Intensive Care

Fever management

Paracetamol is one of the most widely used medicines and is commonly given to Intensive Care Unit (ICU) patients with infections. The MRINZ led a 700 patient clinical trial in 23 ICUs comparing paracetamol to placebo in ICU patients with fever and infection, which was published in the New England Journal of Medicine in 2015. This study demonstrated:

  • Paracetamol was safe and well tolerated in ICU patients with fever or infection
  • The medicine reduced body temperature by around a quarter of a degree
  • Paracetamol neither improved nor worsened patient outcomes
  • Patients spent less time in ICU if they were given paracetamol, but patients who died spent more time in ICU before death if given paracetamol

This study was endorsed by the Australian and New Zealand Intensive Care Society Clinical Trials Group, funded by the Health Research Council of New Zealand, and conducted in partnership with the George Institute for Global Health, Sydney, Australia. Following on from the success of this study, the MRINZ is leading the REACTOR study which is a Randomised Evaluation of Active Control of Temperature vs. ORdinary temperature management in critically ill adults on life support with fever. This study is funded by the Health Research Council of New Zealand and forms one component of a body of work led by Dr Paul Young who was awarded a Clinical Practitioner Fellowship by the Health Research Council of New Zealand to support work in the area of treatment of fever in the critically ill in 2016.

Comment: The fever management programme is one of a number research initiatives which directly challenge core principles of current clinical practice. Other examples include intravenous fluid therapy, oxygen therapy and blood transfusion.

Fluid management

Acutely ill patients are commonly treated with intravenous fluids. We completed the SPLIT study comparing the effectiveness of two commonly used intravenous fluids in 2015. The SPLIT study was published in the Journal of the American Medical Association and was the third highest impact publication in the field of Intensive Care Medicine in 2015. The study enrolled 2278 participants from four New Zealand hospitals and compared the routine use of 0.9% saline for fluid therapy with Plasma Lyte® 148 in ICU patients. The study hypothesis was that routinely using Plasma Lyte® 148 for fluid therapy instead of 0.9% saline would reduce the risk of developing acute kidney failure. Kidney failure which occurs in the setting of acute illness is associated with a high risk of  death and may require treatment with costly kidney dialysis treatments. This study addressed an issue of major global public health significance because more than a million litres of 0.9% saline are administered to patients around the world daily. The study showed that 0.9% saline and Plasma Lyte® 148 led to a similar risk of acute kidney injury but also raised the possibility that in high risk patients Plasma Lyte® 148 therapy may reduce the risk of death. The SPLIT study was funded by the Health Research Council of New Zealand and Baxter Pty Ltd. A follow-up 8800 participant study confirming whether or not Plasma Lyte® 148 therapy reduces mortality compared to 0.9% saline will be conducted by the MRINZ in partnership with The George Institute for Global Health. This follow-up study has been funded by the National Health and Medical Research Council in Australia and the Health Research Council of New Zealand.

Comment: This research programme shows the capacity of well-designed randomised controlled trials to change practice in New Zealand and internationally.


Nutrition therapy is an essential standard of care for all ICU patients who are mechanically ventilated and remain in ICU for more than a few days. There is a substantial and well established dissociation between the recommended calorie requirement and calories actually delivered to ICU patients. Nevertheless, while it remains logical that energy delivery should match energy consumption, the benefits of such matching remain to be confirmed by a robust, high quality clinical trial. We have recently completed enrolling patients into a 4000 patient trial designed to evaluate whether delivery of the full recommended calorie (energy) requirement to critically ill patients improves 90-day survival when compared to standard practice. This trial is funded by the Health Research Council of New Zealand and Australia’s National Health and Medical Research Council.

Comment: This study is an example of the close collaboration between the MRINZ and intensive care colleagues in New Zealand and Australia.

Stress ulcer prophylaxis

The administration of stress ulcer prophylaxis (SUP), either with a Proton Pump Inhibitor (PPI) or a Histamine-2 Receptor Blocker (H2RB) is recommended in international guidelines and incorporated into quality-oriented checklists for care of ICU patients. Our recent data show that PPIs and H2RBs are routinely used for SUP in Australia and New Zealand with the choice of medication probably not based on patient factors, but instead dependent on clinician preference or unit policy. This practice variation reflects the lack of definitive evidence comparing PPIs to H2RBs in the ICU setting. Although data suggest PPIs are more effective at reducing upper GI bleeding risk in ICU patients than H2RBs, it also appears that using PPIs in ICU patients is associated with an increase in pneumonia risk compared to H2RBs, and that PPI use but not H2RB use is associated with increased risk of Clostridium difficile infection. The overall influence of the opposing risks of upper GI bleeding and SUP-related infectious complications on in-hospital mortality is unknown. We are leading a multicentre, multinational trial comparing the safety and efficacy of PPIs vs. H2RBs in 25,000 mechanically ventilated ICU patients. This trial is funded by the Health Research Council of New Zealand and the Intensive Care Foundation.

Comment: This international study which includes sites in New Zealand, Australia, United Kingdom, Ireland and Canada is the largest intensive care randomised trial ever undertaken.

