Human Factors, Decision Making and PSIRF: What Dobble Can Teach Us About Safer Healthcare

What a family card game can teach us about cognitive strategies, work-as-done and safer system design in healthcare, as for NHS and independent healthcare providers implementing PSIRF, understanding Human Factors and decision making is essential for improving patient safety. The way people assess information, prioritise attention and adapt to changing conditions often determines whether systems remain safe, resilient and effective.

It’s the summer holidays, which means the eternal challenge of entertaining the children while still trying to get some work done: the usual Sisyphean task. This morning, we took on Dobble. There are a few ways to play it, but what became interesting, because ergonomists can turn pretty much anything into a study of human behaviour, was the different shortcuts we each adopted. What started as a family card game quickly became a small lesson in decision making, Human Factors and the gap between how a task is described and how people actually get it done.

Same game, different cognitive strategies

Version 1: there’s a single card in the middle and everyone starts with an equal number of cards. The aim is to spot the one symbol on your card that matches a symbol on the centre card. Every pair of cards has exactly one matching symbol, which I think is pretty genius, but that’s for another post. You call out the match, place your card on top of the centre pile, and that card becomes the new reference card.

Version 2: this one is almost the opposite. Each player starts with a single card and the remaining pile sits in the middle. You call out the match between your card and the top card, take that top card and put it to one side, but keep your original card throughout. Whoever collects the most cards wins.

Each card has 8 symbols on it.

So, in game one, I quickly adopted, and stuck to, a shortcut for assessing what was in front of me. In psychology, that is broadly what a heuristic is: a practical rule of thumb that reduces the effort involved in making a judgement, without guaranteeing the right answer. The symbols use a handful of base colours: black, red, blue, yellow, green and purple. Most cards seemed to have two or three symbols in a couple of colours, then one in each of the remaining colours. I used that as a quick way to scan, usually starting with purple, then yellow, then black. Light blue came last because, for me, it had the weakest contrast against the white background. Once I had narrowed things down by colour, I moved on to shape.

Post game, I chatted this through with Katie, who seemed to have adopted a similar approach, but favoured colours in a different order. (for clarity, there’s no colour blindness for us to factor in as to why we had preferences, but still interesting to see this as a difference between us nonetheless)

We then played, and inevitably discussed, game mode 2. Katie took a strong early lead, but I closed the gap in the second half and it became competitive. So what changed? It seemed that we had adopted different search strategies. Katie began with the same rule of thumb as before: filter by colour, then by shape. I used a different one: memorise four shapes on my now-stable reference card, look for those first, then add other distinctive shapes as the card became more familiar. My colour preferences still influenced the shapes I chose first (I have always been a big fan of purple). Katie explained that she switched to a similar strategy partway through, once she had become more familiar with the symbols on her card.

When we compared notes with the children, they described something more like a direct scan of the whole card, waiting for a match to “jump out”. Whether that was genuinely a different process, or simply a different way of describing it, is another question (and probably one that requires more than a conversation over breakfast).

What does Dobble tell us about Human Factors and decision making in healthcare?

Like most Human Factors households, these activities tend to end in some kind of Human Factors discussion. Even with shortcuts that appeared to improve our speed, we still made plenty of mistakes and good outcomes were not guaranteed. The competitive set-up encouraged all of us, children included, to push the boundaries because nobody wanted to come last. The game rewards speed and output, so it should not be surprising when those are the behaviours it produces. The more useful point, though, was that we were all trying to achieve the same outcome through noticeably different cognitive processes.

1) There’s always more than one way to get something done.

Same situation, same pressures, but different approaches, and each one changes where the effort and pressure peak and trough. The second game was particularly interesting for Katie and me. She was trouncing me early doors, to the point that I was thinking, “Boy, have I got this approach wrong.” Katie had speed out of the blocks; I was the slow burn. Once my approach established itself, however, it became a bit of an unstoppable train, particularly through the middle phase as Katie changed strategy. How it would have played out over a longer game is still up for debate, mind you.

2)      The approaches everyone applied, no matter how different they were, all made sense to whomever applied them.

They were all different, they all got the job done (at various efficacies over different timeframes), but they all made sense to us.

