AP analysis doesn’t prove COVID-19 vaccines prevent deaths
https://trialsitenews.com/ap-a…-vaccines-prevent-deaths/
Dr. Ron Brown
June 30, 2021
Heralded across the media, pseudo-epidemiologists at Associated Press (I thought they were mostly journalists over there) apparently skipped the peer-review process and released findings from their flawed analysis claiming extraordinarily high COVID-19 vaccine effectiveness in preventing deaths. In their press release, the pseudo-epidemiologists/journalists removed any doubt that AP is yet another news agency spreading misleading public health information. Even Dr. Rochelle Walensky, Director of the Centers for Disease Control and Prevention, which provided data used in the AP analysis, appeared dazzled by AP claims of vaccine effectiveness, prompting her to pronounce that most deaths from COVID-19 are now entirely preventable.
On June 24, 2021, Carla K. Johnson and Mike Stobbe from AP wrote, “Nearly all COVID-19 deaths in the U.S. now are in people who weren’t vaccinated, a staggering demonstration of how effective the shots have been.” Nearly all COVID deaths in US are now among unvaccinated (apnews.com). Staggering demonstration, indeed! In reality, this claim is more likely a staggering demonstration of limitations and biases inherent in observational studies like the AP analysis.
Unlike randomized controlled trials, which analyze participant data from experiments or interventions organized by researchers, observational studies use data obtained from the real world, where conditions are not controlled by researchers. Observational Studies: Uses and Limitations | SpringerLink. Because observational studies cannot rule out unintended influences on the study results, known as confounding factors, a limitation of observational studies is that they cannot prove causation. The level of evidence from these studies is well below causation demonstrated in randomized controlled trials.
You might think that a lower standard of evidence in the observational AP analysis had little effect when compared to results of COVID-19 mRNA vaccine randomized controlled trials. After all, the AP analysis corroborated the exceptional results found in the randomized trials of very high vaccine efficacy (effectiveness under controlled conditions), which were approximately 95%.
The problem is that reported vaccine efficacy, or relative risk reduction measures in COVID-19 mRNA vaccine trials, did not include absolute risk reduction measures of approximately 1%, which is a more meaningful outcome for clinical and public health applications. Medicina | Free Full-Text | Outcome Reporting Bias in COVID-19 mRNA Vaccine Clinical Trials (mdpi.com). Furthermore, the vaccine trials’ clinical endpoints were laboratory confirmed infections with a mild symptom in otherwise healthy participants, not hospitalizations and mortality in participants with severe infections and underlying health conditions. If the vaccines aren’t really so efficacious after all, even for mild infections, what explains the “staggering demonstration” of vaccine effectiveness to prevent deaths claimed in the AP observational analysis? The answer lies within biases often found in observational studies.
World Health Organization had this to say about biases within observational studies used to assess vaccine effectiveness: “Due to lack of randomization of vaccination in real-world settings, all observational study designs are subject to bias because vaccinated persons often differ from unvaccinated persons in their disease risk, independent of vaccination.” Evaluation of COVID-19 vaccine effectiveness: interim guidance, 17 March 2021 (who.int)
Several critical differences in disease risk potentially affected vaccinated and unvaccinated persons in the AP analysis. People from lower socioeconomic groups have higher COVID-19 incidence and mortality, which they have had even before COVID-19 vaccines were made available to the public. Association of Social and Demographic Factors With COVID-19 Incidence and Death Rates in the US | Health Disparities | JAMA Network Open | JAMA Network. Perhaps due to environmental, economic, social, and lifestyle factors, people in these groups are more susceptible to diseases like COVID-19.
Furthermore, people with lower socioeconomic status are more likely to have vaccine hesitancy. Individual and social determinants of COVID-19 vaccine uptake | BMC Public Health | Full Text (biomedcentral.com). Of relevance, healthy vaccinee bias, the tendency for healthier people to seek out vaccinations, is concordant with the tendency of unhealthier people to avoid vaccinations, regardless of socioeconomic status. Frequency and impact of confounding by indication and healthy vaccinee bias in observational studies assessing influenza vaccine effectiveness: a systematic review (nih.gov)
The association of vaccine hesitancy with increased COVID-19 mortality in the AP observational analysis does not prove causation. The evidence is anecdotal—there is no proof that lack of vaccination had anything to do with higher disease mortality in lower socioeconomic groups and in people with compromised health conditions, within whom higher mortality rates were already evident before vaccination availability. Moreover, not only did the AP analysis confuse causation with correlation in assessing the association of vaccine hesitancy and disease mortality, the flawed findings of the AP assessment were overgeneralized to the entire population.
Experienced epidemiologists do not make these sorts of rookie mistakes. Unlike some journalists, fully trained epidemiologists account for differences in studied groups by adjusting results accordingly for each group. Estimating an accurate adjustment is a difficult task at best, even for the pros. Media reports on findings from the AP analysis make no mention of how limitations and biases in the AP observational analysis were managed and adjusted by the journalists, or even if they were managed at all.