Official Autism Safety Signal (1993)

Post #855

NOTE: This post extends and corrects this prior post.

When CDC data mines VAERS or other passive surveillance for the purpose of pharmaco-vigilance (to verify if a product might be unsafe), it uses a type of dis-proportionality analysis called the Proportional Reporting Ratio (PRR):

It’s a ratio of the share of all adverse event reports (AERs) made up by a particular event, from both a target product and from all other products. If a particular vaccine has twice the proportion of all events coming from a specific AE — such as having autism form 3% of all MMR adverse event reports, but only 0.4% for other vaccines.

In this hypothetical, the PRR is (3 / 0.4 =) 7.5

Three conditions must be simultaneously met:

—PRR of at least 2
—chi-squared statistic of at least 4
—at least 3 adverse event reports (AERs) for the target AE in the target product

Below will be a PRR safety-signal test for MMR vaccines for autism in 1993, using official CDC guidelines. We need 4 numbers to fill in the boxes above: A, B, C, and D.

A (autism AERs for MMR vaccine only)

NOTE: Note how over half of the onset dates above are by Day 2 post-dose

B (all other AERs for MMR - minus autism)

Notice that the “all other” (B) for MMR vaccines becomes (1973 - 164 =) 1809.

C (autism AERs for all other - minus MMR ones)

Notice that “all other autism” (C) is computed as (178 - 164 =) 14.

D (all other AERs for all other jabs - minus MMR)

With 9698 total AERs for all products combined, you subtract the 1973 from MMR and also the 14 autism AERS from “all others” to get D = 9698 - 1973 - 14 = 7711. Here is how it looks when put together and analyzed:

Orange boxes show that all three conditions were met for an official PRR test (AERs = 3 or more; PRR = 2 or more; chi-squared = 4 or more). The formula for the computation of PRR in cell B8 shows up at top.

In 1993, compared to all other vaccine products, autism made up 46 (45.9) times the share of all adverse event reports for the MMR vaccines. This is a strong safety signal, and it occurred well before the pivotal 1998 Wakefield et al. study that had suggested a link between MMR and autism.

This means that this strong safety signal cannot be “explained away” by saying that people were getting all hyped-up about autism and MMR — whipped into a veritable reporting frenzy by reading the Wakefield study — and so they started “over-reporting” it.

Instead, this is an organic, legitimate safety signal which CDC apparently overlooked.

EDIT: Superficial spreadsheet error fixed