p = 0.05). Multivariate step-wise logistic regression analysis demonstrated only the presence of AKI as significantly associated with an ICU length of stay at least 9 days (adjusted OR, 5.2; 95% CI, 0.95–29.2; p = 0.05) and a hospital length of stay at least 10 days (adjusted OR, 4.6; 95% CI, 1.0–21; p = 0.04). Composite Outcome Of the 28 patients meeting the composite outcome, all but one patient had evidence of AKI. Figure 1 demonstrates the breakdown of patients with or without AKI who met the composite outcome. Of the 27 patients who met the composite outcome and had AKI, all developed AKI prior to the composite outcome. Ten patients (37%) had pRIFLE score “Risk,” 10 patients (37%) had pRIFLE score “Injury,” and seven patients (26%) had pRIFLE score “Failure.” Two patients were 1 month old or less at the time of satisfying the composite outcome. Death or transplantation occurred in 61% of patients with AKI and 8% of patients without AKI. Eighteen patients (64%) had AKI at the time of admission with only eight patients (29%) developing AKI 72 hours after admission. Nine patients (32%) were assisted with VAD support (Berlin Heart EXCOR left VAD [Berlin Heart GmbH, Berlin, Germany], three patients; Berlin Heart EXCOR biventricular VAD, three patients; HeartWare [HeartWare, Framingham, MA], two patients; and HeartMate II [Thoratec Corporation, Pleasanton, CA], one patient). Five patients (18%) were supported with ECMO; of the five patients, two were bridged with ECMO to VAD support. All patients supported with ECMO had their cannulas peripherally inserted. Severity of AKI (no AKI vs Risk vs Injury vs Failure) in relationship to the composite outcome was evaluated using the chi-square test with calculation of a lambda value for predicting the composite outcome with knowledge of the severity of AKI or lack of AKI; this analysis demonstrated a p value of 0.002 and a lambda of 0.36 (moderate-to-strong relationship). Multivariate step-wise logistic regression analysis demonstrated that LVEF up to 25% and the presence of AKI were the only factors significantly associated with the composite outcome (Table 3). Univariate variables entered into the multivariate model included age, weight, diagnosis of DCM, heart failure classification, LVEF up to 25%, presence of AKI, and inotrope score at least 10.
DISCUSSION In this study of pediatric patients admitted to the hospital with newly diagnosed primary ventricular dysfunction and ADHF, the prevalence of AKI was high at 77%. Multiple risk factors were associated with AKI, with LVEF less than 25% having the strongest association when all confounders were accounted for. Interestingly, age trended toward significant in the multivariate www.pccmjournal.org
537
Medar et al
Table 1.
Patient Characteristics: Non-Acute Kidney Injury Versus Acute Kidney Injury All (n = 57)
Variable
Age (mo), median (IQR) Male (%)
144 (13–192) 36 (63)
Weight (kg), median (IQR)
40.5 (8.3–68)
Non-AKI (n = 13)
192 (156–204) 9 (82) 64.7 (55–76.6)
AKI (n = 44)
pa
48 (12–192)
0.02b
25 (57)
0.1c
20.1 (8.2–58.5)
0.02b
Raced (%) White
11 (19)
2 (15)
9 (20)
1.0c
Asian
7 (12)
2 (15)
5 (11)
1.0c
Black
20 (35)
3 (24)
17 (39)
0.34c
Hispanic
19 (34)
6 (46)
13 (30)
0.32c
Dilated cardiomyopathy
31 (54)
2 (15)
29 (66)
0.001c
Myocarditis
18 (32)
9 (70)
9 (20)
0.001c
8 (14)
2 (15)
6 (14)
1.0c
Class 1
3 (5)
1 (7)
2 (5)
1.0c
Class 2
11 (19)
7 (55)
4 (9)
0.001c
Class 3
28 (50)
3 (23)
25 (56)
0.05c
Class 4
15 (26)
2 (15)
13 (30)
0.47c
Diagnosis of heart failure (%)
Other Ross/New York Heart Association classification (%)
Left ventricular ejection fraction,b median (IQR)
0.25 (0.18–0.48)
0.45 (0.34–0.51)
0.21 (0.17–0.46)
0.01b
Admission SCr (mg/dL), median (IQR)
0.6 (0.4–0.85)
0.7 (0.7–0.8)
