Characteristics and Outcomes of Children and Young Adults with Sepsis Requiring Continuous Renal Replacement Therapy: A Comparative Analysis from the Worldwide Exploration of Renal Replacement Outcomes Collaborative in Kidney Disease (WE-ROCK)

Abstract

Objective:

Pediatric sepsis-associated acute kidney injury (AKI) often requires continuous renal replacement therapy (CRRT), but limited data exist regarding patient characteristics and outcomes. We aimed to describe these features, including the impact of possible dialytrauma (i.e., vasoactive requirement, negative fluid balance) on outcomes, and contrast them to non-septic patients in an international cohort of children and young adults receiving CRRT.

Design:

A secondary analysis of WE-ROCK, an international, multicenter, retrospective study.

Setting:

Neonatal, cardiac and pediatric intensive care units at 34 centers in 9 countries from 1/12015 to 12/31/2021.

Patients:

Patients aged 0–25 years requiring CRRT for AKI and/or fluid overload.

Interventions:

None.

Measurements and Main Results:

Among 1016 patients, 446 (44%) had sepsis at CRRT initiation and 650 (64%) experienced MAKE-90 (defined as a composite of death, RRT dependence, or >25% decline in eGFR from baseline at 90 days from CRRT initiation). Septic patients were less likely to liberate from CRRT by 28 days (30% vs. 38%, p<0.001), had higher rates of MAKE-90 (70% vs. 61%, p=0.002) and higher mortality (47% vs. 31%, p<0.001) than non-septic patients; however, septic survivors were less likely to be RRT dependent at 90 days (10% vs. 18%, p=0.011). On multivariable regression, pre-CRRT vasoactive requirement, time to negative fluid balance and median daily fluid balance over the first week of CRRT were not associated with MAKE-90; however, increasing duration of vasoactive requirement was independently associated with increased odds of MAKE-90 (aOR 1.16, 95%CI 1.05–1.28) and mortality (aOR 1.20, 95%CI 1.1–1.32) for each additional day of support.

Conclusion:

Septic children requiring CRRT have different clinical characteristics and outcomes compared to those without sepsis, including higher rates of mortality and MAKE-90. Increasing duration of vasoactive support during the first week of CRRT, a surrogate of potential dialytrauma, appears to be associated with these outcomes.

Introduction:

Sepsis impacts 10% of children and young adults admitted to the pediatric intensive care unit (ICU) and is associated with morbidity and mortality(1–3). Outcomes are particularly poor in those who suffer severe acute kidney injury (AKI), which impacts one in five children with severe sepsis(4–6). Because there are no disease-modifying therapies for sepsis-associated AKI, support with continuous renal replacement therapy (CRRT) is often needed(7). Unfortunately, limited data exist regarding best practices for prescribing CRRT to enhance renal recovery and outcomes in children and young adults with sepsis-associated AKI. This paucity of data has resulted in a weak recommendation to consider using CRRT for AKI and fluid overload from the most recent Surviving Sepsis Campaign(8), and a continued knowledge gap that perpetuates practice variation in this population.

The Worldwide Exploration of Renal Replacement Outcomes Collaborative in Kidney Disease (WE-ROCK) is an international, multi-center study that was designed to provide contemporary information regarding the characteristics and outcomes of critically ill children and young adults requiring CRRT for AKI and/or fluid overload(9). Prior to WE-ROCK, the nearly 20-year old Prospective Pediatric Continuous Renal Replacement Therapy Registry (ppCRRT) provided the most recent data (2001–2005)(10), and although one in three patients in that registry had sepsis of whom nearly 60% died(11), no studies looked exclusively at this population. Since ppCRRT, the literature has remained scant, and while data suggest some benefit to use of CRRT in children with septic AKI(12, 13), increasing recognition of the potential harms of renal replacement therapy (RRT) (i.e., so-called “dialytrauma”)(14–16), coupled with the hemodynamic instability of sepsis, highlight the need for a better understanding of how best to prescribe CRRT in this population.

Thus, we had two specific aims: (1) to describe and compare the characteristics and outcomes of septic children and young adults receiving CRRT for AKI and/or fluid overload to patients without sepsis receiving CRRT, and (2) to identify factors associated with major adverse kidney events at 90 days from CRRT initiation (MAKE-90) in septic patients receiving CRRT. We hypothesized that septic patients requiring CRRT would have unique demographic and clinical characteristics and have worse outcomes than those without sepsis, and that specific clinical features, including evidence of possible dialytrauma (i.e., more negative fluid balance and vasoactive requirement), would be associated with MAKE-90.

Materials and Methods:

Study Design and Data Collection:

We performed a secondary analysis of the WE-ROCK international registry of pediatric patients from 34 centers across 9 countries receiving CRRT in the pediatric, neonatal or cardiac ICU from January 1, 2015 to December 31, 2021. Patients were included if they were between 0–25 years of age and received CRRT for AKI or fluid overload; those with previous dialysis dependence, congenital anomalies of the kidney and urinary tract (CAKUT) expected to progress to end stage kidney disease, concurrent extracorporeal membrane oxygenation use, and those who received CRRT for another indication were excluded. Each site received local Institutional Review Board approval for the original study with a waiver of informed consent (Supplementary Appendix 1), and the study was performed in accordance with the ethical standards outlined in the 1964 Declaration of Helsinki and its later amendments. All patients from the registry were included in our secondary analysis. Complete methodological details have been previously published in full (Supplementary Methods)(9). This study is reported following Strengthening the Reporting of Observational Studies in Epidemiology reporting guidelines (Supplementary Appendix 2).

