Study protocol for TARGET protein: The effect of augmented administration of enteral protein to critically ill adults on clinical outcomes: A cluster randomised, cross-sectional, double cross-over, clinical trial

Abstract

Background

It is unknown whether increasing dietary protein to 1.2–2.0 g/kg/day as recommended in international guidelines compared to current practice improves outcomes in intensive care unit (ICU) patients. The TARGET Protein trial will evaluate this.

Objective

To describe the study protocol for the TARGET Protein trial.

Design, setting, and participants

TARGET Protein is a cluster randomised, cross-sectional, double cross-over, pragmatic clinical trial undertaken in eight ICUs in Australia and New Zealand. Each ICU will be randomised to use one of two trial enteral formulae for three months before crossing over to the other formula, which is then repeated, with enrolment continuing at each ICU for 12 months. All patients aged ≥16 years in their index ICU admission commencing enteral nutrition will be eligible for inclusion. Eligible patients will receive the trial enteral formula to which their ICU is allocated. The two trial enteral formulae are isocaloric with a difference in protein dose: intervention 100g/1000 ml and comparator 63g/1000 ml. Staggered recruitment commenced in May 2022.

Main outcomes measures

The primary outcome is days free of the index hospital and alive at day 90. Secondary outcomes include days free of the index hospital at day 90 in survivors, alive at day 90, duration of invasive ventilation, ICU and hospital length of stay, incidence of tracheostomy insertion, renal replacement therapy, and discharge destination.

Conclusion

TARGET Protein aims to determine whether augmented enteral protein delivery reduces days free of the index hospital and alive at day 90.

Trial registration

Australian New Zealand Clinical Trials Registry (ACTRN12621001484831).

1. Introduction

In Australian and New Zealand Intensive Care Units (ICUs), patients who receive organ support have an in-hospital mortality of approximately 25%.¹ Furthermore, rapid and substantial muscle wasting occurs during critical illness, with patients losing up to 18% of muscle mass in the first 10 days of admission.^2,3 This muscle wasting is a major contributor to ‘ICU-acquired weakness’,⁴ which is associated acutely with increased mortality, slower weaning from ventilator support, longer time to discharge alive from ICU and hospital, and higher in-hospital costs.[5], [6], [7] The detrimental effects of ICU-acquired weakness also persist after discharge from the acute care setting, and adversely affect physical function and health-related quality of life.[8], [9], [10] The majority of critically ill patients who receive organ support also receive artificial enteral nutrition (EN), providing a potential therapeutic target for exploration.^11,12

Based on data in health,¹³ it has been suggested that augmenting dietary protein in critical illness could attenuate muscle loss, with subsequent reductions in duration of mechanical ventilation and hospitalisation.[14], [15], [16], [17] While few randomised controlled trials (RCTs) have assessed the impact of augmented protein in critical illness, preliminary data from small RCTs indicate that increasing enteral protein during ICU admission could attenuate skeletal muscle wasting and improve health-related quality of life in survivors.[18], [19], [20] Prospective and retrospective observational studies have also reported that augmenting enteral protein (delivered at ≥ 1.2 g/kg bodyweight/day) is associated with reduced mortality, shorter time to discharge alive, reduced duration of mechanical ventilation, and reduced duration of ICU and hospital admission.[21], [22], [23], [24], [25], [26] However, the impact of unmeasured confounders on the treatment effect in such observational data may be substantial.

Whilst most observational studies suggest benefit with greater protein doses, there are also data to suggest potential harm. A 63-patient cohort study reported a greater loss of ultrasound-derived rectus femoris muscle area in those that received more dietary protein; however, specific protein doses received were not reported.² Existing meta-analyses evaluating greater protein delivery have reported wide confidence intervals (CIs) around all outcomes, such that even if the point-estimate favoured benefit, there remains the possibility that the true effect of protein delivery >1.2 g/kg/day may be harmful.[27], [28], [29]

