Full text loading...
Principles of interpreting endocrine test results
/content/chapter/10.22233/9781910443866.chap4
Principles of interpreting endocrine test results
- Author: Peter A. Graham
- From: BSAVA Manual of Canine and Feline Endocrinology
- Item: Chapter 4, pp 22 - 28
- DOI: 10.22233/9781910443866.4
- Copyright: © 2023 British Small Animal Veterinary Association
- Publication Date: August 2023
Abstract
The interpretation of endocrine laboratory test results can be complicated by the interrelatedness of endocrine systems with other physiological systems and by environmental factors. However, with an understanding of endocrine physiology, the impact of non-endocrine factors and the limitations of diagnostic tests, it is possible to confidently confirm the presence or absence of endocrine disease. This chapter covers diagnostic sensitivity and specificity, the need for caution when interpreting results near reference limits and cut-off values, and the effects of endocrine and non-endocrine factors.
Preview this chapter:
Principles of interpreting endocrine test results, Page 1 of 1
< Previous page | Next page > /docserver/preview/fulltext/10.22233/9781910443866/BSAVA_Manual_Canine_Feline_Endocrinology_5_9781910443866.4.22-28-1.gif/content/chapter/10.22233/9781910443866.chap4
Figures
/content/figure/10.22233/9781910443866.chap4.fig4_4
4.4
Effect of pre-test probability (prevalence) on positive predictive value (PPV) and negative predictive value (NPV) and diagnostic accuracy using the low-dose dexamethasone suppression test for diagnosing canine hypercortisolism. The NPV and PPV are calculated from the 95% sensitivity and 71% specificity reported by
Van Liew et al. (1997
). At low prevalence, PPV is low (5% prevalence is associated with 15% PPV). To have 90% confidence in a positive test result being true requires a pre-test probability of 75%, and a PPV >95% requires a pre-test probability of 85% (purple arrows). However, at such high pre-test probabilities, the confidence in negative results falls. Confidence in a negative test result is >90% at all pre-test probabilities below 60%. © 2023 British Small Animal Veterinary Association
10.22233/9781910443866/fig4_4_thumb.gif
10.22233/9781910443866/fig4_4.png
4.4
Effect of pre-test probability (prevalence) on positive predictive value (PPV) and negative predictive value (NPV) and diagnostic accuracy using the low-dose dexamethasone suppression test for diagnosing canine hypercortisolism. The NPV and PPV are calculated from the 95% sensitivity and 71% specificity reported by
Van Liew et al. (1997
). At low prevalence, PPV is low (5% prevalence is associated with 15% PPV). To have 90% confidence in a positive test result being true requires a pre-test probability of 75%, and a PPV >95% requires a pre-test probability of 85% (purple arrows). However, at such high pre-test probabilities, the confidence in negative results falls. Confidence in a negative test result is >90% at all pre-test probabilities below 60%.