Assembling combinations logically
A doctor may not have thought consciously about which combination of findings were used to make a diagnosis probable. However, a diagnostic lead with a short list of possibilities can be used as a starting point to assemble such a combination of findings in a logical and transparent way.
Example of assembling combinations logically
If a patient presented with severe weight loss then in most cases it could be due to thyrotoxicosis, cancer somewhere, new onset of Type 1 Diabetes Mellitus, tuberculosis or severe depression. If urine testing showed glucose in the urine then this would occur commonly in diabetes mellitus but rarely in each of the other conditions, helping to ‘differentiate’ between the differential diagnoses. Therefore, in a group of patients with the lead of severe weight loss and the differentiator of glucose in the urine a high proportion would turn out to have diabetes.
Combining weak differentiators
A small portion of patients with weight loss and glucose in the urine would have thyrotoxicosis but if the patient had no hand tremor, normal bowel habits and no ‘eye stare’, then a combination of these negative findings would occur less frequently in thyrotoxicosis than in diabetes, making diabetes even more probable and thyrotoxicosis even less probable. If the patient is found to have a single high blood sugars over 11.1mmol/l when a blood sample is taken immediately, then the diagnosis is certain as this is one of the ‘sufficient’ diagnostic criteria for diabetes mellitus - a blood sugar of 11.1mmol/l at any time and glucose in the urine.
Supporting diagnostic impressions with transparent reasoning
If a transparent reasoning process gives the same answer as the doctor’s subconscious impression then it reinforces it. If the reasoning process does not support the subconscious process then it casts doubt upon its validity, suggesting that it should be reconsidered. It is therefore to the advantage of patients that doctors try to justify their subconscious ‘diagnostic impressions’ by being asked to explain them by using a transparent logical reasoning process.
© Huw Llewelyn 2016