0.5 (0.4–0.82)
0.18b
Admission BUN (mg/dL), median (IQR)
14 (11–17)
11 (9–12)
Admission eCCL (mL/min/1.73 m2), median (IQR)
88 (68–116)
100 (96–116)
Highest SCr (mg/dL), median (IQR)
0.8 (0.5–1.1)
0.7 (0.7–0.9)
0.8 (0.5–1.3)
0.43b
Highest BUN (mg/dL), median (IQR)
22 (14–38)
12 (11–14)
28 (20.5–40.6)
0.001b
Lowest eCCL (mL/min/1.73 m2), median (IQR)
65 (46–90)
96.7 (83–116)
60.25 (41.7–74.2)
0.001b
Discharge or outcome eCCL (mL/min/1.73 m2), median (IQR)
96 (70–127)
Highest serum N-terminal pro-brain natriuretic peptide (pg/mL),b median (IQR)
11,717 (3,950–35,000)
96 (83–127) 1,199 (89–3,846)
14.5 (11.7–18)
0.004b
83.4 (67.1–110.6)
0.16b
83.6 (60–109)
0.4b
12,256 (6,663–40,553)
0.02b
Vasoactive medications used (%) Milrinone
30 (53)
1 (7)
29 (44)
0.001c
Dopamine
6 (10)
1 (7)
7 (16)
0.66c
Dobutamine
2 (4)
0
2 (9)
1.0c
Epinephrine
10 (17)
0
10 (23)
0.09c (Continued)
538
www.pccmjournal.org
July 2015 • Volume 16 • Number 6
Cardiac Intensive Care
Table 1. (Continued). Patient Characteristics: Non-Acute Kidney Injury Versus Acute Kidney Injury Variable
Highest inotrope score, median (IQR) Diuretic usef (%)
All (n = 57)
Non-AKI (n = 13)
AKI (n = 44)
pa
6.26 ± 6.37
0.96 ± 2.4
7.83 ± 6.34
0.001e
48 (84)
7 (54)
41 (93)
0.002c
Daily dose of diuretic (mg/kg), median (IQR)
0.63 (0.25–0.73)
2.0 (0.78–4)
0.99b
Nephrotoxic medications (%) Vancomycin
26 (46)
2 (15)
24 (55)
0.02c
Angiotensin-converting enzyme inhibitors
28 (49)
6 (46)
22 (50)
0.06g
Ultrafiltration
2
0
2
Hemodialysis
1
0
1
Renal replacement therapy use
AKI = acute kidney injury, IQR = interquartile range, SCr = serum creatinine, BUN = blood urea nitrogen, eCCL = estimated creatinine clearance. a All comparisons made between non-AKI group and AKI group. b Mann-Whitney U test. c Fisher exact probability test. d Data for all patients not available. e Student t test. f Data pertain to furosemide use only; all patients who received diuretic therapy received furosemide. Less than 10% of patients also received a thiazide diuretic concomitant with furosemide. g Chi-square test.
analysis with a p value of 0.09. It could be hypothesized that a larger sample size may have elicited age as a significantly associated with AKI. It is reasonable to assess that younger patients presumably neonates may be more susceptible to AKI, which may be related to the intrinsic immaturity of the neonatal renal Table 2.
tubules and their inability to adapt to the deficiency of glomerular filtration rate seen in ADHF as well as the state of fluid overload and the nutritional status of the neonate. Only the presence of AKI and a LVEF less than 25% were significantly associated with the composite outcome of cardiac transplant or death.
Clinical Outcomes: Non-Acute Kidney Injury Versus Acute Kidney Injury
Variable
All Patients (n = 57)
Non-AKI (n = 13)
AKI (n = 44)
pa
Risk (n = 19)
pa
Injury (n = 17)
pa
Failure (n = 8)
pa
ICU LOS (d), median (IQR)
10 (4–25)
4 (2–7)
18 0.05b (6.5–32.75)
10 0.19b (6–18.5)
21 (5–50)
0.02b
34 (21.75–45)
0.001b
Hospital LOS (d), median (IQR)
13 (7–27)
7 (6–10)
19 0.047b (7.75–49.25)
12 0.18b (7.5–19)
22 (7–55)
0.022b
37.5 (22.5–54)
0.003b
Mechanical ventilation (d), mean ± sd
9.25 ± 29.1 0.54 ± 1.33
11.8 ± 32.7
0.22c
3 ± 4.82
0.84c 20.4 ± 50.1 0.17c
14.5 ± 19.6
0.017c
Extracorporeal membrane oxygenation (%)
7 (12)
0
7 (16)
0.18d
3 (16)
0.25d
3 (18)
0.23d
Ventricular 10 (18) assist device (%)
0
10 (23)
0.09d
5 (26)
0.06d
5 (29)
0.05d
27 (61)
0.04d
10 (53)
0.02d
10 (59)
Composite outcome (%)
28 (49)
1 (8)
0.006d
2 (25)
0 7 (88)
0.13d
1d < 0.001d
AKI = acute kidney injury, LOS = length of stay, IQR = interquartile range. a All comparisons made to non-AKI group. b Mann-Whitney U test. c Student t test. d Fisher exact test.