From a card game to medicines management: Human Factors in healthcare practice

Dobble is a useful metaphor for everyday work. A task can have one intended outcome without everyone reaching it by the same mental route. Take medicines management. Two experienced professionals might both complete a safe medicines reconciliation, but one works methodically through the record in date order while another starts with high-risk medicines, recent changes or the discrepancies that immediately look odd. Same answer, potentially. Very different process for getting there.

The same goes for completing a finance spreadsheet, reviewing a patient safety incident, preparing a clinic, handing over care or checking equipment. One person follows the formal sequence. Another builds a pattern from experience. Someone else starts with the exceptions because that is where they expect the risk to be lurking. A fourth uses visual cues, memory prompts or a locally created workaround. The finished output might look identical. The thinking behind it almost certainly is not. It often makes me think of those times in my career when a mentor has said to me 'have you to learn the rules before you then learn where to bend them'. Which fundamentally hits at the heart of 'work as imagined', or 'work as described' don't meet the needs of the messy reality that is 'work as done'.

we often overlook that on the surface, a process can look reliable while quietly depending on expertise, familiarity, informal checks or adaptations that never made it into the written procedure. Simply telling people to follow the process misses the point if we have not first understood the task. How do people spot the important cues? What do they do when several things compete for attention? Where do they make trade-offs? And what helps them get it right?

So often, I find myself coming back to the core principals of Human Factors Integration when working with organisations trying to move forward from what has felt like the eternal issue.

Work-as-imagined and work-as-done

This is why user engagement matters, especially when designing ways of working. It also brings to mind Hollnagel’s distinction between work-as-imagined and work-as-done. My imagined version of the Dobble task was different from everyone else’s. Any neat account I wrote down beforehand would probably have been different again from what actually happened. Then the messy reality kicked in: as Katie demonstrated, people adapt as conditions and familiarity change. Would that adaptation ever have appeared in the process description in the first place?

How often, especially when designing ways of working, do we take the time to engage properly with the people who will actually use them? And how often do we settle instead for a nice, neat account of how we want the process to sound?

Why Human Factors matters for PSIRF implementation and patient safety

The Patient Safety Incident Response Framework, or PSIRF, asks healthcare organisations to take proportionate, systems-based approaches to learning and improvement. That means doing rather more than checking whether somebody followed a policy. We need to explore how the conditions of work shaped attention, judgement, decisions and actions, and understand why the approach taken made sense at the time.

Different cognitive strategies are not background noise in a patient safety investigation or learning response. They are part of the system we are trying to understand. Staffing, task design, equipment, interfaces, workload, interruptions, culture and governance all affect which cues stand out, which shortcuts become attractive and which decisions are realistically available. Human Factors gives us a way to examine those interactions without collapsing the explanation into “they did not follow the process”.

For NHS and independent healthcare organisations, this is not just an interesting academic point. Better PSIRF implementation, safer medicines management, stronger clinical governance and useful improvement actions all depend on understanding real work. Design an intervention around only the neat, imagined version of the task and it may add friction, create another workaround or, worse, remove something that was quietly helping the system remain safe.

Designing healthcare work for how people think

Good system design does not require everybody to think in exactly the same way. It makes the intended outcome clear, provides useful cues and controls, supports different levels of experience and reduces the chance that a predictable human limitation becomes a route to harm. Crucially, it also gives people room to explain what they actually do, including the adaptations that keep services functioning when reality refuses to behave like the flowchart.

That is the real value of applying Human Factors in healthcare. It shifts the question from, “Why didn’t they do what we expected?” to, “What shaped the decision that made sense in that context, and what can we change to make safer performance easier next time?”

There are several people with colour-vision deficiencies in my family, although I don’t have that additional factor to manage myself. It would be interesting to understand how different visual needs shape the strategies people adopt, and it is a useful reminder that a method that feels obvious to one person may be unavailable or ineffective for someone else. Whether we are designing a card game, a medicines process or a patient safety response, the people doing the work are not an inconvenient variable. They are the source of the insight we need.

Human Factors, PSIRF and governance support

At Promethean HD, we support NHS and independent healthcare organisations with PSIRF training, Human Factors education, patient safety consultancy and proportionate governance improvement. The focus is practical: understanding work-as-done, strengthening learning responses and designing improvements that have a fighting chance of working in the real conditions of care. Getting it right by design, not luck.

And on that note, I had better start defining the scope of “Dobble: a high five for heuristics”. Sample size: four. Peer review: unlikely.