AUTH/3518/5/21 - Member of the public v AstraZeneca

AstraZeneca press release

  • Received
    27 May 2021
  • Case number
  • Applicable Code year
  • Completed
    07 July 2022
  • No breach Clause(s)
  • Breach Clause(s)
  • Sanctions applied
    Undertaking received
  • Additional sanctions
  • Appeal
    Appeal by the complainant

Case Summary

A concerned member of the public complained about two press releases which appeared on AstraZeneca UK Limited’s website.

The complainant stated that the press releases referred to the use of AstraZeneca’s vaccine in the UK, decisions taken by the UK regulatory authority and the opinions of UK experts on the vaccine and alleged that in both press releases AstraZeneca’s vaccine was referred to as ‘safe’, multiple times.

The complainant stated that a recent peer-reviewed article in the Lancet which pointed out that the efficacy claims being made for the Covid-19 vaccines were based on relative risk reduction (RRR) and not absolute risk reduction (ARR) which was a vastly smaller number. The complainant referred the PMCPA to the two UK press releases including one from AstraZeneca in which efficacy results from its studies were discussed and noted that in both press releases only the RRR results were presented, with no mention of ARR.

The detailed response from AstraZeneca is given below.

The Panel noted AstraZeneca’s submission that the first press release (30 December 2020) was a Regulatory News Release (RNS) and was distributed to appropriate media outlets and posted on the AstraZeneca global corporate website. The second press release (18 March 2021) was a Corporate Business Release (CBR) and was distributed to appropriate media outlets, predominately the same outlets as per the first press release, in common with standard practice for CBRs, the second press release was also posted on the AstraZeneca global corporate website. The Panel did not know the precise role of those individuals listed on the distribution lists but noted that it did not appear that all were based at media outlets associated with a financial and investor audience.
The Panel noted that whilst the broad general public interest in the content of and importance of the press releases was apparent, the press releases did not identify their business importance as required by the relevant supplementary information. In the absence of such a description the impression given was that it was for a broader circulation. In the Panel’s view, there was nothing in the content of either press release that indicated that they were solely for a financial or investor audience; indeed the content of each appeared to be of broader public interest.

The Panel noted that given the nature of certain media outlets, the ultimate audience would go beyond a financial and investment audience and might potentially include members of the public such as tabloid newspaper readers, and it was therefore particularly important to be cautious with reference to whether matters such as the content was balanced.

The Panel noted that the press releases at issue referred to the Covid AstraZeneca vaccine as safe. The Panel noted that the first press release dated 30 December 2020 and titled ‘AstraZeneca’s COVID-19 vaccine authorised for emergency supply in the UK’ stated in the opening paragraph ‘This regimen was shown in clinical trials to be safe and effective at preventing symptomatic COVID-19…’, a quote from a senior member of the vaccine group and an investigator of the Oxford Vaccine Trial that ‘The regulator’s assessment that this is a safe and effective vaccine is a landmark moment’. The second press release dated 18 March 2021 featured the prominent heading ‘UK and EU regulatory agencies confirm COVID-19 Vaccine AstraZeneca is safe and effective’. The Panel considered that the unqualified use, and in the press release dated 30 December, repeated use of the word safe, particularly noting its comments above about the ultimate audience and the weight that the ultimate audience might attach to the authors of the quotations, was such that the press releases were not balanced as required by the supplementary information to Clause 26.2. The Panel also queried whether the repeated use of the word ‘safe’ rendered the press release dated 30 December 2020 promotional in any event given the very broad definition of promotion in the Code. In the Panel’s view, taking into account all of its relevant comments above, AstraZeneca had not established that either press release satisfied the Supplementary Information to Clause 26.2 and therefore the Panel considered that the Code was applicable and, for the reasons set out in its comments above, ruled a breach of the Code in relation to the use of the word safe in each press release.

The Panel noted AstraZeneca’s submission that efficacy data for the vaccine was not included in the main body of the first press release dated 30 December 2020 titled ‘AstraZeneca’s COVID-19 vaccine authorised for emergency supply in the UK’. However, a link to the interim analysis published in the Lancet which supported the MHRA and EMA rolling assessment of data was included in the main body. In the subsection titled ‘AZD1222’, the following statement was included ‘As announced on 23rd November 2020, the primary efficacy endpoint based on a pooled analysis showed that the vaccine was 70.4% (confidence interval: 54.8% to 80.6%) effective at preventing symptomatic COVID-19 occurring more than 14 days after receiving two doses of the vaccine’. According to AstraZeneca, the pre-specified primary endpoints for Covid-19 Vaccine studies defined Vaccine Efficacy as 1 minus the Relative Risk or 1 minus the Hazard Ratio. The Panel noted AstraZeneca’s explanation that the rationale for using a relative reduction in risk was that it was invariant to the level of the underlying risk of infection whereas the absolute risk reduction could change depending on the underlying risk level. Therefore, unlike absolute risk reduction, any differential Vaccine Efficacy on the relative risk scale could be interpreted without being confounded by the observed risk at the time of the analysis as well as for individuals with different risk levels. This was especially important for clinical trials assessing Covid-19 vaccines given the nature of the pandemic where transmission of the virus changed regionally and over time. Therefore, the absolute risk reduction was not presented given it was not generally interpretable because it could change, not because of differential vaccine efficacy but due to the timing of the analyses as well as an individual’s underlying risk of Covid infection.

Whilst noting AstraZeneca’s submission about the difficulties associated with the calculation and inclusion of absolute risk reduction, noting that the results of the study were at a specific timepoint, the Panel considered that the relevant supplementary information, ‘reference to absolute and relative risk’, and compliance with it should be interpreted in light of its associated clause which required that materials etc should not be, inter alia, misleading and material must be sufficiently complete to enable the recipient to form their own opinion of the therapeutic values of the medicine.

The Panel noted that the cited Lancet source presented the efficacy of the vaccine as a relative risk reduction, but also included absolute values required to calculate the absolute risk reduction, in brackets: ‘Overall vaccine efficacy across both groups was 70.4% (95.8% CI 54.8 – 80.6; 30 [0.5%] of 5807 vs 101 [1.7%] of 5829)’. In the Panel’s view, whilst noting AstraZeneca’s submission that the press release did provide some details about the study, including the overall numbers of participants and symptomatic cases, the Panel considered that further details, such as the number of cases and subjects in each arm, would help certain sectors of the ultimate audience to interpret the absolute risk and form their own opinion of the efficacy of the medicine. The Panel considered that in the absence of any explanation in the press release, some readers, such as members of the public, might assume that the efficacy rate was, in effect, an absolute rate and that was not so. A breach of the Code was ruled.

The Panel noted that whilst according to AstraZeneca its Covid-19 vaccine had been granted a temporary authorisation in the UK to permit its supply, the vaccine had not been granted a marketing authorisation and so had not been legally classified as a prescription only medicine when the two press releases at issue were published. The Panel noted that Clause 26.2 only applied to prescription only medicines and on that very narrow technical point, no breach of the Code was ruled. This ruling was upheld on appeal by the complainant.

The Panel noted its comments and rulings above and considered that AstraZeneca had failed to maintain high standards and a breach of the Code was ruled.

The Panel noted the unique circumstances of the Covid-19 pandemic and the trial. The Panel did not consider that the particular circumstances of this case warranted a ruling of a breach of Clause 2 which was a sign of particular censure and no breach was ruled. This ruling was upheld on appeal by the complainant.