O R I G I N A L A RT I C L E S

Study protocol for the Balanced Solution versus Saline in

Intensive Care Study (BaSICS): a factorial randomised trial
Fernando G Zampieri, Luciano C P Azevedo, Thiago D Corrêa, Maicon Falavigna,
Flavia R Machado, Murillo S C de Assunção, Suzana M A Lobo, Letícia K Dourado,
Otavio Berwanger, John A Kellum, Nilton Brandão and Alexandre B Cavalcanti,
for the BaSICS Investigators and the BRICNet

The administration of intravenous (IV) fluids to restore intravascular ABSTRACT

volume is one of the most common interventions in the intensive
care unit. Saline (0.9% sodium chloride) is the most widely Background: The effectiveness and safety of balanced
used crystalloid fluid worldwide.1 Nevertheless, experimental crystalloid fluids compared with saline (0.9% sodium
and clinical studies show that saline has deleterious effects chloride) as a fluid of choice in critically ill patients
on the kidneys, acid–base balance, electrolyte homoeostasis, remain unclear. The effects of different fluid infusion
tissue perfusion, the inflammatory response and coagulation rates on outcomes are also unknown.
parameters.2-5 Objectives: To test the hypothesis that a balanced
Plasma-Lyte 148 is a balanced crystalloid fluid with osmolality crystalloid solution, compared with saline, decreases
closer to human plasma and lower sodium and chloride 90-day all-cause mortality among critically ill patients;
concentrations compared with saline. Observational and small and to test the hypothesis that slow, compared with
randomised studies suggest that balanced crystalloid fluids rapid, infusion rate decreases 90-day mortality in this
might be preferable to saline in critically ill patients.4-6 One of population of patients.
the putative mechanisms for the benefit of balanced solutions is Methods: The Balanced Solution versus Saline
their lower potential to cause hyperchloraemia, which, in turn, in Intensive Care Study (BaSICS) is a pragmatic, 2
could be associated with increased inflammation and acute  2 factorial, randomised controlled trial. A total
kidney injury (AKI).7,8 Recently, the safety and efficacy of volume of 11 000 patients will be recruited from at least
expansion with Plasma-Lyte 148 compared with saline were 100 Brazilian intensive care units. Patients will be
assessed in an exploratory, double-blind, cluster-randomised, randomised to receive Plasma-Lyte 148 or saline,
double-crossover trial (the 0.9% Saline v Plasma-Lyte 148 for ICU and to rapid infusion (999 mL/h) or slow infusion
Fluid Therapy [SPLIT] study).9 In this study, a median infusion of (333 mL/h). Study fluids will be used for resuscitation
2 L of Plasma-Lyte 148 or saline did not affect the risk of AKI episodes (at rapid or slow infusion rates), dilution of
(risk ratio [RR], 1.04; 95% CI, 0.80–1.36; P = 0.77), the need for compatible medications and maintenance solutions.
renal replacement therapy (RRT) (RR, 0.96; 95% CI, 0.62–1.50; Patients, health care providers and investigators will
P = 0.91) or in-hospital mortality (RR, 0.88; 95% CI, 0.67–1.17; be blinded to the solutions being tested. The rate of
P = 0.40). However, the SPLIT trial used a very low dose of study bolus infusion will not be blinded.
fluid (2 L over the ICU stay), did not select fluid used specifically Outcomes: The primary outcome is 90-day all-cause
for resuscitation and studied a relatively low-risk population mortality. Secondary outcomes are: incidence of renal
(mean Acute Physiology and Chronic Health Evaluation [APACHE] failure requiring renal replacement therapy within
II score, 14; overall AKI rate, 9%). 90 days, incidence of acute kidney injury (Kidney
In addition to the type of fluid, there are wide regional Disease: Improving Global Outcomes stages 2 and 3),
variations in infusion rates and volumes prescribed as fluid incidence of non-renal organ dysfunction assessed
boluses. The multicentre Fluid Challenges in Intensive Care study by Sepsis-related Organ Failure Assessment score at
showed that crystalloid fluids represent the most common type of Days 3 and 7, and number of mechanical ventilation-
fluids used in a fluid challenge, typically administered as a bolus free days within the first 28 days after randomisation.
of 500 mL in less than 30 minutes.1 Nevertheless, it has been Results and conclusions: The BaSICS trial will
suggested that fluid bolus infusion rate, a neglected aspect of provide robust evidence on whether a balanced
fluid management, may have important effects on physiological crystalloid, compared with saline, improves important
and clinical outcomes in ICU patients.10 patient outcomes in critically ill patients. BaSICS will
A prospective, randomised study in African children with also provide relevant information on whether bolus
severe infection showed that fluid boluses of saline or 5% infusion rate affects outcomes in this population.
albumin increased mortality compared with standard therapy Trial registration: ClinicalTrials.gov NCT02875873.
without boluses.10 Interestingly, most of the excess mortality
with rapid fluid resuscitation was attributed to cardiovascular Crit Care Resusc 2017; 19: 175-182

Critical Care and Resuscitation • Volume 19 Number 2 • June 2017 175

 O R I G I N A L A RT I C L E S

collapse.11 A rapid fluid bolus may abruptly reduce the

Table 1. Guidelines for volume expansion (fluid
adrenergic tone and/or worsen myocardial compliance, challenge) during the study period
leading to or aggravating haemodynamic instability.12
Despite increasing cardiac output, a fluid bolus (500 mL in These criteria suggest the need for volume expansion (item 1 plus
item 2)
30 minutes) may decrease arterial elastance, compromising
1. At least one sign of hypoperfusion:
arterial blood pressure. In contrast, resuscitation protocols
a. Heart rate > 120 beats/min
based on slower infusion rates and involving reduction of
fluid infusion seem to be safe in specific populations of b. SBP < 90 mmHg or MAP < 65 mmHg or decrease in SBP of >
40 mmHg from baseline
critically ill patients.10,13,14 Therefore, one crucial issue to
c. Capillary refill time > 1 s
be addressed is whether the fluid bolus infusion rate affects
d. Mottling score  2
the outcomes of critically ill adult patients.
We describe a protocol for a multicentre, randomised, e. Lactate level > 2 mmol/L (> 18 mg/dL)

