Hypothyroidism - A Causal Approach to Testing Assumptions against Empirical Results.

There is a controversy in the diagnosis and treatment of hypothyroidism. We propose the disagreement is fueled by statistical paradoxes, and sampling biases that provide different perspectives depending upon the sample selection criteria. The statistical inconsistencies become more apparent when viewed using a causal lens. Foundational hypothyroid research does not reflect the current Levothyroxine treated population. Exploration of empirical data demonstrates an apparent breakdown of the T4 to T3 causal pathway in the treated population. This use case demonstrates the difficulty of translating controlled research into clinical practices for patients with multiple comorbid conditions. We make the case for redundancy in data collection, ongoing attempts to falsify current assumptions and the need for causal approaches to validate the results of controlled research in clinical settings, in order to avoid confirmation bias from statistically insufficient biometrics.