Definitions:

Patients had sepsis if they were being treated for known or suspected infection and met at least 2 Systemic Inflammatory Response Syndrome criteria (Supplementary Methods)(8, 17, 18) in either the 24 hours prior to CRRT initiation, or at ICU admission with CRRT initiated within 1 week of admission(7).

Baseline serum creatinine (SCr) was defined as the lowest SCr (mg/dl) within 90 days prior to admission. If unknown, a value was imputed using body surface area and an estimated glomerular filtration rate of 100 ml/min per 1.73 m², as validated(19). Severity of illness was quantified at admission using Pediatric Risk of Mortality III (PRISM III) score(20), and in the 24 hours prior to CRRT initiation using both Pediatric Logistic Organ Dysfunction 2 (PELOD-2) score(21) and vasoactive-inotropic score (VIS)(22). Fluid balance from ICU admission to CRRT initiation was calculated using the validated formula (Supplementary Methods)(23).

Vasoactive exposure on CRRT was quantified two ways. First, patients were assigned one of two vasoactive change categories over the first 48 hours: (1) never on vasoactives or vasoactives decreased, and (2) vasoactives unchanged or increased. This time point was selected as it was felt that persistent or worsening vasoactive need at 48 hours may be indicative of dialytrauma, as many patients improve with resuscitation and therapy by this time point. Second, the percentage of CRRT days requiring vasoactive support in up to the first week (i.e., the time frame for which we have vasoactive data) was calculated using the following formula:

$$\left[\left(\#\text{of Days Requiring Vasoactives in Day}1-7\text{of CRRT}\right)/\left(\#\text{of Days of CRRT Days}1-7\right)\right]\times 100$$

Outcomes:

The primary outcome was a modified MAKE-90, defined as a composite of (1) death, (2) RRT dependence, or (3) persistent kidney dysfunction (>25% decline in eGFR from baseline) at 90 days from CRRT initiation(24). Secondary outcomes included successful liberation from CRRT on the first attempt within 28 days. Patients were considered successfully liberated if they received no RRT modality for ≥72 after CRRT discontinuation. Additional outcomes included CRRT duration, 28-day ventilator-free days (calculated as 28 − days requiring invasive mechanical ventilation, with patients who died assigned “0”), ICU length of stay (LOS) in survivors, 28-day ICU-free days (calculated as 28 − ICU LOS, with patients who died assigned “0”), and in-hospital mortality.

Statistical Analysis:

Demographics and clinical characteristics were described using medians, interquartile ranges, frequencies, and percentages. Comparisons were performed using Wilcoxon rank sum, Chi-square, or Fisher’s exact test, as appropriate. Comparisons were first made between septic and non-septic patients to assess differences in characteristics and outcomes. Time-to-event analysis using the Kaplan-Meier method was performed to assess differences in cumulative probability of 90-day mortality between septic and non-septic patients who required CRRT.

Subsequent analyses examining factors associated with the development of MAKE-90 were performed exclusively in septic patients who survived >48 hours after CRRT initiation (n=393, Supplementary Figure 1). Patients who died prior to this time were excluded as it is unlikely that CRRT characteristics contributed to their outcome. First, we performed univariate analysis of a priori selected variables felt to be important in development of MAKE-90. Significant variables on univariate analysis (p <0.15) were assessed for collinearity and selected for inclusion in a multivariable regression model, as appropriate. Generalized linear mixed effects regression was used to obtain odds ratios (ORs), 95% confidence intervals (CI), and p-values for MAKE-90. Models included a logit link function to model the binary responses, covariates, and a random site-specific intercept to account for the clustering of patients within centers. Finally, the probability of MAKE-90 and mortality were assessed by increasing number of days requiring vasoactives in the first week of CRRT using a mixed effect logistic regression model adjusting for the same covariates as above, aside from the other measures of vasoactive intensity due to risk of collinearity. Non-linear associations were also tested using natural cubic splines, however, the linear association proved best. A p-value <0.05 was considered significant. All statistical analyses were performed using R Version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria) and the lme4 (v1.1.35.1, Bates et al, 2015) package.

Results:

One thousand and sixteen patients were available in the original registry, of which 446 (44%) had sepsis (Supplemental Figure 1). Fifty-three of 446 patients with sepsis (12%) died within the first 48 hours of initiating CRRT and were excluded from our multivariable analyses (Supplementary Figure 1). Demographic and clinical information for patients by the presence or absence of sepsis are shown in Table 1 and Supplementary Table 1. Septic patients were older, had higher severity of illness, lower platelet counts (51 × 10³/μL [24–126] vs. 71 × 10³/μL [26–160], p<0.001), and higher cumulative percent fluid balance (10% [4–21] vs. 6% [2–16], p<0.001) at CRRT initiation. There were no differences in comorbid conditions, including nephrologic/urologic conditions, between septic and non-septic patients. CRRT dose and modality were no different between groups, however, differences were seen in anticoagulation choice. Citrate was used most frequently in all patients, but septic patients were less likely to receive heparin and more likely to receive no anticoagulation.

