[Clinical Trial] Non-inferiority Trial of Medical Device Clinical Trial

The purpose of the clinical non-inferiority trial of medical devices is to clarify that the trial device is not worse than or non-inferior to the control device, without considering the situation that the test device is superior to the control device.

Even if the effectiveness of the trial device is worse than that of the control device, it is within the clinically acceptable range, and the clinical maximum allowable value is the non-inferiority threshold, provided that the trial device has some other advantages over the control device.

The conclusion of non-inferiority has two meanings: the effectiveness of the trial device is better than the placebo (indirect deduction of the effectiveness of the trial device); or if the effectiveness of the trial device is worse than the control device, the difference is also within the clinically acceptable range.

1. Non-inferiority trial design:

1.1 Non-inferiority trials using simple positive controls: the medical device has a widely accepted positive control, and previous studies can determine the effectiveness and stability of the positive control; it is not suitable for placebo-controlled trials and rare disease studies.

1.2 A three-arm non-inferiority trial is used, both a positive control and a placebo control: due to changes in medical conditions, the effectiveness and stability of the positive control cannot be guaranteed.

2. Determination of non-inferiority margin: ICH El0 recommends that the non-inferiority margin should be determined clinically and statistically and relatively conservative, to ensure that the usual non-inferiority margin should not be greater than the difference in effectiveness observed in the superiority trials of control devices and placebo in historical studies.

3. Statistical assumptions

For different types of metrics and indicators, the expressions of the original hypothesis (H0) and alternative hypothesis (H1) of the non-inferiority trials are different, as shown in the following table. Among them, Δ is the non-inferiority margin, absolute measurement indicators include mean difference and rate difference, and relative measurement indicators include rate ratio, risk ratio, odds ratio, etc. The higher-better indicator is an indicator whose larger value indicates better effectiveness and the lower-better indicator is an indicator whose smaller value indicates better effectiveness.

Table-Null hypothesis (H0) and alternative hypothesis (H1)

* T represents the test group effect, and C represents the positive control group effect.

The non-inferiority trial only needs to perform a hypothesis test α = 0.025 (one-sided) to perform statistical inference. If P≤α, reject H0, it can be inferred that T is not worse than C; Otherwise, it cannot be proved that T is not worse than C.

Confidence Interval Method for Differences Between Groups: The lower limit of the higher-better indicator according to TC's 95% confidence interval on both sides is greater than the non-inferiority margin -Δ, and the lower-better indicator based on the TC's 95% upper limit is less than the non-inferiority margin Δ; It can be concluded that the non-inferiority conclusion of the experiment is established.

According to the requirements in the "Guidelines for the Design of Medical Device Clinical Trials", hypothesis testing is performed on the test results, and it is recommended to use both hypothesis testing and interval estimation methods. Therefore, non-inferiority trials need to satisfy both hypothesis testing and confidence interval methods.

4. Conversion of non-inferiority and superiority inspection

In the non-inferiority trial plan, the conversion between non-inferiority and superiority trial can be defined in advance, that is, the non-inferiority trial is performed first. If the non-inferiority conclusion is established, the superiority trial can be further conducted. If the superior effect conclusion is established, the research conclusion is superior; if the superior effect conclusion is not established, the research conclusion is non-inferior. When the non-inferior effect conclusion is not established, the superior effect trial is no longer conducted, and the research conclusion does not support the non-inferior effect. The above process does not require multiple adjustments.

It is important to note that the conversion between non-inferiority and superiority trials must be clearly defined in the clinical trial protocol.