2026-10-14T20:02:11Z https://api.figshare.com/v2/oai

oai:figshare.com:article/33324973 2026-09-23T13:57:24Z category_19 category_21 category_734 category_931 category_811 category_106 portal_87 item_type_3 month_year_09_2026

Investigating Heterogeneous Treatment Effects with a Latent Class Mixed Model in Average Bioequivalence Eunjung Song (2207200) Bo-Hyung Kim (7853933) Woojoo Lee (3110655) Pharmacology Biotechnology Biological Sciences not elsewhere classified Information Systems not elsewhere classified Mathematical Sciences not elsewhere classified Science Policy Average bioequivalence Heterogeneity Latent class mixed model Subgroup-by-formulation interaction <p>Investigating heterogeneity of treatment effects has received a lot of attention in clinical trials, and this is also important in bioequivalence studies. Average bioequivalence (ABE) using a simple 2× 2 crossover design is currently the most popular method for testing bioequivalence between two drugs, but the heterogeneity issue has been considered as a challenge in performing ABE with 2× 2 crossover design. Considering that factors that can cause heterogeneous treatment effects may not be known a priori or are expensive to measure, we propose to apply the ABE approach based on a latent class mixed model (LCMM), which does not require membership of each subject or the number of subgroups in advance. Our numerical study shows that the standard linear mixed model result should be carefully interpreted when the proposed method detects the heterogeneous treatment effects.</p> 2026-09-23T13:57:24Z Dataset Dataset 10.6084/m9.figshare.33324973.v2 https://figshare.com/articles/dataset/Investigating_heterogeneous_treatment_effects_with_a_latent_class_mixed_model_in_average_bioequivalence/33324973 CC BY 4.0