Table 1:

Demographic and clinical characteristics of children with sepsis receiving continuous renal replacement therapy (CRRT) compared to those without sepsis.

Variable	All (n=1016)	Sepsis (n=446)	No Sepsis (n=570)	p
Demographics
Age, years	8.3 (1.6–15)	10.2 (2.3–15.4)	6.4 (1.3–14.5)	0.003
Sex, female (%)	462 (45)	211 (47)	251 (44)	0.30
Admission Diagnosis, n (%)				<0.001
Shock/Infection/Major Trauma	384 (38)	257 (58)	127 (22)
Respiratory Failure	197 (19)	82 (18)	115 (20)
CNS Dysfunction	40 (3.9)	18 (4.0)	22 (4.0)
Primary Cardiac	123 (12)	28 (6.0)	95 (17)
Other	272 (27)	61 (14)	211 (37)
Comorbidities, n (%)				0.09
0	202 (20)	87 (20)	119 (21)
1	496 (49)	208 (47)	288 (51)
≥2	313 (31)	150 (34)	163 (29)
PRISM III	14 (10–18)	15 (11–20)	14 (9–18)	<0.001
Baseline Creatinine, mg/dl	0.43 (0.27–0.64)	0.42 (0.27–0.63)	0.43 (0.27–0.65)	0.39
Known Baseline Creatinine, n (%)	558 (55)	251 (56)	307 (54)	0.44
CRRT Initiation Data
Time from ICU Admission to CRRT, days	2 (1–6)	2 (1,7)	2 (1,6)	0.56
PELOD-2	5.0 (2.0–8.0)	7.0 (3.0–10.0)	5.0 (2.0–7.0)	<0.001
Receipt of Vasoactives, n (%)	600 (59)	317 (71)	283 (50)	<0.001
VIS	5 (0–20)	10 (0–30)	0 (0–13)	<0.001
Mechanical Ventilation, n (%)	513 (50)	242 (54)	271 (48)	0.70
Platelet Count, × 10 3 /μL	63 (29–146)	51 (24–126)	71 (36–160)	<0.001
%Fluid Balance from ICU Admission	8 (2–18)	10 (4–21)	6 (2–16)	<0.001
Initial CRRT Modality, n (%)				0.21
CVVH	112 (11)	44 (9.9)	68 (12)
CVVHD	114 (11)	43 (9.6)	71 (13)
CVVHDF	767 (76)	353 (79)	414 (73)
mCVVH	8 (0.8)	2 (0.4)	6 (1.1)
SCUF	12 (1.2)	4 (0.9)	8 (1.4)
Initial Anticoagulation, n (%)				0.02
Citrate	609 (60)	275 (62)	334 (59)
Heparin	252 (25)	100 (22)	152 (27)
None	81 (8)	46 (10)	35 (6.2)
Other	71 (7)	25 (5.6)	46 (8.1)
Other CRRT Treatment Data
Median Dose Day 1–7, ml/kg	44 (32–63)	45 (33–60)	44 (32–66)	0.63
Time to First Negative Fluid Balance, days	1 (0–1)	1 (0–1)	1 (0–1)	0.33

Continuous variables reported as median (IQR). p-values were obtained using Pearson’s Chi-squared test, Wilcoxon rank sum test, or Fisher’s Exact Test, as appropriate.

Abbreviations: CNS- central nervous system; PRISM III- Pediatric Risk of Mortality III score; PELOD-2- Pediatric Logistic Organ Dysfunction 2 score; VIS- vasoactive-inotropic score; CRRT- continuous renal replacement therapy; CVVH- continuous veno-venous hemofiltration; CVVHD- continuous veno-venous hemodialysis; CVVHDF- continuous veno-venous hemodiafiltration; mCVVH- modified continuous veno-venous hemofiltration; SCUF- slow continuous ultrafiltration

Among 1016 patients, 650 (64%) experienced MAKE-90, including 384 who died by 90 days (59%), 173 (27%) who had persistent kidney dysfunction without RRT dependence, and 93 (14%) who remained RRT dependent (Table 2). Septic patients were less likely to liberate from CRRT in the first 28 days, required longer duration of CRRT, and had longer ICU LOS in survivors (25 [15–40] vs. 19 days [10–42], p=0.004), fewer ventilator-free days (0 [0–28] vs. 20 days [0–28], p<0.001), and higher incidence of both in-hospital mortality (47% vs. 31%, p<0.001) and MAKE-90 (70% vs. 61%, p=0.002) (Table 2). While MAKE-90 was higher in septic patients, those who survived to 90 days were less likely to be RRT dependent compared to non-septic survivors (10% vs 18%, p=0.01) (Table 2). On time-to-event analysis, the cumulative probability of 90-day mortality was higher for septic patients compared to those without sepsis at all time points (Figure 1, p <0.001).