factorial, clinical trial that will assess the effects of Plasma- f. ScvO2 < 70%
Lyte 148 versus saline as the fluid of choice in critically g. Urinary output < 0.5 mL/kg in past hour
ill patients, as well as the effects of rapid and slow fluid 2. At least one sign of responsiveness to fluid therapy or absence
challenge infusion rates on important patient outcomes. of signs of hypervolaemia:
a. Pulse pressure variation > 13%
b. Increase in pulse pressure > 5% after an expiratory pause of
Methods 15 s
c. Passive leg elevation leads to increased cardiac index (> 10%)
Aims or pulse pressure (> 11%) or MAP (> 5%)
We aim to determine whether a balanced crystalloid c. Respiratory variation in CVP > 1 mmHg
solution (Plasma-Lyte 148) used for fluid resuscitation d. Echocardiographic signs of hypovolaemia
reduces 90-day mortality compared with saline in critically e. CVP  10 mmHg
ill patients. We also aim to determine the effect of rapid f. Absence of clinical signs of hypervolaemia when data for above
administration (999 mL/h) versus slow administration symptoms not available
(333 mL/h) of crystalloid solution on 90-day mortality in SBP = systolic blood pressure. MAP = mean arterial pressure. ScvO2 =
critically ill patients at high risk of renal injury. central venous oxygen saturation. CVP = central venous pressure.

Study design and setting

saline is equally appropriate for patients, with no specific
The Balanced Solution versus Saline in Intensive Care indications or contraindications for any of the fluids or
Study (BaSICS) is a pragmatic, multicentre, 2  2 factorial, for rapid or slow infusion.
randomised controlled trial. Patients will be randomised to
• Not expected to be discharged on the day after admission.
receive Plasma-Lyte 148 or saline, and to rapid (999 mL/h)
or slow (333 mL/h) infusion rates. • At least one of the following risk factors for acute kidney
Patients, health care providers and study investigators will injury:
be blinded to the solutions being tested but not to the rate  age ≥ 65 years
of bolus infusion. Study fluids will be used for resuscitation  hypotension (mean arterial pressure [MAP]
episodes (at rapid or slow infusion rates). These same fluids < 65 mmHg or systolic blood pressure [SBP]
will be used for dilution of compatible medications and < 90 mmHg) or use of vasopressors
maintenance fluids, but only patients who are determined
 sepsis
to require fluid boluses (Table 1) will be enrolled. The
primary outcome is 90-day mortality. We will recruit about  use of invasive mechanical ventilation or
11 000 patients in at least 100 Brazilian ICUs (see Appendix continuous non-invasive mechanical ventilation
1, online at cicm.org.au/Resources/Publications/Journal). (including high-flow nasal cannula) for > 12 h
Patients will be enrolled after ICU admission.  oliguria (< 0.5 mL/kg/h for ≥ 3 h)

Eligibility  serum creatinine level ≥ 1.2 mg/dL (women) or

≥ 1.4 mg/dL (men)
Inclusion criteria  liver cirrhosis or acute liver failure.
To be randomised, patients must meet all the following
inclusion criteria: Exclusion criteria
• Need for fluid resuscitation or plasma expansion (Table 1), We will apply the following exclusion criteria:
and the clinician considers that either Plasma-Lyte 148 or • age < 18 years

176 Critical Care and Resuscitation • Volume 19 Number 2 • June 2017

 O R I G I N A L A RT I C L E S

• acute renal failure treated with RRT or expected to require for the fluid no longer exists. In imminently life-threatening
RRT within the next 6 hours situations (Table 3), patients may receive rapid infusion
• severe electrolyte disturbance (serum sodium level ≤ 120 boluses (999 mL/h) regardless of the infusion rate to which
mmol/L or ≥ 160 mmol/L, serum potassium level ≥ 5.5
they have been randomly assigned.
mmol/L)
• death considered imminent and inevitable within 24 We will suggest that investigators consider the late-
hours strategy criteria of the AKIKI trial to indicate starting RRT
• patients with suspected or confirmed brain death (Table 4),15 because there is no consensus on whether early
• patients receiving palliative care only or late RRT is the most appropriate for critically ill patients.
• patients previously enrolled in BaSICS. We will remind the centres that RRT should not be delayed
Interventions if the late-strategy criteria in AKIKI are met.

Patients will receive Plasma-Lyte 148 or saline

Figure 1. Study design
during their ICU stay (limited to 90 days), by
random assignment. When fluid expansion
is indicated, it will be provided according to
the randomisation group (rates of infusion
of 999 mL/h or 333 mL/h) during the ICU stay
(limited to 90 days) (Figure 1). Whenever possible,
the assigned study fluid and rate should be
adhered to during investigations and procedures
performed while the patient is temporarily
outside the ICU. The volumes and
frequency of administration will be Figure 2. Treatment algorithm for fluid administration for patients who
determined by the primary physician fulfilled inclusion criteria, including need for initial fluid expansion
(Figure 2), but guidelines suggesting
triggers for fluid infusion will be
provided to the study centres (Table
1). Adherence to study fluids and
infusion rate will be assessed on
Days 1 to 3 after randomisation and
on Day 7.
The study fluids will also be
used when crystalloid solutions are
indicated to maintain volume status.
Infusion rates for maintenance
solutions are at the discretion of
health care providers. Medications
and solutions with ingredients that
are compatible with saline and
Plasma-Lyte 148 (eg, sedative drugs,
CRF = case report form.
vasopressors and antibiotics) should
be infused using the study fluids
(Figure 2). A list of medications Table 2. Specific situations in which the assigned study fluid must not be
compatible with both study fluids administered
will be sent to the participating
Situation Recommended alternative fluid
centres (Appendix 2).
In some scenarios (eg, extreme Severe hyperchloraemia (chloride level 120 mmol/L) Plasma-Lyte 148 or Ringer’s lactate
electrolyte disturbances [Table 2]), Severe hypernatraemia (sodium level  160 mmol/L) Plasma-Lyte 148 or Ringer’s lactate
study fluids should not be used; in (if fluid resuscitation is needed)
these scenarios, the physician will Severe hyponatraemia (sodium level 120 mmol/L) Normal or hypertonic saline
be prompted to resume using study Hyperkalaemia (potassium level  5.5 mmol/L) Normal saline
fluids as soon as the contraindication

