Cabotegravir Rilpivirine HIV Treatment Africa Week 96 Results
Okay, here’s a breakdown of the statistical analysis methods described in the provided text, categorized for clarity. I’ll also highlight key aspects and assumptions.
1. Primary Outcome Analysis (virological Suppression at Week 48)
* Endpoint: Viral load <50 copies/mL at week 48.
* Analysis Population: Intention-to-treat (ITT).
* Statistical Test: Non-inferiority test.
* Non-inferiority Margin: 10%.This means the new treatment is considered non-inferior if its suppression rate is no more than 10% lower than the comparator.
* Meaning Level (Alpha): 5% (two-sided).
* Power: 90%.
* Sample Size Calculation: Based on an assumed 94% suppression rate in both groups, the study needed 238 participants (119 per group) to demonstrate non-inferiority.
2. Key Secondary Outcome Analysis (Confirmed Virological Failure)
* Endpoint: confirmed virological failure.
* Analysis Population: Intention-to-treat exposed population.
* Statistical Test: Non-inferiority test (conducted after demonstrating non-inferiority on the primary outcome – hierarchical testing).
* Non-inferiority Margin: 4%.
* Significance Level (Alpha): 5% (two-sided).
* Assumed Failure Rate: 1.7% in both arms (based on previous trials).
* Hierarchical Testing: The secondary outcome was only analyzed for non-inferiority if non-inferiority was established for the primary outcome. This controls for the overall Type I error rate (false positive).
3. Sensitivity analyses
These were conducted to assess the robustness of the findings:
* Model Adjustment: Adjusted for site, sex, and third-drug class.This helps control for potential confounding variables.
* Per-Protocol Population: Excluded participants wiht protocol deviations (withdrawals, missed injections/oral treatment).This assesses the treatment effect under ideal conditions.
* Complete Cases: Used only participants with complete data. This avoids issues with missing data, but can introduce bias if data are not missing wholly at random.
4. Other Outcomes
* Analysis Population: Intention-to-treat.
* Analysis Method: Complete case analysis.
* Statistical Tests: t*-tests or chi-squared (χ2) tests.
* Nature of Analyses: Exploratory and descriptive. Limited to a pre-specified set of parameters.
Key Assumptions & considerations:
* Assumed suppression/Failure Rates: The sample size calculations were based on estimates from previous trials (94% suppression, 1.7% failure). If the true rates are different, the study’s power could be affected.
* Missing Data: The use of complete case analysis for “other outcomes” could introduce bias.
* Hierarchical Testing: The hierarchical approach to testing the primary and secondary outcomes is vital for controlling the overall Type I error rate.
* Non-inferiority vs. Superiority: The study is designed to demonstrate *non-inferiority, not superiority. It aims to show that the new treatment is not substantially worse than the comparator.
Let me know if you’d like me to elaborate on any specific aspect of these methods!