Dr Sreeman Andole is a Stroke Consultant and Geriatrician at Buckinghamshire Healthcare NHS Foundation Trust.
A GP once asked me about a discharge letter I’d sent – and fair point, there was a gap in the details. But his real question wasn’t about the letter at all. It was about a choice I’d made only hours before: whether to start medication for a patient before the official care pathway said I should.
This is what happened. An older man – let’s call him Mr B – had his first ever witnessed seizure. He’d never had epilepsy before. He also had a Clinical Frailty Score of 8, meaning he needed help with almost all daily tasks; a DNACPR order in place; and written confirmation in his care plan that he wanted to stay out of hospital whenever possible.
NICE guideline NG217 sets out the standard approach for suspected epilepsy: generally, you don’t start anti-seizure medicine until a specialist has reviewed the case, unless specific exceptions apply. That makes sense for most people – it stops someone being labelled with a long-term condition and put on lifelong tablets after just one unusual event.
But Mr B wasn’t ‘most people’. If I’d admitted him to wait for a neurology appointment, he would have spent days in hospital – confused, out of his usual surroundings, and at real risk of having another seizure before anyone saw him. So, I started levetiracetam straight away in the emergency department and arranged the specialist referral afterwards.
This is the heart of what’s often called ‘realistic medicine’: guidelines are written around an average patient, based on research that mostly includes people who are younger, fitter and have fewer health problems than the person sitting or lying in front of you. Frail older adults are rarely well represented in the trials that shape these rules. Good clinicians don’t ignore guidance – they use it as a baseline and recognise when the patient in the room doesn’t match the patient in the flowchart.
This isn’t a free pass to do whatever you like. The moment you step away from the standard pathway, you have to be more careful – not less – about what you record, who you speak to, and what arrangements you put in place. I wrote down exactly why I’d made that call. I phoned his care home to explain, rather than just sending a letter. I made sure the neurology referral was sent before I finished my shift. Changing course without that kind of backup isn’t thoughtful care – it’s just taking unnecessary risks.
There’s another, harder point here too: my colleague was right to question it. The discharge letter I’d sent didn’t spell out my reasoning clearly enough – and that was my mistake, not a problem with the guideline itself. Deviating from official advice is only justifiable if you can explain your decision plainly, in writing, so that another doctor reading it with no prior knowledge of the case would understand – and agree. If you can’t do that, you probably shouldn’t have done it.
This tension – between what the rulebook says and what the individual needs – isn’t going away. If anything, it will only get stronger as our population ages and living with multiple long-term conditions becomes normal, not unusual. The 'typical' patient that guidelines were built for is becoming less and less like the people we actually treat every day.
So, I’d ask anyone reading this: when was the last time you set aside a guideline because it was in your patient’s best interests – and did you record your reasoning well enough that a colleague coming to it cold would see why you were right? I’d be really interested to hear other people’s experiences – especially the times when things didn’t go quite as smoothly as they did here.
