30 September 2020 - In considering a submission seeking subsidy of upadacitinib for patients with rheumatoid arthritis in November 2019, the PBAC required a new criterion be met to justify claims of clinical superiority.

Subject to the outcome of AbbVie’s challenge, the PBAC’s requirement for the lower 95% confidence bound around a point estimate of difference to be above the ‘inverse of the non-inferiority margin’ could potentially be used by the Committee to judge whether claims of clinical superiority are valid. The PBAC would require this criterion over and above a determination of statistical significance.

The PBAC has, over the years, considered countless submissions for many medicines across many therapeutic areas with associated claims of clinical superiority. It has accepted many claims and rejected others. Critics would argue that it has rejected too many.

The finding that the PBAC introduced a new test, when there is no evidence in the Public Summary Documents to indicate the PBAC applied it in the past, seems arbitrary and capricious. Could the PBAC argue that it is only following its own Guidelines?

The current version of the PBAC Guidelines (version 5.0) is dated September 2016.

The term ‘inverse of the non-inferiority margin’ does not appear in the Guidelines. Section 2.4.3 makes reference to the noninferiority margin in the context of claims of noninferiority; the Guidelines note that “it is common practice to require that the confidence limits of the difference in treatment effect do not include an a priori stated clinically meaningful difference favouring the comparator.” However, there is no discussion of noninferiority margins in the context of claims of superiority.

A relevant section of the Guidelines concerning claims for superiority is ‘Section 2.4.3 – Outcomes.’ The Guidelines state that the “minimum clinically important difference (MCID) is the smallest difference in a particular outcome that patients perceive as beneficial (or detrimental).” Because statistically significant differences are not necessarily clinically important, Section 2.5.1 of the Guidelines advises that results reporting outcomes be discussed with reference to the MCID.

Consistent with this guidance, the PBAC has frequently required point estimates of difference to exceed an MCID for a claim of superiority to be accepted. However, the Guidelines make no mention that the lower confidence bound should also exceed the ‘inverse of the (pre-specified) non-inferiority margin’ to validate a claim of clinical superiority.