Together with Dr. Marco Alessandria, Dr. Giovanni Trambusti, Dr. Giovanni M. Malatesta and Dr. Alberto Donzelli, we have recently published a very important peer-reviewed scientific study entitled Classification bias and impact of COVID-19 vaccination on all-cause mortality: the case of the Italian Region Emilia-Romagnain which we demonstrate how certain statistical biases have caused an overestimation of the effectiveness and safety of COVID-19 vaccines.
The study analyzed the mortality data by vaccination status in Emilia-Romagna between December 2020 and December 2021, using official data from ISTAT and data obtained from the National Vaccine Registry and the Emilia-Romagna Region. This last source was made accessible thanks to a FOIA request presented by the lawyer Lorenzo Melacarne and released pursuant to art. 5, paragraph 2 of Legislative Decree no. 33/2013. The data, completely anonymized at the source, concerns the entire population, divided by age, and distinguishes between vaccinated (with at least one dose) and unvaccinated. Finally, specific time windows were selected to analyze the mortality trend in relation to the vaccination campaign in the 50-59, 60-69 and 70-79 age groups.
By analyzing the data we identified one statistical bias which can substantially alter real assessments of vaccine efficacy and safety, known as “distortion of the case counting window” (from English case-counting window bias). This distortion, theorized by Fung and co-authors, occurs because people are classified as “unvaccinated” in the first 14 days after vaccination (period considered necessary for the complete development of the immune response). Consequently, any adverse events (with the exception of cases of anaphylactic shock, which has tended to be attributed to vaccination) and deaths from the most varied causes that may occur in this time window are erroneously attributed to the non-vaccinated group, artificially increasing their mortality rate and simultaneously underestimating mortality among those vaccinated. In particular, by analyzing daily data on all-cause mortality and vaccine administration in the Emilia-Romagna Region, we found a clear temporal coincidence between vaccination campaigns and peak deaths among those incorrectly classified as unvaccinated during this critical time window (Figure 1).
Our analysis highlighted significant differences in mortality between the vaccinated and unvaccinated groups during the 14 days post-vaccination during which misclassification occurs. Importantly, these differences cannot be explained by COVID-19 deaths alone, which they accounted for approximately 9% of all deaths in Italy in 2021. Excluding COVID-19-related deaths, the disparity between groups remains significant, indicating systematic misclassification, rather than real vaccine benefit. Although this effect was detected in all the groups analyzed, we observed that the difference decreases with age, probably due to the increase in comorbidities in the elderly, which influence the overall risk of mortality (for further details please refer to the article, published in open access and freely accessible).
Our results suggest an effect “harvest”, whereby vulnerable individuals die shortly after vaccination, but their deaths are incorrectly counted among the unvaccinated. This misclassification hides potential serious adverse events related to vaccination that occur in the short term, such as serious allergic reactions, cardiovascular events or autoimmune responses.
This distortion is potentially widespread internationally and affects all countries that have adopted a similar time window for classifying individuals as vaccinated or unvaccinated. For example, British healthcare practices have counted people as unvaccinated in the first 14 to 21 days after vaccination. This systemic distortion alters vaccine safety profiles, excluding early adverse events from the vaccinated group.
The distortion of the case counting window is related to another phenomenon well known in observational research, the immortal time distortion (from English immortal time bias). Professors Norman Fenton and Martin Neil were among the first to identify how these distortions shift cases and deaths in ways that exaggerate the apparent effectiveness and safety of vaccines, creating misleading temporal categorisations. Prof. Fenton himself defined these manipulations as “cheap makeup” (from English cheap trick) – a statistical illusion that artificially increases the perception of vaccine effectiveness.
In conclusion, our study represents the first study published in the peer-reviewed scientific literature, which analyzes real mortality data by vaccination status, clearly highlighting how not to correct these crucial statistical distortions, such as that of case counting window and teaimmortal timeleads to one overestimation of benefits and to one underestimation of adverse reactions linked to vaccines. Consequently, to ensure accurate assessments and reliable public health decisions, it is essential to correct these biases and have up-to-date and accurate data on the vaccination status of individuals. Finally, based on this evidence, all studies on vaccine effectiveness should be reevaluated taking these aspects into account to ensure a transparent and realistic assessment of the safety and effectiveness of vaccines.
