Healthcare uses many metaphors, but few help us speak honestly about power. Leaders often aspire to stronger patient partnership, yet real practice varies widely. In some settings patients shape strategy; in others their influence appears only through complaints or litigation. Co-production literature shows that although it is increasingly valued, practice remains inconsistent and often unclear about power, process, and purpose.

In our work at University Hospital Southampton, a large University teaching Hospital in the UK, one of our patient partners has developed a simple framework to help teams articulate where patients currently “sit” in relation to involvement in patient safety and improvement work. The aim being it’s memorable rather than diagnostic. A way to open conversations about patient involvement. The metaphor is a car with four possible seats: baby, teenager, passenger, and driver. Each represents a different relationship to the power, voice, and leadership patients have in health care improvement see figure.

Baby in the Backseat – The System Drives, Patients React
Many organisations still rely heavily on complaints, incident reports, or generic feedback as their primary “patient voice”. These mechanisms matter, but they reflect a reactive form of engagement. The co-production evidence shows that when patient involvement happens only retrospectively, the relationship becomes adversarial, and learning remains superficial. 

This is the baby seat: the organisation drives, and the patient is acted upon. It’s a position leaders should recognise not as a failing but as a signal, a sign that patients were not involved early enough in shaping decisions, defining outcomes, or co-creating solutions.

Teenager in the Backseat – Reactive and growing autonomy
Many teams genuinely want to involve patients but feel unsure how. Involvement becomes patchy, dependent on champions, time, or confidence. Studies of patient participation groups show similar variability, often shaped by deprivation, capacity, and organisational readiness. [Gobat et al, 2025]

This is the teenager seat: emerging autonomy but without consistent power. These patients may often feel frustrated that they aren’t being heard when they have much to contribute. The literature identifies limited infrastructure, varied inclusion, and selective use of patient insight as common barriers to meaningful involvement.

For leaders, the question becomes: what are the minimum conditions that allow patient involvement to flourish reliably? Often the answer is simple; build a trained pool of patient partners, clarify expectations in project templates, and make early engagement routine.

Passenger Seat – Strong Partnership & Co-driving
In the passenger seat, patients are invited into early conversations, their insights shape solutions, and their presence is expected rather than exceptional. But the organisation still sets the parameters, agenda, and pace.

Co-production models consistently highlight the importance of shared values, transparent communication, inclusive processes, and attention to diversity of voice. Narrative reviews indicate that although partnership models are growing, many still lack the deeper attention to power and agency required for full co-production. [Robert et al, 2024] 

This seat represents genuine involvement but also the limit of many current approaches.

Driver’s Seat – True Co-Production
The driver’s seat represents shared work and shared power. Patients help define the problem, shape data collection priorities, codesign solutions, and participate in governance. The healthcare coproduction literature describes this not as a luxury but as the natural form of service creation: health services are, by definition, jointly produced by professionals and those who use them. [Robert et al, 2024]

Leading co-production is complex work requiring investment in practices such as power-sharing, trust-building, communication, and co-implementation. But leaders who embrace this discomfort often find that early involvement prevents rework, accelerates adoption, and produces solutions more aligned to what matters. [Boam et al, 2026]

Not Every Journey Needs the Same Driver
One of the most helpful insights the model offers is that the driver’s seat is not always the right seat. In safeguarding, regulatory response, or emergency redesign, the expertise and accountability of clinical teams may need to lead. Conversely, tasks like redesigning letters, deciding meaningful outcomes, or shaping digital tools are often best driven by patients.

The goal isn’t to put patients in the driver’s seat everywhere. It’s to develop leadership discipline around asking the right questions:

• Which seat are patients in for this piece of work?
• Why this seat?
• What would be required to move one seat forward?

This deliberate positioning also helps address equity. Research shows that patient leadership opportunities are less available in more deprived communities, risking amplification of inequalities unless organisations consciously broaden access. [Gobat et al, 2025]

Leadership Behaviours That Make the Model Work
Across the coproduction literature, certain leadership behaviours recur: clarity about purpose, attention to power, inclusive recruitment, and early dialogue. Leaders can anchor these behaviours by embedding the four seat model into project initiation forms, committee papers, and governance reviews. A small icon or tickbox labelled “Baby / Teenager / Passenger / Driver” prompts teams to reflect on the nature of the partnership they intend to create. 

Building capability also matters. The evidence highlights the value of systems that reward contributions, offer training, recognise lived experience, and continually review progress with citizen partners. [Masterson et al, 2024]

And What Happens When the Car Drives Itself?
As artificial intelligence becomes woven into triage, diagnostics, communication, and scheduling, we risk entering a world of driverless care. Algorithms may begin to “steer” decisions traditionally made by clinicians or patients.

Co-production research emphasises that meaningful involvement depends on intention, dialogue, interpretation, and shared understanding, mechanisms that are essential for reciprocal engagement. If these elements are absent, automation may unintentionally push patients back into the baby seat: acted upon, not involved. [Batalden et al, 2016]

Leadership in this future state will require guarding against algorithmic paternalism. The challenge is not to reject automation but to ensure human in the loop governance where patients help define what AI is used for, how fairness is assessed, what outcomes matter, and when humans should override machine judgement.

A Final Thought
The four seat model is simple, imperfect, and deliberately nontechnical, but that is its strength. Leadership is often about finding language that helps people see what was previously invisible. When teams learn to ask, “Which seat are our patients in today?”, they begin to redesign not just services but relationships. And as health systems pursue coproduction and enter an era of automation, it may be more important than ever to make sure patients don’t lose sight of the wheel.

Authors

Dr Kate Pryde
Clinical Director for Improvement & Clinical Effectiveness, University Hospital Southampton (UHS)

Dr Christina Rennie
Clinical Director for Patient Safety, University Hospital Southampton (UHS)

Emlyn Marshall
Quality and Patient Safety Partner, University Hospital Southampton (UHS)

Declaration of Interests:
No interests to declare.