Endocrine Abstracts

Thyroid hormones measure is inevitable in some critically ill patients and may help the management. Alternatively, the physiological changes in the thyroid imbalance during pregnancy also make establishing reference intervals for critical ill challenging.

Objectives: We aimed to compare reference intervals in an obstetric intensive care unit.

Methods: After excluding patients admitted after delivery or pregnant under thyroid disease treatment, we included 109 pregnant women admitted to the ICU between January 2018 and October 2023. We collected data from the mother (age, gestational age, previous diseases, body mass index, blood pressure, APACHE II and the critical events) and the newborn (weight, gestational age – Capurro, Apgar and critical events). Two reference intervals were compared (the assay reference range and the Hoffman-established reference interval). The Cobas Roche Elecsys immunoassay was used to measure the Thyroid Stimulate Hormone (TSH) and the Free Thyroxine (FT4) concentrations. TSH and FT4 were log-transformed before the analysis. Quantile-Quantile Plot and the Hoffman method were calculated using R statistical software. The chi-squared test, the Mann-Whitney and the logistic regression was performed in IBM SPSS Statistics for Windows.

Results: The Hoffman TSH calculated interval was 0.33 to 3.63 mUI/l and the FT4 was 0.79 to 1.29 ng/dL. Considering the participants, 56 (51.4%) on the assay reference and 65 (59,6%) on the Hoffman reference were classified as euthyroid. Conversely, the participants with TSH on the reference or below and a low FT4 reduced from 33 (30.3%) with the assay reference to 11 (10.1%) with Hoffman’s. The assay reference was not able to predict any of the maternal or fetal critical events. On the contrary, Hoffman’s interval predicts coagulation disorders with an odds ratio of 4.4 (P-value: 0.025, CI: 1.201 to 16.45).

Conclusions: Although this work had a selective blood collection for hormone dosages, Hoffman proposed intervals calculated for our pregnant population reduced the patients classified as abnormal and helped predict critical events.