Diagnosis
At the end of your history, you should have a list of diagnoses in order of likelihood - this is called your differential diagnosis. This gets easier the more histories you take. But before you start putting them in order, you have to come up with the list. Unfortunately, this is often just rote learing although if you learn some basic clinical sciences (that's anatomy and physiology), it can be easier to remember symptoms:
Creating a list of possible causes from a symptom
Method A - System-Based
Right at the start of your history, you will get a symptom. Let's take chest pain. What you have to do now is come with all the causes of chest pain. Go...
Not the easiest thing in the world unless you know a little about how to do it (in which case, why the hell are you reading this, doctor/clinical medical student? Shouldn't you be doing something more useful with your time?). For chest pain I generally go by system. Look systems in the chest and see which can cause pain:
- Cardiovascular
- Respiratory
- Upper GI
- Musculoskeletal
- Neurological
Then, you put some diseases that fit into each of these categories that cause chest pain:
- Cardiological
- Respiratory
- Upper GI
- Musculoskeletal
- Neurological
- Thoracic radiculopathy
Method B - The Surgical Sieve
It is quite intimidating having a consultant staring at you, asking for a differential diagnosis of loss of proprioception in the left index finger. And the above system only really works if you're fairly familiar with the symptom and have a way of categorising the differential. The Surgical Sieve is for those symptoms where you're totally stuck.
Essentially it is an acronym with various types of disease. The one I use is:
- Vascular
- Inflammatory/Infectious
- Neoplasitc
- Traumatic
- Autoimmune
- Metabolic
- Endocrine
- Degnerative
- Iatrogenic/idiopathic
- Congenital
I have no idea what VINTA means and I'm pretty sure I just misheard a consultant telling us about the surgical sieve. Other popular ones are VITAMIN C and VITAMIN D - it doesn't matter. If your symptom is jaundice, come up with a vascular cause, an inflammatory cause, an infectious cause etc. and you'll do better than if you just randomly try and remember causes.
Closed Questions
This is the clever bit and which consultants are really good at. That's why they're consultants and get paid lots of money. And to think I've given away their secret on this webpage. Mwahahahaha! You have your list of possible diagnoses for chest pain (just because it's easy):
- MI
- Angina
- Pericarditis
- Pneumonia
- Pulmonary Embolism
- Hiatus Hernia
- Mallory-Weiss Tear
- Reflux
- Rib Fracture
- Costochondritis
So let's take. You then ask some closed questions about it that would make it an MI (how long have you had the pain? Where abouts is it? etc.). If you get answers that point to MI, all well and good. If not, you move to your next possible and start asking specific questions. The MI example is a bit rubbish because the questions are fairly broad but the answers are potentially clinically diagnostic.
It is important to make sure you exhaust the list. If you think somebody has angina because they have left-sided chest pain, that's fine. But if they have a fever and the pain is pleuritic but you don't ask, you haven't eliminated important negatives. Fever and pleuritic chest pain suggest pneumonia. The example is fairly obvious but make sure you've got everything and got a differential before you walk away from the patient.
Differential Diagnosis
When you asked the first open question about the main problems, you immediately should have come up with a list of potential differentials. For instance, diarrhoea would give you:
- Gastroenteritis
- Inflammatory bowel disease
- Irritable bowel syndrome
- Bowel cancer
- Metastatic disease
- Coeliac disease
- Hyperthyroidism
- Iatrogenic
- Idiopathic
If you're a first-year and haven't heard of some of this stuff, it's fine. If you're a final year or a gastro reg, yes, I've probably missed stuff but I'm trying to illustrate a point so shut up for a second.
You have to order your differential as you go through the history and generally come up with two or three possible diagnoses. For instance, blood in the stool would make IBD, cancer and gastroenteritis more likely but pretty much rule out hyperthryoidism and irritable bowel. A fever would make infection (and possibly drugs) more likely whilst making a IBS and coeliac less likely. The more questions you ask, the more accurate your list gets.
Eventually, you'll get a description of the diarrhoea, something like "2 day history of bloody diarrhoea associated with a fever and after trying to eat a raw chicken breast", which makes gastroenteritis the most likely. "2 month history of intermittent bloody diarrhoea" makes inflammatory bowel disease and bowel cancer more likely. Etc., etc... you get my point. You need to make a list with most likely stuff at the top and then act appropriately.