On the interpretation of diagnostic tests: a common logical fallacy.


5%). Recoveries were similar when samples were sonicated before or after addition of perchloric acid. The compounds were determined in seven 250ML portions of a pooled whole-blood sample; the within-day CVs were 4.3% for 5-HT and 2.1% for TRP. Determinations made on 20-ML samples of pooled wholeblood (prepared by adding 10 ML of ascorbic and perchloric acid) gave CVs of 9.7% for 5-HT and 6.8% for TRP. Portions of a specimen of whole blood stored at -80 #{176}C were analyzed during six months; the CV was 10% for 5-HT, with no significant loss of 5-HT or TRP. Recently, we have found 5hydroxytryptophan (100-500 g/L added) to be an excellent internal standard. It is eluted at 5 mm, is completely resolved from 5-HT, and is analytically recovered from plasma and whole blood in yields intermediate to those of 5-HT and TRP. Whole blood and plasma samples analyzed to date have had values of <1 to 500 g/L, with a mean of -‘200 Mg/L. The absolute detection limit of 5 pg for 5-HT allows a concentration of 1 ng of 5-HT per milliliter to be measured with a 20-ML injection. The method is more rapid than present LC-F methods (2,3) for TRP in plasma and is simpler than the LCamperometric (electrochemical) methods (4,5) for 5-HT in plasma, and it allows both compounds to be easily determined in either plasma or whole blood. Because nearly all of the serotonm in blood is bound to platelets, measurements in plasma are complicated by difficulties in obtaining plasma with high and consistent yields of platelets. By measuring 5-HT in whole blood, one can assume that the entire platelet population has been assayed, and inter-laboratory results will be more comparable.

Cite this paper

@article{Dix1981OnTI, title={On the interpretation of diagnostic tests: a common logical fallacy.}, author={Douglas E. Dix}, journal={Clinical chemistry}, year={1981}, volume={27 5}, pages={776-7} }