While international critical care nutrition guidelines recommend delivery of dietary protein to achieve at least 1.2 g/kg/day,^11,12 this does not occur in current practice.³⁰ Point- and period-prevalence data indicate that patients in Australia and New Zealand, as well as globally, receive protein doses of 0.6–0.69 g/kg of actual body weight/day^31,32. A parallel-group, blinded, patient-randomised, clinical trial was conducted across six ICUs in Australia and New Zealand to determine the feasibility of achieving augmented enteral protein doses in critically ill adults receiving invasive organ support using a very high protein formula compared to usual care.³³ One hundred and sixteen patients were randomised to receive isocaloric enteral formulae with augmented protein (100 g/1000 mL, intervention) or standard protein (63 g/1000 mL, comparator) with the aim of achieving protein doses representative of international guidelines (≥1.2 g/kg/day) in the intervention group compared to doses representative of ‘usual care’ in the comparator group (<1.0 g/kg/day). This strategy demonstrated that it was feasible to deliver protein doses representative of international guidelines, compared to usual care with a clear difference in the protein dose delivered (1.52 (standard deviation: 0.52) vs. 0.99 (0.27) g/kg ideal body weight (IBW)/day; mean difference 0.53 (95% CI 0.38 to 0.69)) with no difference in energy delivery between groups.³³ This study informed the proposed TARGET Protein trial.

The primary hypothesis of the TARGET Protein trial is that augmenting dietary protein to critically ill adult patients when compared to usual care will increase the number of days free of the index hospital and alive at day 90.

2. Methods and analysis

2.1. Trial management

The TARGET Protein trial is an investigator-initiated collaboration between the TARGET Protein Investigators.

The principal sponsor of the TARGET Protein trial is the Central Adelaide Local Health Network – Royal Adelaide Hospital, Adelaide Australia, which will undertake the following roles: coordination of human research ethics approval, oversight of data collection, building and monitoring of study database, and organising the technical logistics of trial enteral formulae.

This trial is funded by a Medical Research Future Fund Grant - Rare Cancers Rare Diseases and Unmet Need Streams and an Australian and New Zealand Intensive Care Foundation Project Grant. Nutricia Australia Pty Ltd have provided an unrestricted grant to support the costs of the trial enteral formulae. Representatives from Nutricia Australia Pty Ltd were provided the protocol prior to trial commencement; however, neither Nutricia Australia Pty Ltd nor their representatives had input into trial design, conduct of the trial, and will have no input into data analysis or interpretation, or drafting of the manuscript. Nutricia Australia Pty Ltd will be provided with a pre-publication manuscript for review.

This trial was endorsed by the Australian and New Zealand Intensive Care Society Clinical Trials Group (ANZICS CTG) on 10 September 2021, with all publications approved prior to submission. The trial was prospectively registered with the Australian New Zealand Clinical Trials Registry on 1 November 2021 (ACTRN12621001484831). This protocol follows the Standard Protocol Items: Recommendations for Interventional Trials SPIRIT 2013 statement.³⁴

2.2. Trial design

TARGET Protein is a cluster randomised, cross-sectional, double cross-over, pragmatic clinical trial enrolling critically ill adult patients from eight ICUs across Australia and New Zealand. The trial commenced at four sites (South Australia, New South Wales, and New Zealand) on 23 May 2022, with four sites in Victoria commencing on 23 August 2022. Each site will participate in the trial for 12 months such that recruitment will cease on 23 May 2023 for the initial sites and on 23 August 2023 for the subsequent sites. All patients in an individual ICU (cluster) who meet eligibility criteria will receive the same trial enteral formula across a three-month period. After a three-month period, the ICU will then administer the alternative trial enteral formula for all new eligible patient admissions over the next three months. Participants will continue to receive the treatment that they were originally assigned if they remain in the ICU following a crossover period. The process is then repeated so each ICU (cluster) crosses over twice. There will be no washout between cluster periods (Fig. 1).

Fig. 1.

Trial design.

2.3. Rationale for trial design

For augmented enteral protein dosing to affect patient outcomes, it is recommended to be commenced as soon as practicable after tracheal intubation or where oral intake is not possible,^11,12 which also ensures significant expose to the augmented protein intervention. This requires the trial enteral formula to be administered as standard care at each participating site. A cluster RCT where the participating ICU (cluster) is randomised to augmented dietary protein or usual dietary protein is the most feasible trial design to reduce selection bias and recruit significant number of patients to detect a difference in the primary outcome.