Critical Care and Resuscitation • Volume 19 Number 2 • June 2017 177

 O R I G I N A L A RT I C L E S

in which F = 0.742 (female patients) and B = 1.21 (black

Table 3. Specific situations in which study fluids may
be administered at rapid infusion rate (999 mL/h), patients). If a patient is classified as having a KDIGO score
regardless of randomised assignment of 2 at randomisation, they will be excluded from the
sample for the evaluation of this outcome.
Situation 1: severe arterial hypotension (SBP < 80 mmHg or MAP
< 50 mmHg) • New respiratory, hepatic, cardiac, neurological and
or coagulation dysfunction, assessed by Sepsis-related
Situation 2: diagnosis of haemorrhagic shock with active bleeding Organ Failure Assessment scores at Days 3 and 7.17
requiring aggressive volume replacement
• Mechanical ventilation-free days at 28 days after
SBP = systolic blood pressure. MAP = mean arterial pressure. randomisation.

Table 4. Indications for starting renal replacement Tertiary outcomes

therapy Tertiary outcomes are:
Renal failure (KDIGO stage 2 or 3) combined with one of • ICU and in-hospital mortality due to any cause
the following: • length of ICU stay
 Serum potassium level > 6 mEq/L
• length of hospital stay.
 pH < 7.15, in a context of pure metabolic acidosis or
mixed acidosis with PaCO2 > 50 mmHg and no possibility of
increasing minute volume Other exploratory outcomes
 Hypervolaemia with respiratory impairment requiring oxygen • Comparison of serum chloride levels between the four
delivery > 5 L/min (for patients on spontaneous ventilation) or study groups over time.
inspired oxygen fraction > 50% (for patients receiving invasive
mechanical ventilation or non-invasive ventilation) • Quality of life (utility) assessment after 6 months,
 Serum urea level > 240 mg/dL through the EuroQol five dimensions health state (EQ-5D)
questionnaire, which will be administered to a selected
KDIGO = Kidney Disease: Improving Global Outcomes.
subsample of about 1100 patients.

Outcomes Randomisation
Outcomes are similar for both factors in BaSICS (saline We will allocate patients in a 1:1:1:1 ratio to receive
versus Plasma-Lyte 148, and rapid versus slow infusion Plasma-Lyte 148 as slow infusion, Plasma-Lyte 148 as rapid
speed). infusion, saline as slow infusion or saline as rapid infusion.
The randomisation list will be generated with online
Primary outcome
software using random permuted blocks, stratified by
The primary outcome is 90-day all-cause mortality. centre according to fluid type (A to F) and infusion speed.
Therefore, each block will have 12 patients. Block size will
Secondary outcomes
not be disclosed to research personnel.
Secondary outcomes are: Research personnel will randomise patients via a web-
• Renal failure requiring RRT within 90 days based central, automated randomisation system, available
• AKI (classified as Kidney Disease: Improving Global 24 hours per day, maintained by the Research Institute
Outcomes [KDIGO] stage ≥ 2 at Days 3 and 7 after HCor, São Paulo, Brazil. The assigned study group will only
randomisation).16 For diagnosis of AKI, serum creatinine be disclosed after patients have been registered in the web-
level and urine output will be assessed using the following based randomisation system.
criteria: twofold or higher increase in serum creatinine
Blinding
level from reference level, or urine output level < 0.5 mL/
kg/h for ≥ 12 hours. The reference creatinine level will be Patients, health care providers, data collectors, outcome
the lowest of the randomisation creatinine and previous assessors and statisticians will be blinded to the study fluids
creatinine levels (the most recent value available in the (Plasma-Lyte 148 or saline). Plasma-Lyte 148 and saline
previous 6 months and before current admission). If no will be macroscopically identical and will be available in
previous creatinine value is available, it will be estimated identical plastic containers of 500 mL, identified with codes
using the Modification of Diet in Renal Disease equation: A, B, C, D, E and F. Blinding to fluid bolus infusion rate is not
feasible. Thus, this intervention will remain open to those
Creatinine level = 75/(186  [age – 0.203]  F  B) – 0.887 involved in patient care.

