By Peter A. McCullough, MD, MPH
Dr Panagis Polykretis shared this paper titled Classification bias and impact of COVID-19 vaccination on all-cause mortality the case of the Italian region Emilia-Romagna.pdf. This analysis focussed on the study’s analysis of classification bias—specifically the “case counting window bias”—and its impact on interpreting COVID-19 vaccination effectiveness in real-world mortality data. Review was assisted by Copilot AI. [Classifica…ia-Romagna | PDF]
Background and Motivation
The study investigates biases in real-world assessments of COVID-19 vaccine effectiveness, with a particular focus on the “case counting window bias.” This bias arises when individuals are classified as “unvaccinated” for a period (typically 14 days) after receiving a vaccine dose, under the assumption that the immune response has not yet developed. Events (infections, hospitalizations, deaths) occurring during this window are attributed to the unvaccinated group, potentially distorting mortality statistics and vaccine effectiveness estimates. Previous research highlighted similar biases, such as the “immortal time bias” and “healthy-adherer bias,” which can also affect observational studies. [Classifica…ia-Romagna | PDF]
Objectives
The paper aims to:
Document the presence and impact of the case counting window bias in the Italian region of Emilia-Romagna.
Quantify differences in all-cause mortality between vaccinated and unvaccinated groups within the 14-day post-vaccination window.
Explore whether temporal patterns in vaccination and mortality data suggest misattribution of deaths due to this bias.
Data Sources and Methods
Researchers used data from three institutional sources:
ISTAT (Italian National Institute of Statistics): Provided daily all-cause mortality and population data by age group.
ANV (Anagrafe Nazionale Vaccini): Supplied daily vaccine administration data.
Emilia-Romagna Region: Offered anonymized records of vaccination and mortality for vaccinated individuals, obtained via a FOIA request.
The study period spanned December 27, 2020, to December 31, 2021. The analysis focused on age groups 50–59, 60–69, and 70–79, where trends in vaccination and mortality rates were most clearly aligned. Daily death incidence rates (per 100,000 people) were calculated for vaccinated and unvaccinated groups. Statistical tests (Mann-Whitney U, Shapiro-Wilk) and regression analyses. [Classifica…ia-Romagna | PDF]
Key Findings
1. Significant Mortality Differences
Within the defined time windows, the unvaccinated group consistently showed higher all-cause mortality rates than the vaccinated group (both COVID-19 and non-COVID-19 death). These differences were statistically significant across all age groups analyzed (p < 0.0001). [Classifica…ia-Romagna | PDF]
2. Temporal Patterns and Regression Analysis
Exponential regression models revealed strong associations between the number of vaccine administrations and all-cause death incidence in the unvaccinated group, with the best fit in the oldest age group (70–79, R² = 0.659). Kernel density estimates showed that peaks in vaccine administration were followed by peaks in unvaccinated mortality, suggesting a temporal overlap. However, the “tail” of increased mortality in the unvaccinated group lasted longer than the vaccination surge, especially in older age groups. [Classifica…ia-Romagna | PDF]
3. Attribution of Deaths and Bias Impact
The study argues that the observed mortality differences cannot be explained solely by COVID-19 deaths (which accounted for only 9% of total deaths in Italy in 2021). Nor are there plausible biological reasons for vaccines to reduce non-COVID deaths so dramatically. Instead, the authors attribute the discrepancy to the case counting window bias: deaths occurring within 14 days of dosing in the vaccinated are counted as “unvaccinated,” artificially inflating mortality in that group and deflating it in the vaccinated group. This misclassification has lead to erroneous conclusions about vaccine effectiveness and safety. In other words, it has led to the false narrative that COVID-19 vaccination saved “millions of lives.” [Classifica…ia-Romagna | PDF]
4. Broader Implications
The paper notes that similar classification practices are used in other countries (e.g., UK Office for National Statistics), sometimes with a 21-day window. This suggests the bias may be widespread in observational studies of vaccine effectiveness. The authors call for greater transparency and publication of mortality data by vaccination status to enable more accurate analyses. [Classifica…ia-Romagna | PDF]
Limitations
Population estimates relied on linear interpolation between annual data points.
Differences in data collection methods (e.g., region of vaccination vs. residence) could introduce errors, but these were assumed to be balanced across regions.
The study could not analyze sex-specific mortality due to data limitations.
The number of daily unvaccinated deaths was not directly available.
Conclusions
The study provides evidence that the case counting window bias significantly affects real-world estimates of COVID-19 vaccine effectiveness on all-cause mortality. By shifting deaths occurring shortly after vaccination to the unvaccinated group, this bias can distort epidemiological results and public health decisions. The authors recommend correcting for this bias in future research and urge the release of more granular mortality data by vaccination dose time and cumulative administration. All deaths should be counted from the very first moment of injection and attributed to the vaccinated group. [Classifica…ia-Romagna | PDF]
Please subscribe to FOCAL POINTS as a paying ($5 monthly) or founder member so we can continue to bring you the truth.
Peter A. McCullough, MD, MPH
FOCAL POINTS has partnered with Patriot Mobile to defend your medical freedom. Join Patriot Mobile today!
This situation has been known pretty much since it started. Too bad we had to wait four+ years to have it officially reported.
Amazing how they can twist things to suit their purposes...