Statistics: Reactions By NUMDAYS

VAERS Exploratory Analysis.
NUMDAYS til reaction.
Statistics on the Number of Days after a vaccination, that a reaction occurs.

First, please see analysis of data anomalies:

Field Analysis of NUMDAYS

Field Analysis Of AGE_YRS and NUMDAYS

More Anomalies With Vax_Date And Onset_Date

And some associated views to remove some data anomalies:

View V_AD_NUM_YRS_GT_AGE_YRS

View V_AD_VAX_ONSET_DT_EXACT_YEAR_MULTIPLE

Note: Even using these two views, there are still a lot of data anomalies and data entry errors. It’s not perfectly clean data. For instance:

—-

VAERS_ID: 819826
Vax_date: 1941-09-05
Onset_date: 2019-06-18
Numdays: 28,410

Symptom_text:
Swollen injection site Sore \” \” itching red bumbs over body

Comment:

Vax_Date could be birth date (78 years old).

Or, the VAX_Date could have actually been 2018-09-05 and the ONSET_DATE 2019-06-18, about 9 months later.

But, from the Symptom_Text, you would think that the NUMDAYS til reaction was quite soon after the vaccination.

What is correct?

Another clerical/data entry error.

—-

use VAERS_2019
go

Select convert ( int, vd.NUMDAYS )  as NUMDAYS
, count(*)  Count
from dbo.[VAERSDATA] vd
left outer join dbo.[V_AD_NUM_YRS_GT_AGE_YRS]  ny
  on vd.vaers_id = ny.vaers_id 
left outer join dbo.[V_AD_VAX_ONSET_DT_EXACT_YEAR_MULTIPLE] ym
  on vd.vaers_id = ym.vaers_id 
where vd.numdays is not null
and  NY.VAERS_ID is null
and  ym.VAERS_ID is null
group by convert ( int, vd.NUMDAYS )  
order by 1 
go

—–

—-

Statistics:

n = 585,971

Already on Day 0, 47% of all reactions have started
By Day 1, 70% of all reactions have started
By Day 3, 80% of all reactions have started
By Day 11, 90% of all reactions have started
By Day 33, 95% of all reactions have started

The remaining 5% of reactions are found in 2,891 (2924 minus 33) rows.
A very long tail. (Results edited for readability.)

The pattern is very clear. Most reactions start very soon after vaccination.

If the reaction was not caused by the vaccine, what was it caused by?