Pediatric Critical Care Medicine
www.pccmjournal.org
539
Medar et al
Figure 1. Flowchart of patient classification by acute kidney injury (AKI) and composite outcome.
The prevalence of AKI demonstrated in the current review is higher than that previously reported by Price et al (9). Possible reasons for the discrepancy between the report by Price et al (9) and the current study may be related to differences in the definition of AKI (absolute rise in SCr ≥ 0.3 from baseline vs the pRIFLE criteria; the earlier time frame studied 2005–2008 vs 2008–2012). In addition, a significant portion of the cohort (26%) described by Price et al (9) had prior admissions for heart failure, and thus, the baseline creatinine may have been higher due to chronic heart failure. Our finding of the prevalence of AKI of 77% is also higher than the prevalence of AKI in other pediatric populations studied, such as after cardiopulmonary bypass (CPB), 3–62% (17–22); sepsis, 55% (23); and perinatal asphyxia, 30% (24). This may be attributable to the potentially indolent or occult injury occurring prior to the patient presenting with overt symptoms of heart failure rather than hyperacute discrete points of injury in the case of CPB, to some extent sepsis, and asphyxia. We also demonstrated that a significant portion of the cohort of patients with AKI (59%) during their index admission exhibited evidence of AKI at the time of discharge (including mortality) or at the time of cardiac transplantation if it occurred during the index admission. This finding has important potential implications for patients receiving chronic medical therapy with potentially nephrotoxic medications to treat ADHF or following cardiac transplantation. It has been demonstrated in the
Risk Factors for Transplantation or Mortality Table 3.
Multivariate Analysis
Adjusted OR
95% CI
p
< 0.001
Left ventricular ejection fraction ≤ 25%
16.7
3.39–42.5
Presence of acute kidney injury
19.1
2.26–160.12 < 0.001
OR = odds ratio.
540
www.pccmjournal.org
literature (25) that renal insufficiency (creatinine > 2 mg/dL) prior to transplantation is independently associated with increased mortality in patients following cardiac transplantation. As a result, can we hypothesize that those patients with milder degrees of AKI at the time of cardiac transplantation may also be at a greater risk for posttransplantation morbidity and mortality related to the presence of AKI compared with those patients pretransplantation with no evidence of AKI. Cardiorenal syndrome (CRS), type 1 in ADHF, presents a complex pathophysiologic state in which cardiac and renal dysfunctions coexist and potentially exacerbate each other. Multiple factors contribute to the development of this syndrome, such as diminished cardiac output, diminished renal perfusion, neurohormonal factors, fluid retention, increased systemic vascular resistance, and medications, all of which are integral components of patients with ADHF. The description of CRS in children has mostly been in the setting of CPB with just one study in the population of ADHF (9, 26). Our study is one of the few to allude to CRS in a non-CPB cohort of children with ADHF with structurally normal hearts. In their study, Price et al (9) described infusions of nesiritide and dopamine as well as admission BUN and creatinine as significant risk factors for development of worsening renal function. In our study, we demonstrated that in patients with ADHF, a LVEF less than 25% was independently associated with the development AKI. In our study, AKI was associated with a 19.1-fold increase in mortality and/or transplantation, as compared with a 10.2-fold increase in mortality in the study by Price et al (9). A common dilemma facing the clinician when caring for patients with ADHF is determining the appropriate time to escalate or modify therapy. Typically, these decisions are based on clinical variables and assessment of end-organ function. Guidelines for clinical variables or thresholds have been suggested in the literature for indications for heart transplantation (27). However, assessment of end-organ function continues to remain somewhat of an enigma for the clinician. “How bad is too bad?” Specifically, the gold standard for assessment of kidney function has been SCr, which traditionally has been widely used, inexpensive, reproducible, and