Добавил:
Sekretar
kiopkiopkiop18@yandex.ru
t.me/Prokururor I Вовсе не секретарь, но почту проверяю
Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз:
Предмет:
Файл:Ординатура / Хирургия / Библиотека им академика М.И. Перельмана / Книга_5224_Библиотеки_им_академика_М_И_Перельмана.pdf
40 Indications and Timing of Renal Replacement Therapy 465
chemokine ligand 14 [CCL14]) have shown association with receipt of RRT . The
quality of available evidence for use of these and other biomarkers for decisions to
start or delay RRT is expanding; however, it is not sufficient to support routine use in
clinical practice [
15, 31, 32]. In a systematic review and meta-analysis of 41 studies
(n = 15,928), the pooled AUCs for predicting receipt of RRT for urine and serum
NGAL were 0.72 (95%CI 0.64–0.80) and 0.76 (95%CI 0.71–0.80), respectively,
while serum creatinine and cystatin C had pooled AUCs of 0.76 (95%CI 0.73–0.80)
and 0.77 (95%CI 0.73–0.81), respective ly [
31]. Urine biomarkers interleukin-18,
cystatin C, and cell cycle arrest markers [TIMP-2]•[IGFBP7] showed pooled AUCs
of 0.67 (95%CI 0.61–0.73), 0.72 (95%CI 0.58–0.87), and 0.86 (95%CI 0.79–0.93),
respectively [31]. Recent data have characterized the urinary CCL14 for the prediction of persistent severe AKI [33–36]. There are challenges with the interpretation of
many of these biomarkers for predicting receipt of RRT. For example, many studies
are small, susceptible to selection bias, and confounded by case-mix and acuity, use
single biomarker measures and variable thresholds for discrimination, and often do
not fully integrate into models of baseline prediction using common clinical
The
32].
use
a dynamic response to a “standardized” furosemide stress test (FST) to
of
variables [
“interrogate” tubular cell function can aid in predicting the likelihood of worsening
or persistent AKI [37–40
].
This point-of-care assessment can be integrated into
available clinical information when considering and planning for RRT. The principle
of the FST is relatively simple. Furosemide is a largely lipid insoluble organic anion
transported to the kidneys highly bound to albumin. To gain access to the tubular
lumen, furosemide is secreted across the proximal tubule via organic anion transporters (OAT) 1 and OAT3 on the basolateral member. Following secretion into the
tubula
r lumen, furosemide inhibits active chloride transport throughout the thick
ascending limb (TAH) of the loop of Henle by competitively binding the 2NaKCl
cotransporter. The FST comprises a single intravenous (IV) dose of furosemide
(1.0 mg/kg for diuretic naïve patients and 1.5 mg/kg for prior diuretic exposure)
along with replacement of resulting urine output (1 mL crystalloid for each 1 mL
urine 6 h to mitigate risk of iatrogenic volume depletion). A threshold urine output of
>200 mL in 2 h following IV furosemide administration would suggest kidney
tubular function is relatively intact (sensitivity 87%; specificity 84%) [
37–39].
The
FST can be leveraged as a screening tool to identify patients whom AKI is likely to
persist or worsen. In a small clinical trial, among critically ill patients with AKI, only
14% of FST responsive patients were started on RRT, whereas 78% of those not
responding enrolled in a delayed strategy started RRT or died [40]. Recently, the
FST has been used in concert with urinary CCL14 to estimate decision support for
41].
starting RRT in a single center cohort of patients undergoing cardiac surgery [
In
this study, the primary endpoint was the development of a conventional indication
for starting RRT (e.g., severe hyperkalemia, diuretic-resistance volume overload,
metabolic severe acidosis, serum urea >53 mmol/L, oligoanuria). Among the
208 enrolled postoperative patients, 108 (52%) had a negative FST (urine output
<200 mL in 2 h), of which 82% fulfilled the primary endpoi nt and 71% received
RRT. These patients had significantly higher urinary CCL14 levels (1.34 ng/
466 S. M. Bagshaw and R. Wald
mL vs. 6.47 ng/mL, p < 0.001). The inclusion of both the FST and urinary CCL14
coupled with a baseline clinical model improved the AUC from 0.71 to 0.87 to
discriminate the primary endpoint. These observations reinforce how novel strategies, such as the FST or biomarkers, may be integrated to better inform decision
support on likelihood and planning for RRT [
42].
