The Protective Effect of Coronavirus Disease 2019 (COVID-19) Vaccination on Postacute Sequelae of COVID-19: A Multicenter Study, 2022, Zisis et al

Full title: The Protective Effect of Coronavirus Disease 2019 (COVID-19) Vaccination on Postacute Sequelae of COVID-19: A Multicenter Study From a Large National Health Research Network

Abstract

Background
Coronavirus disease 2019 (COVID-19) vaccines have been proven to decrease the severity of acute-phase infection; however, little is known about their effect on postacute sequelae of COVID-19 (PASC).

Methods
Patients with confirmed COVID-19 diagnosis and minimum age of 18 years with 3-month follow-up postdiagnosis between 21 September 2020 and 14 December 2021 were identified from the TriNetX Research Network platform. The primary outcomes consisted of new-onset or persistent symptoms, new-onset diagnoses, and death and were compared between vaccine and no-vaccine groups.

Results
At baseline, 1 578 719 patients with confirmed COVID-19 were identified and 1.6% (n = 25 225) completed vaccination. After matching, there were no differences (P > .05) in demographics or preexisting comorbidities. At 28 days following COVID-19 diagnosis, the incidence of hypertension was 13.52 per 1000, diabetes was 5.98 per 1000, thyroid disease was 3.80 per 1000, heart disease was 15.41 per 1000, and mental disorders was 14.77 per 1000 in the vaccine cohort. At 90 days following COVID-19 diagnosis, the relative risk of hypertension was 0.33 (95% confidence interval [CI], .26–.42), diabetes was 0.28 (95% CI, .20–.38), heart disease was 0.35 (95% CI, .29–.44), and death was 0.21 (95% CI, .16–.27). Differences in both 28- and 90-day risk between the vaccine and no-vaccine cohorts were observed for each outcome, and there was enough evidence (P < .05) to suggest that these differences were attributed to the vaccine.

Conclusions
Our data suggest that COVID-19 vaccine is protective against PASC symptoms, new onset of health conditions, and mortality.

Open access, https://academic.oup.com/ofid/article/9/7/ofac228/6582238

 

They report a risk ratio of 0.32 (.25–.42) for new diagnosis of malignant neoplasm (cancer) for vaccinated versus unvaccinated people within 28 days of a Covid test and 0.23 (.17–.32) within 90 days. This seems pretty damning because cancer doesn't just show up in 28 or 90 days (most cancers take years to develop). So the four times higher risk of diagnosis of cancer for the unvaccinated after Covid is likely due to bias in who got vaccinated or had vaccination recorded in the health record (vaccination is oddly low in their data). This was actually the outcome in which vaccination was supposedly the most protective out of all the diagnoses and symptoms in the study (except thyroid disease and death at 90 days), meaning all the other less strong results are probably bunk too.

So there's some pretty bad bias in this study and it doesn't tell us anything about PASC or Long Covid and vaccination. It's odd that they even highlight cancer in their discussion as if that would be a reasonable outcome 28 or 90 days after Covid. There could be some reason that cancer would be more likely to be diagnosed after more severe Covid (which would be more likely for the unvaccinated) like being in the hospital, but that seems like a stretch to account for the risk being four times higher.

I wrote a little more about this here:

https://twitter.com/user/status/1547293015877619713