validated across diverse patient populations. However, SCr may be a limited tool in the assessment of end-organ damage; it is well known that SCr is a marker of renal function rather than a marker of renal injury or damage. Our study may lend credence to the notion that SCr may have limitations as a tool for risk assessment of patients with ADHF. Nonetheless, until more specific and sensitive markers of renal injury become established and validated, we must remain cognizant of subtle changes in our current armamentarium of biochemical markers (i.e., SCr, estimated glomerular filtration rate, casts, fractional excretion of sodium, filtered high-molecularweight proteins, and tubular proteins or enzymes). Our study did demonstrated no significant difference between median admission and peak SCr levels between patients with and without AKI (0.5 vs 0.7, p = ns; 0.8 vs 0.7, p = ns; respectively), and this may be attributable several factors. First, we must keep in mind that several aspects may alter the value of creatinine, such as age, body habitus, and the nutritional and clinical status of July 2015 • Volume 16 • Number 6
Cardiac Intensive Care
the patient. Second, a rise in creatinine can take several days in manifest and may make appreciation of a decrease in renal function difficult. Third, the etiology of a rise or fall in SCr may be multifactorial and not necessarily indicate intrinsic renal injury or recovery. Further complicating the diagnosis of AKI is that an acute drop in GFR will lead to an elevation in SCr, but the rate of rise may be difficult to predict due the limitations mentioned previously. In addition, it may take days for SCr to equilibrate; thus, a true rise in creatinine or nadir in estimated GFR may not be immediately apparent, hindering the chances of making an early diagnosis of AKI and modifying treatment. There are several limitations to this study. This is a singlecenter study. This was a retrospective study with the inherent limitations related to data accuracy and missing data points, as well as the ability to evaluate only short-term clinical outcomes. We did not evaluate urine output as a determinant of AKI. We chose primarily to exclude this portion of the pRIFLE criteria secondarily to the unreliability of retrospectively collecting hourly urinary output measurements from the medical record. Furthermore, wide spread and expeditious use of diuretic therapy in this population would potentially limit the utility of hourly urine output as a determinant of AKI. An additional limitation to the study may be evident in the various etiologic phenotypes of ADHF, which may hinder our ability to generalize the effects of AKI in this cohort. In conclusion, more than half of children with ADHF demonstrate AKI. The risk is even greater in those patients with the worst cardiac function. AKI was associated with longer ICU and hospital lengths of stay, as well as longer durations of mechanical ventilation. Patients admitted with ADHF who either had AKI at admission or developed it during their admission were more likely to have a cardiac transplantation or die. This study supports the need for additional work on AKI in ADHF in the pediatric population with larger prospective multicenter studies.
REFERENCES
1. Rossano JW, Kim JJ, Decker JA, et al: Prevalence, morbidity, and mortality of heart failure-related hospitalizations in children in the United States: A population-based study. J Card Fail 2012; 18:459–470 2. Fonarow GC, Yancy CW, Heywood JT; ADHERE Scientific Advisory Committee, Study Group, and Investigators: Adherence to heart failure quality-of-care indicators in US hospitals: Analysis of the ADHERE Registry. Arch Intern Med 2005; 165:1469–1477 3. Gheorghiade M, Zannad F, Sopko G, et al; International Working Group on Acute Heart Failure Syndromes: Acute heart failure syndromes: Current state and framework for future research. Circulation 2005; 112:3958–3968 4. Jhund PS, Macintyre K, Simpson CR, et al: Long-term trends in first hospitalization for heart failure and subsequent survival between 1986 and 2003: A population study of 5.1 million people. Circulation 2009; 119:515–523 5. Adams KF Jr, Fonarow GC, Emerman CL, et al; ADHERE Scientific Advisory Committee and Investigators: Characteristics and outcomes of patients hospitalized for heart failure in the United States: Rationale, design, and preliminary observations from the first 100,000 cases in the Acute Decompensated Heart Failure National Registry (ADHERE). Am Heart J 2005; 149:209–216 6. Krumholz HM, Chen YT, Vaccarino V, et al: Correlates and impact on outcomes of worsening renal function in patients > or =65 years of age with heart failure. Am J Cardiol 2000; 85:1110–1113