Table 2:

Outcomes in children with sepsis receiving continuous renal replacement therapy (CRRT) compared to those without sepsis.

Variable	All (n=1016)	Sepsis (n=446)	No Sepsis (n=570)	p
28-day CRRT Liberation Status, n (%)				<0.001
Liberated, n (%)	349 (34)	134 (30)	215 (38)
Liberation not attempted, n (%)	370 (36)	193 (43)	177 (31)
Reinstituted, n (%)	297 (29)	119 (27)	178 (31)
CRRT Duration, days	6 (3–14)	7 (4–15)	5 (3–13)	0.026
28-day Ventilator-Free Days	16 (0–28)	0 (0–28)	20 (0–28)	<0.001
28-day ICU-Free Days	0 (0–11)	0 (0–6)	0 (0–14)	<0.001
ICU LOS in Survivors, days	22 (12–41)	25 (5–40)	19 (19–42)	0.004
In-Hospital Mortality, n (%)	388 (38)	209 (47)	179 (31)	<0.001
MAKE-90, n (%)	650 (64)	309 (70)	341 (61)	0.002
90-day mortality, n (%)	384 (38)	209 (47)	175 (31)	<0.001
Persistent kidney dysfunction, n (%)	173 (33)	76 (36)	97 (31)	0.22
RRT dependence, n (%)	93 (15)	24 (10)	69 (18)	0.011

Continuous variables are reported as median (IQR). p-values were obtained using p-values were obtained using Pearson’s Chi-squared test or Wilcoxon rank sum test, as appropriate.

Abbreviations: CRRT- continuous renal replacement therapy; ICU- intensive care unit; LOS- length of stay

Figure 1:

Kaplan-Meier curve demonstrating cumulative probability of 90-day mortality for patients with and without sepsis requiring continuous renal replacement therapy.

Table 3 outlines results for the analyses examining the association of specific features with MAKE-90 in septic patients who survived >48 hours from CRRT initiation (n=393, Supplementary Figure 1). Septic patients who experienced MAKE-90 were older, more frequently had pre-existing comorbidities, had lower baseline SCr, were more likely to have a known baseline SCr, more often had their vasoactive infusions maintained or increased over the first 48 hours of CRRT (24% vs. 13%, p=0.009) and required vasoactive support for a greater percentage of time in the first week of CRRT (50% vs. 38%, p=0.006). Those experiencing MAKE-90 were more frequently admitted for respiratory failure, central nervous system (CNS) dysfunction, or a primary cardiac indication (p<0.001). On multivariable regression, admission diagnosis (respiratory failure, CNS dysfunction, and primary cardiac compared to a reference of shock, infection, or major trauma), the presence of comorbidities, having a known baseline SCr, and increasing percentage of time requiring vasoactive support in the first week of CRRT were independently associated with MAKE-90 (Table 3). Notably, the association between MAKE-90 and 48-hour vasoactive trend was not maintained on multivariable regression; there were also no significant associations on univariable or multivariable analysis with any of the measures of fluid balance and MAKE-90 (Table 3). Additionally, there were no differences in median VIS on CRRT Days 0–1 in those with MAKE-90 compared to those without, but patients with MAKE-90 had higher median VIS on CRRT Days 3–7 (Supplementary Table 2).

Table 3:

Univariate and multivariable analyses examining the association between demographic and clinical variables and the development of MAKE-90 in children with sepsis requiring CRRT.

