Taking a history is a skill. It takes a long time to develop this skill. History taking is part of the process of clinical reasoning. I think of clinical reasoning of being the process whereby you compare the information you have at hand – from your history, examination and investigations – to that ‘encyclopaedia’ you have in your brain (and dare I say it – supplemented by the odd bit of googling) – and try to find the best match. The process of clinical reasoning is about pulling together your knowledge, and your diagnostic skills. For me at least, it was something that I only even really started thinking about in the final stages of medical school, and takes years to really get a grasp on. In the early stages of medical school I was much more focussed on building my “encyclopaedia”.
As a medical student taking a history can be a laborious, time consuming process, with long, meandering conversations including the weather, the recent form of the local football team, stories of Doris next door who has a young Labrador, and the state of the local park. Then a consultant can waltz past, ask four questions in under 30 seconds, and then tell you the diagnosis.
But don’t worry, you will get better! And all it takes is a little knowledge, and some practice. It’s all about how know what questions are relevant to ask in a given situation. And that comes with experience.
So, in my best almostadoctor bullet point list style, here are my top tips:
Almost nobody is a classical presentation
- Annoying isn’t it. You spend five years at medical school reading about classical presentations, then, when you get out there in the real world, nobody really fits. If they did, it would be easy. But they don’t. And it isn’t. You have to try to ‘best match’ them to a diagnosis. And you’ll get it wrong. Even consultants get it wrong. So don’t beat yourself up about it. Try to learn what symptoms will almost rule out a particular diagnosis, and which symptoms just make it more or less likely.
- For example, diarrhoea in a patient with abdominal pain doesn’t rule out appendicitis. But it makes it less likely. But now say, that they have bloody diarrhoea, which has been going on for 3 months, now appendicitis is now very unlikely – almost ruled out (but remember nothing is every 100% certain in medicine!).
Don’t decide on your diagnosis too early!
- Getting to your diagnosis isn’t a race. In fact, it’s not even the goal of your history taking. That’s right. The purpose of taking a history is NOT to get to the diagnosis. The goal of your history taking is to make a list of differential diagnoses so that you can focus your examination and investigations. This is the process of clinical reasoning. History taking is an important part of clinical reasoning, but it isn’t the whole story.
- There is a very good podcast on clinical reasoning – called IM clinical reasoning – if you want to learn more about practising this as a skill.
- It’s very, very easy to fall into the trap of only asking questions to cover one diagnosis. You hear the first two or three symptoms, and you think ‘Aha I know what this is! I’m a bloody genius! Can’t wait to show of to my Consultant’, and you quickly think of questions to ask that confirm your hypothesis, they fit, and then.. boom! “Yes! I’m a genius!” I’ve done it. I think we’ve all done it. Then you go and present to your Consultant. Expecting a pat on the head and a box of Quality Street. Then he says ‘and have they had any diarrhoea?’. And your heart sinks. Oh Shit. I don’t know. What shall I say? Do I guess?! Do I admit it? Do I quickly pull a ‘lightbulb moment face’ turn around with my finger raised, march back to the patient, ask them, and then return with the answer?! Then, the answer is not what you want to hear. Bollocks. Then it dawns on you. You haven’t even attempted to rule out if Mr. Jones has had any symptoms of diverticulitis, because you were far too busy congratulating yourself on being the next House.
- Remember tip no. 1. Ask questions that help to rule out or rule in possible diagnoses – and use this to build up a list of possible diagnoses in your head. The next step in clinical reasoning is then to work out which investigations help you to narrow down that list even further.
Let the patient talk (but not too much)
- I don’t know about you, but at medical school we had this very firmly drummed into us. Start with ‘open questions’. I usually start with something very ambiguous, like: “So, Mr. Bond, tell me what’s been happening?” (Top Tip. Don’t call all of your patients ‘Mr. Bond’.)
- Often, the patient will give you lots of relevant information in their opening few sentences. If this is a planned appointment, and even if its not, the patient has probably rehearsed in their head what they are going to say to you, when they see you. So let them get it all out. Quite often, they’ll probably say something that you would have forgotten, or that will jog your memory about the sorts of questions you should be asking. When you’re first learning to take a history, this can seem like a lot of information to hold in your head, because questions will pop into your mind all the time. You will latch onto something that you know lots about and what to head down that avenue. But resist the temptation! Let them talk. At least for the first minute or two.
- Don’t let them get off topic too much. A little is fine. Build a rapport. But if you’re losing control, bring them back to the matter at hand.
It’s all about pattern recognition
- You are a machine. You are trained to retain information. Like an encyclopaedia. You are trained on how to extract information from a patient. And you are trained on how to compare this information with the encyclopedia stored in your head, and to find the best match. That is your job. At some point in our lifetime, someone might invent a computer powerful enough to do this better than we do it at the moment. Maybe I will be that person, then, I can quickly hop onto my private yatch and retire to my new estate in the Bahamas. Because it will make me an absolute fortune. But in the meantime, I have a more modest goal. Trying to perfect my own history taking technique.
- Your encyclopaedia changes over time, based on your experience, and your knowledge, and other people’s knowledge, as medical science advances, and best practice changes. You try not to make the same mistakes twice.
- But just because your job description is best suited to a machine, doesn’t mean you have to be like a machine. The art of extracting the information from the patient requires personality. Rapport. Interpersonal skills. I just try to have a ‘chat’ with my patients. Sometimes this is easy. Sometimes this is more difficult. But I get the best results when we are both relaxed and enjoying the conversation. We are going on a journey towards the same goal together. They want to get better. I want them to get better. Everyone wants them to get better. So lets do it! Try not to fall into your ‘medical student voice’. Its impersonal and can be alienating. You know that voice. It’s the one you overhear your colleagues doing when they’re talking to patients.
- The pattern recognition part requires that you have enough knowledge about the topic to ask the right questions, and that you can match up the answers to the differential. So, there’s a lot going on. If you don’t have the knowledge, you can’t ask the right questions. If you have the knowledge, but you don’t ask the right questions, you don’t get the right answers.
So there we have it. Practice lots, and good luck!
Image from flickr. Used under a cc license.