The Art and Science of Clinical Decision Making

Clinical decision making is an important skill physicians utilise in their daily work. It may not be well taught in medical school as a distinct process, but knowing how physicians think, come to a diagnosis or generate a list of differentials, and make a decision on testing or treating, is an important first step in figuring out how the process work and how to make the best decisions.
Take this real case as an example:
A middle aged man with a past history of hypertension and old stroke with good recovery presented to the ED after he was punched repeatedly in the face and head. There was no loss of consciousness nor amnesia, but patient complained of dizziness and 2 episodes of vomiting at presentation. There were no neck complaints or other injuries.
GCS was 15, there were bilateral periorbital swelling, abrasions on the face as well as minor scalp hematoma and lacerations. There was no neurological deficit and cervical spine was examined normal. CT head was reported as no acute intracranial hemorrhage or infarct and CT face showed right orbital floor and left medial orbital wall fractures. ENT and EYE doctors on duty both saw the patient, were happy with their evaluations, and patient was discharged with close followup in the SOC.
However, patient represented 36 hours later with persistent dizziness and 3 further episodes of vomiting. At the morning of presentation, wife found patient slightly drowsy and he also complained of pain around both eyes and generalised weakness. Right pupil could be seen and was reactive and brisk. Left pupil was obscured by eyelid swelling. Neurological exam was repeated, determined to be patient's baseline and neck remained supple. A repeat plain CT brain was performed, and reported with findings as before; there were no suspected intracranial bleeding, delayed or otherwise.
Should we now proceed to dispose of this patient as post-concussion syndrome with admission or observation? Or is there something else?
Post-concussion syndrome is a possibility here, and the diagnosis is made based on a heuristic known as 'pattern recognition'. It doesn't require much thinking, and we draw upon our past experience or exposure to similar cases or patterns of presentation.
In this case, a patient presenting with minor head injury as the initial problem, now re-presenting with probable neurological complaints of dizziness and vomiting, must be having a condition related to the head injury in the 1st place (or so we assume).
But, this type of heuristic thinking may be fraught with certain biases, like anchoring bias and premature closure.
So, the ED team switched to another type of heuristics, this time using analytical thinking. We simply sat down and thought about the other differential diagnoses that were possible with this patient's re-presentation, and worked through the patient's symptoms, signs, and probabilities for each of the differential. Of course, we had a bit of help from Google's friendly search and Uptodate.
Not surprisingly, one of the differentials came out as a prime suspect. The gummed up left eye of this patient, in which we could not see the pupil, had a hazy cornea. We called EYE to come put a tonometer on patient's left eye as we suspected traumatic secondary glaucoma. The pressure in patient's left eye measured 80 mmHg. The diagnosis was made, and patient was immediately started on eyedrops and azetalomide.
If we had admitted the patient as before, he might not have the diagnosis made until many hours later when the respective specialties perform the reviews as inpatient. Who knew if the patient's sight might or might not have been compromised.
Therefore I urge all readers to read the following article on "The Art and Science of Clinical Decision Making", examine your thought processes and clinical decisions as you continue  in your daily practice. You will be surprised and amply rewarded.