Variable	Univariate Analysis (n=393)			Multivariable Analysis (n=393)
	MAKE-90 (n=260)	No-MAKE-90 (n=133)	p	aOR	95% CI	p
Age, years	11 (3–16)	7 (2–13)	0.031*	1.04	1.0–1.07	0.06
Sex, female (%)	124 (48)	61 (46)	0.73
Admission Diagnosis, n (%)			<0.001*
Shock/Infection/Major Trauma	127 (49)	99 (74)		Ref	--	--
Respiratory Failure	60 (23)	11 (8)		3.12	1.50–6.47	0.002
CNS Dysfunction	12 (5)	2 (2)		5.84	1.19–28.6	0.029
Primary Cardiac	22 (9)	3 (2)		4.54	1.27–16.2	0.020
Other	39 (15)	18 (14)		1.63	0.82–3.25	0.16
Comorbidities, yes (%)	227 (87)	83 (63)	<0.001*	2.81	1.55–5.09	<0.001
Baseline Serum Creatinine, mg/dl	0.40 (0.23–0.65)	0.48 (0.35–0.62)	0.036*	0.78	0.48–1.28	0.33
Known Baseline, yes (%)	168 (65)	52 (39)	<0.001*	1.95	1.18–3.22	0.009
PRISM III	14 (11–19)	16 (11–21)	0.18
PELOD-2 at CRRT Initiation	6 (3–10)	6 (3–8)	0.49
VIS at CRRT Initiation	10 (0–26)	6 (0–27)	0.50
Time from ICU Admission to CRRT, days	2 (1–8)	2 (1–4)	0.22
%Fluid Balance from ICU Admission to CRRT	9 (4–20)	11 (4–24)	0.21
Initial CRRT Modality			0.94
CVVH, n (%)	25 (9.6)	15 (11)
CVVHD, n (%)	24 (9.2)	13 (9.8)
CVVDHF, n (%)	206 (79)	104 (78)
mCVVH, n (%)	2 (0.8)	0 (0)
SCUF, n (%)	3 (1.2)	1 (0.8)
Vasoactive Change 1st 48h CRRT			0.009*
Never/Decreased, n (%)	197 (76)	116 (87)		0.61	0.32–1.18	0.14
Unchanged/Increased, n (%)	62 (24)	17 (13)		Ref	--	--
Percentage of Day 1–7 of CRRT Requiring Vasoactives	50 (13–100)	38 (0–76)	0.006*	1.01	1.00–1.01	0.016
CRRT Dose Day 1–2, ml/kg	43 (32–58)	44 (32–60)	0.74
CRRT Dose Day 1–7, ml/kg	43 (33–59)	45 (32–60)	0.92
Time to First Negative Fluid Balance, days	1 (0–1)	1 (0–1)	0.77
Median Fluid Balance Day 1–2, ml/kg/day	−6 (−19,7)	−7 (−21,10)	0.60
Median Fluid Balance Day 1–7, ml/kg/day	−4 (−13,3)	−7 (−16,4)	0.087*	1.00	1.00–1.00	0.80

Continuous variables reported as median (IQR).

^*Variables with p<0.15 on univariate analysis were included in the multivariable model, which also adjusted for center using mixed-effects logistic regression.

Abbreviations: PRISM III- Pediatric Risk of Mortality III score; PELOD-2- Pediatric Logistic Organ Dysfunction 2 score; VIS- vasoactive-inotropic score; CRRT- continuous renal replacement therapy; CVVH- continuous veno-venous hemofiltration; CVVHD- continuous veno-venous hemodialysis; CVVHDF- continuous veno-venous hemodiafiltration; mCVVH- modified continuous veno-venous hemofiltration; SCUF- slow continuous ultrafiltration

Finally, increasing number of days of CRRT on vasoactives was independently associated with commensurate increases in the probability of MAKE-90 (Figure 2A, Supplementary Table 3, aOR 1.16, 95%CI 1.05–1.28, p=0.003) and in-hospital mortality (Figure 2B, Supplementary Table 4, aOR 1.20, 95%CI 1.1–1.32, p<0.001) in septic patients. Patients who never required vasoactive medications had a 57% probability of MAKE-90 compared to 82% in those who required vasoactives the entire first week of CRRT (Figure 2A). Similarly, patients who did not require vasoactives had a 27% probability of in-hospital mortality, and that probability increased to 63% if they required vasoactives each of the first 7 days of CRRT (Figure 2B).

Figure 2:

Probability of Major Adverse Kidney Events at 90 days from continuous renal replacement therapy initiation (A) and in-hospital mortality (B) as a function of increasing number of CRRT days requiring vasoactive support in the first week. Curves were generated using a mixed effect logistic regression model adjusting for age, admission diagnosis, the presence of comorbidities, baseline serum creatinine (mg/dl), whether their baseline serum creatinine was known or imputed, and median daily fluid balance over the first week of CRRT.

Discussion:

In this contemporary, large-scale, international study of children and young adults requiring CRRT for AKI and/or fluid overload, we demonstrate that sepsis is a common inciting event (44% of all patients) that is associated with different clinical characteristics and outcomes compared to non-septic patients. Specifically, septic children and young adults requiring CRRT were older, had greater severity of illness, more commonly required vasoactive medications, and had more fluid accumulation at CRRT initiation. They were less likely to liberate from CRRT in the first 28 days and had greater incidence of mortality and MAKE-90, though patients who survived were less likely to be RRT dependent at 90 days compared to non-septic survivors. Despite these differences, there were no differences in the prescribed dose of CRRT between septic and non-septic patients, an area requiring further investigation. Finally, increasing duration of vasoactive support on CRRT was independently associated with MAKE-90 and mortality in septic patients, representing a potential tool for bedside prognostication.

The finding that septic children and young adults requiring CRRT have higher severity of illness than those without sepsis is unsurprising, given the strong associations between sepsis-associated AKI and mortality(4–6). In this study, nearly half of septic patients requiring CRRT died, compared to just one-third of non-septic patients. Interestingly, while MAKE-90 was also more common in septic patients, the difference between these groups was less pronounced as septic survivors were less likely to be RRT dependent at 90 days. Thus, these data suggest that if a child survives their initial septic insult, the likelihood of renal recovery sufficient to liberate from RRT by 90 days is higher than those who required CRRT for other reasons. The reason behind this difference is unclear; notably, there were no differences in the number or type of baseline comorbid conditions that might decrease the likelihood of renal recovery in non-septic compared to septic patients. However, it is possible that AKI etiology (for which we lack granular data) was different between groups, impacting the risk for ongoing sequelae. Another possibility is mitigation of the sepsis-induced inflammatory response by CRRT resulting in attenuation of ongoing injury and increased likelihood of renal recovery(25, 26), though no direct evidence exists to support this hypothesis as these markers were not measured. Exploring the reasons behind this difference in RRT dependence at 90 days is an important area of future research.