Rationale for an Early Strategy to Starting RRT
There is strong physiological rationale for why earlier RRT initiation in critically ill
patients with severe AKI, even in the absence of urgent indications, should confer
benefit in selected circumstances [43] (Table 40.2). Earlier RRT can theoretically
facilitate timely correction of electrolyte and metabolic and acid-base derangements,
along with mitigating or preventing complications of fluid accumulation
40.2). Earlier RRT can prevent AKI complications from occurring and may
(Table
facilitate weaning from organ support [e.g., invasive mechanical ventilation) and
short ICU stay [10, 44]. Whether RRT has a role in modulating inflammation and
immune function in sepsis and other vasoplegic states (e.g., post-cardiac surgery) is
biologically plausible; however, it remains largely unproven to favorably impact
clinical outcomes [45, 46].
Rationale for a Conservative Strategy to Starting RRT
A clinically meaningful proportion of critically ill patients that develop severe AKI
will recover kidney funct ion without receipt of RRT. From this, a reasonable
inference is that conservative strategy of careful “watching and waiting” can translate into selected patients avoiding RRT [
relative fixed risks associated with RRT (e.g., complications related to catheters,
anticoagulation, blood contact with extracorporeal circuit, circuit loss, intradialytic
hypotension, etc.). This is particularly relevant for patients who may have started on
RRT for marginal indications, where the risks exceed benefit [49]. The avoidance of
RRT could also logically translate into sparing health resources and cost savings.
However, there is uncertainty whether delays to starting RRT among those with a
high likelihood of receiving or deriving benefit from RRT could exert attributable
risk due to the prolonged consequences of severe persistent AKI. In the AKIKI trial,
patients allocated to the delayed RRT strategy had biochemical evidence of worsening AKI (e.g., higher serum creatinine, urea, and potassium; lower serum bicarbonate and pH) at the time of starting RRT compared to patients allocated to the
early strategy [
icantly more interventions to manage the fluid and metabolic complications of AKI
(e.g., hyperkalemia, metabolic acidosis) [50]. Similarly, in the STARRT-AKI trial,
patients allocated to the standard strategy showed worsening biochemical evidence
Moreover, patients in the delayed strategy also received signif-
50].
47, 48]. RRT avoidance would mitigate the
40 Indications and Timing of Renal Replacement Therapy 467
of AKI and a greater toxic milieu upon starting RRT compared with those in the
accelerated strategy [48]. In the IDEAL-ICU trial, 17% of patients with septic shock
allocated to the delayed strategy were started on RRT for emergent indications
[51]. Observational data have suggested that the accumulation of AKI-related
complications may exert an incremental risk for worse outcome [10]. In the
AKIKI-2 trial, 278 critically ill patients with severe persistent AKI without urgent
indications for RRT were randomly allocated to either start RRT immediately
(reflecting the delayed strategy in AKIKI) or subjected to further RRT deferral
(“more delayed”), commencing only for serum urea >50 mmol/L or an urgent
indication [
through day 28 between the groups (12 days for delayed vs. 10 days for more
delayed; p = 0.93). However, in a prespecified analysis, 60-day mortality was higher
in the more delayed strategy compared to the delayed strategy (55% for more
delayed vs. 44% for delayed; adjusted HR 1.65 [95% CI, 1.09–2.50], p = 0.07) [52].