Timing of initiation of kidney dialysis in ICU patients with acute kidney injury

Acute kidney injury (AKI) is a common and devastating complication of critical illness. Once AKI is established, treatment is largely supportive and no intervention has been found to restore kidney function or improve overall survival. Renal replacement therapy (RRT), usually in the form of haemodialysis, is frequently needed to manage patients with severe AKI. Such patients have an inhospital mortality that consistently exceeds 50% with delays in RRT initiation implicated as a possible contributor. The optimal timing of RRT initiation is an existing knowledge gap and a clear priority for investigation. The STARRT-AKI trial will evaluate whether earlier / pre-emptive / accelerated RRT initiation is associated with enhanced survival as compared to a conservative strategy for initiation of RRT, which is driven by conventional indications and clinician judgment. This multinational trial is being led in New Zealand by the MRINZ and is funded by the Health Research Council of New Zealand.

Comment: The STARRT-AKI trial will provide high level evidence to guide clinicians in deciding the optimal time to commence kidney dialysis in critically ill patients with acute kidney injury.

Duration of antibiotic therapy in ICU patients with bacteraemia

Up to 50% of antibiotic use is inappropriate, with excessive durations of treatment the greatest contributor, to leading contributing to antibiotic resistance. Shorter duration antibiotic therapy (≤7 days) has been demonstrated to be as effective as longer duration in a range of infections and also in post hoc subgroup analyses of trials of infections commonly complicated by bacteraemia (blood stream infections). However, high-grade randomised trial evidence is lacking for the treatment of patients with bloodstream infection, which affects 15% of critically ill patients. The MRINZ is leading the NZ component of the BALANCE trial which is a a large multi-centre international study to determine if critically ill patients with bloodstream infection can be treated with a shorter rather than longer duration of antibiotic therapy. As well as contributing to the international effort, the NZ health system will benefit from participation in the BALANCE trial because local recruitment will provide specific data on NZ patients for whom there is uncertainty regarding antibiotic duration. Moreover, study participation, and will ensure the will ultimately enhance rapid translation of results into practice, enabling NZ’s health system to reduce costs and complications with no loss of treatment effectiveness. The NZ contribution to the BALANCE trial is funded by the Health Research Council of New Zealand.

Comment: Antibiotic resistance is one of the major global threats to mankind, and this study will inform as to whether shorter antibiotic courses may be one initiative which reduces the risk of antibiotic resistance.

Community acquired pneumonia

Pneumonia is a common reason for admission to hospital, and it is the most common site of severe infection causing organ failure and a need for intensive care. However, for many aspects of treatment we do not know which approaches lead to the best outcomes. In conjunction with colleagues in Europe, UK, USA and Australia we have developed new research design which can simultaneously compare treatment options in multiple “domains” of care. Initially we will assess up to 5 alternatives for antibiotic treatment, the use of steroids to reduce lung damage, and an extended duration of treatment with a macrolide (a class of antibiotics) which may also reduce lung inflammation. Later in 2018 alternative approaches to support for patients who require mechanical assistance with breathing will be added to the study, and other questions such as those which may arise in a respiratory illness pandemic can easily be added. The design will be integrated within the usual clinical processes and patients benefit from the knowledge that is gained as the trial proceeds. Analysis using Bayesian statistics undertaken at monthly intervals will inform the treatment allocations for new participants, and each domain continues until a clear answer is determined and then the result becomes usual patient care. This contrasts to the usual design where only a single hypothesis is tested, analysis does not occur until a fixed sample size is reached, and answers are often indeterminate. MRINZ is co-ordinating New Zealand’s participation in this trial which is funded by the Health Research Council of New Zealand, Australia’s National Health and Medical Research Council and a European Union FP7 grant.

Comment: The potential exists for this novel study design to be applied to other research questions both within and beyond intensive care medicine.

Blood transfusion

Blood transfusion is commonly required in ICU patients. The TRANSFUSE trial, which was led in New Zealand by Dr Colin McArthur, was published in the New England Journal of Medicine on September 27th 2017. This trial included 5000 intensive care patients from Australia, New Zealand, Finland, Ireland and Saudi Arabia and investigated the effect of the age of transfused red blood cells on critically ill patients’ outcomes. The trial found the transfusion of older stored red blood cells is safe and surprisingly, associated with fewer side effects.

Comment: This study has major implications for the New Zealand Blood Service as it reduces the requirement for fresh blood, which was previously thought to be the case, and which put considerable strain on resources.

Anti-fibrinolytic treatment in severe trauma

Bleeding causes most early deaths from major trauma, and is often exacerbated by abnormalities in blood clotting. Tranexamic acid (TXA) reduces abnormal clot breakdown, and is known to reduce bleeding in other settings such as major surgery. There is some evidence from lower income countries that early prophylactic use of TXA improves outcomes after traumatic injury. However, as it is uncertain if the possible benefits outweigh the risks (dangerous blood clots) in advanced trauma care systems with other rapid and effective treatments to stop bleeding and improve blood clotting, TXA is rarely used prophylactically in New Zealand. In this study led by Australian researchers, severely injured patients at risk of abnormal blood clotting will be randomly allocated to receive TXA or placebo commenced by ambulance staff at the scene and completed in hospital, in addition to all usual treatment. Patients will be assessed for abnormal excessive blood clotting while in hospital and followed for 6 months to determine if treatment with TXA improves survival and/or reduces disability. This bi-national trial is being led in New Zealand by the MRINZ and is funded by the Health Research Council of New Zealand and Australia’s National Health and Medical Research Council.

Comment: If effective without additional risk, prophylactic TXA will become standard treatment in ambulance services and Emergency Departments.