Jab fears explained: a base rate fallacy

Episode Summary

Executive Summary: The episode explains why rising counts of COVID infections among vaccinated people do not necessarily mean vaccines are failing. It uses the base rate fallacy to show that as vaccination coverage grows, more cases can appear in vaccinated people even when vaccines greatly reduce individual risk and continue protecting against severe disease.

Main Topics: Base rate fallacy explained (Priority: 5/5): The program introduces base rate bias as the statistical mistake of ignoring how large the vaccinated population has become when interpreting infection counts. Why more breakthrough cases can look alarming (Priority: 5/5): As more people get vaccinated, a larger share of the remaining population is vaccinated, so infections will increasingly occur in that group even if vaccines are effective. Israel as a real-world example (Priority: 4/5): A reported figure that 40-50% of new cases were among vaccinated people is shown to be misleading without considering that about 85% of adults were vaccinated. Infection rates vs. case share (Priority: 5/5): The transcript distinguishes between the percentage of infections among vaccinated people and the actual infection rate within vaccinated and unvaccinated groups. Need for surveillance data (Priority: 4/5): The episode argues that understanding vaccine performance nationally requires tracking vaccination status among infected people, which is incomplete in the U.S. but stronger in places like Israel and the U.K. Vaccines still protect against severe disease (Priority: 5/5): Despite breakthrough infections, the evidence cited indicates vaccines remain effective at preventing hospitalizations and deaths, with additional support from lab studies.

Key Arguments: A larger share of infections among vaccinated people can be expected when vaccination coverage is high; this is a denominator problem, not proof of vaccine failure. The statement 'half of the infected people were vaccinated' is not the same as 'half of vaccinated people were infected.' In a hypothetical 100-person community where 85 are vaccinated, 50% of cases being vaccinated still corresponds to a much lower infection rate among the vaccinated than the unvaccinated. To assess whether vaccines are losing effectiveness or if a variant is escaping immunity, public health authorities need three numbers: how many are vaccinated, how many are infected, and how many infected people are vaccinated. The U.S. lacks systematic tracking of vaccination status among cases, limiting national-level efficacy estimates outside clinical trials. Israel and the U.K. have better surveillance systems and provide evidence that vaccines continue to work against severe disease. Breakthrough infections do happen because vaccines are not perfect, but they do not negate broad vaccine effectiveness. Epidemiologists triangulate surveillance data with lab and petri-dish studies to assess immune protection against variants such as Delta.

Data Points: Vaccinated adults in Israel: About 85% - Used to explain why many new cases could still appear among vaccinated people. New COVID cases among vaccinated in Israel: About 40-50% - Reported by Israel's Health Ministry Director General, then analyzed as potentially misleading without context. Community vaccination rate in hypothetical example: 85 out of 100 people vaccinated - Dr. Caitlin Jetalina's example showing how case shares can mislead. New cases in hypothetical example: 4 cases - Used to illustrate base rate fallacy in interpreting breakthrough infections. Vaccinated infections in hypothetical example: 2 of 85 vaccinated people infected - Shows an infection rate of 2% among vaccinated people. Unvaccinated infections in hypothetical example: 2 of 15 unvaccinated people infected - Shows an infection rate of 13% among unvaccinated people. Infection rate among vaccinated in example: 2% - Calculated from 2 infections among 85 vaccinated people. Infection rate among unvaccinated in example: 13% - Calculated from 2 infections among 15 unvaccinated people.

Pivotal Quotes: "The vaccines work really well, but we know that they're not perfect." — Dr. Caitlin Getalina: Explaining why breakthrough infections can still occur even when vaccines are effective. "This is very different than half of vaccinated people were infected." — Dr. Caitlin Getalina: Clarifying the common misunderstanding behind the Israeli case numbers. "What that means is if a new variant comes along, for example, Delta, we want to ensure that the vaccines continue to work." — Dr. Caitlin Getalina: Describing why ongoing surveillance data is essential.

Implications: Listeners should interpret breakthrough-case headlines carefully and always check denominators. Public health agencies need better tracking of vaccinated infections to monitor variants and vaccine performance, especially for protection against severe outcomes.