Though septic patients had different clinical features than non-septic patients, few of these differences were associated with MAKE-90. In contrast to previous literature(12, 13), there was no difference seen in the time to CRRT initiation from ICU admission in patients who had MAKE-90 compared to those who did not. However, the median time to CRRT initiation in septic patients in this cohort was relatively short at 2 days, which would fall within the “early initiation” categorization of the previous literature(12, 13). Additionally, while fluid overload at CRRT initiation has been demonstrated to be associated with poor outcomes in critically ill children(27), there was no difference in percent fluid balance at CRRT initiation between patients with and without MAKE-90. Given the median percent fluid balance at CRRT initiation was relatively low in patients with sepsis, it is possible that practice changes adopted since the ppCRRT data were published has mitigated fluid overload as a key driver of outcomes in these patients(27).

While direct evidence of intradialytic hypotension (i.e., dialytrauma) was not captured by this dataset(15, 16, 28, 29), duration of vasoactive support was independently associated with MAKE-90 and mortality in septic children. Our work suggests that while nearly 3 in 4 children and young adults with septic AKI requiring CRRT will experience MAKE-90, including over half of whom will die, this risk is modulated by the need for and duration of vasoactive support. While some of this can likely be attributed to severity of illness, even after adjusting for relevant covariates, the differences in probability of MAKE-90 and mortality between patients who never required vasoactives and those requiring vasoactive support for the entire first week of CRRT are staggering (25% and 36%, respectively). Interestingly, initial intensity of vasoactive support (by pre-CRRT VIS and change in VIS over the first 48 hours) were not associated with either outcome, highlighting the possibility that the initial severity of shock may be less important than duration of ongoing need for vasoactives. While future work incorporating granular blood pressure data and reasons for vasoactive requirement are needed to better understand these findings, these data provide a framework to consider when counseling patients and families at the bedside (i.e., regarding the probability of poor outcomes as need for vasoactive support persists) and highlights the need for future studies examining how vasoactive medications should (or should not) be used to facilitate fluid removal in this population.

This study has several strengths. It is the first large, multi-center, international study to examine children and young adults with sepsis requiring CRRT, describing and contrasting their characteristics and outcomes to those without sepsis, and identifying clinical features that may impact outcomes. The multinational and multicenter nature of the study makes the results more generalizable across global populations. However, this work also has limitations. This was a retrospective study, and there is the possibility of unintended bias. The lack of blood pressure information to better quantify intradialytic hypotension, as well as reasons for and degree of vasoactive medication changes, are important gaps to be filled in future work. The study was designed prior to the newly released pediatric sepsis consensus criteria and thus relies on a previously used definition, though the incidence of organ failures in this cohort alleviates concerns about a high misclassification rate(17, 30). We lack data regarding the sepsis-induced inflammatory response, a key-driver of septic AKI that likely contributes to treatment response heterogeneity and is an important area of future research (7, 31–33). Baseline SCr was known in only 55%, and the association between known baseline and MAKE-90 raises the possibility that the incidence of MAKE-90 was underestimated. Other long-term kidney health data, such as incidence of hypertension and proteinuria, are lacking. Finally, data were missing for some patients; however, only patients with complete data were included in the multivariable regression analyses, and the overall proportion of other missing data for the cohort was below 5%.

Conclusions:

In a large, multi-national, multi-center study of children and young adults requiring CRRT for AKI and/or fluid overload, nearly half of all subjects had sepsis as the inciting event, and these patients had different clinical characteristics and outcomes compared to those without sepsis. Duration of vasoactive support over the first week of CRRT was independently associated with greater mortality and MAKE-90, and future studies are needed to understand the interaction between vasoactive duration and poor outcomes in this population.

Key Points:

Question:

Do children and young adults with sepsis receiving continuous renal replacement therapy (CRRT) have different characteristics and outcomes compared to non-septic patients and does dialytrauma contribute?

Findings:

Secondary analysis of an international, multicenter study of 1016 patients aged 0–25 years who received CRRT from 2015–2021. We determined that 44% had sepsis, that these patients had worse outcomes than non-septic patients, and that increasing duration of vasoactive support during CRRT was independently associated with poor outcomes.

Meaning:

Children and young adults with sepsis undergoing CRRT experience poor outcomes. Further study is needed to optimize CRRT use in this population.

Consortia-on behalf of the WE-ROCK Investigators:

The following individuals served as collaborators and investigators for the WE-ROCK studies. They collaborated in protocol development and review, data analysis, and participated in drafting or review of the manuscript, and their names should be citable by PubMed.

