1. What is clinical reasoning and why is it important? (A whirlwind tour!)

2. What is ‘clinical reasoning’? Clinical reasoning describes the thinking and decision-making processes associated with clinical practice

3. It is a clinician’s ability to … a)make decisions based on the available clinical information that includes history, clinical examination and (sometimes) test results – often against a backdrop of clinical uncertainty b)formulate and communicate these effectively with the patient and/or their carers

4. Clinical reasoning Clinical skills (history, examination, communication) Use and interpretation of diagnostic tests Understanding cognitive biases and human factors Critical thinking (metacognition) Patient centred evidence-based medicine Shared decision making

5. Why is clinical reasoning important? Diagnosis is wrong 10-15% of the time Diagnostic error is more likely to lead to harm than other types of error Two thirds of the root causes of diagnostic error involve errors in reasoning – most commonly when the available data is not synthesised correctly

7. Causes of diagnostic error No fault errors System failures Human cognitive error 1.Knowledge gaps 2.Misinterpretation of diagnostic tests 3.Cognitive errors and biases

8. Understanding diagnostic tests

9. There’s no such thing as a perfect test Tests can lie! All tests are affected by the following: Normal values Factors other than disease which influence test results Operating characteristics Sensitivity and specificity Prevalence of the disease in the population

10. Sensitivity and specificity Sensitivity is the ability to detect true positives Specificity is the ability to detect true negatives No test has 100% sensitivity and specificity! DiseaseNo disease Positive testA (True pos) B (False pos) Negative testC (False neg) D (True neg) Sensitivity = A/(A+C) x 100 Specificity = D/(D+B) x 100

11. The probability of a disease depends on the clinical (pre-test) probability plus the sensitivity and specificity of the test An elderly lady is admitted following a fall. She had hurt her left hip and was unable to weight bear. On examination, the left hip was extremely painful ++ to move. Her X-ray (shown) is normal. Is there a fracture?

12. The most fundamental principle in clinical decision making is that the interpretation of new information depends on what you believed* beforehand

13. ‘Prior probability’ and ‘posterior probability’

14. Small group work - quick quiz!

15. Prevalence of disease If a test to detect a disease whose prevalence is 1:1000 has a false positive rate of 5%, what is the chance that a person found to have a positive result actually has the disease, assuming you know nothing about the person’s symptoms and signs? (45% of Harvard doctors said 95%)

16. Answer PresentAbsentTotal Actual disease19991000 Test +15051 Test -0949 False positives 50/1000 But only 1/1000 actually has the disease Means the chance of a positive result with disease = 1 out of 51 or 2% The importance of understanding prevalence (or denominator).

17. Predictive values Predictive values are the combination of sensitivity, specificity and prevalence Sensitivity and specificity are characteristics of the test – the population does not change this But we are interested in the Q, ‘What are the chances that a person with a positive test result truly has a disease?’ DiseaseNo disease Positive testA (True pos) B (False pos) Negative testC (False neg) D (True neg) Positive PV = A/(A+B) x 100 Negative PV = D/(D+C) x 100

18. Example of an imperfect test for something that is not very common … A cab was involved in a hit-and-run at night. Two cab companies operate in the city, the Green and the Blue. 85% of the cabs in the city are Green and 15% are Blue. A witness identified the cab as Blue. The Court tested the witness under the circumstances that existed on the night of the accident and concluded that the witness correctly identified the colour 80% of the time. What is the probability that the cab was actually Blue? (The most common answer is 80%)

19. Answer GreenBlue Actual8515 Witness + 68 (80% of 85)12 (80% of 15) Witness - 173 12/100 times the witness will correctly identify a Blue cab as Blue 17/100 times the witness will incorrectly identify a Green cab as Blue There is therefore a 12+17=29% chance the witness will identify the cab as Blue This results in a 12/29 or 41% chance that the cab identified as Blue is actually Blue. Base rate neglect (prevalence neglect)

20. A test result by itself is not the answer (unfortunately patients don’t know this!) Tests must be interpreted in the light of CLINICAL PROBABILITY You must also know something about the CHARACTERISTICS of the test in question And if the PREVALENCE of the disease is very high or very low in the patient’s group – this affects the predictive value of the test

21. Cognitive biases and human factors

22. A bat and a ball cost £1.10 in total. The bat costs £1 more than the ball. How much does the ball cost?

23. Cognitive biases are subconscious errors that lead to perceptual distortion, inaccurate judgement and illogical interpretation of information, prevalent in everyday life: 'To err is human'.

