Testing the virtual clinic

Have you ever sat in a GP clinic wondering if a symptom is something serious?

That consultation can involve carefully piecing together clues – your symptoms, test results and history – to build a fuller picture of whether something needs urgent attention.

Early detection can save lives, and digital health innovations are helping to ensure the decision-making process works at its best.

AI scribes can reduce administrative tasks, keeping doctors focused on patients instead of screens.

Decision aids – digital tools that can automatically flag rare risks by connecting patient information with the latest research – can spot factors that a busy GP may not be aware of, such as early cancer markers. In theory, this all means better care and better communication.

But in busy clinics, even a well-designed tool can stumble if it doesn’t fit the workflow.

That’s why at the University of Melbourne Centre for Digital Transformation of Health, we built the Validitron SimLab – a purpose-built clinical simulation lab where new digital tools are tested in realistic scenarios before reaching real patients or staff.

Healthcare workers, trained actors and patient advocates participate in simulated consults designed to mirror real-life care in a safe, controlled space.

Through simulation, we can observe how tools work in practice, identify barriers early, and refine the design, so that technology helps – not hinders – patient care.

It’s all about making digital health robust, reassuring and ready for the real world.

Testing is key to safe, digitally-enhanced care

Thousands of digital health tools exist, yet uptake in clinics remains low when they disrupt workflow, add clicks, or feel like a “black box”.

The Validitron SimLab is a purpose-built clinical simulation lab where new digital tools are tested in realistic scenarios before reaching real patients or staff. 

The potential is enormous – particularly for early detection, where timely prompts at the first point-of-care (usually a general practitioner or GP clinic) can make a life-changing difference. But those prompts must be trusted, usable and relevant.

Clinical simulation lets us test and refine these innovations safely and efficiently.

The Validitron’s exam rooms, GP workstations and integrated electronic medical record systems recreate a real GP clinic or hospital environment.

Then, our recording and observation facilities allow multidisciplinary experts to study a digital health tool’s usability and workflow.

These teams bring together expertise in human factors (how people interact with systems, and where errors can creep in), user experience (UX), digital implementation science, equity, codesign and patient and community involvement.

Testing cancer risk alerts in practice

Early cancer detection is critical, yet symptoms like unintended weight loss can be hard to interpret as they can be caused by a wide range of issues, including depression, chronic illness, or even poverty.

We tested a clinical decision support system (CDSS) embedded in GP electronic records to flag cancer risk when patients presented with unintended weight loss.

In our simulated consults with GPs, trained actors and cancer survivors, we observed how the tool affected GP decision-making and communication.

Before testing, GPs told us they thought the tool could support their reasoning and communication with patients.

But for them to actually use it, it needed to blend seamlessly into their workflow, work within short consult times and feel trustworthy.

Simulation allowed us to identify usability and workflow issues that could and should be fixed before real-world use.

For example, whether the tool could disrupt communication between GP and patient, highlighting a need for communication-focused training before implementation.

By catching these issues early, when changes are simple and inexpensive, we can streamline a tool’s path from prototype to practice.

Flagging critical but rare pancreatic cancer risks early

We also tested a tool designed to prompt GPs about pancreatic cancer risk factors.

Pancreatic cancer can be difficult to catch early due to vague, nonspecific symptoms and low prevalence, and only 14 per cent of patients survive five or more years after diagnosis.

Through simulation, GPs said the tool was easy to use, unobtrusive and helpful in prompting appropriate investigations.

They also highlighted important cautions: to avoid overtesting in very low-risk cases, to minimise unnecessary patient anxiety, and to ensure every suggestion is backed by clear evidence.

Simulation is especially important for AI-enabled tools, which evolve quickly and must be validated just as quickly. 

They preferred step-by-step investigation pathways that allow them to escalate only when needed.

To support this, the tool needs to match prompts to risk level, present stepwise options, link directly to guidelines and clearly outline the benefits and downsides of each recommendation.

Simulation highlighted these needs early, helping ensure the tool strikes the right balance – offering support without over-medicalising situations, and offering prompts without putting pressure on GPs.

Supporting, not replacing, clinicians

As new tools and AI systems enter healthcare, it’s important that they strengthen the clinician-patient relationship, not get in the way of it or replace it.

When designed well, digital tools act like an extra safety net – helping clinicians stay up to date with changing guidelines and reducing administrative load so they can spend more time with patients.

Simulation is especially important for AI-enabled tools, which evolve quickly and must be validated just as quickly.

Testing them in realistic, low-risk environments helps us see how they work in real practice, whether they’re safe and intuitive and how well they fit into day-to-day workflows.

This protects patients, builds trust and ensures these tools enhance practice rather than deskilling clinicians.

Testing protects patients, builds trust and ensures digital tools enhance practice rather than deskilling clinicians. 

From testing to practice

Everything we learn through simulation feeds directly into the next stage of development – guiding design iteration, informing rollout plans and shaping the training clinicians need to use tools confidently.

For partners across industry, research, health services and government, the Validitron provides a way to de-risk innovation.

By testing ideas early, building rapid prototypes, and generating evidence, we help decision-makers invest wisely and roll out tools ready for real-world use.

Ultimately, our goal is simple: to get fit-for-purpose tools into clinicians’ hands faster – and ensure they truly support the patients and healthcare systems they’re designed to enhance.

Interested in learning more about simulation? Take this 5-minute microlearning module

This article was written by Kit Huckvale and Javiera Martinez-Gutierrez of the University of Melbourne. It was published by Pursuit.