Emily Ahern CPNP, DNP¹, Ayse Akcan Arikan MD², Issa Alhamoud MD³, Rashid Alobaidi MD, MSc⁴, Pilar Anton-Martin MD, PhD⁵, Shanthi S Balani MD⁶, Matthew Barhight MD, MS⁷, Abby Basalely MD, MS⁸, Amee M Bigelow MD, MS⁹, Gabriella Bottari MD¹⁰, Andrea Cappoli MD¹⁰, Eileen A Ciccia MD¹¹, Michaela Collins BA¹², Denise Colosimo MD¹³, Gerard Cortina MD¹⁴, Mihaela A Damian MD, MPH¹⁵, Sara De la Mata Navazo MD¹⁶, Gabrielle DeAbreu MD⁸, Akash Deep MD¹⁷, Kathy L Ding BS¹⁸, Kristin J Dolan MD², Sarah N Fernandez Lafever MD, PhD¹⁶, Dana Y Fuhrman DO, MS¹⁹, Ben Gelbart MBBS²⁰, Katja M Gist, DO MSc¹², Stephen M Gorga MD, MSc²¹, Francesco Guzzi MD²², Isabella Guzzo MD¹⁰, Taiki Haga MD²³, Elizabeth Harvey MD²⁴, Denise C Hasson MD²⁵, Taylor Hill-Horowitz BS⁸, Haleigh Inthavong BS, MS², Catherine Joseph MD², Ahmad Kaddourah MD, MS²⁶, Aadil Kakajiwala MD, MSCI²⁷, Aaron D Kessel MD, MS⁸, Sarah Korn DO²⁸, Kelli A Krallman BSN, MS¹², David M Kwiatkowski MD Msc²⁹, Jasmine Lee MSc²⁴, Laurance Lequier MD⁴, Tina Madani Kia BS⁴, Kenneth E Mah MD, MS¹⁵, Eleonora Marinari MD¹⁰, Susan D Martin MD³⁰, Shina Menon MD³¹, Tahagod H Mohamed MD⁹, Catherine Morgan MD MSc⁴, Theresa A Mottes APRN⁷, Melissa A Muff-Luett MD³², Siva Namachivayam MBBS²⁰, Tara M Neumayr MD¹¹, Jennifer Nhan Md, MS²⁷, Abigail O’Rourke MD⁸, Nicholas J Ollberding PhD¹², Matthew G Pinto MD³³, Dua Qutob MD²⁶, Valeria Raggi MD¹⁰, Stephanie Reynaud MD³⁴, Zaccaria Ricci MD¹³, Zachary A Rumlow DO³, María J Santiago Lozano MD, PhD¹⁶, Emily See MBBS²⁰, David T Selewski MD, MSCR³⁵, Carmela Serpe MSc, PhD¹⁰, Alyssa Serratore RN, MsC²⁰, Ananya Shah BS¹⁸, Weiwen V Shih MD^1,18, H Stella Shin MD³⁶, Cara L Slagle MD¹², Sonia Solomon DO³³, Danielle E Soranno MD³⁷, Rachana Srivastava MD³⁸, Natalja L Stanski MD¹², Michelle C Starr MD, MPH³⁷, Erin K Stenson MD^1,18, Amy E Strong MD, MSCE³, Susan A Taylor MSc¹⁷, Sameer V Thadani MD², Amanda M Uber DO³², Brynna Van Wyk ARNP, MSN³, Tennille N Webb MD, MSPH³⁹, Huaiyu Zang PhD¹², Emily E Zangla DO⁶, Michael Zappitelli MD, MSc²⁴

¹Children’s Hospital Colorado, University of Colorado School of Medicine, Aurora, CO, USA

²Baylor College of Medicine, Texas Children’s Hospital, Houston, TX, USA

³University of Iowa Stead Family Children’s Hospital, Carver College of Medicine, Iowa City, IA, USA

⁴Univeristy of Alberta, Edmonton, Canada

⁵Le Bonheur Children’s Hospital, Memphis, TN, USA

⁶University of Minnesota, Minneapolis, MN, USA

⁷Ann and Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL, USA

⁸Cohen Children’s Medical Center, Zucker School of Medicine, New Hyde Park, NY, USA

⁹Nationwide Children’s Hospital, The Ohio State University College of Medicine, Columbus, OH, USA

¹⁰Bambino Gesù Children Hospital, IRCCS, Rome, Italy

¹¹Washington University School of Medicine, St. Louis Children’s Hospital, St. Louis, MO, USA

¹²Cincinnati Children’s Hospital Medical Center; University of Cincinnati College of Medicine, Cincinnati, OH, USA

¹³Meyer Children’s Hospital, IRCCS, Florence, Italy

¹⁴Medical University of Innsbruck, Innsbruck, Austria

¹⁵Stanford University School of Medicine, Palo Alto, CA, USA

¹⁶Gregorio Marañón University Hospital; School of Medicine, Madrid, Spain

¹⁷King’s College Hospital, London, England

¹⁸University of Colorado, School of Medicine, Aurora, CO, USA

¹⁹University of Pittsburgh Medical Center Children’s Hospital of Pittsburgh, Pittsburgh, PA, USA

