Background
Challenges to conventional approach
-
Cardiac output may be preserved, increased or depressed in patients with sepsis, although there is a paucity of human studies undertaken in the ‘non-resuscitated’ state [6]
-
Elevated serum lactate in sepsis is produced aerobically, probably in response to adrenergic stimulation, as an adaptive response to increase bio-energetic efficiency, rather than because of tissue hypoxia [12]
-
Organ failure in sepsis involves cellular dysfunction unrelated to hypoxia/hypoperfusion, including structural mitochondrial changes and reduced oxygen consumption [13]
-
A key role is played by the microcirculation, including the vascular endothelium, in the pathogenesis of sepsis. These microcirculatory alterations are not related to macrocirculatory indices such as blood pressure [14]
Is liberal intravenous fluid administration associated with harm?
Why might intravenously administered fluid be harmful?
-
Tissue oedema – leading to increased requirement for ventilatory support, increased translocation of gut organisms, and increased renal venous pressure compromising perfusion [5]
-
Opening of shut-down capillary beds (so called ‘hibernating circulation’) leading to ‘flooding’ of the systemic circulation with cytokine-rich blood, exacerbating systemic inflammation [5]
-
Degradation of the glycocalyx layer lining the luminal wall of the vascular endothelium [20]. The glycocalyx plays a critical role in vascular homeostasis by maintaining endothelial cells in a quiescent state. Loss of integrity of the glycocalyx is a critical step in endothelial cell activation and propagation of the systemic inflammatory state. Glycocalyceal damage occurs as a direct effect of fluid administration intravenously [21]
Alternative approach
Rationale for a randomised trial and the need for pilot data
Methods/design
Aim, design and setting
Scientific title | Fluid-restricted versus fluid-liberal resuscitation in sepsis; a randomised controlled pilot trial |
---|---|
Short title | REstricted Fluid REsucitation in Sepsis-associated Hypotension (REFRESH) trial |
Health condition | Sepsis |
Ethics | HREC/15/Austin/486 (Austin Health, Victoria) & HREC/15/114 (South Metropolitan Health Service, WA) |
Protocol version | Version 3, October 2016 |
Funding |
Grants: |
Emergency Medicine Foundation (EMSS-229R24-2015-KEIJZERS) | |
University of WA/University of Queensland Bilateral Research Collaboration Award | |
Royal Perth Hospital Medical Research Foundation | |
Industry: | |
Nil | |
Primary sponsor | Investigator-initiated and -driven study. Centre for Clinical Research in Emergency Medicine, University of Western Australia. Chief investigator and study contact: Dr. Stephen Macdonald, Royal Perth Hospital, PO Box X2213, Perth, WA, Australia |
Email: stephen.macdonald@uwa.edu.au | |
Background | Sepsis is a common condition with high morbidity and mortality. Patients with sepsis can develop hypotension (low blood pressure) due to a combination of factors. The traditional first-line treatment of sepsis-associated hypotension is to give a rapid, large volume of intravenously administered fluid (fluid bolus). Emerging clinical data and advances in the understanding of fluid physiology suggest that fluid bolus may be associated with worse patient outcomes. This randomised pilot clinical trial will compare a restricted fluid-volume approach, including early use of vasopressor drugs if required, against standard (liberal fluid-volume) care to assess the feasibility and safety of this approach. An embedded laboratory study will measure the differences in a range of relevant blood markers (such as activation of the vascular endothelium) to provide mechanistic plausibility for a restricted-volume approach. Together these data will be used to inform the design of a large multicentre trial to assess clinical outcomes |
Hypothesis | That a volume-restricted approach to sepsis-associated hypotension is clinically feasible; that this approach results in reduced systemic inflammation and associated biomarkers of endothelial cell activation |
Study aims | 1. To investigate the feasibility of delivering volume-restricted resuscitation in sepsis-associated hypotension |
2. Compare the total volume of fluid administered at 6 h and 24 h | |
3. To assess effects on the biomarkers of the inflammatory response, endothelial activation and related pathways of interest | |
