Vaccine scientists and the public health community cautiously and occasionally will admit that vaccines can cause adverse reactions just like "any other medication or biological product." Although experts are less willing to openly disclose the fact that adverse reactions can and do include death, one has only to look at reports to the U.S. Vaccine Adverse Event Reporting System (VAERS) to see that mortality is a possible outcome. From 1990 through 2010, for example, VAERS received 1,881 reports of infant deaths following vaccination, representing a 4.8% mortality rate. Moreover, analysts acknowledge that VAERS, as a passive surveillance system, is subject to substantial underreporting. A federal government report from 2010 affirms that VAERS captures only about 1% of vaccine adverse reports.
On the international frontier, the public health community—with the World Health Organization (WHO) in the vanguard—previously used a six-category framework to investigate and categorize serious adverse events following immunization (AEFI), including death. Guided by this tool, public health teams examined temporal criteria and possible alternative explanations to determine whether the relationship of an AEFI to vaccine administration was "very likely/certain," "probable," "possible," "unlikely," "unrelated," or "unclassifiable."
In 2013, the WHO's Global Advisory Committee on Vaccine Safety discarded the prior tool, ostensibly because users "sometimes [found it] difficult to differentiate between 'probable,' 'possible,' and 'unlikely' categories." The WHO enlisted vaccine experts to develop a "simpler" algorithm that would be more readily "applicable" to vaccines. The resulting four-category system now invites public health teams to classify an AEFI as either "consistent," "inconsistent," or "indeterminate" with a vaccine-related causal association or as "unclassifiable." Despite the patina of logic suggested by the use of an algorithm, "the final outcome of the case investigation depends on the personal judgment of the assessor" [emphasis added], especially (according to the tool's proponents) when the process "yields answers that are both consistent and inconsistent with a causal association to immunization."