2.4. Setting and population

Sites are eligible for inclusion if they are an ICU capable of treating mechanically ventilated critically ill patients. All eligible patients in the ICU at the time will receive the trial enteral formula. Eligible patients are those aged ≥16 years prescribed EN during their index admission to ICU or prescribed EN for the first time in ICU during subsequent admissions. Patients that meet all inclusion criteria and none of the exclusion criteria will be eligible. Inclusion and exclusion criteria are listed in Table 1.

Table 1.

Eligibility criteria.

Inclusion criteria
1. ≥16 years of age
2. Receiving/about to commence EN during their index admission to ICU or receiving/about to commence EN for the first time in ICU during subsequent admissions
Exclusion criteria
1. At the commencement of EN, the treating clinician considers the trial enteral formula to be contraindicated
2. The patient has received greater than 12 h of non-trial enteral formula
3. The patient has previously participated in the TARGET Protein trial

2.5. Allocation generation and randomisation

Site randomisation was computer generated by the trial investigator responsible for data management (independent of the study statistician), with concealment maintained until one month prior to trial commencement. As the trial is open label, sites were notified one month prior to commencing to facilitate trial logistics and order trial enteral formula in preparation. To reduce the risk of demand for trial enteral formulae being greater than supply, the trial commencement was staggered: two groups of four sites commenced three months apart; hence, the randomisation was conducted as two strata of blocks of four, to ensure each period was balanced.

2.6. Intervention and comparator

During the intervention cluster periods (augmented dietary protein), all eligible patients receive the enteral formula ‘Nutrison Protein Intense®’ which contains 100 g protein per 1000 ml. During the comparator cluster periods (usual dietary protein), all patients receive the enteral formula ‘Nutrison Protein Plus®’, which contains 63 g protein per 1000 ml. The protein sources are similar between formulae, each derived from whey, casein, pea, and soy protein. The osmolarity of the formulae are the same and both are fibre-free (Table 2).

Table 2.

Trial enteral formula product information.

Nutritional composition per 1000 ml	Nutrison protein intense (Intervention)	Nutrison protein plus (Comparator)
Energy, kcal	1260	1250
Protein, g	100	63
Carbohydrates, g	104	142
Fat, g	49	49
Water, ml	810	810
Sodium, mg	1160	1110
Osmolarity, mOsmol/kgH2O	340	340

2.7. Study procedures

Following the treating clinician's decision to commence EN, the participant receives the trial enteral formula to which the ICU is currently randomised. The goal rate for trial enteral formula is set by the site's treating clinician (intensivist or dietitian) consistent with usual practice for that site. To ensure patient safety and prevent excess delivery, sites received additional education regarding the use of prescribing nutrition to IBW. All aspects of nutrition management, other than the choice of formula, are according to individual unit practice. The rate at which trial enteral formula is commenced and incremented, and strategies to increase nutrient delivery (e.g., prokinetic drugs, post-pyloric tubes), are at the discretion of the treating team. Trial enteral formula is administered when clinically indicated until either the patient reaches day 90, discharged from ICU, or dies. If a patient develops a clinical requirement for a non-trial enteral formula (e.g., calorie dense formula for fluid restriction or semi-elemental formula for malabsorption), the trial enteral formula is discontinued and replaced by non-trial enteral formula and this is treated as a protocol deviation. As soon as the patient no longer requires the non-trial enteral formula for clinical requirements, the trial enteral formula will be re-commenced. Patients discharged and readmitted to the ICU within the index hospital admission and within 90 days of study enrolment, still requiring EN, will be recommenced on trial enteral formula as per the initial treatment allocation. Patients who are enrolled in TARGET Protein at a participating site who are then transferred to another TARGET Protein participating site will be excluded from the trial at the receiving site. They will not be required to continue the trial enteral formula they were originally allocated to at the first site, as this will be a new index hospital admission. The rational for this relates to the availability of trial enteral formula, as sites may be randomised to the alternate trial enteral formula.