52]. There was no difference in the primary endpoint of RRT-free days
RRT Replacement Therapy and Clinical Outcomes
The growth in utilization of RRT likely reflects temporal changes to patient demographics, multi-morbidity (e.g., chronic kidney disease), and innovations in medical
therapies (e.g., cancer therapeutics) and surgical interventions (e.g., complex cardiac
or hepatobiliary surgery, transplantation). While RRT adds complexity and costs to
the bedside care of critically ill patients, trends have shown modest reductions in
mortality among those receiving RRT in selected clinical contexts [10, 53].
There is uncertainty for selected patients whether receipt of RRT significantly
modifies outcomes or whether RRT, as an organ support intervention, is more of a
surrogate for burden of premorbid disease (e.g., chronic kidney disease, cardiovascular disease, cancer), physiological reserve, and severity of critical illness. Select
studies have implied that the receipt of RRT may exert a direct hazard for death
[54, 55] or not show significant association with outcomes when compared with
critically ill patients not receiving RRT [
tainly susceptible to bias and limited generalizability due to differences in
populations (case-mix, illness acuity), confounding by indication and uncontrolled
sources of bias (practice variation, information bias) [56]. Indeed, patient-, clinician-
11].
These observational studies are cer-
and outcome, particularly variation in decision-making to offer and start RRT
11].
[
There may also be important effects of the type and size of ICU and volume
of RRT performed and outcomes, with smaller ICUs, those with less experience and
those treating fewer patients showing greater adjusted risk for mortality [55, 57].
468 S. M. Bagshaw and R. Wald
Current Clinical Practice Guideline Recommendations
Several organizations have published clinical practice guideline statements focused
on timing of initiation of RRT in critically ill settings [58– 60] (Table 40.3). In 2012,
the Kidney KDIGO consortium made two statements regarding the timing of RRT
initiation in AKI, based on expert opinion [58]. The first was to start RRT
“emergently when life-threatening changes in fluid, electrolyte, and acid-base bal-
ance exist.” The second asked clinicians to consider the “broader clinical context, the
presence of conditions that can be modified with RRT, and trends of laboratory
tests—rather than single BUN and creatinine thresholds alone—when making the
decision to start RRT.” Though the second statement may be perceived as vague and
provide clinicians with a wide scope of subjective parameters to inform their
decision-making, it was a reasonable reflection of contemporary bedside practice
given the available evidence. In 2013, the National Institute for Health and Care
Excellence (NICE) in the United Kingdom published recommendations that are
similar to KDIGO [59]. The NICE guidelines acknowledged the paucity of evidence
from RCTs on when to start RRT in both critically ill children and adults and
emphasized the need to develop and evaluate tools, such as clinical risk prediction
scores or novel point-of-care tests (e.g., novel kidney damage biomarkers). In 2015,
the French Intensive Care Society (SRLF) published recommendations for RRT in
ICU settings, including statements on when to start RRT [60]. Each of these
organizations acknowledged the limitations in current evidence and associated
clinical uncertainty, and each called for additional evidence from RCTs to inform
practice and update guidelines. Since their publication, several RCTs have been
reported [47, 48, 50, 51, 61]. Future iterations of these and any other clinical practice
guidelines should aim to integrate the principles of shared decision-making with
patients and families in decisions on if and when to escalate support with RRT [15]
(Fig. 40.2).
Clinical Trial Evidence on Timing of Starting RRT
Several high-quality randomized trials recently aimed to better inform practice by
generating new knowledge on when to start RRT in critically ill patients with AKI
[40, 50, 51, 62–64].
