01-02-22 - What is 'normal'?

Question 1

Learning Outcomes

Answer 1

• Explain how science and medicine reached their current understanding of what is a “reference” and what is considered “normal”
• Define what is meant by “norm referencing”, “reference intervals”, “reference limit”, “reference range” and “clinical decision limits”
• Define what is meant by “norm referencing”, “reference intervals”, “reference limit”, “reference range” and “clinical decision limits”
• Explain how to apply the principles of Bayesian thinking to recalibrate the “reference” or “normal” in your clinical and scientific reasoning
• Illustrate, using specific examples, the history leading to current medical and scientific understanding and practice
• Illustrate, using specific examples, where use of “normal” parameters would be appropriate or inappropriate to use

Question 2

What do norms tend to be?

What if the norm is ‘wrong’? What are 6 examples of this?

Answer 2

• Norms tend to be parameters derived from a study of a large set of individuals
• We assimilate information, find or create our own ‘norm reference’
• We then compare our ‘measured value’ to our defined ‘norm’ and act accordingly
• Sometimes the info, or population that we base our ‘norm’ off of is incorrect/inappropriate to compare with our own measurement

• Examples:

1) What happens if the sample wasn’t collected properly? –a mistake in the method
2) What if the populations studied are different? – South Africa/UK omicron
3) What happens if one piece of data overrides others unfairly? – our friend said it, so it must be true
4) What if the norm is based on falsehood? – MMR
5) What if our skin isn’t the same colour? – signs of disease missed
6) What if we know “this type of person” would never do “that”? – diagnosis discounted

Question 3

What does Bayesian Reasoning Involve?

Answer 3

• Bayesian reasoning involves incorporating conditional probabilities and updating these probabilities when new evidence is provided.
• This means that you can’t just get a theory and take it to be true if it fits the evidence.
• You need to also look at alternative hypotheses and see if they explain the evidence better
• Our problem is that when we have symptoms that are correlated with something bad, we immediately assume that it’s the bad thing that has happened, and panic.
• In that process we don’t consider alternate reasonings, and then do a Bayesian analysis.
• Is it a horse or a zebra?
• When you hear hooves, you think horses, not zebras

Question 4

What are 9 characteristics protected by the Equality Act 2010?

Answer 4

• 9 characteristics protected by the Equality Act 2010:

1) age
2) disability
3) gender reassignment
4) marriage and civil partnership
5) pregnancy and maternity
6) race
7) religion or belief
8) sex
9) sexual orientation

Question 5

What are 8 interpretations of normal?

Answer 5

• Interpretations of normal:

1) Determined statistically
2) Most representative of its class
3) Most encountered
4) Wild type: Most suited to survival and reproduction
5) Harmless “carrying no penalty”
6) Most often aspired to
7) The most perfect of its class
8) “Used to imply that the patient has no physiological derangement and/or that the distribution follows a Gaussian distribution” – where most people fall within 2 standard deviations of the mean of the curve

Question 6

What do reference intervals give?

What are the reference limits?

What is the reference range?

Answer 6

• Reference intervals gives the lower and upper extremes (x to y) – represents the possible fluctuations which can be seen within the normal range
• The reference limits are the upper and lower extremes of the reference interval
• The reference range refers to the difference between 2 values
• The reference range refers to the difference between these 2 values

Question 7

Where might we see two Gaussian distributions with no overlap?

What are the 4 limitations of this?

Answer 7

• We may see two gaussian distributions with no overlap in the same population e.g those with a disease and those without, with 95% of individuals falling within 2SD of the mean of each curve
• Limitations of this:

1) May (or may not) be totally segregated
2) The distribution (s) may be skewed
3) The curve may not follow a perfect shape
4) ‘Normal’ is considered +-2 SD from the mean, which is arbitrary (based on random choice) in this case

Question 8

What is an example of when 2 Gaussian curves may overlap?

Answer 8

• Two Gaussian curves may overlap e.g patients with and without a disease
• The test will usually identify most of those with and without the disease
• There will be results that don’t fit under the curve, and some that are at the extremes of the curve e.g false positives and negatives

Question 9

What is partitioning?

Why do we measure creatinine?

Why do we need to partition creatinine?

What are 2 issues that can arise in determining how to apply these reference values for your patient?

Answer 9

• Partitioning is a way of splitting numbers into smaller parts to make them easier to work with
• Creatinine is a useful measure of kidney function, but is also produced by muscle
• The larger muscle mass, the greater the amount of creatinine made
• Due to puberty and the large increase in muscle mass in men compared to women, we need to partition creatinine levels into male and female to ensure we aren’t making erroneous understandings of where we should put limits based on biological variability

• Issues that may arise:

1) Someone who goes to the gym a lot and has elevated muscle mass compared to the average person
2) Someone who transitioned genders