2.8. Study outcomes

The primary outcome is ‘number of days free of the index hospital and alive at day 90’. Secondary outcomes relate to clinical outcomes. All primary and secondary outcome definitions are listed in Table 3.

Table 3.

Study outcomes.

Primary outcome
•Number of days free of the index hospital and alive at day 90. This will be calculated as 90 days and subtracting all days admitted to the index hospital after commencement of trial enteral formula, plus any days readmitted to the index hospital within 90 days. As patients may be readmitted for part of a day (e.g. for dialysis), a readmission will only be counted when the patient remains in hospital at 2400 h. Patients will be considered alive if they are alive at discharge from the index hospital, and there is no evidence of death before day 90. In Australia, death after hospital discharge will be ascertained using data linkage with the Australian National Death Index.35,36 At the New Zealand site, the Adult Patient Database is linked to the New Zealand national death registry to record death after hospital discharge. Patients who die during the 90-day period, will be assigned a value of 0.
Secondary outcomes
•Days free of the index hospital at day 90 in survivors. This will be calculated only for those that survived to day 90. The outcome will be the same as the primary outcome and calculated as 90 days and subtracting all days admitted to the index hospital after commencement of trial enteral formula, plus any days readmitted to the index hospital within 90 days. As patients may be readmitted for part of a day (e.g., for dialysis), a readmission will only be counted when the patient remains in hospital at 2400 h •Alive at day 90. Patients will be considered alive if they are alive at discharge from the index hospital, and there is no evidence of death before day 90. Health service records will be used to provide evidence of patient status, including if the patient was discharged alive from hospital but known to have a health event prior to day 90. Health events include rehabilitation hospital admission, outpatient appointment attended, pathology test completed, other medical investigation, day admission, eg dialysis, chemotherapy. •Duration of invasive ventilation (hours) •Duration of admission to ICU (days) •Duration of admission to hospital (days) •Incidence of tracheostomy insertion •Incidence of renal replacement therapy •Discharge destination

2.9. Sample size and power

The sample size was calculated using The Shiny Cluster Randomised Trial Calculator (https://clusterrcts.shinyapps.io/rshinyapp/) and data from TARGET Calories.¹ In TARGET Calories, the standard deviation for the primary outcome being used in TARGET Protein (days free of the index hospital and alive at day 90) was 11.3 days. Assuming an exchangeable correlation structure, a coefficient of variation for cluster size of 0.5 – [based on 8 ICUs (i.e. 8 clusters) each enrolling a mean of 160 patients/3 month period (95% CI, 80 to 240 patients per cluster per period)] – and the default settings for cluster trials of an intracluster correlation coefficients of 0.02 (with a lower extreme of 0.001 and upper extreme of 0.1) 60 to 65 patients per period per cluster would provide 80% power, and ∼80 patients per cluster per period would provide 90% power to find a mean difference of one day for the primary outcome. Assuming a parametric distribution, a sample size of 2560 from the 8 clusters would provide 90% power to detect a mean difference of 1 day free of the index hospital and alive at day 90. However, the primary outcome (days free of the index hospital and alive at day 90) is unlikely to be normally distributed and parametric calculations are optimistic. For this reason, the sample size was inflated by 15% to allow for the likely skewed distribution of data, such that ∼3000 patients from the 8 clusters should provide 90% power.³⁷

