There are several established programs in Australia and New Zealand that make it possible for registrars and early career researchers to conduct ICU research — without building infrastructure from scratch. Each program works differently: some hold existing datasets you can apply to access; others allow you to collect prospective data at your site; others let you join a running study as a site investigator.
This guide explains what each program is, how it works, and how to get involved.
Comparing the programs
| ANZICS APD | NCCR | Open ICU | PPP | EXCEL | PRIMELab | |
|---|---|---|---|---|---|---|
| Prospective | — | ✓ | ✓ | ✓ | ✓ | ✓ |
| Retrospective | ✓ | ✓ | — | — | — | ✓ |
| Multicentre | ✓ | ✓ | ✓ | ✓ | ✓ | — |
| Can add data | — | ✓ | ✓ | ✓ | — | ✓ *** |
| Can request existing data | ✓ | ✓ | — | — | ✓ | ✓ |
| Independent project | ✓ | ✓ * | ✓ | ✓ | ✓ ** | ✓ |
| Join as site investigator | — | ✓ | ✓ | ✓ | — | — |
| Ethics required | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ **** |
| Central ethics available | — | — | — | ✓ | — | — |
| Suitable for first project | ✓ | ✓ | ✓ | ✓ | — | ✓ |
* Requires sites to agree to collect additional data variables.
** Only relevant if your site runs ECMO.
*** Preclinical/animal study design, not a clinical data field.
**** Animal ethics approval, not human research ethics.
A large database of ICU patient records collected across Australia and New Zealand. Over 200 ICUs contribute data on every adult patient admitted — demographics, diagnoses, severity scores, basic interventions, and outcomes. Data goes back to the 1990s, with over 3 million records. There is also a paediatric version (ANZPICR) and an annual survey of ICU resources (CCR).
Request a de-identified dataset and analyse it to answer a research question. You choose the question; ANZICS provides the data. You are responsible for your own analysis and write-up — ANZICS does not provide statistical support, though they can sometimes connect you with others working in similar areas.
Contact ANZICS CORE before anything else — before ethics, before formal applications. They will confirm whether your project is feasible and point you to others who have done similar work. There is also an annual Datathon which is a practical way to see how the data works before committing to a project.
Once feasibility is confirmed: submit a formal data request → committee review → sign a data use agreement → receive data.
Your own institution. Low or Negligible Risk is usually sufficient. Do the feasibility check first — do not submit ethics before you know the project is viable. De-identified aggregated data, and data from an individual site, can be returned to that site without restriction, other than approval from the ICU director or delegate.
This type of project takes longer than most people expect. You need either some background in data analysis or a supervisor who does. Data linkage studies — for example, linking to death registries or hospital administrative data — are significantly more complex and usually require dedicated funding.
Has the rate of renal replacement therapy in ICU patients changed over the past decade, and does this vary by hospital type?
The APD captures renal replacement therapy as a binary intervention variable for every adult ICU admission. A researcher requests a de-identified extract of adult admissions over a defined time period, specifying the variables needed — admission year, renal replacement therapy, hospital type, illness severity, and outcome. No new data collection is required. The researcher analyses the extract and writes up the findings independently.
A research infrastructure platform built on top of the ANZICS APD. It allows researchers to design their own study using ICU data — you write the question, design the data collection form, and the platform connects your study to ICUs across Australia and New Zealand. It can be retrospective (using existing APD data) or prospective (collecting new variables at participating sites). Current focus is sepsis, but the platform is broader. Explicitly designed to support early and mid-career researchers.
Design and lead an independent study using NCCR infrastructure. Request existing data. Collect new variables prospectively at sites that agree to participate. Join an active sub-study or working group. Expressions of interest for project support are accepted — projects should be completable within 18 months, and EMCR involvement is actively encouraged.
Contact Andrew Udy or Alex Poole directly.
Your own HREC approval required. No central ethics is available through NCCR.
What proportion of ICU patients with sepsis receive appropriate source control within six hours, and does this vary by hospital type?
Source control timing is not captured in the standard APD dataset. Using NCCR, a researcher designs an electronic case report form that adds this as a new variable. This can be done retrospectively — extracting the variable from medical records — or prospectively, with sites entering the data in real time for each new sepsis admission. Sites that agree to participate enter the additional field via the NCCR platform, alongside the existing APD data they already submit. The researcher receives a combined dataset containing the new variable linked to existing APD fields across all participating sites.
A new multicentre research network set up specifically to give early-career clinicians the opportunity to conduct prospective multi-site research. One study runs at a time. Currently small — six sites participated in the first study — but actively growing and seeking new investigators across all disciplines.
Get involved in designing a study from the beginning — including writing the protocol, choosing outcome measures, and building the data collection form. Collect prospective data at your site. Join the study management team. The second study is currently in planning.
Contact George Walker directly. The network is actively seeking clinicians interested in the next study.
Study-dependent — confirmed when each study is planned. No central ethics currently available.
No published output yet from the first study. This is a ground-floor opportunity — low existing infrastructure, high hands-on involvement. Well suited to registrars who want to understand the full research process rather than join an established study at the data collection stage.
What are the treatment thresholds and management strategies for new-onset atrial fibrillation in ICU, and do they vary between centres?
