1 -> Tom Yengst
Science Fellow, Children's Health Defense and OpenVAERS
2 -> The Humanity Projects Team
Review Date: February - 2023
A - Overview
In this section the distribution of hospitalisations following dose 2 of the COVID-19 vaccine is detailed. Hospitalisations are relatively rare events that, as we noted in part 2, occur in less than 5% of the V-Safe app registrants.
Definition of a hospitalisation event:
The definition of a hospitalisation event that we use in the following analysis is a health visit that resulted in:
- An emergency room or emergency department visit;
- An outpatient clinic or urgent care clinic visit;
- A hospitalisation.
We investigate how hospitalisation events are related to the inoculations, and in particular how the rate of hospitalisations changes over time. As noted in part 2 "lost work or school" events typically occured in the first 5 days after inoculation, with most events having been a temporary, mild, or moderate condition.
As shown below, hospitalisations tend to have a much wider distribution over time, with about 25% occurring from 6 months to 1 year after dose 2. Consequently, these events appear to be drawn out over time. Many of the hospitalisations refer to emergency room care where internment was not necessary. The more serious health events are likely to be spread out over time, which are consistent with the anecdotal evidence of blood clots or strokes occurring in healthy individuals more than a year after the inoculations.
B - Time Series of Dose 2 Inoculations and Hospitalisation Events
When plotting the time series of dose 2 inoculations and hospitalisations for the sample of individuals who suffered hospitalisation events following dose 2, these events were drawn out over time even though most inoculations occurred in the first 6 months of 2021 (from 1-Jan-2021 to 1-Jun-2021), as shown in the figure below.
Each hospitalisation refers to only one individual irrespective of the number of hospitalisation events that he/she incurred. For individuals with multiple hospitalisation events, the last hospitalisation event on record (max date) was selected. The total number of hospitalisations was 326,468 of a total of 7,418,097 individuals who received dose 2, resulting in a hospitalisation rate of 4.4% for the period from 1-Jan-2021 to 31-Jul-2022.
By looking closely at the time series of dose 2 inoculations and hospitalisations, we observe that the pattern of inoculations seems to repeat over time. This is due to a feature of the V-Safe app that we'll explain below. To better understand the relationship, we plot the correlation chart between the dose 2 inoculations performed from 1-Jan-2021 to 1-Jun-2021 and the hospitalisations 1 year (365 days) ahead.
We can clearly observe that the correlation is very high, about 90%. Additionally, the regression slope coefficient of .25 indicates that the hospitalisation events occurring 1 year from the dose 2 inoculation amount to 25% of all hospitalisations.
Histogram of hospitalisations
In order to better understand the previous charts, we plot the histogram (count) of the time elapsed since the dose 2 inoculation and a hospitalisation event (see figure below). The x-axis corresponds to the days since inoculation while the y-axis refers to the total count of hospitalisation events.
There is a discrete distribution of days elapsed from dose 2 to a hospitalisation event. The explanation for this distribution is a feature of the V-Safe app that tracks these events over time by prompting the user for the occurrence of a hospitalisation event. The time schedule of the app prompts seem to be consistent with the following algorithm:- Day 1, 2, 3, 4, 5, 6, 7 (Week 1)
Hospitalisation rates per 10,000
To compute the hospitalisation rate we perform the following steps to the raw hospitalisation event histogram:
1 - Assume an uniform distribution over the prior period for each of the event counts.
2 - Divide the obtained distribution of events by the total number of events (consequently obtaining the daily rate at each day since dose 2 inoculation)
3 - Perform a 30-day moving average to smooth the distribution.
4 - Compute the hospitalisation rate per 10,000 (for a given day since a dose 2 inoculation).
The chart below shows the hospitalisation rates per 10,000 on a given day following dose 2 inoculation.
The daily hospitalisation rate following dose 2 inoculations decreased over time from a maximum rate of about 2.4 per 10,000 during the first 3 months to a residual value of about 0.5 per 10,000 after 7 months.
The distribution also shows that hospitalisation events are high for the first 3 months and decrease thereafter. It should be noted that the monitoring of Serious Adverse Events (SAEs) in the Pfizer clinical trial for the Covid-19 vaccine stopped 1 month after dose 2, as they were assumed to be negligible. Here, we can observe that the rate of hospitalisation events remains high for at least 3 months after dose 2 and then declines slowly over time. The decline in the rate of hospitalisation events is slow, so that 25% of these events occur in the period between 6 month and 1 year after dose 2.
The monthly rate of hospitalisation events during the first 3 months can be estimated to be (30 x 2.4 = 72) per 10,000. This value is comparable to the rate of Severe Adverse Events (SAEs) found during the Pfizer clinical trials, which was 119.5 per 10,000 in the vaccine cohort, measured from dose 1 to 1 month after dose 2. The monthly rate of SAEs is calculated by dividing the previous rate by 2, resulting in 119.5/2 = 59.75 per 10,000.
The SAEs in the Pfizer trial is reviewed in Part 6 of the disabilities work.
Relative rates of hospitalisations
In this step we computed the relative decay in the hospitalisation rate over time. For this purpose we divided the distribution of hospitalisations over time which we obtained previously by its maximum value (which was the 30 day average rate of the first 30 days following dose 2 inoculation). This calculation provides a normalised rate of hospitalisation over time where the maximum rate is normalized to 1 and over time the relative rates correspond to a percentage of the maximum.
The figure below shows the relative rate of hospitalisations over time. For the first 3 months, the rate of hospitalisations is close to 1 and slowly decays thereafter. From 4 to 6 months the rate of hospitalisations was about 50% the initial rate, from 6 months to 1 year was about 24%. The V-Safe app did not provide data after 1 year.