4. To record any differences in requirement for organ support and in clinical outcomes | |
Study design | • Multi-centre (Armadale, Perth/Austin, Melbourne/Fiona Stanley, Perth/Gold Coast/Royal Hobart/Royal Perth/Sir Charles Gairdner, Perth/The Prince Charles, Brisbane) |
• Randomised controlled/un-blinded/feasibility | |
• Interventional | |
Setting | 8 emergency departments (EDs) in Australia |
Inclusion and exclusion criteria |
Inclusion criteria: |
1. Suspected infection and
| |
2. Systolic blood pressure (SBP) < 100 mmHg, despite 1000 ml intravenously administered isotonic crystalloid administered over not more than 60 minutes and
| |
3. Study intervention can be administered within 2 hs of inclusion criteria being met | |
Exclusion criteria: | |
1. Hypotension thought due to, or contributed to by, a non-sepsis cause (e.g. arrhythmia, haemorrhage) | |
2. Clinical requirement for fluid replacement (e.g. gastrointestinal losses) | |
3. Transfer from another hospital | |
4. More than 2000 ml intravenously administered fluid has been given (pre-hospital, in ED, or both) | |
5. Likely requirement for immediate surgery | |
6. Age < 18 years | |
7. Pregnancy (confirmed or suspected) | |
8. Patient in extremis or death deemed imminent and inevitable | |
9. Patient wishes or comorbidities such that either fluid loading or vasopressor support is not considered clinically appropriate | |
Intervention |
Intervention arm: |
Commence vasopressor infusion (± maintenance IV fluid) if SBP < 90 mmHg or mean arterial pressure (MAP) < 65 mmHg). Reassess hourly next 6 h and administer further 250-ml IV fluid bolus if required | |
Comparator arm: | |
Give a second 1000-ml fluid bolus plus further 500-ml boluses as clinically indicated until judged to be euvolaemic. Commence vasopressor infusion (± maintenance IV fluid) if SBP < 90 mmHg or MAP < 65 mmHg). Reassess hourly next 6 h and administer further 500-ml IV fluid bolus if required | |
Primary outcome measure | Total volume of intravenously administered fluid (including pre-randomisation) at 6 h |
Secondary outcome measures | • Mortality (all-cause) at 90 days post enrolment |
• ICU length of stay | |
• Hospital length of stay | |
• Hospital-free days to day 90 | |
• Organ failure | |
o Cardiovascular | |
▪ Requirement for vasopressors | |
▪ Duration of vasopressor requirement (h) | |
▪ Vasopressor-free days to day 28 | |
o Respiratory | |
▪ Requirement for ventilation (NIV/IPPV) | |
▪ Duration of ventilator support (h) | |
▪ Ventilator-free days to day 28 | |
o Renal | |
▪ Requirement for renal replacement therapy (RRT) | |
▪ Duration of RRT | |
▪ RRT-free days to day 28 | |
▪ AKIN score to day 7 | |
Laboratory/mechanistic outcome measures | Peak values, as well as patterns of expression over time (T0–T24), will be analysed for a range of biomarkers and compared between the groups including: |
• Atrial natriuretic peptide | |
• Troponin | |
• Inflammatory cytokines | |
o IL-6, IL-10, resistin | |
• NGAL (a biomarker of renal injury) | |
• Endothelial cell activation biomarkers | |
o sVCAM, sICAM, sE-Selectin, sFlt-1 | |
• Soluble markers of glycocalyx degradation | |
o Heparan sulphate, syndecan-1, hyaluronan | |
Feasibility outcome measures | • Proportion of eligible patients enrolled |
• Randomisation errors | |
• Compliance – protocol violations | |
• Proportion with completely recorded data | |
• Proportion with complete study blood sampling | |
Sample size | 50 per arm/100 total |
Randomisation | Eligible participants in whom a consent process has been commenced will be randomised to either the fluid-volume restricted or the fluid-liberal (standard-care arm) |
Randomisation procedure | • Eligibility assessment and randomisation conducted by clinical staff with support by research personnel (in person or by telephone) |
• Computer-generated randomisation sequence | |
• Real-time web-based randomisation | |
• 1:1 randomisation in blocks of 2 or 4 | |
Blinding | • Un-blinded to patients and to treating team |
• Blinding of laboratory staff and data analysts | |
Data collection methods | • Paper CRF completed for each participant |
• CRF populated by research staff in real time or from clinical record | |
• Subsequent data entry into secure database by trained research assistant | |
• Audit of data entry against source documentation | |