There was no existing dataset to draw on. Registrars across six sites collaborated to design the study, agree on definitions, and build the data collection form. Variables collected included heart rate thresholds triggering treatment, pharmacological agents used, vasopressor requirements, and rates of reversion to sinus rhythm. Each registrar prospectively identified and enrolled eligible patients at their own site and entered data in real time. The combined multisite dataset was pooled centrally for analysis.
On a designated day each year, all patients present in participating ICUs at 10am are enrolled and data is collected for 24 hours. Multiple research questions are investigated simultaneously on the same study day. Run by the ANZICS Clinical Trials Group in partnership with The George Institute for Global Health. Around 44 ICUs and 550 patients participate per study day. Between 2009 and 2016, the program produced 13 published papers and contributed to 18 subsequent grant applications.
There are two options. The first is to participate as a site — collect data at your ICU on the study day and gain authorship on the resulting publication. The second is to propose an independent research question to be included in an upcoming study day. If approved, the variables you specify are added to the data collection form and gathered across all participating sites on your behalf.
Contact the ANZICS Clinical Trials Group or The George Institute for Global Health to confirm the next study day and current process for submitting a research question.
A central ethics approval (historically held at Royal Prince Alfred Hospital, Sydney) covers most participating sites, with a waiver of consent. This substantially reduces the administrative burden compared to running your own study independently.
What proportion of ICU patients are receiving enteral nutrition, and are prescribed targets being met?
A researcher applies to have this question included in the next PPP study day, specifying the variables to be collected — whether enteral feeding is running, the prescribed rate, and the actual rate delivered. On the study day, data collectors at each participating ICU complete the standardised form for every patient present at 10am, drawing directly from the ICU chart. No additional patient contact or consent is required. All data is entered into a central REDCap database at the George Institute. The researcher receives a single national dataset without having recruited or consented a patient themselves.
A prospective registry of every patient receiving ECMO in Australia and New Zealand, run by the Australian and New Zealand Intensive Care Research Centre (ANZIC-RC) at Monash University. Captures approximately 91% of all adult ECMO cases. Collects data on patient selection, indications, complications, costs, and patient-reported outcomes at 6 months. Active until 2027.
If your site runs ECMO, it likely already contributes data to EXCEL. Investigators at participating sites can apply to access the registry dataset and use it to answer specific research questions. The registry also serves as a platform for embedded clinical trials and investigator-initiated sub-studies.
Contact the Chief Investigator or Project Manager at ANZIC-RC, Monash University.
EXCEL is available to investigators at ECMO centres — sites that already contribute data to the registry. If you work at one of these centres, the data is already being collected. You need a question and access, not a new study setup. If you are unsure whether your site participates, contact the team directly.
What is the 6-month functional outcome of patients who receive V-V ECMO for respiratory failure in Australia and New Zealand?
EXCEL prospectively enrols every patient who receives ECMO at participating centres. Site investigators enter data at the time of the ECMO run — indication, complications, ICU and hospital outcomes — and survivors are contacted at 6 months to complete patient-reported outcome measures. A researcher applies to access the existing registry dataset, which already contains this information for all enrolled patients. No new data collection is required.
A different category from the other programs on this page — PRIMELab is one of the largest preclinical (large-animal) research facilities in Australia, based at The Prince Charles Hospital in Brisbane and led by A/Prof Gianluigi Li Bassi. It is part of the Critical Care Research Group (CCRG). Rather than working with patient data, PRIMELab tests the safety and efficacy of new medical and surgical interventions in animal models before they reach clinical trials — covering areas such as severe lung injury, sepsis, ECMO, and heart and lung transplantation.
Join as a research fellow and learn advanced preclinical techniques — modelling critical illness in large animal models and testing interventions firsthand. This includes hands-on experience with surgical and medical procedures, prospective study design, data analysis, and publication. There are also opportunities to apply for grants and collaborate with national and international researchers. Several working groups run concurrently, covering ARDS, ECPR, sepsis, VV/VA-ECMO, and heart transplantation — fellows are typically exposed to several areas before choosing one to focus on.
Read the Research Prospectus on the CCRG website first, then make contact directly.
Animal ethics approval is required for all studies — this is different from the human research ethics process used by the other programs on this page. Fellows are taught how to navigate the animal ethics approval process as part of their research training.
This is laboratory-based research, not clinical data analysis — the time commitment and skill set are different from the other programs on this page. It suits trainees wanting hands-on experience in translational science, with direct exposure to the techniques that have historically shaped major advances in critical care (e.g. lung-protective ventilation, ECMO, inhaled nitric oxide). Many findings progress on to assessment using larger clinical registries, including via CCRG's connections to international networks such as the Extracorporeal Life Support Organization.
See Hoe LE et al. demonstrated that donor heart ischaemic time can be safely extended beyond 9 hours using hypothermic machine perfusion in a sheep model — the first Australian study of its kind, which prompted adoption of this storage method for donor hearts in Australia (J Heart Lung Transplant, 2023).
A separate study from the group developed the first heart transplantation model accurately replicating the inflammatory response to brain injury prior to donation and transplantation (Intensive Care Med Exp, 2021).
Other support available
ESSIG supports trainees and registrars with an interest in sustainability, resource conservation, or implementation projects in the ICU — including projects that can count toward formal CICM training requirements. This is not a dataset or registry; it is a standing offer of mentorship and project support from the group.
ESSIG members are available to discuss project ideas before you commit to a topic, and can convene a panel discussion with group members if useful. They are also interested in hearing about sustainability initiatives already underway at your hospital.