Do childhood vaccines save lives, or reduce serious illness and its effects?

The evidence shows that several childhood vaccines reduce hospitalizations and deaths from their target diseases, but the data is dominated by modeling and observational studies, not randomized trials, and the net effect on all-cause mortality is rarely measured.

The strongest evidence comes from pneumococcal and rotavirus vaccines. In England, after the pneumococcal vaccine was introduced, invasive pneumococcal disease in children dropped to about a quarter of its pre-vaccine rate, and pneumococcal pneumonia fell to about a fifth [3]. Rotavirus vaccination in the US cut rotavirus-coded hospitalizations from 15 per 100,000 children to about 5 per 100,000 [2]. A global modeling study estimated that rotavirus vaccination prevented about 139,000 deaths between 2006 and 2019, mostly in low-income countries [1].

Measles vaccination is credited with the largest number of lives saved. One modeling study estimated that measles vaccine alone prevented about 93.7 million deaths globally between 1974 and 2024 [1]. When coverage drops, outbreaks return: in 2023, global measles deaths reached about 107,500, with a 200% increase in cases in Europe [1].

But there are important limits to this evidence. Nearly all of it comes from modeling or ecological before-after comparisons, not from randomized controlled trials that compare vaccinated children to unvaccinated children. The studies do not measure all-cause mortality in the vaccinated versus the unvaccinated. A vaccine that prevents one cause of death could theoretically trade it for another, and this kind of analysis would miss that. The pneumococcal vaccine also shows a pattern of serotype replacement: as vaccine-type strains decline, non-vaccine strains increase, which can partially offset the benefit [4, 5].

The evidence for benefit is strongest for measles, rotavirus, and pneumococcal vaccines in settings where the diseases are common. The evidence is weaker for vaccines targeting diseases that are already rare in a given population. The studies that exist were not designed to measure net harm, and the absence of that measurement is a real gap.

My call: the evidence supports that these vaccines reduce targeted disease and death, especially in high-burden settings, but the lack of randomized trials and all-cause mortality tracking means the net effect is less certain than the headlines suggest. Confidence: moderate.