2.10. Statistical analysis

For the primary outcome (days free of the index hospital and alive at day 90) and the first secondary outcome, we will use individual patient-level data and fit a quantile mixed effects model to compare the median response between treatment groups. The mixed effects model will include treatment group, period, and the stratification variable used in randomisation, delayed start (“Group 1 – Commencing May 2022” and “Group 2- Commencing August 2022”), as fixed effects. Cluster (ICU) will be included as a random effect and assumed to be normally distributed with mean zero and variance components ${\sigma }_C^2$ (5). The effect of treatment comparisons will be presented as a difference in medians (95% CI); the 95% CIs will be calculated using the block-bootstrap method. For the binary secondary outcomes (i.e., alive at day 90, incidence of tracheostomy insertion and incidence of renal replacement therapy), we will fit generalised estimating equations with a logarithmic link function, an exchangeable working correlation matrix, and robust standard errors using the ICU as the clustering unit. Finally, time to event secondary outcomes (i.e. duration of invasive ventilation, admission to ICU and admission to hospital) will be analysed by fitting Cox regression models with covariate adjustment for period and delayed start, with stratification by ICU, and robust standard errors clustered at ICU level (for any residual within-ICU correlation) to estimate cause-specific hazard ratios and CIs, with patients dying prior to discharge (or extubation) censored at their time of death. Sensitivity analyses will be conducted to investigate the impact of missing data and competing risk of death. Sub-group analyses will be presented for the following subgroups: invasive mechanical ventilation, age 70 years and older, body mass index of 35 kg/m² or above and acute kidney injury (AKI) prior to receiving trial enteral formula (note that the last subgroup analysis was added following publication of the EFFORT trial on 25 January 2023³⁸). Acute kidney failure is defined as receiving new renal replacement therapy between hospital admission and commencement of trial enteral formula. A detailed statistical analysis plan will be publicly available prior to final database lock.

2.11. Data collection and management

A secure password protected internet-based electronic case report form (eCRF) is being used to record all patient data using Research Electronic Data Capture (REDCap) software. REDCap is a secure good clinical practice-compliant web-based server.³⁹ Study participants will be followed up until either death or until 90 days after enrolment, censored at hospital discharge. If consent for participation is withdrawn, all data are destroyed unless consent is obtained for retention of collected data. Patients that meet all inclusion criteria are recorded on an electronic screening log, with patients who meet one or more exclusion criteria recorded. Data collection will be from two sources: (1) The Australian New Zealand Intensive Care Society Adult Patient Database (ANZICS APD) (Table 4);⁴⁰ (2) local site medical records, which will be entered into the REDCap eCRF (Table 5). Linkage between these two databases will occur using the statistical linkage key, a unique patient identifier which considers the patient's family name, given name, date of birth, and gender. The ANZICS APD was set up in 1992, currently holds >2 million patient episodes, and all participating sites contribute data to the ANZICS APD for every admission.⁴⁰

Table 4.

Data obtained from the Australian New Zealand Intensive Care Society Adult Patient Database.

Baseline data
Demographics — age (years), sex (male/female), weight (kg), height(m)
Acute Physiology And Chronic Health Evaluation (APACHE) II score
ICU admission diagnosis from APACHE III-J score
Elective, emergency, or non-surgical admission category
Clinical frailty score within 2 months preceding their first admission to ICU within the current hospital stay
Biochemistry data (highest and lowest within 24 h of ICU admission): blood urea, serum creatinine, serum albumin, serum potassium, blood glucose
Ventilation type on ICU admission (day 1), categorised as non-ventilated or ventilated
ICU and hospital source of admission
Post intervention outcome data
Mechanical ventilation hours in ICU
Tracheostomy insertion in ICU
Need for renal replacement therapy in ICU
Duration of ICU and hospital admission
ICU and hospital mortality
ICU readmissions within the same hospital admission
ICU and hospital readmissions to the index hospital
Discharge destination (home, another acute hospital, rehabilitation, nursing home/chronic/palliative care, ICU in another hospital, other destination)

Table 5.

Data collection — Obtained from local site medical records and entered into study electronic case report form.

TARGET Protein trial enteral formula delivery	Daily volume of trial enteral formula delivered, collected on study days 1–5, day 10, day 20, day 30 and day 90. (Study day 1 commences on the day TARGET Protein trial enteral formula is commenced and concludes at the end of the calendar day (2400 h), i.e., will represent a partial day).
Biochemical data	Blood urea, serum albumin, serum phosphate, and blood glucose closest to 0800 h on study days 1,5, and day 10.
Adverse and serious adverse event data	Collected whilst participants are receiving trial enteral formula up until 48 h post cessation of trial enteral formula.
Last known patient status	At any time up to study day 90, as per medical records, to ascertain whether the patient is alive at day 90 without direct contact.

