It's easier to intubate than to ask

It's 7pm on a Friday and the ward registrar calls, worried. A 78-year-old man admitted with pneumonia is deteriorating, and they think he needs to be intubated. From here, one path is far easier than the other. Accept the referral, find the bed, put in the tube and the lines. The patient is sick, but this is what we're trained for.

The other path is often more complex and starts with a question: who is this man, and what would a good outcome look like to him? He can't tell you. He's acutely delirious, exhausted, working hard just to breathe. His family need an interpreter, and you've been on hold for the last 15 minutes while they source one. His GP of twenty years closed two hours ago. The home team's registrar has never met him and would rather wait for their consultant. Somewhere in the hundreds of pages of his electronic record there may be a documented conversation about his values, but there's no easy way to find it. None of these obstacles is anyone's fault, and none of them is rare. They're simply what stands between you and the conversation everyone agrees you should be having.

Intensive care takes this seriously, and on most nights somebody chases the interpreter, tracks down a daughter, and pieces the man together before any irreversible decision is made. But it's messy, slow and hard. It consumes time on a Friday evening that was busy with MET calls and red buzzers and admissions. It only happens because a clinician decides this matters more than everything else demanding their attention. And herein lies the first problem: anything that runs on individual determination gives way exactly when the stakes are highest. The busiest night, the sickest patient, the most stretched registrar. I think a lot about how to make quality care easier to deliver and errors harder to commit. This scenario is a good example of where we're failing: it is easier to intubate than to ask.

---

The outcomes that matter

The central idea behind values-based healthcare is both simple and difficult to ignore: care should be judged by the outcomes that matter to the person receiving it, not just the ones that are easy for us to count (read: length of stay, CLABSIs). Which means we have to actually know what matters to the people we care for. Not assume it, and not reconstruct it afterwards from a distressed family staring at grandma, who said she never wanted to return to hospital, now sedated on a ventilator.

There's promising early work on tools for systematically eliciting and documenting patients' values, such as PerEmpo,¹ and it's exactly the right place to start. But a tool and a training session won't be enough, because training doesn't change the conditions the work is done under. The registrar who did the workshop still faces the same Friday evening: the delirious patient, the interpreter queue, and a workload with no room in it for an unhurried conversation.

---

What taking it seriously looks like

Now look at occupational violence, something hospitals have started taking genuinely seriously. Genuine EMR integration with nudges and forcing functions, dedicated teams of specially trained nurses, backed by proper clinical governance frameworks. Purpose-built behavioural assessment spaces in emergency departments. Duress alarms on the walls. If we put that much effort into values-based healthcare, what could it look like? A patient's goals, hopes and fears explored by dedicated teams whose rostered, funded job is this conversation. Facilitated by interpreters specifically trained in these interactions. Involving their GP, carers and loved ones through established MDT mechanisms. Documented natively in the EMR and surfacing automatically when it counts.

---

From values to a decision

Then there's another challenge. Suppose we build all of it, and every patient arrives with a genuinely good advance care directive. "My goal is to take my granddaughter to the playground. She's four. If I couldn't do that, or needed someone else's help with everyday activities, that wouldn't be a life I'd accept." Somebody still has to translate that into a decision about tonight's intubation, and that translation is only as good as our estimate of prognosis. I've written before about how noisy that step is. In one study, 146 intensivists estimated the survival of the same elderly patient being considered for ICU admission,² and their estimates ranged from zero to 95 per cent. Even when two intensivists gave similar survival estimates, they often made opposite admission decisions. The disagreement wasn't just about prognosis. It was about what the prognosis meant.

To truly start delivering values-based healthcare we need genuine structural investment in the task of eliciting our patients' goals, hopes and fears. We also need a more structured, consistent and evidence-based method for translating those values into treatment recommendations, so that the plan depends on who the patient is, not on which consultant answered the phone. None of these challenges arose because clinicians don't care. They care enormously. Most nights, caring is what carries this work. But caring is not a system. Systems produce their defaults. If we want values-based care to become one of them, it can't be easier to intubate than to ask.

---