More Nivel worries

The herziene Nivel-rapport maakt opnieuw het nodige los. Natuurlijk wordt er weer geroepen dat is aangetoond dat vaccins werken maar met zulke bizarre cijfers en rammelende methodologie is helemaal niets aan te tonen. Misschien dat het iets zou kunnen zeggen over de patiëntenpopulatie van die 400 huisartsen maar ook dan moeten ze opnieuw beginnen. Maar eerst wat anders, waar ik niemand over hoor.

The privacy trump card

In the Netherlands, it is not necessary to Explicit consent before you enter data Anonymized used for scientific research. This is because anonymized data can no longer be traced back to an individual, which means that it falls outside the scope of the General Data Protection Regulation (GDPR).

Even if you have objected to the use of your data, i.e. have given non-consent to the question 'may your data be used for scientific research', your data may still be used anonymously. If data can no longer be traced back to a person, that data is not covered by the GDPR, nor by the EU privacy directives. And on top of that, there are also exceptions in science and statistics.

An exception has been made for the Covid vaccines. For privacy reasons, there is still no meaningful data. To be able to screen this, a 'nonconsent' option has even been created. Hardly anyone knows why. It is exactly in line with not serializing the vaccination vials. As a result, it is no longer possible to trace how they got from the incubation tank into the arm of the threatened patient.

So, if something turns out to be wrong with the stuff later, it is impossible to find out how they were treated and for privacy reasons, the healthcare industry and government can keep the patient cards close to their chest.

If some cards do need to be laid (WOO requests), the data is first neutered with data obfuscation techniques.

While anonymizing while retaining information is not that difficult after all. The covid data is now deliberately impoverished by offering them unreadable for computers in low resolution, and by adjusting dates of death in such a way that they can no longer be linked to a vaccination date. So you still don't know anything.

However, anonymization can be done very easily without sabotaging a research goal.

• We do not examine newborns so the date of birth and time can be removed, age on January 1 is sufficient
• Voor iedereen boven de 90 wordt een randomisatie van plus/min 1 jaar op de geboortedatum toegepast.
• The date of vaccination and the date of death are crucial. On days with very few deaths, these together may be unique to 1 individual and therefore traceable. This link can be broken by shifting both dates by the same number of days, for example with a random number between 7 and -7. Or if necessary 14 and -14. An anonymization specialist probably knows more tricks.

None of this happens. Privacy is far too important to medical professionals. After all, if anyone should be trusted by patients, it's the doctor. Woe betide us if that basis of trust disappears...

Vaccinated people could therefore indicate that their data should not be used, no one knows why. To what extent you can forbid science to look at certain facts, I wonder at all. It would mean the death of science – assuming it is still alive. The Nivel report bears little evidence of this.

As a result of the privacy move, the non-consenters are part of the 7% that is not listed in the CIMS vaccination registration system. So there is no note "non-consent", no, nothing has been filled in. The jab is missing, including patient information. Died or not? Unknown. Did something go wrong while catching up on the registration backlog? Such a percentage of 7% already makes the research worthless, especially in view of the research goal.

So much for data management. Once again.

But: Nivel to the rescue, they were able to limit the disaster!

The 'small percentage' (7%, as mentioned a disastrous data gap) of missing vaccinated people has been partially recovered. This probably included data from people who had not given permission to the GGD to make their data available; After all, doctors are very concerned about the privacy of their patients.

Anyway, through the general practitioners, personal confidants of those non-consent patients, their data could still be retrieved. In their practice records, they actually found patients who had reported their vaccination status to their GP and were missing from the CIMS register.

The patients knew that this could happen because there are posters and flyers in the waiting rooms of the affiliated Nivel doctors that inform about it. That's tacit consent or something like that – yes, I still like one like that. That information should be in every waiting room, because anonymized data can of course always be used.

Non-consent and privacy concerns are played out very selectively. If vaccination readiness is at stake, they weigh very heavily. Vaccination promotion simply avoids it.

The part of the patients whose vaccination status is not reflected in the GP data could still be contacted to update their file. De artsen hebben die gegevens. Dan komen de non-consenters er vanzelf uit – of ze doen alsnog mee omdat de huisarts het vraagt. Daarmee zouden de data aangevuld of misschien wel gecompleteerd kunnen worden. Bij de huisartsen zelf is er immers geen enkel probleem met non-consent. Maar het sop is de kool niet waard zonder WLZ-data (zie verderop). Maar wel leuk voor een intern bulletin onder huisartsen. Arts & Auto, Medisch Contact?