To ensure compliance with trial enteral formula, random audits are conducted. The audits are conducted on random weekdays when sufficient research staff are available and involve a researcher independent to the Management Committee recording the formula that is hung at the bedside at the time of the audit. These audits will be conducted every six weeks: midway through a cluster period and in the week following cluster crossover. The formula visualised will be cross-checked against the electronic medical record documented formula, and case report form to ensure compliance and source data accuracy.

The project data management conforms to the National Health and Medical Research Council (NHMRC) of Australia guidelines on data management.⁴¹ The local site principal investigators will be responsible for data management at their site. The TARGET Protein Chief Investigator and the coordinating centre (Central Adelaide Local Health Network – Royal Adelaide Hospital) will be responsible for the management of pooled non-identifiable data.

2.12. Data Safety Monitoring Board

A Data and Safety Monitoring Board (DSMB) has been formed. The DSMB will be an advisory to the Chief Investigator and, through them, to the TARGET Protein Management Committee. A single review by the DSMB will occur when all eight participating sites have completed two cluster periods. No interim analyses are pre-specified for this trial for assessment of the efficacy of the interventions. Mortality data will not be provided to the DSMB due to the study design (cluster crossover) and staggered start, four of the eight sites will have only 90 days of recruitment left when all sites have completed two crossover periods. With delays anticipated for complete entry into the ANZICS APD and subsequent data linkage, 90-day mortality will not be available until after recruitment has been completed. As both trial enteral formulae are approved for use by relevant Australia and New Zealand authorities and are currently administered to critically ill patients in routine clinical care, the lack of mortality data for the DSMB does not represent a substantial risk to participants.

2.13. Adverse and serious adverse events

It is recognised that the patient population in the ICU will experience aberrations in laboratory values, signs, and symptoms due to the severity of the underlying disease and the impact of standard treatments in the ICU. These do not necessarily constitute adverse events (AEs) or serious adverse events (SAEs) unless they are considered related to study treatment or, in the principal investigator's clinical judgement, are not recognised events consistent with the patient's underlying critical illness and/or chronic diseases and expected clinical course. Accordingly, the reporting of AEs will be restricted to events that are considered related to study treatment (possibly, probably, or definitely). AEs are defined as any untoward medical occurrence in a patient or clinical investigation subject administered an investigational intervention and will be collected from commencement of trial enteral formula until 48 h post cessation of trial enteral formula. The patient will be followed until the event is resolved or explained. Frequency of follow-up is left to the discretion of the site principal investigator.

The baseline mortality of intensive care patients is high due to the illness that has necessitated their ICU admission. ICU patients will frequently develop life-threatening organ failure(s) unrelated to study interventions and despite optimal management. Therefore, consistent with ICU international best practice and, as outlined above, deaths that are part of the natural history of the primary disease process or expected complications of critical illness will not be reported as serious adverse events (SAEs) in this study.⁴² Events already defined and reported as study outcomes (e.g., mortality) will not be reported separately as adverse or SAEs unless they are considered to be causally related to the study intervention or are otherwise of concern in the site principal investigator's judgement. SAEs will be reported up to day 90 post-study enrolment, including SAEs that are considered related to trial enteral formula, namely mesenteric ischaemia and abdominal compartment syndrome at the discretion of each site's principal investigator.^43,44

2.14. Protocol deviations

Pre-specified protocol deviations will be those related to delivery of non-trial enteral formula categorised under the following: (1) clinical decision: when the treating clinician believed the patient would subsequently benefit from a specific formula; (2) unable to locate trial enteral formula: when clinical staff cannot locate assigned trial enteral formula and delivered a different formula; (3) error at commencement of EN: when the assigned trial enteral formula was not commenced but instead a different formula was erroneously administered; (4) error on subsequent administration of EN: when this occurred after commencement of EN, e.g., on readmission to ICU.

2.15. Ethics approval

Human research ethics approval has been granted in Australia by the Central Adelaide Local Health Network Human Research Ethics Committee (2021/HRE00248) on 3 September 2021 and in New Zealand by Northern B Health and Disability Ethics Committee (2021 FULL 11097) on 14 December 2021. Minor protocol amendments which consisted of minor wording changes, improvements in clarity around inclusion criteria, data sharing, and the primary outcome definition (as documented above) were approved on the 19 May 2022 (protocol version 6.0, dated 18 May 2022) prior to the commencement of recruitment on the 23 May 2022. Minor protocol amendments were made after commencement of recruitment, the current protocol is version 8.0, dated 16 September 2022.