178 Critical Care and Resuscitation • Volume 19 Number 2 • June 2017

 O R I G I N A L A RT I C L E S

Timeline, data collection and management Statistical analysis

Trained research personnel at the sites will collect data in We will prepare a detailed statistical analysis plan before
a web-based case report form. Data will be collected and patient enrolment begins, which is intended to be published
analysed independent of adherence to the study protocol or made available electronically. All analysis will be conducted
to allow intention-to-treat analyses. Data will be collected on the intention-to-treat principle. The effect of type of fluid
on: the day of randomisation (Day 0), Days 1 to 3, Day 7, (Plasma-Lyte 148 v saline) and the infusion rate (rapid v slow
ICU discharge and hospital discharge. Investigators will infusion) on the primary outcome will be assessed with Cox
telephone the patient or relatives to obtain 90-day follow- proportional hazards regression and Kaplan–Meier curves,
up data. They will also collect 180-day follow-up data and will be presented as hazard ratios with 95% confidence
on vital status and health-related quality of life from the intervals and P values. Binary secondary outcomes will be
subsample of about 1100 patients. compared using 2 tests and presented as risk ratios with
We will use the following procedures to ensure data 95% CI and P values. Results for continuous outcomes will
quality: be expressed as mean differences, 95% CIs and P values,
• Investigators will attend a training session before the calculated with the independent t test, or Wilcoxon rank sum
beginning of the study to standardise procedures, test in the case of non-normal distribution. We will analyse
including data collection. the available dataset for secondary and tertiary outcomes.
• Investigators will be able to contact the coordinating Subgroup analysis
centre to resolve problems that may arise. The effect of the type of fluid (Plasma-Lyte 148 v saline)
• Data cleaning will be applied continuously to identify and infusion speed on the primary outcome will be assessed
inconsistencies and missing data. The centres will be using Cox proportional hazards regression in the subgroups
notified of any inconsistencies and missing data and of patients with and without the following characteristics
prompted to resolve them. at baseline:
• The coordinating centre will review detailed reports on • sepsis
screening, inclusion, follow-up and data consistency and • AKI KDIGO Stage 1
completeness, on a weekly basis. The coordinating centre • admitted from the surgical block
will take immediate action to resolve any problems.
• traumatic brain injury
• Centres will be monitored throughout the study and
• APACHE II score ≥ 25
onsite visits will take place, if there is a need, from the
recruiting centre. A trained professional will be appointed • patients receiving > 1.0 L of saline in the 24 hours before
by the coordinating centre to monitor the participating randomisation.
centres. During the monitoring visits, all information will We will infer a subgroup effect if the P value for
be considered strictly confidential. We do not intend to homogeneity of treatment effects is < 0.05.
conduct audits at the participating centres except for As exploratory analysis, we will assess the subgroup of
usual monitoring. patients who received at least 6 L of fluid in the first 3 days
after ICU admission. We will also perform a dose-response
Statistical methods analysis to assess whether there is a minimum volume to be
infused to obtain a positive association between balanced
Sample size
solution used and outcome. This analysis is sustained by
We will enrol 11 000 patients in the trial. We estimate that previous research suggesting that the potential benefit from
there will be a mortality rate of 35% within 90 days in balanced solutions may only occur in aggressively resuscitated
the control groups (saline or rapid infusion), based on the patients,5 but will be considered strictly exploratory.
eligibility criteria applied to the databases of two Brazilian
multicentre studies.18,19 The study will have a statistical Interim analyses and data monitoring committee
power of 89% to detect a hazard ratio of 0.90 or less for We have set up a data monitoring committee (DMC)
90-day mortality, with an alpha level of 0.05. with the primary aim of helping to ensure the safety of
We do not expect a significant interaction between the patients in the trial by protecting them from avoidable
two interventions (Plasma-Lyte 148 versus saline and slow harm. A group of independent experts has been appointed
versus rapid infusion rate). Nevertheless, if such interaction to the DMC: Gordon Guyatt (Chair), methodologist and
exists, the study will have a power of 80% to detect a trialist; Niall Ferguson, triallist and intensivist; and Stephen
positive hazard ratio of 0.835, considering that Plasma-Lyte Walter, statistician. All have extensive experience in trial
148 and slow infusion speed will result in lower mortality. methodology.

Critical Care and Resuscitation • Volume 19 Number 2 • June 2017 179

 O R I G I N A L A RT I C L E S

The DMC will initially prepare a charter specifying details of • Sites will receive study fluids and reimbursement for
the committee, its operation, meeting schedule and the trial patient inclusion.
stopping rules. The rules will be based on the principles listed • Investigators will be acknowledged in all publications
below. We will conduct interim analyses after recruitment of and will be offered the opportunity to lead substudies.
about 25%, 50% and 75% of the study cohort. Based on • We will prioritise promotion of patients’ enrolment using
the interim analyses and occasionally on external evidence, the several strategies (phone contacts, emails, newsletters,
DMC will decide whether there is evidence beyond reasonable weekly enrolment reports and others) in the daily
doubt to support that the experimental treatment (Plasma- activities of the Research Institute HCor.
Lyte 148 v saline; and slow v rapid infusion rate) increases 90- • We will act according to the 12 components for marketing
day mortality, with P < 0.01. The members of the DMC may clinical trials proposed by Francis and colleagues.22
consider requesting to continue the study for an additional 3
months to confirm such effects (especially if P is between 0.01 Trial organisation and funding
and 0.001). If the safety criterion is met in one of the factors The Research Institute HCor is the sponsor and coordinator
analysed in the study, the corresponding factorial arm will be of the study, in association with the Brazilian Research
stopped and the study will continue with the remaining one. in Intensive Care Network (BRICNet). The trial structure
In the case of evidence of superiority of Plasma-Lyte includes the following groups: the coordinating centres,
148 over saline, or of slow infusion over rapid infusion, investigators, steering committee and DMC.
with P < 0.001 in the interim analysis conducted after the The study will be conducted as part of and funded by
recruitment of 50% or more patients, the DMC may also the Program to Support Institutional Development of the
consider stopping the study. However, the DMC will request Universal Health System from the Brazilian Ministry of
to continue the study for an additional 3 months and for the Health. Baxter Latin America will provide the fluids and
analyses to be repeated to confirm the difference between transport logistics used in the trial. Baxter will not be
the treatments and to allow stabilisation of estimated effect. involved in any aspects of trial design, execution, analysis or
In accordance with the DMC, we decided that the interpretation of the results.
study will not be discontinued because of a benefit in the
first interim analysis (after the recruitment of 25% of the Safety
sample). The reasons for this decision were: All unexpected serious adverse events related to the study
• Early discontinuation of randomised trials because of interventions must be reported to the Research Institute
a benefit tends to produce biased estimates of effect HCor within 24 hours. An unexpected serious adverse event
(overestimation of the true effect), which may lead to directly related to the study is defined as any event meeting
erroneous medical guidelines and decisions, especially these two criteria:
with a small number of events.20 • any fatal or life-threatening event (immediate risk of
• According to the ethical principle of non-maleficence, a death), or any event that causes sequelae or permanent
new treatment should not be used until there is clear, disability, or that extends hospitalisation
objective evidence that it is beneficial. • the attending physician believes the event is related to
• Clinical practice does not usually change unless there the patient’s inclusion in the BaSICS trial.
is convincing evidence of the advantages of the new
Serious adverse events will be considered as “related to
treatment, which will be undermined if the study is
the study” if the attending physician believes that the event
discontinued early due to a benefit.21
was probably caused by the fluid and/or the rate of infusion
Recruitment strategy used in the study and follows a plausible time sequence
Recruiting 11 000 critically ill patients is clearly challenging. after the administration of the fluid.
We anticipate that we should recruit this sample size within Ethics and dissemination
48 months in 100 ICUs. We will use several strategies to
ensure adequate recruitment: Ethics approval and registration
• Submission to all ethics committees will be supported This study protocol, version 1.5, from August 2016, has
by staff at the Research Institute HCor, with the aim of been approved by the Research Ethics Committee of the
shortening time to study approval. Coordinating Centre (Hospital do Coração/Associação do
• Adequate training will be provided to sites in investigator Sanatório Sírio) (CAAE: 57395816.6.1001.0060). The study
meetings and during site initiation visits. will not begin at the participating centres until approval
• Study procedures will be kept as simple as possible and has been obtained from the responsible local internal
we will foster a culture of embedding study activities in review board for each participating ICU. All amendments
the routines of participating ICUs. to the protocol must be approved by the internal review

