A Continuous Outcome Non-Inferiority Trial is a clinical trial method that aims to show that a new treatment is non-inferior to an existing treatment in terms of a continuous outcome, where an acceptable difference is defined. Continuous outcomes include numerical values such as cholesterol levels, pain intensity or weight loss, which can be measured on a scale.

Novustat is your experienced partner and advises you on all aspects of statistics, from data analysis to the interpretation of results, and helps you to gain maximum insight from your data sets. Contact us without obligation for further information!

Free online calculator to calculate the sample size

Use our online calculator to quickly and easily calculate the optimal sample size for your non-inferiority study with continuous results.

Goal and concept

The aim of a non-inferiority trial is to show that the new treatment may not be better, but within a predetermined acceptable difference to the existing treatment. This type of trial is particularly useful if the new treatment offers advantages such as lower cost, fewer side effects or ease of use , which could compensate for possible losses in efficacy.

Calculation of the sample size

Calculating the sample size is crucial to ensure that the study has sufficient statistical power. This helps to achieve valid and reliable results. The acceptable difference, standard deviation of the results, significance level and power are taken into account. Our online calculator helps you to precisely determine the required sample size.

Application and requirements

An example of a continuous outcome non-inferiority trial would be the comparison of two cholesterol-lowering drugs. Suppose the new drug lowers LDL cholesterol by an average of 20 mg/dL and the existing drug by 22 mg/dL. The study shows that the new drug is not worse if a difference of up to 5 mg/dL is considered clinically irrelevant. Such a study requires clear definitions of the non-inferiority threshold and robust statistical analyses in order to obtain valid results.

Notes on sample size calculation for non-inferiority studies with continuous outcomes

The calculation of the sample size is based on the expected difference between the groups, the standard deviation of the results, the significance level and the statistical power. These factors ensure that the sample is large enough to provide valid and meaningful results. It ensures that the sample is large enough to provide valid and reliable results. These parameters ensure that the study is robust enough to demonstrate non-inferiority.

Significance level (alpha)

The significance level (α) defines the risk of an error of the first kind, i.e. the probability of finding a difference even though none exists. An alpha of 0.05 is usually selected, which corresponds to a 5% risk .

Power (1-Beta)

The power (1-β) indicates how likely it is that a true non-inferiority is correctly detected. A power of 80% to 90% ensures that the study is sensitive enough to detect a clinically relevant difference.

Standard deviation (σ)

The standard deviation (σσ) indicates the variability of the measured results in the groups. A higher variability requires a larger sample in order to provide precise results.

Non-inferiority limit (Δ)

The non-inferiority threshold (ΔΔ) defines the maximum acceptable difference between the groups at which the new treatment is still considered non-inferior. For example, a difference of 5 mg/dL in cholesterol levels could be considered acceptable.

Formula for calculation

The sample size is calculated using the following formula:

n=2⋅f(α,β)⋅σ2Δ2n = \frac{2 \cdot f(\alpha, \beta) \cdot \sigma^2}{\Delta^2}

Here, f(α,β)f(\alpha, \beta) stands for the Z values, which are determined by the significance level and the power. σ\sigma for the standard deviation of the results and Δ\Delta for the non-inferiority limit. This formula ensures that the sample is large enough to provide valid and statistically sound results in your non-inferiority trial.