• September 2022: Minister of Health, Welfare and Sport, Ernst Kuipers, requests VWS data from RIVM and CBS. They refuse 'for privacy reasons'.
• September 2022: Kuipers Data Privacy Impact Assessment (DPIA) to investigate this (source)
• February 2023: The Dutch Data Protection Authority sees no objection to the sharing of anonymized data (source) (Why the 'asessment' at all... (source))
• April 2023: Kuipers has taken note of the data. New position: more openness about the corona approach endangers the functioning of the state. (source)

Representativeness

The study was conducted on the patient population of 400 general practitioners, totaling almost 1 million people. The vaccination rate there was 84%, which is slightly higher than the national vaccination rate of 82%. Relatively healthy people, after all, are vaccinated more often. That may also be true because the WLZ (nursing homes, for example) are not included, those residents are not registered with general practitioners.

But wait a minute: although there is only a very small part of the population in those nursing homes, almost half of the mortality comes from there! So you leave out a crucial driver of excess mortality into your research into the link between vaccination and excess mortality.

This research population is, in view of the subject, rather unfortunate in its choice.

Withdraw then, I would say. At least if you pretend to be able to claim something about the total population.

Determination of excess mortality

The fact that Nivel realised that they were analysing a special, non-representative group is also evident from the new mortality probabilities that they determined themselves, both for unvaccinated and vaccinated people. Apparently, the normal CBS mortality probabilities were inadequate.

In other words, the study did not use national averages. There is something to be said for that:

• Bij het landelijk gemiddelde zit ook WLZ (m.n. verpleeghuizen) dus het verwachte NIVEL sterftecijfer zou logischerwijs behoorlijk omlaag moeten.
• Furthermore, the vaccinated are slightly healthier than the average (because that includes the unvaccinated), so that risk of mortality could be reduced even further
• For the unvaccinated, the expectation must then be raised sharply to compensate for the lowered expectation in that large group of vaccinated people. Otherwise, we will not achieve total mortality. And that's what happened.

But slightly down in vaccinated people...? Not really. The expected mortality will increase considerably in each group! Almost one and a half times as many among the vaccinated aged 76-80! And among the unvaccinated, they even expect 178% compared to the national mortality rate. Even the long-term care residents pale in comparison. We are on the eve of a huge wave of deaths! Fortunately, it's about 2021. That was not so bad, also in the GP practices. Excess mortality has been limited to about 15% since mid-2021.

They estimated the mortality probabilities based on their own patients. Let's hope this doesn't say anything about the effectiveness of GP care. But how does that expectation work out? The shaded bars are the expectation, the solid bars are the mortality measured afterwards.

Bij 76-80 is de sterfte onder ongevaccineerden bijna 3,5x zo hoog als de verwachting. Met droge ogen constateert Nivel hieruit “een oversterfte van 233%”. En dat over de periode van een jaar…! Dan zit er echt ergens een schroefje los. Dan heb je geen benul van wat cijfers zijn en wat HVE is.

So there are two more reasons to withdraw the study:

• With a mortality rate of 24.7%, this results in 22,158 deaths per year. Almost all 90,000 unvaccinated people over the age of 81 should have died by now. This is not reflected in the CBS figures.
• De niet-gevaccineerden zijn dus ofwel verkeerd ingeschat (en dan zou ik meteen de verwachting van de gevaccineerden nog eens kritisch bekijken) ofwel de overlijdens zijn totaal verkeerd gemeten. Er is geen reden te bedenken (behalve het Healthy Vaccinee Effect) waarom niet-gevaccineerden 3,5x zo vaak overlijden omdat anderen zijn geprikt.

If they had known this in advance, it could have nicely increased the willingness to vaccinate. If that's not a big stick, I don't know what is.

So thousands of extra unvaccinated people are dying (3.5 times as many as expected) because others have been vaccinated. In that sense, "you prick for someone else" is a bit more relative than we were initially led to believe. Any drug with such an adverse reaction to others should be immediately withdrawn from the market.

Have your grandparents been a bit stubborn? Have they perhaps not had a shot? No!? Well then go and say goodbye quickly!

Found: old easter egg

I went looking for the difference in deaths in the non-WLZ population between not vaccinated and vaccinated. In week 25 (April) of 2021, mortality varied widely. In the left (or on mobile the top) graph, the ratio is 7.57 per 100K against 24.4 per 100K. That's a factor of 3.2. Unvaccinated people died 3.2 times as often as vaccinated people after two jabs. (Read at Herman Steigstra how that was possible.)

Then my eye fell on the graph that deals with those who had only had one jab. I even found a factor of 4.5 there! If we follow Nivel's reasoning a bit (although we didn't come up with our own expectation), 1 shot worked even better than two! The biggest difference in deaths, and this too shortly after the vaccinations have been administered. It doesn't get any better than that!

Ik ben bang dat al deze RIVM en CBS rapporten lijden aan dezelfde dataproblematiek en vooral: het negeren van het HVE…

Edit: achteraf is op enkele plekken de term ‘HVE’ aan de tekst toegevoegd. Het fenomeen werd wel omschreven maar niet expliciet benoemd, waardoor dit artikel niet naar boven kwam bij het zoeken naar ‘HVE’.