180 Critical Care and Resuscitation • Volume 19 Number 2 • June 2017

 O R I G I N A L A RT I C L E S

board of each participating centre. This protocol conforms names of other trial committees and all collaborators will be
to the Standard Protocol Items: Recommendations for listed after the text.
Interventional Trials recommendations,23 and has been
registered in ClinicalTrials.gov (NCT02875873). Conclusions
The BaSICS trial will provide robust evidence as to whether
Informed consent the use of a balanced crystalloid fluid reduces 90-day
Whenever possible, prospective written informed consent mortality compared with saline in critically ill patients
will be requested before randomisation from all eligible receiving fluid resuscitation. BaSICS will also provide relevant
patients, or from their legal representatives when the patients information on whether the fluid bolus infusion rate affects
are unable to provide consent due to communication or outcomes in this population of patients.
cognitive limitations. If no legal representative is available
at the time, the patient will be randomised, and written Competing interests
consent from the patient or legal representative will be We declare that Baxter Latin America will donate fluids and
sought as soon as possible afterwards (deferred consent). provide distribution logistics for the BaSICS trial. Baxter will
not be involved in any aspect of the study design, analysis or
Confidentiality
decision to publish the results. Luciano Azevedo and Murillo de
Each patient and research centre will be identified by Assunção received personal grants for the advisory board from
a unique number in the electronic case report form. Baxter Brazil. Otavio Berwanger has received research grants for
Information obtained from medical records must be investigator-initiated clinical trials from AstraZeneca, Amgen,
handled as confidential data by the research centres; it must Boehringer-Ingelheim, Roche and Bayer, which are unrelated to
be kept in restricted access locations, and anonymity must the topic of this study.
be ensured on interim and final reports.
Author details
Data sharing
Fernando G Zampieri, Clinical Triallist 1
We intend to make the study database available to other Luciano C P Azevedo, Intensivist 2,3
researchers. In the first 2 years after the publication of the Thiago D Corrêa, Intensivist 4
main manuscript, the investigators will perform analyses Maicon Falavigna, Clinical Triallist 5
for substudies proposed by investigators from the BaSICS
Flavia R Machado, Professor, Intensive Care 6
collaborative group. In this phase, the database will be kept
Murillo S C de Assunção, Intensivist 4
in conditions of restricted access by the study coordinators,
Suzana M A Lobo, Professor, Intensive Care, and Head,
and third parties will have access to it only with previous Intensive Care Division 7
authorisation of the BaSICS steering committee, which will
Letícia K Dourado, Clinical Triallist 1
analyse each research proposal and statistical analysis plan.
Otavio Berwanger, Clinical Triallist 1
The steering committee will provide the requested data
John A Kellum, Professor 8
if the proposal does not conflict with future or ongoing
Nilton Brandão, Clinical Triallist 9
substudies of investigators from the collaborative group.
Alexandre B Cavalcanti, Clinical Triallist 1
After 2 years, all data collected during the BaSICS trial
for the BaSICS Investigators and the BRICNet
will be publicly available on a free-access platform. We
emphasise that specific data that may identify a patient, 1 Research Institute HCor, Hospital of Coração, São Paulo, Brazil.
such as initials or date of birth, will not be made publicly 2 Intensive Care Unit, Hospital Sírio-Libanês, São Paulo, Brazil.
available. Each individual requesting access to the database 3 Emergency Medicine Discipline, University of São Paulo, São
must formally commit to notifying the steering committee Paulo, Brazil.
of the study in the event that any database information 4 Intensive Care Unit, Hospital Israelita Albert Einstein, São
identifying a patient is found. Paulo, Brazil.

An individual patient data meta-analysis grouping the 5 Hospital Moinhos de Vento, Porto Alegre, Brazil.
data from this study and the Plasma-Lyte 148 versus Saline 6 Anesthesiology Intensive Care Unit, Hospital São Paulo,
Study (NCT02721654) is planned. Federal University of São Paulo, São Paulo, Brazil.
7 Faculty of Medicine, São José do Rio Preto, Brazil.
Dissemination 8 Center for Critical Care Nephrology, University of Pittsburgh,
The BaSICS steering committee will publish the study Pittsburgh, Penn, USA.
findings, whatever they are. The main manuscript will 9 School of Medicine, Federal University of Health Sciences,
be submitted by the writing committee on behalf of the Porto Alegre, Brazil.
research group (BaSICS investigators and the BRICNet). The Correspondence: [email protected]

Critical Care and Resuscitation • Volume 19 Number 2 • June 2017 181

 O R I G I N A L A RT I C L E S

References PLoS One 2015; 10: e0143062.

