Review Date: January - 2023
We review the peer-reviewed paper, published in Vaccine Journal on 9/2022, that investigates the Serious Adverse Events (SAE) in the Pfizer and Moderna clinical trials prior to the unblinding of the participants. The paper reference is:
J. Fraiman et al, "Serious adverse events of special interest following mRNA COVID-19 vaccination in randomized trials in adults", Vaccine, Vol. 40, Issue 40, 22 September 2022, pp 5798-5805
Our review of the paper is significant in the context of our V-Damage project as it establishes a direct link between our analysis of disabilities (using data from the Bureau of Labor Statistics) and the original phase III clinical trials for the Covid-19 mRNA vaccines.
Summary of the results:
The authors find that the Pfizer and Moderna mRNA Covid-19 vaccines were associated with an excess risk of serious adverse events of special interest.
For the Pfizer trial, the rate of ASEs of special interest was 27.7 per 10,000 for the vaccine group while 17.6 per 10,000 foe the placebo group. The difference was of 10.1 per 10,000, which corresponded to a 57% increase in risk of an a SAE of special interest relative to the baseline.
For the Moderna trial the rate of ASEs of special interest was 57.3 per 10,000 for the vaccine group while 42.2 per 10,000 foe the placebo group. The difference was of 15.1 per 10,000, which corresponded to a 36% increase in risk of an a SAE of special interest relative to the baseline.
The authors also investigated Severe Adverse Events as a whole, and found an increased risk for the Pfizer vaccine of 36%. The equivalent increased risk for the Moderna vaccine was 6%.
All SAEs ocurring after the 2nd vaccination dose were counted, with a median monitoring period of around >= 2 months. The authors also state that Pfizer trial SAEs may be under-reported as Pfizer did not report SAEs in participants beyond 1 months following dose 2. The study is also limited in the sense that it does not report SAEs between dose 1 and dose 2.
The authors perform a secondary analysis of serious adverse events reported in placebo-controlled, phase III randomised clinical trials of Pfizer and Moderna mRNA Covid-19 vaccines in adults (NCT04368728 and NCT04470427). They then classified each SAE against the Brighton Collaboration definitions of adverse events of special interest. Analysing only those events “of scientific or medical concern specific to the sponsor’s product or program” seeks to discount random medical events occurring after vaccination that are not likely to be relevant to the Covid-19 vaccines.
The Pfizer and Moderna trials were performed in adults, aged 16+ for Pfizer and 18+ for Moderna.
The Pfizer and Moderna trials were expected to follow the participants for two years. However, within weeks of the emergency authorisation, the sponsors began a process of elective unblinding of participants, and those who had received the placebo were offered the vaccine. The unblinding compromised the post-authorisation data. Consequently, the authors used the interim datasets that were the basis for emergency authorisation in December 2020, approximately 4 months after the start of the clinical trials.
The authors chose SAE tables that presented information by SAE type and that met the FDA's requirement for a median safety follow-up time of at least 2 months after dose 2.
Definition of an SAE:
The authors use a definition for an SAE that is consistent with regulatory expectations, and also the ones used in the Moderna and Pfizer clinical trials. A SAE is defined as one that meets one of these conditions:
- Life threatening at the time of the event
- Inpatient hospitalisation or prolongation of existing hospitalisation
- Persistent or significant disability/incapacity
- A congenital anomlay/birth deffect
- Medically important event, based on medical judgment.
The table below, copied from the table 2 of the paper, summarises the findings when analysing all ASEs, and those of special interest, in the Pfizer and Moderna clinical trials.
|Total Events (per 10,000)|
|Trial||Vaccine||Placebo||Risk Difference. (95% CI)||Risk Ratio (95% CI)||Serious adverse events|
|Pfizer||127 (67.5)||93 (49.5)||18.0 (1.2 to 34.9)||1.36 (1.02 to 1.83)|
|Moderna||206 (135.7)||195 (128.6)||7.1 (-23.2 to 37.4)||1.06 (0.84 to 1.33)|
|Combined||333 (98.0)||288 (84.8)||13.2 (-3.2 to 29.6)||1.16 (0.97 to 1.39)||Serious adverse events of special interest|
|Pfizer||52 (27.7)||33 (17.6)||10.1 (-0.4 to 20.6)||1.57 (0.98 to 2.54)|
|Moderna||87 (57.3)||64 (42.2)||15.1 (-3.6 to 33.8)||1.36 (0.93 to 1.99)|
|Combined||139 (40.9)||97 (28.6)||12.5 (2.1 to 22.9)||1.43 (1.07 to 1.92)|
Limitations of the study:
The authors warn of several limitations of their study, which mostly stem from the fact that the raw data from the Covid-19 vaccine trials are not publicly available, namely:
- Pfizer's trial did not report SAEs occurring more than 1 month after dose 2, which may have led to an underestimation of SAEs in the Pfizer trial.
- For both trials, the limited follow up period prevented formal harm-benefit analysis.
- All the SAEs in the authors' analysis met the regulatory definition of a SAE, but many adverse events that the patient may themselves judge as serious may not meet the regulatory threshold.
- The decision about which SAEs to classify as events special interest requires subjective clinical judgement, and this is challenging in the absence of detailed clinical information about the SAE. An important limitation was the authors' lack of access to the individual participant data which forced them to use a conservative adjustment to the standard errors. The 95% confidence intervals (CE) are an approximation because they did not know which patients had multiple adverse events.
The pre-authorisation period (before unblinding) for the Pfizer and Moderna clinical trials, from August 2020 to December 2020, is the most reliable data we have for estimating the Covid-19 vaccine efficacy and possible harms. Unfortunately, the clinical trials were electively unblinded and the unblinded placebo participants offered the vaccine (which most accepted) after a very short period, rendering the analysis of medium-term harms almost impossible.
Additionally, as mentioned above in the limitations of the paper, the lack of access to the individual participant data limits the scope of any analysis.
Even with these severe limitations, researchers across the world are trying to put together the vaccine harm puzzle using different sources of data, including those that are available from the clinical trials. This paper points to strong evidence of harm in respect to a rate of SAEs of special interest above placebo control which is statistically significant, and likely corresponds to a safety signal. There is also a rate of overall SAEs above that for placebo control for the Pfizer vaccine. This data alone might have warranted caution over the decision of opening the vaccine rollout to the whole population.
We find the results very interesting, as they corroborate our findings when investigating the change in disabilities in the population, civilian labour force and employed population, using the BLS (Bureau of Labor Statistics) survey on employment and disabilities.
The rate of excess of SAEs of special interest of 12.5 per 10,000 in the vaccinated individuals versus the placebo group for the approximately 2 month median monitoring period, is of the same order of magnitude that we find in the rate of rise in disabilities in the civilian labour force aged 16-64 in the USA, which corresponds to roughly the same age-group as the participants in the clinical trials.
In part 5 of the disabilities analysis, we perform simulations of the increase in disabilities from early 2021 to late 2021 (before the rollout of the boosters) where we use the rate of SAEs of special interest found in this paper, and the data we have on the vaccine rollout across the 16-64 population. As we don't have the results for the phase III clinical trials after December 2020, we had to make an assumption as to the rate of SAEs of special interest being persistent over time.