Data collected | • All fluids administered pre-randomisation |
• All fluids administered post randomisation – 6 h (recorded hourly) | |
• All fluids administered over 6–24 h | |
• Urine output to 24 h | |
• Vital signs at enrolment and hourly until 6 h | |
• Antimicrobials – agent, dose and time administered | |
• Corticosteroid administration | |
• Source of sepsis (suspected/proven) | |
• Charlson Comorbidity Score | |
• Microbiological/serological results from samples obtained in first 24 h | |
• Sequential Organ Failure Assessment (SOFA) score at admission and at 24 h | |
• Acute Physiology and Chronic Health Evaluation (APACHE) II score (peak first 24 h) | |
• Acute Kidney INjury (AKIN) score up to 7 days | |
• Study blood sampling at enrolment (T0), 3 h (T3), 6 h (T6) and 12–24 h (T24) | |
Statistical analysis | • Intention-to-treat analysis |
• Statistician blinded to treatment allocation | |
Data and safety monitoring | • DSMC – emergency physician, intensive care physician, statistician |
• SAE reporting to study coordination centre within 24 h | |
• DSMC review all SAE | |
• Periodic data review by DSMC and recommendations to PI in event of study conduct of SAE issues | |
• Trial coordinator will monitor trial conduct at each site at regular intervals | |
Ethics and governance | • HREC approvals obtained for all sites |
• Clinical governance (site-specific authority) procedures in place for all sites | |
• Any subsequent protocol amendments to be submitted to lead ethics site with appropriate dissemination to participating study sites | |
• The study will be conducted in accordance with GCP | |
Consent | • Where possible, prospective informed consent will be obtained |
• Due to the time-critical nature of the condition, approval in place for verbal consent to randomise and initiate care followed by formal written consent process | |
• Some participants will lack capacity so consent will be sought from next of kin | |
• Where next of kin not immediately available, provision in place for enrolment under initial waiver of consent (or procedural authorisation) followed by delayed consent to continue in the trial | |
Confidentiality | • All data will be de-identified |
• Data (paper and electronic) will be stored securely | |
Dissemination | • Trial results to be presented at relevant scientific meetings and published in peer-reviewed journal |
Study status | • Opened to recruitment July 2016 (Protocol V2) |
• Protocol amended October 2016 (Protocol V3) after 6 participants enrolled | |
• As at 22 June 2017, 42 participants enrolled at 6 of 7 sites active | |
• Anticipated to complete enrolment of 100 participants within 12 months |
Participant characteristics
Inclusion criteria
Exclusion criteria
Screening and randomisation
Study interventions
General management – all participants
Restricted-volume (±early vasopressor) arm (see Additional file 3 : Figure S2)
Standard treatment arm (see Additional file 4 : Figure S3)
Study measurements
-
All fluids administered pre-randomisation
-
All fluids administered post-randomisation – 6 h (recorded hourly)
-
All fluids administered over 6–24 h
-
Urine output to 24 h
-
Vital signs at enrolment and hourly until 6 h
-
Antimicrobials – agent, dose and time administered
-
Corticosteroid administration
-
Source of sepsis (suspected/proven)
-
Charlson Comorbidity Score [28]
-
Microbiological/serological results from samples obtained in first 24 h
-
Sequential Organ Failure Assessment (SOFA) score [29] at admission and at 24 h
-
Acute Physiology and Chronic Health Evaluation (APACHE) II score [30] (peak first 24 h)
-
Enrolment (T0), 3 h (T3), 6 h (T6) and 12–24 h (T24)
-
Plasma and serum samples (total 5 ml) at each sampling time point
-
Blood gas (arterial or venous) including lactate and haemoglobin at each sampling time point
-
Blood samples should not to be taken during fluid-bolus administration and should be taken from the opposite arm from any maintenance fluid being administered
Strategies to ensure protocol compliance and data integrity
Outcome measures
Primary outcome
-
A clinically and statistically significant difference in mean total fluid volume administered at 6 h post randomisation (including pre-randomisation). In addition pre- and post-randomisation volumes will be reported separately. See ‘Statistical methods and sample size calculation’ below.