1 Cecconi M, Hofer C, Teboul J-L, et al. Fluid challenges in 14 Silversides JA, Major E, Ferguson AJ, et al. Conservative fluid
intensive care: the FENICE study: a global inception cohort management or deresuscitation for patients with sepsis or
study. Intensive Care Med 2015; 41: 1529-37. acute respiratory distress syndrome following the resuscitation
2 Zhou F, Peng Z-Y, Bishop JV, et al. Effects of fluid resuscitation phase of critical illness: a systematic review and meta-analysis.
with 0.9% saline versus a balanced electrolyte solution on Intensive Care Med 2016; 43: 155-170.
acute kidney injury in a rat model of sepsis. Crit Care Med
15 Gaudry S, Hajage D, Schortgen F, et al. Initiation strategies for
2014; 42: e270-8.
renal-replacement therapy in the intensive care unit. N Engl J
3 Rochwerg B, Alhazzani W, Sindi A, et al. Fluid resuscitation in
Med 2016; 375: 122-33.
sepsis: a systematic review and network meta-analysis. Ann
Intern Med 2014; 161: 347-55. 16 Kellum JA, Lameire N. Diagnosis, evaluation, and management
4 Shaw AD, Schermer CR, Lobo DN, et al. Impact of intravenous of acute kidney injury: a KDIGO summary (Part 1). Crit Care
fluid composition on outcomes in patients with systemic 2013; 17: 204.
inflammatory response syndrome. Crit Care 2015; 19: 334. 17 Vincent JL, Moreno R, Takala J, et al; Working Group on Sepsis-
5 Zampieri FG, Ranzani OT, Azevedo LCPC, et al. Lactated Ringer Related Problems of the European Society of Intensive Care
is associated with reduced mortality and less acute kidney
Medicine. The SOFA (Sepsis-related Organ Failure Assessment)
injury in critically ill patients: a retrospective cohort analysis.
score to describe organ dysfunction/failure.. Intensive Care
Crit Care Med 2016; 44: 2163-70.
Med 1996; 22: 707-10.
6 Raghunathan K, Shaw A, Nathanson B, et al. Association
between the choice of IV crystalloid and in-hospital mortality 18 Soares M, Bozza FA, Angus DC, et al. Organizational
among critically ill adults with sepsis. Crit Care Med 2014; 42: characteristics, outcomes, and resource use in 78 Brazilian
1585-91. intensive care units: the ORCHESTRA study. Intensive Care
7 Neyra JA, Canepa-Escaro F, Li X, et al. Association of Med 2015; 41: 2149-60.
hyperchloremia with hospital mortality in critically ill septic 19 Cavalcanti AB, Bozza FA, Machado FR, et al; Writing Group for
patients. Crit Care Med 2015; 43: 1938-44.
the CHECKLIST-ICU Investigators and the Brazilian Research
8 Nadeem A, Salahuddin N, El Hazmi A, et al. Chloride-liberal
in Intensive Care Network (BRICNet). Effect of a quality
fluids are associated with acute kidney injury after liver
improvement intervention with daily round checklists, goal
transplantation. Crit Care 2014; 18: 625.
setting, and clinician prompting on mortality of critically ill
9 Young P, Bailey M, Beasley R, et al. Effect of a buffered
patients: a randomized clinical trial. JAMA 2016; 315: 1480-90.
crystalloid solution vs saline on acute kidney injury among
patients in the intensive care unit. JAMA 2015; 314: 1701-10. 20 Montori VM, Devereaux PJ, Adhikari NKJ, et al. Randomized
10 Maitland K, Kiguli S, Opoka RO, et al. Mortality after fluid trials stopped early for benefit: a systematic review. JAMA
bolus in African children with severe infection. N Engl J Med 2005; 294: 2203-9.
2011; 364: 2483-95. 21 Mueller PS, Montori VM, Bassler D, et al. Ethical issues in
11 Maitland K, George EC, Evans JA, et al. Exploring mechanisms stopping randomized trials early because of apparent benefit.
of excess mortality with early fluid resuscitation: insights from
Ann Intern Med 2007; 146: 878-81.
the FEAST trial. BMC Med 2013; 11: 68.
22 Francis D, Roberts I, Elbourne DR, et al. Marketing and clinical
12 Monge García MI, Guijo González P, Gracia Romero M, et al.
trials: a case study. Trials 2007; 8: 37.
Effects of fluid administration on arterial load in septic shock
patients. Intensive Care Med 2015; 41: 1247-55. 23 Chan AW, Tetzlaff JM, Altman DG, et al. SPIRIT 2013 statement:
13 Aung NM, Kaung M, Kyi TT, et al. The safety of a conservative defining standard protocol items for clinical trials. Rev Panam
fluid replacement strategy in adults hospitalised with malaria. Salud Publica 2015; 38: 506-1 4. ❏

182 Critical Care and Resuscitation • Volume 19 Number 2 • June 2017

Appendix
This appendix was part of the submitted manuscript and has been peer reviewed. It is posted as
supplied by the authors.