²⁰Royal Children’s Hospital, University of Melbourne, Murdoch Children’s Research Institute, Melbourne, Victoria, Australia

²¹University of Michigan Medical School, C.S. Mott Children’s Hospital, Ann Arbor, MI, USA

²²Santo Stefano Hospital, Prato, Italy

²³ Osaka City General Hospital, Osaka, Japan

²⁴Hospital for Sick Children, Toronto, Ontario, Canada

²⁵NYU Langone Health, Hassenfeld Children’s Hospital, New York, NY, USA

²⁶ Sidra Medicine and Weil Cornel Medicine, Qatar, Doha, Qatar

²⁷Children’s National Hospital, Washington, DC, USA

²⁸Westchester Medical Center, Westchester, NY, USA

²⁹Lucile Packard Children’s Hospital, Palo Alto, CA, USA

³⁰Golisano Children’s Hospital at University of Rochester Medical Center, Rochester, NY, USA

³¹Seattle Children’s Hospital, University of Washington, Seattle, WA, USA

³²University of Nebraska Medical Center, Children’s Hospital & Medical Center, Omaha, NE, USA

³³Maria Fareri Children’s Hospital at Westchester Medical Center, Valhalla, NY, USA

³⁴Hopital Bicetre, APHP Université Paris-Saclay, Kremlin-Bicetre, Val de Marne, France

³⁵Medical University of South Carolina, Charleston, SC, USA

³⁶Children’s Healthcare of Atlanta, Emory University, Atlanta, GA, USA

³⁷Indiana University School of Medicine, Riley Hospital for Children, Indianapolis, IN, USA

³⁸Mattel Children’s Hospital at UCLA, Los Angeles, Ca, USA

³⁹Children’s of Alabama/University of Alabama at Birmingham, Birmingham, AL, USA,

Acknowledgements: The follow individuals contributed to data collection, cleaning and management: T. Christine E. Alvarez MHI RN¹, Elizabeth Bixler BS², Erica Blender Brown MA, CRA³, Cheryl L Brown BS¹, Ambra Burrell BA⁴, Anwesh Dash BS⁵, Jennifer L Ehrlich RN MHA⁶, Simrandeep Farma HBSc⁷, Kim Gahring RN BSN, CCRN⁸, Barbara Gales RN⁹, Madison R Hilgenkamp¹⁰, Sonal Jain MS¹¹, Kate Kanwar BA MS⁴, Jennifer Lusk BSN RN, CCRN⁸, Christopher J. Meyer BA AA¹, Katherine Plomaritas BSN RN¹², Joshua Porter BS⁵, Jessica Potts BSN RN¹³, Alyssa Serratore BNurs, GDipNP(PIC), RN, MsC¹⁴, Elizabeth Schneider BS⁵, Vidushi Sinha BS⁵, PJ Strack RN,BSN,CCRN¹⁵, Sue Taylor RN¹⁶, Katherine Twombley MD³, Brynna Van Wyk MSN, ARNP CPNP⁶, Samantha Wallace MS¹⁷, Janet Wang BS⁵, Megan Woods BS⁵, Marcia Zinger RN¹⁸, Alison Zong BS⁵

¹Cincinnati Children’s Hospital Medical Center, Cincinnati, OH, USA

²Baylor College of Medicine, Texas Children’s Hospital, Houston, TX, USA

³Medical University of South Carolina, Charleston, SC, USA

⁴Nationwide Children’s Hospital, Columbus, OH, USA

⁵University of Tennessee Health Science Center College of Medicine, Memphis, TN, USA

⁶University of Iowa Stead Family Children’s Hospital, Carver College of Medicine, Iowa City, IA, USA

⁷Hospital for Sick Children, Toronto, ON, Canada

⁸Children’s Hospital Colorado, Aurora, CO, USA

⁹Mattel Children Hospital at UCLA, Los Angeles, CA, USA

¹⁰University of Nebraska Medical Center, Children’s Hospital & Medical Center, Omaha, NE, USA

¹¹Seattle Children’s Hospital, Seattle, WA, USA

¹²University of Michigan, C.S. Mott Children’s Hospital, Ann Arbor, MI, USA

¹³Children’s of Alabama/University of Alabama at Birmingham, Birmingham, AL, USA

¹⁴Royal Children’s Hospital, Melbourne, VIC, Australia

¹⁵Children’s Mercy Hospital, Kansas City, MO, USA

¹⁶King’s College Hospital, London, England

¹⁷Indiana University School of Medicine, Riley Hospital for Children, Indianapolis, IN, USA

¹⁸Cohen Children’s Medical Center, New Hyde Park, NY, USA

Data Availability:

De-identified summary data are available through the WE-ROCK collaborative. Data dictionaries, in addition to study protocol, the statistical analysis plan will be made available upon request. More information about the process and available data can be obtained by contacting the corresponding author (NLS). The data from the WE-ROCK collaborative will be made available to researchers who provide a methodologically sound proposal for use in achieving the goals of the approved proposal following an application process and execution of a data-use agreement as required by the Institutional Review Board at the Cincinnati Children’s Hospital Medical as part of the approval of this collaborative study.