Are non mRNA Covid vaccines safe ?

The short version: the evidence for non-mRNA COVID-19 vaccines shows real, documented harms that are not rare enough to dismiss, and the safety data is built on designs that hide the full picture.

The evidence here covers adenovirus-vectored vaccines (AstraZeneca, J&J) and one inactivated virus vaccine (CoronaVac). The most serious documented harm is vaccine-induced immune thrombotic thrombocytopenia (VITT), a clotting disorder with low platelets. The WHO global database found about 0.21 thrombotic events per million doses overall [2]. But that number comes from passive surveillance, which catches only a fraction of real events. The 11 initial VITT cases reported in the New England Journal of Medicine all had moderate-to-severe thrombocytopenia and abnormal clotting; 5 had disseminated intravascular coagulation [2]. The mechanism is now understood: antibodies against platelet factor 4 (PF4) activate platelets through the FcγRIIa receptor, the same pathway as heparin-induced thrombocytopenia but with stronger binding [2, 9]. Onset is typically 5 to 20 days after vaccination [2].

Other documented harms include immune thrombocytopenia (ITP) without clotting, with one case showing a platelet count of 2 x 10^9/L (normal is 150-450) one week after AstraZeneca [8]. Myocarditis has been reported after adenovirus-vectored vaccines as well, with one case showing subnormal heart function three months after a single dose [5]. A systematic review found new-onset arthritis after both adenovirus and inactivated virus vaccines, with cases appearing days to weeks after vaccination [7]. The Brazil cohort study showed that adenovirus-vectored vaccines produce more systemic reactions than inactivated virus: 25.8% fever for ChAdOx1 versus 2.3% for CoronaVac, 30.9% muscle pain versus 5.9% [3].

The inactivated virus vaccine (CoronaVac) has a milder acute reaction profile but the evidence here does not include any long-term safety data for it. The adenovirus-vectored vaccines have been linked to lymphoma appearing within days of vaccination in case reports [21], though these are individual cases, not population studies.

The key design problem: every safety study here compares vaccinated people to other vaccinated people, or uses passive surveillance. No study in this evidence set compares vaccinated to unvaccinated over time. That means the true rate of harms is unknown. The manufacturer of AstraZeneca and J&J hold liability shields, so the market force that would push for safer products does not exist.

My call: non-mRNA COVID-19 vaccines carry documented risks of serious clotting disorders, immune platelet destruction, heart inflammation, and autoimmune reactions that are not vanishingly rare. The safety evidence is built on passive surveillance and vaccinated-to-vaccinated comparisons, which means the true harm rate is likely higher than reported. Confidence: moderate.