2.16. Consent

Most eligible patients will lack capacity to provide prior consent due to their underlying medical condition requiring unplanned/emergency life support measures. For Australian sites, the trial has been approved for a waiver of consent to enrol participants according to guidelines set out in Sections 2.3.10 and 2.3.12 of the NHMRC National Statement.⁴⁵ For New Zealand, the trial has been approved for a waiver of consent to commence the intervention with consent to continue. In New Zealand, under Right 7 (4) of the code of Health and Disability Services Consumers' Rights, where a legal surrogate is not recognised, the patient may be enrolled without prior consent if enrolment is in the patients' best interest.⁴⁶ The views of other persons in respect of a patient's likely wishes in relation to trial participation will be sought by asking their family/whānau/friend as soon as it is clinically reasonable to do so. Although this will be the primary approach, if this is not possible prior to enrolment (e.g., due to family/whānau/friend not being present or due to discussion prior to enrolment not being possible for other reasons), enrolment will be without prior discussion, but family/whānau/friend views will be sought as soon it is reasonable to do so. Participants who recover sufficiently to consent for themselves will be free to decide whether to consent to use of data that have already been collected and whether to continue to participate in the study.

2.17. Co-enrolment

The ANZICS CTG policy on co-enrolment will be followed. Studies involving another nutrition intervention in ICU for >24 h will be reviewed by the TARGET Protein Management Committee on an individual basis.

2.18. Competing trials

After TARGET Protein began, the EFFORT trial was published.³⁸ The EFFORT trial evaluated the effect of increased protein (≥2.2 g/kg/day), as either enteral or intravenous protein (including enteral protein supplementation), to ‘usual dose’ protein (≤1.2 g/kg/day) to ‘nutritionally high-risk’ adult ICU patients. The trial was ceased with less than one third of the planned cohort enrolled due to slower than anticipated recruitment. At that time, 1301 patients had been recruited (645 in the increased protein group and 656 in the usual dose group). The primary outcome of the trial was changed before recruitment was completed, and there was no reported difference in the primary outcome of ‘time to discharge alive from hospital up to 60 days after ICU admission’. Two of the eleven sub-group analyses favoured usual care: these were patients with AKI and those with high Sequential Organ Failure Assessment (SOFA) score (≥9) on ICU admission. No statistical adjustment was made for multiple analyses. Due to the considerable uncertainty as to the inferences from these data, the TARGET Protein Management Committee agreed to continue the TARGET Protein trial as planned, with an amendment to the intended sub-group analyses to include patients with AKI prior to commencing trial enteral formula. Given SOFA score is not routinely collected as a part of the ANZICS APD dataset, and the measure of illness severity collected as part of the ANZICS APD dataset in our trial, the APACHE II score, did not show harm; there was no change to the sub-group analyses for illness severity.

The PRECISE study is an additional trial that aims to evaluate high enteral protein compared to standard enteral protein provision on functional recovery (health related quality of life) following ICU admission.⁴⁷ The estimated completion date for this trial is December 2023.

2.19. Data sharing statement

Non-identifiable individual participant data that underlie the results reported in this trial will be made available after three years following publication and ending seven years after publication. Availability will only be made to researchers who provide a written proposal for data evaluation that is judged to be methodologically sound by a committee approved by the TARGET Protein Investigators. If the proposal is approved, access data requestors will be required to sign a data access agreement prior to accessing data.

2.20. Dissemination

The results of this trial will be disseminated in an international peer-reviewed journal and presented at national and international critical care and nutrition scientific meetings.

2.21. Summary

TARGET Protein is a cluster randomised, cross-sectional, double cross-over, pragmatic clinical trial enrolling critically ill adult patients from eight ICUs across Australia and New Zealand, which will determine whether augmented enteral protein delivery will reduce days free of the index hospital and alive at day 90 compared to usual care.

