Confounding by indication occurs in an observational study when some patients have conditions that indicate a higher need to be prescribed the tested treatment than other patients. This could cause patient response to the treatment to differ based on patient conditions, thus confounding the study results: Confounding by Indication. For example, in a recently published November 2021 Israel study, a group of people with more serious conditions had higher priority to receive the Pfizer COVID-19 vaccine earlier in January and February compared to another group without serious conditions who received the vaccine later in March and April. Outcomes of both groups were then tested in June and July, and the incidence of breakthrough infections was found to be 51% higher in the Early Vaccinees group compared to the Late Vaccinees group, which increased to 126% higher when comparing the earliest vaccinees in January to the latest vaccinees in April: SARS-CoV-2-breakthrough infections to time-from-vaccine.

Subscribe to the Trialsitenews "Vaccine" Channel

No spam - we promise

These study results give the appearance that the effectiveness of the Pfizer vaccine waned to a greater extent in the Early Vaccinees group over a period of six months compared to four months in the Late Vaccinees group. Even more vaccine waning was implied by the higher infection rates and longer difference in time-from-vaccine between the January and April vaccinees. But the authors of the study cautioned, “As chronically ill patients were given priority for vaccination, confounding by indication should be considered when interpreting the study results.”

In reality, vaccinees in the earlier group were at higher risk for COVID-19 infection to begin with. Therefore, due to confounding by indication, this study does not prove that effectiveness of the Pfizer vaccine wanes over time.

Furthermore, with an absolute risk reduction of only 0.7% in the clinical trial, the Pfizer vaccine has hardly any efficacy at all. Therefore, any waning effectiveness of the vaccine expected within the population is practically nil. The same applies to the Moderna vaccine with an absolute risk reduction of only 1.1%: Outcome Reporting Bias in COVID-19 Vaccine Trials.

Observational studies of vaccine effectiveness, like the Israel study, lack the controlled experimental conditions of randomized controlled trials. Effectiveness studies cannot prove causation and are highly susceptible to confounding and selection bias, even when researchers attempt to adjust the study results to control for known confounding factors like age, sex, comorbidities, etc.

On the other hand, randomization in a vaccine clinical trial neutralizes all known and unknown confounding factors by equally distributing these influences between the vaccine and placebo groups, so the only differences in the group outcomes are due to the efficacy of the vaccine alone.

Unfortunately, reporting clinical trial outcomes is also suspect to bias, which cannot be controlled through randomization if the trial’s absolute risk reduction is not reported, as occurred in the COVID-19 mRNA vaccine clinical trials.