A lab director has reportedly been reassigned and other heads at the U.S. Center for Disease Control are likely to roll after an incident earlier this month in which at least 75 staff members may have accidentally been exposed to live samples of anthrax being transported from one lab to another. The samples were supposed to have been inactivated before they were transferred—and it is unclear why they were not.

Naturally there will be an investigation. But there is a bigger issue at play here beyond just why proper procedure was not followed. And it’s one that Scientific American last raised in 2008 with "Postal Anthrax Aftermath: Has Biodefense Spending Made Us Safer?" and, at greater length, in 2007 with "Laboratory Letdowns."

Put another way: which is the more likely threat to public safety—a series of accidental releases of deadly organisms from the high-level biodefense labs that have proliferated in the wake of the anthrax attacks of 2001 or a single, but much bigger, intentional release by an actual terrorist network?

So far, the lab releases of biological agents have proven far more deadly: the 2001 attack, which killed five people, was itself probably carried out using a research strain of anthrax, probably from a U.S. laboratory. In 1978, a medical photographer in England died of smallpox, which had somehow escaped from a laboratory at the University of Birmingham. At least three people died in 1971 after smallpox was released (accidentally or on purpose, no one knows) from a Soviet-era bioeweapons lab in what is now Kazakhstan.

Update (6/24/2014): As Kelly Hills pointed out on Twitter, my list doesn't even include the Sverdlovsk accidental release of anthrax, which killed 70-100 people, in 1979.

A 2013 report from the General Accounting Office shows that the U.S. still does not have a good sense of how many high security biological agent laboratories it needs. In the rush to protect the country from biothreats after the 2001 attacks, no one bothered to do figure out what the actual need was so we may already have too many (or too few) high-containment labs. Meanwhile, auditors at the U.S. Department of Agriculture found significant lapses in safety protocols in their 2012 report, available here.