24. Cognitive biases Social biases Memory biases Decision making biases Probability / belief biases

25. Confirmation bias

26. Small group work – case history

27. Subconscious errors Anchoring When we fix on a particular bit of information, leading us to think in a constrained way Confirmation bias Tendency to look for confirming evidence to support our initial hypothesis rather than looking for disconfirming evidence to refute it Diagnostic momentum Tendency for a particular diagnosis to stick despite lack of supporting evidence Search satisficing From the words ‘satisfy’ and ‘sufficient’ - when we stop searching because we have found something that fits or is convenient, instead of systematically looking for the best alternative

28. Diagnostic momentum Tendency for a particular diagnosis to stick despite lack of supporting evidence ‘Like a boulder rolling down a mountain, the diagnosis gathers momentum, crushing all else in its path’ Usually involves several intermediaries, including the patient Often starts as an opinion, not necessarily medical, and passed with increasing certainty from one person to the next Diagnostic labels become particularly ‘sticky’ once a patient has been seen by a consultant Wife: ‘I’m worried you’re having a heart attack’ Patient: ‘It feels like I’m having a heart attack’ Paramedic: ‘52-year-old male with possible ACS’ Nurse: ‘You know that man with ACS in cubicle 12?’ Doctor: Documents ‘possible ACS’ in notes

29. ‘Bias should be considered a normal operating characteristic of the human brain – biases are everywhere and have the potential to influence almost every decision we make.’ Croskerry P. Bias: a normal operating characteristic of the diagnosing brain. Diagnosis. 2014; 1: 23-7

30. The human brain is wired to: – Jump to conclusions – See patterns that do not exist – Miss things that are obvious Our estimate of probability is also poor (as we have already seen)

31. FINISHED FILES ARE THE RESULTS OF YEARS OF SCIENTIFIC STUDY COMBINED WITH THE EXPERIENCE OF YEARS

35. What is going on here? Human factors – ‘to err is human’ Type 1 thinking

36. Human factors Human factors training encompasses: – Understanding error: ‘Swiss Cheese’ and the limitations of human performance – Creating work environments that make it easy for people to do the right thing – How to communicate safely in teams All pilots have to pass a human factors exam, but many clinicians have a poor understanding of human factors

38. The science of decision making Cognitive psychology deals with human thinking, reasoning and decision making Dual process theory describes how the human brain has two distinct types of processes when it comes to decision making (supported by psychology, fMRI studies etc.) – intuitive and analytical (type 1 and type 2)

39. Dual process theory System 1 Intuitive, heuristic (patterns) Automatic, subconscious Fast, effortless Low/variable reliability Vulnerable to error Highly affected by context High emotional involvement Low scientific rigour System 2 Analytical, systematic Deliberate, conscious Slow, effortful High/consistent reliability Less prone to error Less affected by context Low emotional involvement High scientific rigour

41. DIAGNOSTIC ERROR EXTERNAL FACTORS Distractions Cognitive load Decision density Time pressures Ambient conditions Insufficient data Team factors Patient factors Poor feedback EXTERNAL FACTORS Distractions Cognitive load Decision density Time pressures Ambient conditions Insufficient data Team factors Patient factors Poor feedback INTERNAL FACTORS Knowledge Training Beliefs/values Emotions Sleep/fatigue Stress Affective/physical illness Overconfidence Risk-taking behaviour INTERNAL FACTORS Knowledge Training Beliefs/values Emotions Sleep/fatigue Stress Affective/physical illness Overconfidence Risk-taking behaviour COGNITIVE ERRORS/BIASES Use of intuitive (system 1) decision- making processes COGNITIVE ERRORS/BIASES Use of intuitive (system 1) decision- making processes Factors increasing the likelihood of diagnostic error

42. Personality and context Personality type and other individual characteristics (e.g. experience/specialty) influence decision making Decisions are also made in context (noise, interruptions, sleep deprivation, multi-tasking, unfamiliar problem etc.) ‘Comfortably numb’*

43. ‘Good doctors are not those who don’t make mistakes; good doctors are those who expect to make mistakes and act on that expectation.’ James Reason

45. Further resources www.clinical-reasoning.org www.internalmedicineteaching.org