Have ONS admitted to problems with their data?

Were vaccinated called unvaccinated?

There have been numerous papers and much public health data where the vaccinated have been described as unvaccinated for two or three weeks which causes a double bias in the data. The higher rate of covid in the first few weeks after injection amplifies the problem. The vaccinated look like they have less disease and the unvaccinated look worse off.

In contrast, the ONS have been careful in their deaths by vaccination status data to split out the first few weeks after injection from the period afterwards. However, there is a glaring problem with their data. With the rollout of the first dose, the unvaccinated see a huge spike in mortality which cannot be explained by the “healthy vaccinee effect” alone. Norman Fenton et al have clearly shown how misclassification alone could create the effect seen, even if only placebo had been given. 

The important point is what happened between December 2020 and April 2021 when the vaccine data was first published. There must have been huge political pressure to produce some numbers. This was the flagship policy! They wanted to show how their efforts at organising a fast rollout had paid off. For UKHSA to have refused to publish for four months suggests there must have been real issues with this data. Were there redundant old records, duplicates, missing vaccine records? They waited until it was good enough and then, finally, published. The idea that during that period, accurately collecting data on people who had been vaccinated and had already died was prioritised is wishful thinking.

Evidence of the problem includes the fact that the government issued guidance on what to do if someone turned up for their second dose and wasn’t on the system. 

In a recent Subject Access Request they were communicating about how many deaths were missing from the vaccine database (NIMS). 

Figure 1: Email referring to Dr Clare Craig admitting to incomplete records in vaccine database (NIMS)

However, the ONS have only admitted to fewer than 2,044 deaths being misclassified. 

Figure 2: Email from ONS admitting to misclassified deaths

Where someone did die and their vaccine record was not in the system, it would be very hard to put right in retrospect. The truth is we will probably never know how many vaccinated people did die in those early months of 2021.