Proposal for delta check limits of frequently requested hormones using real-world data.

Abstract

Introduction: Research on delta check limits (DCLs) for hormones is limited, yet some laboratories apply arbitrary DCLs. We aimed to propose DCLs for commonly requested hormones.

Materials and methods: This study analyzed 59,657 paired results for adrenocorticotropic hormone (ACTH), cortisol, parathyroid hormone (PTH), prolactin, insulin, testosterone, and thyroglobulin from five Korean university hospitals. Delta check limits were established using the absolute delta difference (absDD) and absolute delta percent change (absDPC) with 5% cutoff for inpatients/emergencies (IE), outpatients (O) and both (combined; mean of them). Proportions outside the DCLs were compared across groups.

Results: Using absDD and absDPC, each group's DCLs showed 4.3% to 6.4% of values outside the DCLs, aligning with the 5% cutoff (excluding group IE for insulin, testosterone, and thyroglobulin due to < 1000 data pairs). Delta check limits of absDD differed between groups for ACTH, cortisol, PTH, and prolactin, while for absDPC, differences were seen only for ACTH and prolactin. Cross-validation revealed IE and O groups differed outside DCLs of absDD for ACTH, cortisol, and PTH, but only ACTH with absDPC. Combined DCLs of absDD showed ACTH and cortisol exceeded limits in 7.2% and 9.0% in IE, but only 2.6% and 0.6% in O. With absDPC, ACTH differed (10.4% in IE, 2.8% in O), while cortisol, PTH, and prolactin ranged from 4.0% to 6.1%.

Conclusions: Combined DCLs of absDPC are recommended for cortisol, PTH, and prolactin, while ACTH requires separate DCLs on clinical settings. These DCLs from real-world data provide a foundation for establishing DCLs of hormones in clinical laboratories.