Variability of Thyroid Measurements from Ultrasound and Laboratory in a Repeated Measurements Study

Area of Science:

• Endocrinology and Metabolism
• Medical Imaging and Diagnostics
• Epidemiology

Background:

• Measurement variability is a key concern in medical research, potentially introducing bias in epidemiological studies.
• Quantifying measurement error is crucial for probabilistic sensitivity analyses.
• This study investigated the variability of thyroid-related outcomes from ultrasound and laboratory tests.

Purpose of the Study:

• To quantify the intraindividual variation of thyroid-related outcomes measured by ultrasound and laboratory analyses.
• To provide data for probabilistic sensitivity analyses in future epidemiological research.
• To assess the reliability of different thyroid markers for population studies.

Main Methods:

• Twenty-five volunteers underwent monthly examinations over one year.
• Ultrasound measured thyroid volume, goiter, and nodules.
• Laboratory tests included urinary iodine, TSH, fT3, fT4, and thyroglobulin.

Main Results:

• Urinary iodine concentrations exhibited the highest coefficient of variation (56.9%), while free thyroxine (fT4) had the lowest (6.3%).
• Intraclass correlation coefficients (ICCs) were highest for thyroglobulin (0.90) and thyroid volume (0.87), indicating good reliability.
• Cohen's kappa values for goiter and nodules were 0.64 and 0.70, respectively.

Conclusions:

• The study provides essential variation measures for thyroid outcomes, applicable to epidemiological data analysis.
• Serum thyroglobulin demonstrates low intraindividual variation, highlighting its potential as a reliable marker for population iodine status.
• These findings aid in minimizing bias and improving the accuracy of thyroid-related epidemiological research.