Study protocol for the Balanced Solution vs. Saline in Intensive Care Study (BaSICS): a factorial

randomized trial assessing balanced crystalloid versus 0,9% sodium chloride and rapid versus slow

infusion rate in critically ill patients

ELECTRONIC APPENDIX

Appendix 1 – List of Participating Centers (in alphabetical order of site investigator)

Hospital Site Investigator

Hospital Municipal Vila Santa Catarina Airton Leonardo Manoel de Oliveira
Instituto Nacional De Cardiologia Alexandre Rouge Felipe
Hospital Universitário Regional de Maringá - HUM Almir Germano
Hospital do Trabalhador Álvaro Rea Neto
Hospital Marcelino Champagnat Álvaro Rea Neto
Hospital Universitário Cajuru Álvaro Rea Neto
Instituto de Neurologia de Curitiba Álvaro Rea Neto
Hospital Vita Batel Álvaro Rea Neto
Hospital Municipal Moysés Deutsch Ana Helena Vicente Andrade
Hospital do Servidor Público Municipal de Sâo Paulo Ana Helenir Benaglia
Hospital Primavera André Luis Veiga de Oliveira
Hospital da Cidade André Luiz Nunes Gobatto
Hospital Guilherme Alvaro André Scazufka Ribeiro
Hospital HOME Antonio Aurélio de Paiva Fagundes
Jr.
Hospital Estadual Roberto Chabo Astor Bruno Ferreira de Mello
Hospital da Luz Bruno Adler Maccagnan Pinheiro
Besen
Hospital Universitário - Universidade Federal de Juiz de Fora Bruno do Valle Pinheiro
Hospital Samaritano Bruno Franco Mazza
Hospital Estadual Getúlio Vargas Bruno Gonçalves
 Hospital Anchieta Bruno Jardim Grossi
Instituto Estadual do Cérebro Paulo Niemeyer Cássia Righy Shinotsuka
Hospital Moinhos de Vento Cassiano Teixeira
Hospital Paulistano Cesar Biselli Ferreira
Hospital Universitário Regional do Norte do Paraná - UEL Cintia Magalhães Carvalho Grion
Hospital Nossa Senhora das Neves Ciro Leite Mendes
Hospital Samaritano João Pessoa Ciro Leite Mendes
Hospital Unimed João Pessoa Ciro Leite Mendes
Vitória Apart Hospital Claudio Piras
Hospital Regional de Mato Grosso do Sul Rosa Pedrossian Claudnei Menezes de Rezende
Hospital Municipal Irmã Dulce Daniela Boni
Instituto de Infectologia Emilio Ribas II - Fundação do ABC Daniela Boni
Prevent Senior Instituto Prevent Senior - IPS Daniella Cabral de Freitas
Hospital das Clínicas da Universidade Federal de Goiás Denise Milioli Ferreira
Fundação Amaral Carvalho Éderson Roberto de Mattos
Hospital Adventista de Belém Edgar Brito Sobrinho
Hospital do Coração de São Paulo Edson Romano
Hospital São Francisco da Santa Casa de Porto Alegre Eraldo de Azevedo Lúcio
Sociedade Literária e Caritativa Santo Agostinho – Hospital São José Felipe Dal Pizzol
Hospital Alemão Oswaldo Cruz Fernando Godinho Zampieri
Hospital São Paulo, Universidade de São Paulo Flávia Ribeiro Machado
Hospital SEPACO Flávio Geraldo Rezende de Freitas
Hospital Geral De Vitória Da Conquista Geovani Moreno Santos Junior
Santa Casa de Misericórdia de Vitória da Conquista Geovani Moreno Santos Júnior
Hospital Universitário São Francisco na Providência de Deus Giovana Colozza Mecatti
Centro Hospitalar Unimed Glauco Adrieno Westphal
Hospital de Ilhéus Heitor Portela Póvoas Filho
Hospital São Paulo UNIFESP Hélio Penna Guimarães
UTI - Moléstias Infecciosas - HCFMUSP Ho Yeh Li
Hospital Nereu Ramos Israel Silva Maia
Hospital Santa Paula Joao Geraldo Simoes Houly
Hospital do Servidor Público Estadual João Manoel Silva Junior
Santa Casa de Misericórdia de São João Del Rei Jorge Luiz da Rocha Paranhos
Hospital Sírio Libanês José Mauro Vieira Junior
 Hospital das Clinicas da Faculdade de Medicina de Botucatu Laercio Martins De Stefano
Hospital Distrital Evandro Ayres De Moura Lanese Medeiros Figueiredo
Hospital Quinta D´Or Laura Brasil Herranz
UTI Pronto Socorro - HCFMUSP Leandro Utino Taniguchi
Hospital Geral de Palmas Leonardo Guimarães Castro Boa
Sorte
Hospital de Clínicas de Uberlândia Liliane Barbosa da Silva Passos
Hospital Geral Cleriston Andrade Lúcio Couto de Oliveira Junior
Unidade de Apoio Cirúrgico - HCFMUSP Luiz Marcelo Sá Malbouisson
Hospital Monte Klinikum Marcelo Jorge Jacó Rocha
Hospital Universitário Evangélico de Curitiba Marcelo Oliveira Santos
Hospital Municipal Evandro Freire Márcio Duarte Viçoso Barcellos
Hospital São João de Deus - Fundação Geraldo Correa Marcone Lisboa Simões da Rocha
Casa de Saúde Pinheiro Machado Marcus Alexandre da Silva Bezerra
Hospital Municipal Dr Mário Gatti Marcus Vinicius Pereira
Hospital Evangélico Cachoeiro do Itapemirim Marlus Muri Thompson
Hosp Providência de Apucarana Mateus Dias de Moura
Hospital e Maternidade São Vicente Meton Soares de Alencar Filho
Hospital das Clínicas - UFPE Michele Maria Gonçalves de Godoy
Hospital SAMUR Miquéias Martins Lima Silva
Hospital Universitáio Cassiano Antônio Moraes - UFES Paula Frizera Vassalo
Complexo Hospitalar de Mangabeira Governador Tarcísio Burity Paulo Cesar Gottardo
Santa Casa de Maringá Paulo Roberto Aranha Torres
AC Camargo Cancer Center Pedro Caruso
Hospital do Câncer - UOPECCAN Péricles Duarte
Hospital Universitário de Cascavel - HU do Oeste do Parana Péricles Duarte
Hospital Estadual Jayme dos Santos Neves Priscilla de Aquino Martins
Hospital Santa Cruz Rafael Botelho Foernges
Hospital de Clinicas – Universidade Estadual de Campinas Renan Alves da Cruz
Hospital Nossa Senhora dos Prazeres Ricardo Rath de Oliveira Gargioni
Conjunto Hospitalar do Mandaqui Roberto Camargo Narciso
Hospital Municipal Souza Aguiar Roberto Seabra Lannes
Hospital Maternidade São José - UNESC - Fundação Social Rural de Rodrigo Cruvinel Figueiredo
Colatina
Hospital São Lucas - PUCRS Sergio Baldisserotto
 Hospital Geral do Grajaú Sergio Elia Mataloun
Hospital Norte D’Or Sergio Teixeira Sant’Anna Junior
Fundação Faculdade Regional de Medicina de São José do Rio Preto Suzana Margareth Ajeje Lobo
Irmandade Santa Casa de Porto Alegre Thiago Costa Lisboa
Hospital Albert Einstein Thiago Domingos Correa
Hospital das Clínicas Universidade Federal de Minas Gerais Vandack Alencar Nobre Junior
Real e Benemérita Associação Portuguesa de Beneficência Viviane Cordeiro Veiga
Hospital Novo Atibaia Walter Carlos Girardelli Baptista
Hospital e Pronto Socorro 28 de Agosto Wilson de Oliveira Filho
Hospital das Clínicas Ribeirão Preto - FMUSPRP Wilson José Lovato
 Appendix 2 – Medications compatible with both saline and Plasma-Lyte148