The Early Ver
Patients With Acute Kidney Injury (ELAIN) trial was a single-center German RCT of
231 critically ill patients that evaluated whether early RRT, defined as starting RRT
within 8 h of fulfilling stage 2 AKI, would improve survival as compared to delayed
RRT, defined as starting RRT within 12 h of developing stage 3 AKI or upon the
development of an urgent indication [65] (Table 40.4). Eligible patients were
predominantly surgi cal, had a plasma NGAL >150 ng/mL (as a surrogate of tubular
damage), and had one of the following: sepsis, volume overload, nonrenal organ
sus Late Initiation of Renal Replacement Therapy In Critically Ill
40 Indications and Timing of Renal Replacement Therapy 469
Table 40.3 Summary of clinical practice guideline statements for starting RRT in critically ill
patients with AKI
Organization Recommendations
Kidney Disease: Improving Global
Outcomes (KDIGO) [58]
National Institute for Health and
Care Excellence (NICE) [
French I
(SRLF) [
ntensive C
60]
59]
are Society
(i) Initiate KRT emergently when life-threatening
changes in fluid, electrolyte, and acid-base balance exist
(not rated)
(ii) Consider the broader clinical context, the presence
of conditions that can be modified with KRT, and trends
of laboratory tests—rather than single BUN and creatinine thresholds alone—when making the decision to
start KRT (not rated)
(i) Discuss any potential indications for renal replacement therapy with a nephrologist, pediatric nephrologist, and/or critical care specialist immediately to ensure
that the therapy is started as soon as needed
(ii) Refer adults, children, and young people immediately for KRT if any of the following are not responding
to medical management:
• Hyperkalemia
• Metabolic acidosis
• Complications of uremia (i.e., pericarditis, enceph-
alopathy)
• Fluid overload
• Pulmonary edema
(iii) Base the decision to start KRT on the condition of
the adult, child, or young person as a whole and not on
an isolated urea, creatinine, or potassium value
(i) KRT
threatening situations (hyperkalemia, metabolic acidosis, tumor lysis syndrome, refractory pulmonary edema)
(expert opinion; strong agreement)
(ii) The available data are insufficient to define optimal
timing of initiation of KRT outside life-threatening situations (expect opinion; strong agreement)
(iii) In children, fluid and sodium overload probably of
above 10% and very probably of above 20% should be
consi
replacement therapy (expert opinion; poor agreement)
(iv) “Early” initiation of KRT means at KDIGO stage
2 or within 24 h after onset of acute renal failure of
which reversible seems unlikely (expert opinion; poor
agreement)
(v) “late” initiation of KRT means over 48 h after onset
of acute renal failure, KDIGO stage 3, or when a lifethreatening situation arises because of acute renal failure
(expert opinion; poor agreement)
should be initiated without delay in life-
as one of the criteria for initiation of renal
dered
dysfunction, or need for vasoactive support. All patients in the early strategy started
RRT, as well as 91% in the delayed strategy, mostly triggered by AKI progression
[64]. The median difference from randomization to starting RRT was less than 1 day
(21 h [IQR] 18 – 24). Mortality at 90 days was reduced by 15.4% in the early
470 S. M. Bagshaw and R. Wald
Fig. 40.2 Proposed algorithm for initiation of RRT in critically ill patients with AKI
compared with delayed strategy (39.3% vs. 54.7%; HR 0.66; 95% CI, 0.45–0.97).
ELAIN also found that the early strategy confer red a greater likelihood of kidney
recovery and RRT independence and shorter stay in ICU and hospital, compared
with a delayed strategy. ELAIN found the benefits of the early strategy on a
composite of major adverse kidney events (MAKE) to be durable [66].
The Artificial
Kidney
Initiat
ion
K
idney Injury (AKIKI] trial was a French
in
multicenter RCT that evaluated a delayed strategy of RRT initiation compared
with an early strategy on 60-day survival in 620 critically ill patients with severe
AKI who were receiving mechanical ventilation and/or vasoactive support [50]
(Table 40.4). Patients in the early strategy started RRT within 6 h of fulfilling
stage 3 AKI, and those in the delayed strategy only started RRT for medically
refractory complications (e.g., oliguria or anuria for ≥72 h following randomization,
uremia, hyperkalemia, metabolic acidosis, volume overload). There was no difference in 60-day all-cause mortality between the strategies (49.7% vs. 48.5%,
p = 0.79). Only 51% of patients in the delayed strategy received RRT compared
with 98% in the early strategy. The median difference for starting RRT between
strategies was 57 h (IQR 25–83) among those who received RRT. RRT-free days
were greater (19 vs. 17 days, p < 0.001), and the occurrence of catheter-related
bloodstream infections was lower (5% vs. 10%, p = 0.03) with the delayed strategy.
