November 21, 2025

Outline

  • Introduction
  • Methods
  • Results
  • Concluding Remarks
  • References

Introduction

Background of the Study

  • Hypertension remains a major public health challenge, and identifying effective interventions to reduce is critical for preventing long-term complications.

  • Worldwide, about are living with hypertension, yet fewer than one in five have the condition under control.

  • Because hypertension significantly elevates the risk of cardiovascular disease, identifying intervention strategies that lead to meaningful is a critical public-health priority.

  • By employing a structure, this study mimics real-world treatment settings to compare how different interventions influence SBP reduction, thereby offering insight into effective hypertension management.

Introduction (Cont.)

Purpose of the Study

  • To evaluate whether the varies meaningfully across the four treatment groups in our simulated multi-center clinical trial. By comparing these interventions, the study aims to identify which approaches may offer the greatest potential benefit for improving hypertension management.

Rationale

  • Understanding which treatment leads to the greatest is essential for effective hypertension management.

  • Identifying the most effective intervention supports , improves patient outcomes, and informs future .

Research Questions

  1. What do the reveal about the distribution of SBP reduction across the four treatment groups?

  2. Is there a among the four treatments based on the overall regression/ANOVA F-test?

  3. Which treatment groups show in SBP reduction according to post-hoc comparisons?

Methods

Study Design

  • The data was generated from a comparing four treatment strategies for lowering systolic blood pressure (SBP).

  • The independent variable was the (Placebo, Drug A, Drug B, Lifestyle Changes).

  • The dependent variable was the measured after treatment.

Statistical Analysis: Descriptive

  • Computed (mean, standard deviation) for each treatment group to summarize central tendency and variability.

  • Used a to visualize distributional patterns and compare SBP reduction across treatments.

Methods (Cont.)

Statistical Analysis: Inferential

  • A was fitted with treatment as a categorical predictor, placebo as the reference level, and SBP reduction as the outcome. This modeling strategy is equivalent to conducting a for comparing group means.

  • The from the model was used to determine whether there were significant overall differences in average SBP reduction among the four treatment groups.

  • were computed for each treatment group and plotted with their corresponding 95% confidence intervals to visualize differences in estimated SBP reduction.

  • Model assumptions were evaluated by examining a and performing a on the residuals to assess whether the error terms followed an approximately normal distribution.

  • were performed using the Tukey method to identify which specific treatment groups differed from one another.

Results: Descriptive Statistics

  • The descriptive analysis revealed clear differences in across the four treatment groups.

  • The group showed very small reductions in systolic blood pressure.
  • produced moderate reductions.
  • demonstrated the .
  • produced reductions higher than Placebo but smaller than Drug A and Drug B.

Results: Descriptive Statistics (Cont.)

  • The boxplot showed well-separated distributions, with having the highest median and widest reduction range.

Results: Inferential Statistics

  • The plot of estimated marginal means showed clear differences in the mean for each group, with the 95% confidence interval for each group not even coming close to overlapping.

Results: Inferential Statistics (cont.)

  • The summary table for the linear regression model shows that Drug B has the highest impact on SBP reduction, as shown by it having the highest value of estimate. We also see from this model that all of its estimates for the individual treatments’ impacts are significant, since the p-value for each is practically 0. Finally, the R-squared values are large, indicating the model is a strong fit to the data.

Results: Inferential Statistics (cont.)

  • The ANOVA table shows a high variability between the group means, indicating a significant difference in their values, meanwhile it shows a low variability between values in each group, indicating the values are not subject to much error.

Results: Inferential Statistics (cont.)

  • This table is a summary of statistics that compare the groups to each other. As the F-statistic is far greater than the critical value, it falls in the rejection region, giving us strong evidence that at least one pair of group means is not equal. This is supported by the miniscule p-value.

Results: Inferential Statistics (cont.)

  • The QQ plot affirms our assumption that that the residuals (differences between actual values and predicted values) follow a normal distribution.

Results: Post-hoc comparisons

  • This table shows the estimated difference between each pair of each groups’ means. For all pairs, the p-value for the hypothesis that the means are equal is extremely low, indicating that there is strong evidence to suggest the means are all different. The absolute value of the difference between Drug B and Placebo is the highest, and the absolute value of the difference between Drug B and Drug A is the lowest.

  • Note: Post-hoc tests are purely exploratory.

Clinical Conclusion

  • In conclusion, every statistical model we used to describe the data or perform inferences on it all showed that Drug B was the most effective in reducing systolic blood pressure, followed by Drug A. Lifestyle was the least effective method.

References

Thank you page

  • Thanks to Dr. Emmanuel Thompson, professor of Mathematics at SEMO.