Financial support

Project support is provided by Medical Research Future Fund Grant ($1,992,218AUD), an Australian and New Zealand Intensive Care Foundation Project Grant ($100,000AUD), and Nutricia Australia Pty Ltd have provided an unrestricted grant to support the costs of the trial enteral formulae ($290,883AUD). Personal support for individual investigators includes: MJS is supported by a University of Adelaide Research Scholarship. EJR is supported by a NHMRC Emerging Leadership Investigator Grant. AMD is supported by an NHMRC Leadership Investigator Grant. This research was conducted during the tenure of a Clinical Practitioner Research Fellowship from the Health Research Council of New Zealand held by PY. The Medical Research Institute of New Zealand is supported by Independent Research Organisation Funding from the Health Research Council of New Zealand. PY reports receiving advisory fees from Baxter Healthcare and from AM Pharma (not related to this project).

Credit author statement

M.J. Summers, L.S. Chapple, R. Bellomo, M.J. Chapman, S. Ferrie, M.E. Finnis, C. French, S. Hurford, N. Kakho, A. Karahalios, M.J. Maiden, S.N. O’Connor, S.L. Peake, J.J. Presneill, E.J. Ridley, A. Duy-Tran, P.J. Williams, P.J. Young, S. Zaloumis, A.M. Deane contributed to the conception and design of the trial. M.J Summers, L.S. Chapple and A.M. Deane drafted the manuscript. All authors critically revised the manuscript and approved the final manuscript. A.M. Deane, M.J. Chapman, S.L. Peake, L.S. Chapple, J.J. Presneill, R. Bellomo, P.J. Young, E.J. Ridley, A. Karahalios, A. Duy-Tran and S.N. O’Connor were responsible for funding acquisition.

Conflict of interest

LSC has received lecture fees from Nutricia and Fresenius Kabi. EJR has received lecture fees from Baxter Healthcare, Nestle and Nutricia. AMD is employed by an institution that has received lecture fees from Baxter Healthcare for his time. All other investigators report no competing interests or financial support. Rinaldo Bellomo, John P. Young and Adam M Deane declare a conflict of interest as Editors or Editorial Committee members of this journal.

APPENDICES

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

SITE INVESTIGATORS

Austin Hospital - Rinaldo Bellomo, Christine Choong, Glenn Eastwood, Kate Hamilton, Leah Peck, Helen Young.

Royal Adelaide Hospital – Imogen Asser, Nerissa Brown, Lee-anne Chapple, Sarah Doherty, Mahni Foster, Kathleen Glasby, Rhea Louis, Fiona McDonald, Stephanie O'Connor, Mark Plummer, Justine Rivett, Matthew Summers.

Royal Melbourne Hospital – Deborah Barge, Alice Barrese, Kathleen Byrne, Adam Deane, Kate Fetterplace, Olivia Gigli, Jeffrey Presneill, Brianna Tascone, Kym Wittholz.

Royal Prince Alfred Hospital – Suzie Ferrie.

Sunshine Hospital – Samantha Bates, Craig French, Haindavi Muppa, Giang Nguyen, Stephanie Schembri.

The Queen Elizabeth Hospital – Sandra Peake, Patricia Williams, Catherine Kurenda, Srilatha Vemparala.

University Hospital Geelong – Stacey Hawker, Michelle Horton Lynda Jaques Nima Kakho, Matthew Maiden, Jemma Trickey.

Wellington Regional Hospital – April Aguilar, Melissa Butt, Mary Rose Sol Cruz, Kirsha Delaney, Dushanka Hettige, Sally Hurford, Reece Latonio, Eden Lesona, Joradee Marmol, Leanlove Navarra, Shan Qiu, Nola Wood–Harris, Paul Young.

DATA AND SAFETY MONITORING BOARD MEMBERS

Associate Professor Todd Rice (chair) - Experienced clinical trialist in critical care nutrition

Professor Nilesh Mehta - Experienced clinical trialist in critical care nutrition

Assistant Professor Michael Harhay - Biostatistician with experienced in design, analysis and reporting of cluster randomised trials