There were no differences in any other prespecified outcomes. AKIKI also reported a
post hoc analysis focused on the 60 (10%) patients with premorbid CKD—there was
suggestion of heterogeneity in treatment effect on 60-day mortality, with those
allocated to early RRT having a greater risk of death [67]
The Initiation of
Dialysis EArly versus Late in the Intensive Care Unit (IDEAL-
.
ICU) was a French multicenter randomized trial of early versus delayed RRT
strategies that aimed to enroll 864 patients with septic shock and AKI [
68]. In
IDEAL-ICU, patients fulfilling RIFLE-Failure AKI criteria within the first 48 h of
40 Indications and Timing of Renal Replacement Therapy 471
Table 40.4 Summary of recent randomized trials evaluating the starting of renal replacement
therapy in the setting of acute kidney injury in the intensive care unit
STARRT-
Feature ELAIN AKIKI IDEAL-ICU
AKI
AKIKI-2
Country Germany France France International France
No. of sites 1 31 24 168 39
No. of
231 620 488
a
3019 278
participants
Setting/
population
ARR for sam-
Mixed medical/surgical
ICU
(94.8%
surgical)
Mixed medical/surgical
ICU
(79.7%
medical)
Mixed medical/surgical
ICU
(septic shock)
Mixed medical/surgical
ICU
18% 15% 10% 6%
Mixed medical/
surgical ICU
(58.0% septic
shock)
ple size
calculation
Control
55% 55% 55% 44%
group
mortality
Interventions:
Early
(accelerated)
KDIGO
stage 2
(within 8 h)
KDIGO
stage 3
(within 6 h)
RIFLE-failure
(within 12 h)
KDIGO
stage 2
(within
–
12 h)
Delayed
(conservative)
KDIGO
stage 3
(within
12 h)
Specific
criteria/
emergent
indications
Specific
criteria
48–60 h after
eligibility or
emergent
Specific
criteria/
emergent
indications
KDIGO stage
3 with oliguria
>72 h; urea
40–50 mmol/L
(within 12 h)
indications
More
delayed
Urgent indications; urea
>50 mmol/L
Time
25.5 h 57.0 h 43.9 h 25.0 h 30.0 h
difference
Received
90.8% 51.0% 62.0% 61.8% 79%
RRT in
delayed
RRT
modality
CRRT Physician
discretion
Physician
discretion
Physician
discretion
Physician
discretion
(Initial IHD
55%)
SOFA score
~16.0 ~10.9 ~12.3 ~11.7
at enrollment
Primary
endpoint
Early
90-day
mortality
60-day
mortality
90-day
mortality
39.3% 48.5% 58.0% 43.9% 12 days
90-day
mortality
RRT-free
(day 28)
days
(accelerated)
Delayed
54.7% 49.7% 54.0% 43.7% 10 days
(conservative)
(continued)
472 S. M. Bagshaw and R. Wald
Table 40.4 (continued)
STARRT-
Feature ELAIN AKIKI IDEAL-ICU
Effect
estimate
Kidney
recovery
Early
HR, 0.66
(95% CI,
0.45–0.97)
RRT depen-
dence at
90 days
HR, 1.03
(95% CI,
0.82–1.29)
RRT depen-
dence at
60 days
RR, 1.08
(95% CI,
0.90–1.30)
RRT depen-
dence at
90 days
53.6% 2.0% 2.0% 10.4% 4.0%
AKI AKIKI-2
b
RR, 1.00
(95% CI,
0.93–1.09)
RRT depen-
dence at
90 days
p = 0.93
RRT depen-
dence at
60 days
(accelerated)
Delayed
38.7% 5.0% 3.0% 6.0% 2.0%
(conservative)
Effect
estimate
Adverse
events
Early
OR, 0.55
(95% CI,
0.32–0.93)
RR, 0.53
(95% CI,
0.20–1.41)
b
RR, 0.83
(95% CI,
0.28–2.46)
Aggregate CRBSI Emergent
RRT
75.0% 10.0% – 23.0% 44.0%
b
RR, 1.74
(95% CI,
1.24–2.43)
c
Aggregate 60-day
RR 2.0
(95% CI,
0.19–8.25)
mortality
(accelerated)
Delayed
(conservative)
Effect
estimate
