What method compares mean blood pressures between a new drug and an old drug?

Statistical Method for Comparing Mean Blood Pressures Between Drug Groups

The appropriate statistical method to compare mean blood pressures between a new drug and an old drug is Student's t-test, assuming the data follows a normal distribution. If normality cannot be confirmed, then a non-parametric alternative such as the Mann-Whitney U test (also called Wilcoxon rank sum test) should be used.

Determining the Appropriate Statistical Test

The selection of the statistical test depends on several key factors:

1. Normality Assessment

Before applying a t-test, the normality of blood pressure data should be evaluated:

• Shapiro-Wilk test is commonly used to assess normality of blood pressure measurements 1, 2
• If both samples pass the normality test, the parametric t-test is appropriate
• If normality is rejected, the non-parametric Mann-Whitney U test should be used instead

2. Independent Samples t-test

When comparing two independent groups (new drug vs. old drug):

• The independent samples t-test is the standard method for comparing means between two groups
• This test assumes that both samples come from normally distributed populations with unknown but equal variances
• If variances are unequal, a modified version (Welch's t-test) should be used

3. Special Considerations for Blood Pressure Data

Blood pressure measurements have specific characteristics that affect statistical analysis:

• Blood pressure typically follows a circadian rhythm, which may need to be accounted for if using ambulatory measurements 3
• Reproducibility of blood pressure measurements varies depending on measurement method (clinic, home, or ambulatory) 4
• Home blood pressure measurements have shown superior reproducibility compared to both clinic and ambulatory measurements, potentially reducing required sample sizes in comparative trials 4

Practical Implementation

When designing and analyzing a study comparing blood pressure response between two drugs:

1. Calculate sample size based on:

   • Expected difference in blood pressure between groups
   • Estimated standard deviation of blood pressure measurements
   • Desired statistical power (typically 80-90%)
   • Significance level (typically α = 0.05)

2. Collect data ensuring:

   • Standardized measurement protocols
   • Adequate number of measurements per patient
   • Appropriate timing relative to drug administration

3. Analyze data:

   • Test for normality of distribution
   • Apply t-test if normality is confirmed
   • Use Mann-Whitney U test if normality is rejected

Advantages of Modern Approaches

Ambulatory blood pressure monitoring (ABPM) offers advantages over conventional clinic measurements for drug efficacy studies:

• Provides a profile of blood pressure over the 24-hour period
• Detects white coat hypertension
• Reduces placebo effect and regression to the mean
• Increases statistical power, potentially reducing required sample size 5

When using ABPM data, analysis may need to account for the time-course of drug effects, including peak and trough effects, which requires specialized statistical approaches 3.

In conclusion, while the t-test remains the fundamental method for comparing mean blood pressures between drug groups, careful attention must be paid to the distribution of the data and the measurement methodology to ensure valid conclusions about drug efficacy.