Erratum: Korf B. Pitfalls in the interpretation of molecular diagnostic tests. J Clin Lab Anal 10:368–374, 1996
- Publisher
- John Wiley and Sons
- Year
- 1998
- Tongue
- English
- Weight
- 6 KB
- Volume
- 12
- Category
- Article
- ISSN
- 0887-8013
No coin nor oath required. For personal study only.
✦ Synopsis
The calculation on page 372 referring to the risk of cystic fibrosis in a fetus with echogenic bowel and one detected cystic fibrosis mutation was done incorrectly. This fetus has a prior risk of having cystic fibrosis, given the finding of echogenic bowel, of 0.02 (as was correctly calculated at the top of the page). The probability that this fetus would have one detected and one undetected cystic fibrosis mutation if the fetus has cystic fibrosis is (2)(0.1)(0.9), whereas the probability of having one detected mutation if the fetus does not have cystic fibrosis is (0.04)(0.9) [i.e., the carrier frequency times the detection rate]. Hence the likelihood of the fetus having cystic fibrosis given echogenic bowel and this DNA finding is (0.02)(2)(0.9)(0.1) = 0.0036, and the likelihood of not having cystic fibrosis given these findings is (0.98)(0.04)(0.9) = 0.03528. The relative likelihood of cystic fibrosis is therefore 0.0036/[0.0036+0.03528] = 0.0925, about 9%. This is lower than the figure quoted in the paper, ~15%, but is still much higher than the risk that would be calculated based on the use of DNA data without taking the echogenic bowel into account. The overall point is therefore not changed, but the change in the risk should be noted.