Pediatric Critical Care Medicine
7. Smith GL, Vaccarino V, Kosiborod M, et al: Worsening renal function: What is a clinically meaningful change in creatinine during hospitalization with heart failure? J Card Fail 2003; 9:13–25 8. Dries DL, Exner DV, Domanski MJ, et al: The prognostic implications of renal insufficiency in asymptomatic and symptomatic patients with left ventricular systolic dysfunction. J Am Coll Cardiol 2000; 35:681–689 9. Price JF, Mott AR, Dickerson HA, et al: Worsening renal function in children hospitalized with decompensated heart failure: Evidence for a pediatric cardiorenal syndrome? Pediatr Crit Care Med 2008; 9:279–284 10. Brion LP, Fleischman AR, McCarton C, et al: A simple estimate of glomerular filtration rate in low birth weight infants during the first year of life: Noninvasive assessment of body composition and growth. J Pediatr 1986; 109:698–707 11. Schwartz GJ, Brion LP, Spitzer A: The use of plasma creatinine concentration for estimating glomerular filtration rate in infants, children, and adolescents. Pediatr Clin North Am 1987; 34:571–590 12. Staples A, LeBlond R, Watkins S, et al: Validation of the revised Schwartz estimating equation in a predominantly non-CKD population. Pediatr Nephrol 2010; 25:2321–2326 13. Akcan-Arikan A, Zappitelli M, Loftis LL, et al: Modified RIFLE criteria in critically ill children with acute kidney injury. Kidney Int 2007; 71:1028–1035 14. Ross RD: The Ross classification for heart failure in children after 25 years: A review and an age-stratified revision. Pediatr Cardiol 2012; 33:1295–1300 15. Criteria Committee, New York Heart Association: Diseases of the Heart and Blood Vessels. Nomenclature and Criteria for Diagnosis. Sixth Edition. Boston, MA, Little, Brown, 1964, pp 114 16. Hoffman TM, Wernovsky G, Atz AM, et al: Prophylactic intravenous use of milrinone after cardiac operation in pediatrics (PRIMACORP) study. Prophylactic Intravenous Use of Milrinone After Cardiac Operation in Pediatrics. Am Heart J 2002; 143:15–21 17. Tóth R, Breuer T, Cserép Z, et al: Acute kidney injury is associated with higher morbidity and resource utilization in pediatric patients undergoing heart surgery. Ann Thorac Surg 2012; 93:1984–1990 18. Li S, Krawczeski CD, Zappitelli M, et al; TRIBE-AKI Consortium: Incidence, risk factors, and outcomes of acute kidney injury after pediatric cardiac surgery: A prospective multicenter study. Crit Care Med 2011; 39:1493–1499 19. Sethi SK, Goyal D, Yadav DK, et al: Predictors of acute kidney injury post-cardiopulmonary bypass in children. Clin Exp Nephrol 2011; 15:529–534 20. Zappitelli M, Bernier PL, Saczkowski RS, et al: A small post-operative rise in serum creatinine predicts acute kidney injury in children undergoing cardiac surgery. Kidney Int 2009; 76:885–892 21. Chiravuri SD, Riegger LQ, Christensen R, et al: Factors associated with acute kidney injury or failure in children undergoing cardiopulmonary bypass: A case-controlled study. Paediatr Anaesth 2011; 21:880–886 22. Aydin SI, Seiden HS, Blaufox AD, et al: Acute kidney injury after surgery for congenital heart disease. Ann Thorac Surg 2012; 94:1589–1595 23. Duzova A, Bakkaloglu A, Kalyoncu M, et al; Turkish Society for Pediatric Nephrology Acute Kidney Injury Study Group: Etiology and outcome of acute kidney injury in children. Pediatr Nephrol 2010; 25:1453–1461 24. Durkan AM, Alexander RT: Acute kidney injury post neonatal asphyxia. J Pediatr 2011; 158:e29–e33 25. Ojo AO, Held PJ, Port FK, et al: Chronic renal failure after transplantation of a nonrenal organ. N Engl J Med 2003; 349:931–940 26. Jefferies JL, Goldstein SL: Cardiorenal [corrected] syndrome: An emerging problem in pediatric critical care. Pediatr Nephrol 2013; 28:855–862 27. Ammirati E, Oliva F, Cannata A, et al: Current indications for heart transplantation and left ventricular assist device: A practical point of view. Eur J Intern Med 2014; 25:422–429
www.pccmjournal.org
541