Clinical reasoning errors

Traps for the unwary - common reasoning mistakes made by junior residents

• Not accounting for the patient's unique vocabulary in determining what is being said e.g. 'dizzy' meaning lightheaded and not vertigo, 'sudden' meaning occurred today versus abrupt and severe,
• Leading the patient e.g. 'so the pain was crampy?' when the patient meant fluctuating in intensity
• Not identifying a major change in the chronology of the illness and symptoms particularly when there is an acute deterioration (e.g. ischaemic stroke with haemorrhagic transformation)
• Not maintaining a broad enough differential diagnosis
• Moving on to an investigation strategy before even thinking about the differential diagnosis
• Ordering tests but not anticipating what to do with unexpected results
• Seizing upon the first possibility that matches the clinical information and taking a cursory history (see The Eureka Syndrome)
• Not looking for a plausible cause for the patient's condition (e.g. 'food poisoning' despite sharing the same food as other asymptomatic people)
  An inadequate understanding of pathophysiology or disordered anatomy and symptomatology e.g. syncope due to TIA
• Not treating the cause (e.g. transfusion for anaemia but not treating the slowly bleeding ulcer)
• Not making a proper assessment of the severity of the condition resulting in delayed treatment or incorrect disposition (e.g. asthma imminently going into respiratory failure)
• Not reacting promptly when there is a change in the patient's condition or inadequate response to treatment (e.g. hypotension following GTN infusion for ACS - ?cardiogenic shock) 
• Assuming there is only one problem going on when there may be multiple (e.g. CCF and COAD)

"You see only what you look for; you recognise only what you know"

• This quote from the famous neuroradiologist Merrill Sosman is equally applicable to clinical reasoning as it is to radio-diagnosis.
• Several reasoning errors are commonly committed by even experienced clinicians.   The most common one goes along these lines.....
  "When you hear hoofbeats.  I think horses" - In many cases this is true but occasionally you will meet a zebra.  Often there are features that suggest it isn't a horse but the clinician neglects to make a search for these and terminates the assessment and analysis (premature closure).  
• On the other hand, the novice clinician sees zebras everywhere; lacking an intimate knowledge of the range of diagnostic possibilities or the practical experience to recognise common disease patterns.  
• Both of these errors reflect a haphazard system of approaching an undifferentiated complaint.  
• An example is a patient who presents with upper abdominal pain and the doctor only considers GIT causes and not cardiac or respiratory ones.   The prudent historian will ensure that such situations she would include a screening cardiac and respiratory assessment to identify this possibility.
• A solid grounding in pathophysiological mechanisms and anatomical relationships goes a long way to developing a systematic and meticulous approach to the diagnostic process.   

The Eureka syndrome

• Is the epidemiology of the disease consistent with the age, sex and demographics of the patient e.g. Tertiary syphilis in a 25 year Western European native.
• Is the time course of the symptoms consistent with the known pathogenesis and natural history of the condition (see The chronology suggests the pathology) e.g. 2 months of occasional PR bleeding - ?acute food poisoning
• Is the time course of the clinical data consistent with a known complication of the disease e.g. New ST elevation 3 days post MI - ?ventricular aneurysm
• Does the patient possess known risk factors for acquiring the disease e.g. Aortic aneurysm in a young person with no history of hyperlipidaemia, diabetes, smoking, hypertension or family history of vascular disease
• Are the physical findings and investigations consistent with the diagnosis e.g. pericardial rub and normal cardiac enzymes in acute MI
• Can the clinical features be equally explained by an alternative diagnosis e.g. pleurisy vs pulmonary embolus
• Does any of the clinical information contradict what is known about the disease e.g. normal lung function tests and emphysema
• Is the condition responding to the treatment in the time frame that is expected e.g. no response to a course of antibiotics for 'ear infection'
• Can all of the observations be explained in mechanistic fashion (i.e. deranged anatomy and physiology, homeostatic compensatory mechanisms, progression of disease, symptoms and signs, investigation results, response to treatment) e.g. massively swollen leg with normal leg ultrasound - ?small non-visible clot