Secondary outcomes: feasibility
-
Proportion of screened patients eligible
-
Proportion of eligible patients enrolled
-
Enrolment rate (i.e. number of enrolments per month per site)
-
Protocol compliance
Secondary outcomes: process-of-care
-
Admission rate to ICU versus general ward
-
Rates and timing of CVC insertion
-
Peripheral versus central administration of vasopressors
Secondary outcomes: clinical
-
Mortality (all-cause) at 90 days post enrolment
-
ICU length of stay
-
Hospital length of stay (acute)
-
Hospital-free days to day 90 (acute)
-
Organ failure
-
o
Cardiovascular-
▪ Administration of vasopressors
-
▪ Duration of vasopressor requirement (h)
-
▪ Vasopressor-free days to day 28
-
-
o
Respiratory-
▪ Requirement for ventilation (non-invasive ventilation (NIV)/intermittent positive-pressure ventilation (IPPV))
-
▪ Duration of ventilator support (h)
-
▪ Ventilator-free days to day 28
-
-
o
Renal-
▪ Requirement for renal replacement therapy (RRT)
-
▪ Duration of RRT
-
▪ RRT-free days to day 28
-
▪ Acute Kidney Injury Network (AKIN) score to day 7 [31]
-
Laboratory/mechanistic outcomes
-
Atrial natriuretic peptide
-
Troponin
-
Inflammatory cytokines
-
o
Interleukin (IL)-6, IL-10, resistin
-
-
Neutrophil gelatinase-associated lipocalin (NGAL; a biomarker of renal injury)
-
Endothelial cell activation biomarkers
-
o
sVCAM, sICAM, sE-Selectin, sFlt-1
-
-
Soluble markers of glycocalyx degradation
-
o
Heparan sulphate, syndecan-1, hyaluronan
-
Statistical methods and sample size calculation
Trial administration and Data Safety Monitoring Committee
Adverse events
-
Extravasation of peripherally administered noradrenaline infusion (whether or not any tissue injury occurred)
-
CVC-related complications
-
Bleeding
-
Inadvertent arterial puncture
-
Pneumothorax
-
Thrombosis
-
Infection
-
Arterial catheter-related complications, including thrombosis
-
Any other event that is considered to be of concern by the site investigator
-
Serious adverse events
-
Results in death
-
Is life-threatening
-
Requires inpatient hospitalisation or prolongation of existing hospitalisation
-
Results in persistent or significant disability/incapacity
-
Is an important medical event, which may require intervention to prevent one of the previously listed outcomes
Adverse event and serious adverse event reporting
Strategies to achieve adequate enrolment
Discussion
Significance
Limitations
Stakeholder engagement and dissemination of findings
Trial status
Acknowledgements
Funding
Availability of data and materials
Authors’ information
Ethics approval and consent to participate
-
South Metropolitan Health Service, Western Australia (Reference 15-114, 13 October 2015) for Western Australian Sites
-
Austin Health, Victoria (HREC/15/Austin/486, 20 April 2016) for all other sites