Abbreviations: AKI acute kidney
68.5% 3.0% 17% 16.5% 55.0%
RR, 1.18
(95% CI,
0.86–1.61)
b
RR, 1.35
(95% CI,
1.08–1.68)
b
– RR, 1.40
(95%
CI,
1.21–1.62)
HR 1.65
(95% CI,
1.09–2.50)
injury, ARR absolute risk reduction, CKD chronic kidney disease,
d
ICU intensive care unit, KDIGO Kidney Disease: Improving Global Outcomes, RRT renal replacement therapy, SOFA Sequential Organ Failure Assessment, CRBSI catheter-related bloodstream
infection
a
Terminated prematurely due to futility
b
Calculated. Not provided in primary publication
c
Reported for patients allocated to the delayed KRT strategy only
d
Adjusted by Simplified Acute Physiology Score (SAPS) 3, mechanical ventilation, catecholamine
infusion, sepsis status, time between ICU admission and acute kidney injury
onset of septic shock were eligible. The early strategy was defined as starting RRT
within 12 h of eligibility, whereas the delayed strategy was defined by RRT being
deferred for ≥48 h (but no more than 60 h) from the onset of RIFLE-Failure AKI,
unless patients developed AKI complications and urgent indications to start RRT
(Table 40.4). The trial was terminated prematurely due to futility after enrolment of
488 patients (56.5%). Mortality at 90 days was similar (58% in the early vs. 54% in
51]. T
the delayed) [
he majority (97%) of patients in the early strategy received RRT,
whereas only 62% received RRT in the delayed strategy. In 17% of patients in
delayed strategy, emergent RRT was started prior to 48 h in response to complications. Mortality at 90 days in these patients was 68%, implying there may be risk to
the expectant follow-up of AKI and the protocolized delay of RRT initiation in
selected patients [10, 50].
40 Indications and Timing of Renal Replacement Therapy 473
The STandard versus Accelerated initiation of Renal Replacement Therapy in
Acute Kidney Injury (STARRT-AKI) trial was a large multinational collaboration
across 15 countries and 168 sites designed to compare a strategy of accelerated
(early) to standard (delayed) RRT initiation in critically ill patients with severe AKI
(aligned with KDIGO stage 2) who did not have an urgent indication for starting
RRT [
69]. The STARRT-AKI trial was pragmatic in its approach to starting RRT
and was unique among trials by integrating individual clinician equipoise into the
eligibility criteria [69]. The design enabled clinicians to exclude patients who
required immediate RRT or those with a high likelihood of imminent kidney
recovery and thus a low likelihood of ever receiving RRT. Patients in accelerated
strategy started RRT within 12 h of fulfilling eligibility. The standard strategy
comprised of a “watch and wait” approach, whereby RRT was discouraged unless
patients developed an urgent indication (e.g., serum potassium ≥6.0 mmol/L, pH
≤7.20 or serum bicarbonate ≤12 mmol/L, volume overload, or the persistence of
AKI for ≥72 h) [69]. The trial randomized 3019 patients, of whom 2927 were
included in the modified intention-to-treat analysis. Patients were predominantly
medial (67%) and commonly had CKD (44%), and 58% had sepsis (Table 40.4).
