Objective This study aimed to evaluate the effectiveness of various statistical methods based on the complementary sensitivity analysis framework recommended by the European Medicines Agency (EMA) in Alzheimer’s disease clinical trials, providing a reference for statistical practice that meets international regulatory standards for similar studies.

Methods Based on longitudinal data from a phase Ⅲ clinical trial for the treatment of Alzheimer’s disease, the change from baseline in Alzheimer’s Disease Assessment Scale-Cognitive Subscale (ADAS-Cog) score from baseline at week 26 (with a decrease indicating cognitive improvement) was used as the primary outcome measure. This study employed a mixed-effects model with repeated measures (MMRM) under the assumption of missing at random (MAR) as the primary analysis, while using critical point analysis, rank-based analysis, and slope-based analysis for sensitivity verification. A systematic comparison of the efficacy estimation results from the four analytical methods was conducted to compare the consistency and robustness differences in research conclusions across different analytical strategies.

Results The primary analysis results from the MMRM showed that the improvement in ADAS-Cog score from baseline at week 26 in the treatment group was superior to that in the placebo group (LSMD=-3.22, 95%CI: -4.55 to -1.88, P<0.001). Both rank-based analyses and slope-based analyses indicated significant differences between groups (both P<0.001). Critical point analyses revealed that the treatment effect was negative under most reasonable missing hypothesis scenarios, with the effect value only potentially turning positive in extreme non-random missing scenarios. There was no overall reversal of the efficacy conclusion.

Conclusion The combination of MMRM, critical point analysis, rank-based analysis, and slope-based analysis can effectively address the challenge of missing data in clinical trials for progressive diseases. This strategy fully complies with international regulatory standards, verifies efficacy conclusions from multiple perspectives, and enhances the credibility and scientific validity of clinical trial results. It can provide a practical reference for statistical analysis of missing data in similar clinical trials for progressive diseases.