Drug Manufacturer Concentration

Aciclovir Hospira 25 mg/ml

Adrenaline Aspen 12 mg/100 ml

Amikacin DBL 40 mg/ml

Atracurium DBL 0.5 mg/ml

Atropine Pfizer 0.4 mg/ml

Benzylpenicillin CSL 2400 mg/50 ml

Calcium Chloride Baxter 40 mg/ml

Calcium Gluconate Baxter 40 mg/ml

Caspofungin MSD 70 mg/100 ml

Cefoxitin Hospira 20 mg/ml

Ceftazidime Hospira 100 mg/ml

Cephazolin Hospira 2000 mg/50 ml

Ciprofloxacin Bayer 2 mg/ml

Clindamycin Pfizer 900 mg/50 ml

Clonidine Boehringer Ingelheim 20 mcg/ml

Cloxacillin Teva 100 mg/ml

Cyclophosphamide Baxter 8 mg/ml

Dexamethasone Aspen 4 mg/ml

Digoxin Aspen 0.25 mg/ml

Dobutamine Hospira 5 mg/ml

Dopamine Hospira 3.2 mg/ml

Ephedrine Hospira 5 mg/ml

Ergometrine Hospira 200 mcg/5 ml

Esmolol Phebra 10 mg/ml

Esomeprazole Astra Zeneca 0.4 mg/ml

Fentanyl Hospira 10 mcg/ml

Flucloxacillin Hospira 40 mg/ml

Fluconazole Pfizer 200 mg/100 ml

Foscarnet Clinect 24 mg/ml

Frusemide Sandoz 10 mg/ml

Gentamicin Pfizer 10 mg/ml

Glyceryl Trinitrate Hospira 30 mg/50 ml

Glycopyrrolate Aspen 0.2 mg/ml

Granisetron Hospira 0.05 mg/ml

Heparin Pfizer 1000 units/ml

Hydralazine Link 2 mg/ml

Hydrocortisone Pfizer 100 mg/2 ml

Hydromorphone Mundipharma 2 mg/ml

Imipenem/Cilastatin MSD 5 mg/ml

Isoprenaline Hospira 1 mg/100 ml

Ketamine Hospira 2 mg/ml

Labetalol Sandoz Canada 5 mg/ml

Lignocaine Hydrochloride Pfizer 8 mg/ml

Lincomycin Pfizer 2 mg/ml

Magnesium Sulfate Baxter 0.4 mmol/ml

Meropenem Hospira 40 mg/ml

Metarminol Montrose 0.2 mg/ml

Metoclopramide iNova 5 mg/ml

Metoprolol AstraZeneca 1 mg/ml

Metronidazole Hospira 5 mg/ml

Midazolam Pfizer 1 mg/ml

Milrinone Sanofi 300 mcg/ml

Morphine Sulfate Hospira 1 mg/ml

Moxifloxacin Bayer 1.6 mg/ml

Naloxone Hospira 0.4 mg/ml

Neostigmine AstraZeneca 0.5 mg/ml

Noradrenaline Hospira 16 mg/100 ml

NovoRapid insulin Novo Nordisk 1 unit/ml

Ondansetron GSK 1 mg/ml

Oxytocin Aspen 1 unit/ml

Pancuronium AstraZeneca 2 mg/ml

Paracetamol Pfizer 10 mg/ml

Parecoxib Pfizer 40 mg/2 ml

Pethidine Hospira 10 mg/ml

Piperacillin/Tazobactam Pfizer 4500 mg/50 ml

Potassium Chloride Baxter 0.5 mmol/ml

Potassium Dihydrogen Phosphate Baxter 0.5 mmol/ml

Protamine Sanofi 10 mg/ml

Rocuronium Hospira 10 mg/ml

Salbutamol GSK 0.05 mg/ml

Sodium Nitroprusside Hospira 0.6 mg/ml

Sugammadex MSD 25 mg/ml

Suxamthonium AstraZeneca 2 mg/ml

Syntometrine Novartis 1 mL/4 mL

Thiopentone Link 50 mg/ml

Tramadol Sandoz 50 mg/ml

Tranexamic acid Pfizer 100 mg/ml

Trimethoprim/Sulfamethoxazole Hospira 1 mg/25 ml

Vancomycin Hospira 20 mg/ml

Verapamil Abbott 2.5 mg/ml

Voriconazole Pfizer 5 mg/ml