A Response to Defenses Against our Critique on Cohort Composition in the Kim et al. COVID-19 Vaccine Study (Biomark Res, 13:114, 2025)

Short Communication Rebuttal to Defenses Citing “Internally Matched Cohorts, Time-to-Event Models, and Person-Time Denominators” - A Response to Defenses Against our Critique on Cohort Composition in the Kim et al. COVID-19 Vaccine Study (Biomark Res, 13:114, 2025) Author: Marco Roccetti Affiliation: University of Bologna, Department of Computer Science and Engineering, Bologna, Italy Correspondence: [email protected] Introduction and Disclaimer The details of our quantitative critique on the methodology of the Kim et al. study [1] can be found in our preprints [2-6]. Our analysis is not focused on whether COVID-19 vaccines are biologically associated with Serious Adverse Events (SAE) or cancer risk. Instead, our core finding is that the cohort used in the scrutinized study is fundamentally flawed and unreliable due to a severe methodological error in participant selection. I will now explain, using quantitative metrics, why the study's reported association is a statistical artifact and not a genuine biological signal. The$Core Thesis: External Validity Vitiates Internal Results The central argument is that the study by Kim et al. failed the fundamental requirement of external validity (Generalizability – STROBE Item 21) [7]. A failure of external validity, demonstrably caused by severe selection bias, renders any subsequent statistical attempts at internal validity (such as Propensity Score Matching) irrelevant. The Paradox The cohort's overall Crude Incidence Rate (CR) for cancer was 26% lower than the established national average rate for South Korea (40.78 vs. 55.02 per 10,000 in 2022) [8]. This drastic suppression means the cohort does not represent the target population's true baseline risk. Similar suprression holds if one consider previous years [9,10]. The Quantitative Evidence of Asymmetric Bias The source of this suppression is a pronounced, asymmetric underrepresentation in the highrisk, non-vaccinated subgroup: • Overall Demographic Deficit: The total cohort shows a relative deficit larger than 32% in the high-risk demographic group (individuals aged >= 65) compared to the South Korean national standard of 18.0% [11]. • The Critical Asymmetry: The most crucial finding is the incidence deficit within the non-vaccinated elderly participants (>= 65 years old): This subgroup's observed cancer incidence rate was 85.2 per 10,000, compared to the official national rate of 155.2 per 10,000 [12]. This constitutes an unprecedented larger than 45% undercount of expected cancer cases in the reference group of unvaccinated individuals. The Mechanism of the Artifact The 45% deficit in the non-vaccinated elderly group is the specific mechanism that mathematically generated the spurious association: The selection process (likely a flawed Propensity Score Matching procedure) led to the non-vaccinated reference group being disproportionately composed of younger, healthier, lower-risk individuals (the Healthy User Bias) [13]. By artificially suppressing the baseline cancer risk in the reference group, the lesssuppressed rate observed in the vaccinated group was made to falsely appear as an excess risk. Addressing the Rebuttals: Precision Cannot Correct Bias Critics argue that the study's use of internally matched cohorts, time-to-event models, and person-time denominators validates their findings. This argument is unsound because it confuses precision (Internal Validity) with truthfulness (External Validity). • Internal vs. External Validity: These sophisticated techniques address Internal Validity (precision and balancing within the cohort). However, as demonstrated, the cohort was fundamentally compromised by selection bias, a failure of External Validity. • The Failure of Denominators: A fundamental flaw in the denominator (the population at risk) cannot be corrected by advanced statistical modeling. If the population denominator is missing 45% of the expected cancer cases in the highest-risk stratum, the resulting incidence rate will be inherently flawed. Precision cannot correct for bias. Conclusion The defense of the Kim et al. study based on internal validity metrics is insufficient. Our quantitative analysis demonstrates that the signal of increased cancer risk is the likely consequence of severe, asymmetric selection bias that suppressed the true baseline risk. A methodologically rigorous and demographically balanced cohort would predictably show no statistically important difference in cancer incidence between groups. References [1] Kim HJ, et al. (2025) 1-year risks of cancers associated with COVID-19 vaccination: a large population-based cohort study in South Korea. Biomark Res. 13(114). DOI: 10.1186/s40364-025-00831-w [2] Roccetti M. (2025) A Critical Note on Contradictions in South Korean Cancer Incidence Rates: The Paradox of Crude Rates Derived from the Kim HJ et al. Cohort (Biomark Res, 13:114, 2025) Showing Concurrent Increases in the Vaccinated and Overall Decrease. Preprint n. 202510.0883, Preprints.org. DOI:10.20944/preprints202510.0883.v1. [3] Roccetti M. (2025) Addendum to “A Critical Note on Contradictions in South Korean Cancer Incidence Rates: The Paradox of Crude Rates Derived from the Kim HJ et al. Cohort (Biomark Res, 13:114, 2025)” - On Possible Sampling Bias and Inverted Propensity Score Matching. Preprint n. 202510.1664, Preprints.org. DOI:10.20944/preprints202510.1664.v1. [4] Roccetti M. (2025) Methodological Considerations on the External Validity of the Kim HJ et al. COVID-19 Vaccination Study (Biomark Res 13:114, 2025): A Quantitative Analysis. Preprint n. 17658026. Zenodo.org. DOI: 10.5281/zenodo.17434738. [5] Roccetti M. (2025) A Biostatistical Reappraisal Unveiling the Mechanism Behind Apparent Cancer Risk Signals in a COVID-19 Vaccinated Cohort. Preprint n. 17651590. Zenodo.org. DOI: 10.5281/zenodo.17508346. [6] Roccetti M. (2025) Inferential Z-Test Validation of Dual Structural Bias in Cancer Risk Assessment within Large COVID-19 Vaccine Cohorts. Preprint n. 17688859. Zenodo.org. DOI: 10.5281/zenodo.17434738. [7] Elm E, Altman DG, Egger M, et al. (2007) The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. PLoS Med.4(10):e296. DOI: 10.1016/j.jclinepi.2007.11.008. [8] Park EH, Jung K-W, Park NJ, et al. (2025) Cancer Statistics in Korea: Incidence, Mortality, Survival, and Prevalence in 2022. Cancer Res Treat. 57(2):312-330. DOI: 10.4143/crt.2025.264. [9] Kang MJ, Jung K-W, Bang SH, et al. (2023). Cancer Statistics in Korea: Incidence, Mortality, Survival, and Prevalence in 2020. Cancer Res Treat., 55(2):385-399. DOI: 10.4143/crt.2023.447. [10] Park EH, Jung K-W, Park NJ, et al. (2024). Cancer Statistics in Korea: Incidence, Mortality, Survival, and Prevalence in 2021. Cancer Res Treat., 56(2):357-371. DOI: 10.4143/crt.2024.253. [11] World Bank. Population ages 65 and above (% of total population) - Korea, Rep. World Population Prospects, United Nations (UN). [Accessed 2025 Nov 2]. Available from: https://data.worldbank.org/indicator/SP.POP.65UP.TO.ZS?locations=KR. [12] Statista. South Korea: Cancer crude incidence rate by age, 2022. Statista; 2024. [Accessed 2025 Nov 2]. Available from: https://www.statista.com/statistics/1440818/southkorea-cancer-crude-incidence-rate-by-age/. [13] Chemaitelly H, Ayoub H, Coyle P, et al. (2025) Assessing healthy vaccinee effect in COVID-19 vaccine effectiveness studies: a national cohort study in Qatar. eLife 2025; 14:e103690. DOI: 10.7554/eLife.103690.