Among
[IQR] of 6.1 [3.9–8.8] h after full eligibility, whereas in the standard strategy, 61.8%
started RRT after a median [IQR] of 31.1 [19.0– 71.8] h. There was no difference in
all-cause 90-day mortality between the group (43.9% in the accelerated
strategy vs. 43.7% in the standard strategy (risk ratio [RR], 1.00; 95% CI,
0.93–1.09)). There were also no differences in survival across prespecified subgroups, including those with sepsis, CKD, and surgical status and within declines of
severity of illness scores. An important finding in STARRT-AKI was that among
survivors at 90 days, RRT dependence was more common in the accelerated strategy
(10% in the accelerated strategy vs. 6% in the standard strategy; RR 1.74, 95% CI
1.24–2.43). Adverse events, in particular RRT-associated hypotension and
hypophosphatemia, were more common in accelerated strategy (23.0% vs. 16.5%;
RR, 1.40; 95% CI, 1.21–1.62).
patients
randomized to the accelerated strategy, 97% started RRT a median
Implications for Practice
The findings of the STARRT-AKI trial, supported by those of the AKIKI and
IDEAL-ICU trials and meta-analyses, provide persuasive evidence that a delayed
start of RRT, characterized by “watch and wait” strategy and starting when
confronted with worsening, persistent, or medically refractory complications of
AKI, should now be recommended as the standard [47, 48, 50, 51]. However,
there remains some uncertainty on the safety of protracted deferral of RRT in the
presence of persistent AKI. The findings of STARRT-AKI, AKIKI, and IDEALICU may not be applicable to those prolonged and unresolving AKI (> 3–4 days).
However, this was the focus of the AKIKI-2 trial, where extended delay was shown
to potentially confer harm [
52].
474 S. M. Bagshaw and R. Wald
A key strategy for implementing evidence and harmonizing best practice is to
minimize nonessential variation. A recent controlled study in ICUs at the Brigham
and Women’s Hospital implemented a Standardized Clinical Assessment and Management Plan (SCAMP) for critically ill patients with AKI for decision support for
starting RRT [
70]. The SCAMP defined criteria for RRT initiation which comprised
a series of specific indications (pH <7.2, potassium >6.5 mmol/L, toxin ingestion,
volume overload, FiO
> 0.7, urine output <100 mL/24 h, uremic symptoms).
2
SCAMP and a “sham” control were applied in alternative periods over 1 year. There
was comparable likelihood of receiving RRT with both strategies and no difference
in mortality (41% vs. 47%). However, duration of stay in both ICU and hospital was
shorter during SCAMP periods compared to control. In addition, SCAMP implementation was associated with lower RRT utilization compared with control, driven
by patients where RRT was perceived by clinicians to be non-beneficial.
As with all organ support interventions in ICU settings, clinicians should aim to
adopt a shared approach to decision-making for starting RRT, considering prognosis, potential for recovery and/or harm, and patient/family preferences for care [15],
along with consideration of the evolving and dynamic nature of critical illness and
AKI. Any perceived benefit to starting RRT should be balanced with the clinical
context, the risk for harm, and the resource implications and within the context of the
patient’s and family’s preferences for care [15, 49].
Existing Knowledge Gaps and Future Research
There have been substantial advances in our understanding for when to consider
starting RRT in critically ill patients with severe AKI; however, there remains
existing knowledge gaps that can be focused on future research. The se include the
following:
1. Better understandin g of strategies for starting RRT in critically ill children,
recognizing fewer children routinely receive RRT and die in ICU compared
with adults [ 71, 72]
2. Further development and evaluation of clinical risk modeling, leveraging elec-
tronic health record alerting, machine learning, and new disease markers to
improve precision in the selection of patients most likely to benefit from
starting RRT
3. Interrogation of the interactions between fluid accumulation, RRT strategies, and
organ support and outcomes [73]
4.
Evaluation
manage AKI complications, RRT strategy, and outcomes [44, 74–76]
of the interactions between prolonged untreated AKI, therapies to