L3 Reference intervals Flashcards
For determining RI, outliers are excluded in the post-analytical phase.
For example, unhealthy individuals that are discovered during the test are excluded.
In addition, statistical modality can be done to establish a central 95%.
Briefly explain the Gaussian distribution and non-Gaussian distribution, as well as examples of applying them.
- Gaussian distribution (normal distribution)
- parametric method (+/- 2SD)
- e.g. Na+, K+, Ca2+, Albumin - Non-Gaussian distribution
- Non-parametric method (2.5-97.5 percentile)
- e.g. Bilirubin, CK
- i.e. 5% of normal patients may be outside RI
What are clinical decision levels?
Give an example.
Range/threshold passing which a different medical decision is recommended.
e.g. TDM/ Cardiac troponin level
Which type of cardiac troponins are used in PWH?
What is the cut-off?
cTnT rather than cTnI
Higher cutoff, one-tailed: 99th percentile with <10% coefficienct of variation (CV = SD/mean x 100%)
> lower sensitivity but high specificity
What form of Vitamin D do we measure?
storage form - Total 25-OH vitamin D
bone-related disease is the only confirmed associated factor
What are the 2 limitations of RI?
- Analytical variation (SD1) - repeated measurements of a single specimen 20 times
- Biological variation (SD2) - repeated measurements at regular intervals in a healthy individual
What is the definition of a significant change from the previous result in repeated measurement of a specimen/ individual?
(Value2-value1) >2.77 SD
where SD = square root (SD1^2+SD2^2)
A hypertensive male with plasma K+ of 5.5 mmol/L on admission. The next day, plasma K+ is 6.2 mmol/L. Is there a significant change?
For K+, the biological and analytical variations are 0.10 and 0.19 mmol/L, respectively.
(3 marks)
- SD = sqaureroot (SD1^2+ SD2^2)
= squareroot (0.1^2 + 0.19^2)
= 0.22 - 2.77 x 0.22 = 0.61 mmol/L
- 6.2-5.5 = 0.7 mmol/L
where 0.7 > 0.61, thus there has been a significant change in plasma potassium
