Categorise or die


You’ll probably hear this one a lot. It makes sense. There are an entire universe of conditions that can cause a lot of overlapping symptoms, so when the patient presents with one or two symptoms, it’s not massively helpful to list random conditions in the hope that one of them is right. Methodically working through what system is likely to be affected, what pathology might underlie the symptoms, and what from the history is likely to have set it off, will put you in very good stead, even if you don’t know the exact condition. Medical notes need a differential diagnosis of a few possible options anyway, and I was once told that some surgeons care more about how you classify than what you actually know…


Hat pin

“Get a hat pin”

A consultant I had for rheumatology carried a hat pin around with him. I asked him why, and he said, “It’s useful for loads of things, testing blind spots, sensation, colour vision, instruction following, small object perception… and it holds your hat on” 

Basically this point is to say you should build up a set of tools that you like to use, and get used to using them. Usually people have their own stethoscope, pen torch, some have tendon hammers. Whatever you like to have on you, get used to using it, and get proficient.


Muscle, Skin, Joint

Nerves crossing a joint supply:

  • Muscle
  • Skin
  • Joint

This is a reminder that whenever doing any examination of a joint, you must consider any nerves that are crossing that joint, and what they supply. Especially important for situations such as a dislocated shoulder, where you must test sensation over the deltoid to see if the axillary nerve has been damaged prior to relocating the shoulder, or you could have a sizeable lawsuit in your lap.


What else?

What else is there?

Rather than a morose outlook on life itself, this nugget was a great little for when you think you have worked out what the problem is.

What other problems or lesions are there?

Congenital anomalies especially often travel in pairs or groups. Trauma patients may have multiple fractures that aren’t immediately obvious because they’re on so much ketamine. Chest X rays very often have multiple lesions, and it is tempting to stop searching when you think you’ve found one lesion (satisfaction of search is an entire area of research into radiological error!) A fractured rib often has a pneumothorax in its wake!

Don’t assume anything, ever, but especially don’t assume that the problem you have found is the only problem.


Say what you see


When clerking a new patient, taking the history and performing an examination, it is very tempting to try and work out what pattern the patient fits into best – making your own diagnosis is the whole point of doing medicine, and of course, is good practice.

However this risks a tunnel-visioned approach that can skew your judgement further down the line. As a junior, your job is to provide the senior with all the information they need to make a diagnosis. If you have your own ideas about what is going on, by all means mention it afterwards, but don’t remove or alter any information because it doesn’t fit your diagnosis.

Instead, get really good at simply describing lesions and signs. Not only will this make you better at differential diagnosis yourself, it will also make your consultant happy, as they can then confidently make the diagnosis using your description, rather than assuming you got it right first time.


Go for a swim

Immerse yourself.

The best advice I ever had at medical school – spend as much time as you can just being around patients on the wards. Books have lots of wonderful information in them, but they’re a very ineffective way of learning. as the material is abstract with no experiential hooks for your brain to fish them out of your memory. However the first patient that you see with rheumatoid arthritis is going to show you at least five things about the condition that you didn’t know before, and you won’t have to revise them ever again because you were there.

Also swimming is good for you health, do that too.

Look at the normal side first


We are blessed with two of many aspects of our anatomy – legs, eyes, nostrils and nipples to name the four most important. (maybe not…)

This comes in very handy when trying to identify anomalies, as one can simply compare the pathological side with the normal one. Look at the normal side first in order to get a reference – what’s normal for the patient – and to steady your nerves a little in exams, then examine the side of interest and see how it compares.


Use your nut

Your clinical opinion is always more reliable than a test.

More precisely, the clinical opinion of your consultant. A low CRP does not mean no infection. A raised D-dimer does not mean a definite pulmonary embolism. A result that doesn’t fit with your clinical suspicion should make you question your suspicion, but should not override it. Machines go wrong, more often than they should, and it’s your job to take the information in the context of the clinical picture. 

If the patient has a potassium of 12, and they’re having a sandwich, they don’t have a potassium of 12, your machine has gone wrong.


Ask your patient what they want

You as the expert clinician with multiple years of dedicated training under your belt may be incredibly confident that you know what the patient needs, but they know what they want.

Usually what your patient wants is rapid, painless recovery, which you usually aren’t able to give them, however just being able to express it to someone who will listen already makes your patient feel better, and may help you figure out what the problem is…

“Is there anything in particular you’d like us to do for you?”

“I’d like something for this rash” 

*shows me rash I had completely missed*


Reassess, reassess, reassess

Medicine is an evolving, mobile and fluctuant creature. It is never finished, and neither is your assessment of your patient.

What it is not: See patient, diagnose, start treatment

What it is: See patient, form likely differential, begin most useful treatment, reassess, tweak and alter treatment to optimise response, reassess, reassess, reassess!

The old ABCDE assessment is really:

A, try something, A,B, try something, A, B, C, try something, A, B, C….. etc

If in any doubt, go back and start again.

Inspection, inspection, inspection

One of my favourite things about medicine is that you can get your Sherlock on and deduce vast amounts of information without uttering a word. Look at the surroundings. Look at your patient. Do they look unwell? What have they got with them? 

I was once asked by an incredibly charismatic neurosurgeon to examine a child who had had a large tumour removed from their occipital lobe, without asking the patient any questions. I plunged into the best neurological exam I could summon, trying desperately to remember the various aspects of tone, power, reflexes, sensation, and by the end of my seven minute performance I genuinely thought I had done a pretty good job. 

“How old is your patient?”


“I wasn’t allowed to ask – I don’t know”

Look around”

The patient’s bay was full of birthday cards, presents, cakes and balloons all with the number 5 gleaming mockingly at me… I hadn’t seen a single one.

Open your eyes. Look around, there’s lots to see.

Stop the bleeding

This one is a little less silly than it sounds. Medicine is getting all the more complex as technology advances and we get better at testing for different conditions. This means a lot of people pay excessive attention to the numbers, and not to their patient, often forgetting the basic principles especially in an emergency situation.

Don’t forget the basics, and use your common sense. Keep the blood on the inside, keep the air moving in and out and your patient probably will survive until someone can help you.