Experimental feature

Listen to this article

Experimental feature

As a GP trainee, I’ve had to read many lugubrious books. In these books, history-taking (or “the consultation”) is condensed into various models that share a view of the patient interview as a highly structured and productive exercise, in which you ask “open” questions (“How are you feeling?”) and “closed” questions (“Have you ever had pneumonia?”) and do things such as “visit the patient’s agenda”, “elicit his ideas, concerns and expectations” and carry out the “chunk and check” (breaking down information into shorter clauses before finding out if your patient has understood it).

It often goes awry. Trainee GPs rotate through jobs in both hospital and primary care: during my GP surgeries, I film many of my consultations so that I can see where it’s going wrong. The patient can’t fathom what you want to know; or you don’t realise what she is trying to tell you; or you may literally not understand each other. In this last case you’ll have a translator adding an extra sentence to everything you say.

I recently worked in a hospital ward that dealt with new admissions. One morning I went to clerk a 70-year-old man who had arrived during the night. He had been roaming the ward introducing himself to everyone; the nurses called him Graham.

Graham was skeletal, and I could see that his weight loss was new as his pyjama trousers were pleated and secured with a safety pin but questioning him was proving unfulfilling: “What did you eat yesterday?”

“Something from the Co-op.”

“What kind of something?”

“Well, it came in a box that you put in the microwave. Four minutes. Well, it wasn’t what I was expecting. I wasn’t expecting that sort of a taste at all.”

“So what’s brought you in here?” I asked, switching to the kind of open question that’s meant to yield informational gold – although I can’t tell you how many patients will answer this with “a car!”, “an ambulance!” or “you did, Doctor!”

“Oh no, no. No, no. That is a complicated business,” Graham said, bucking convention. “Have you asked Derek? And have you, by the way, indoctrinated this banana? Do you really think I want to put my teeth into such a thing? I’m not at all sure about that. It’s not what one needs. Where are my slippers?”

Once you’ve slid far enough away from the normal conventions of information exchange, you obviously need to forget models and start thinking about delirium, or dementia, and check that your patient is actually oriented (“Who is the prime minister?” etc). But as the day wore on, and the antibiotics we had given Graham for a urine infection started to work, his delirium settled and he became more rational. He was still evasive, though: any attempt to summarise, signpost or establish a rapport was stymied by his obsession with unfriendly foodstuffs.

Most consultations do not fit neatly into a model – sane or insane, coherent or nonsense. The most lucid patients can come out with extraordinary non-sequiturs, and in fact so do the doctors. “Any difficulty breathing?” we ask, and then: “Do you have any pets?” Even apparent randomness has a logic somewhere. In Graham’s case, his obsessive fears eventually led to a brain scan, which revealed multiple metastatic tumours from undiagnosed lung cancer. Consultation models teach techniques that can be useful, but not the flexibility that you need to have truly meaningful conversations.

Sophie Harrison is a hospital doctor in South Yorkshire. Names have been changed to protect identity. This column appears fortnightly.

To comment on this article please post below, or email magazineletters@ft.com

Get alerts on FT Magazine when a new story is published

Copyright The Financial Times Limited 2019. All rights reserved.

Follow the topics in this article