introduction and objective:
In Poland, the assessment of nocturnal GH secretion has gained the status of screening test; however, this procedure is not included in international recommendations. The aim of the study was to assess the accuracy and predictive value of the test of nocturnal GH secretion as a screening procedure in diagnosing GHD, and to check the adequacy of the cut-off value for GH peak in this test on the level of 10 ng/ml.

Material and Methods:
The analysis comprised the data of 1,000 children with short stature. In all the patients, GH secretion was assessed in a screening test (after falling asleep) and in 2 stimulating tests (reference tests), with simultaneous assessment of IGF-I secretion before stimulating tests. The indices of screening test accuracy, likelihood ratios and predictive values were assessed. The cut-off level of GH peak after falling asleep, ensuring its 95% sensitivity, was calculated in ROC curve analysis.

Results:
Sensitivity of the screening test was 70.4%, while the specificity – 61.2%, positive likelihood ratio – 1.842, negative likelihood ratio – 0.482, positive predictive value – 0.462, negative predictive value – 0.812. The sensitivity of the test of GH secretion after falling asleep is too low with respect to the requirements for screening test. The ROC curve analysis showed 95% sensitivity for the screening test on the level of 19.0 ng/ml; however, with a very low specificity – below 25%, thus making this test completely useless as a screening procedure.

Conclusions:
The obtained results strongly contradict the opinion that the assessment of GH secretion after falling asleep should be a screening test in diagnosing GHD in children with short stature.