Lab accident

I wrote the post on Texas A&M that appears below while sitting in an airport lounge, sans connection. In the car on the way home from the airport late last night I heard on the BBC World Service the following story, to which I found links provided by two kind readers upon being reconnected (hat tip PM and KR).

Consider this BBC story a preamble to the Texas A&M post that follow it:

The strain of foot-and-mouth disease found at a Surrey farm has been identified, [the Department for Environment, Food and Rural Affairs (Defra)] has said.

The strain in infected cattle is identical to that used for vaccines at the Institute for Animal Health at Pirbright, three miles from the farm.

Defra could not say the laboratory was the source but has increased the size of the protection and surveillance zones covering farms in the area.

An urgent assessment of bio security has begun at the institute.

The strain is not one normally found in animals but is used in vaccine production and in diagnostic laboratories.

In a statement [Defra] said: "The present indications are that this strain is a 01 BFS67-like virus, isolated in the 1967 foot-and-mouth disease outbreak in Great Britain."

The strain was used in a vaccine batch manufactured on 16 July by a private pharmaceutical company Merial Animal Health.

The firm shares Pirbright with the government's Institute for Animal Health (IAH), which conducts research into foot-and-mouth and where the strain is also present. (BBC)

Foot and Mouth Disease is not a public health problem but it is a disease of extraordinary importance to the livestock industry and is taken with great seriousness. Strict controls and movements in and out of affected areas with intense biosecurity measures are the normal response, so the release of FMD into the livestock community from a research or vaccine laboratory is a major event.

With that as introduction, here is the post I wrote earlier for today:

The biodefense laboratory snafu (see here, here, here and here) at Texas A&M got a little worse this week when it was found that some of the personnel working with "select agents" (potential biowarfare agents) didn't have clearance to do this work. Previously Texas A&M lab workers had become exposed, in one case resulting in illness (Brucellosis) and in the other with showing signs of having been infected (antibodies to Q fever), although no illness was reported. Is this a sign of unusual incompetence or laxity at Texas A&M? Yes and no. Clearly there were errors made, erros that resulted in exposure and illness. But Texas A&M isn't unusually lax or incompetent. They are just normally so. In this sense, scientist defenders of Texas A&M's record are probably right.

Infectious-disease researchers say that it's impossible to prevent all lab accidents and that these numbers are relatively low. The fact that so few cases involved select agents, and that hardly any resulted in illness, is a sign that Texas universities take quick action and carefully follow protocol when accidents occur, the researchers argue.

"They're very rare - and most of them aren't even the kinds of exposure we're required to report," said Dr. Stanley Lemon, who directs the federally funded Galveston National Laboratory at the UT Medical Branch. "When they happen, we take them very seriously. ... The biggest hazard in a laboratory like these is to the lab worker, and it's being stuck by a needle. We work extremely hard to make the environment as safe as possible for these individuals."

CDC says there are 350 facilities authorized to do work with select agents, but "just 15" select agent exposure incidents have occurred since 2006. Fifteen exposures in 18 months doesn't sound like so little to me. That's almost one a month. The Bush administration is pushing research on these agents (at the opportunity cost of research on agents that really do cause disease in the US population), so the more laboratories that do this work the more chance of an accident that does some real damage. In an important sense, this "defensive" work makes us less safe, not more safe.

As the Texas A&M story shows, these accidents are not at all rare. They are relatively common and occur with some regularity. Many years ago I sat on several institutional biosafety committees as part of my professional responsibilities. The most arrogant and contemptuous people of biosafety regulations were the principal investigators themselves. They didn't do most of the work but they set the tone for their laboratories and they were often also responsible for the pressure on technicians, postdocs and graduate students to get the work done as fast as possible in a very competitive environment. This led to cutting corners and from there, to lab accidents.

Here are some examples discovered by the Dallas Morning News of other incidents:

• In February, a lab worker at UT Southwestern Medical Center in Dallas was cut by a pair of scissors used on a mouse infected with HIV. The worker went on antiviral medication immediately and never fell ill.
• A lab worker at A&M cut a finger in June while slicing the lymph node of a cow with tuberculosis. The employee suffered no health problems.
• An employee at the University of Texas Medical Branch pricked herself with a needle used to treat an anthrax-infected mouse in July 2006. The employee took antibiotics and did not got sick.
• Four people working in a lab at the UT Health Science Center in Houston in May discovered a leak of a liquid they believed was anthrax-tainted. The workers, who feared they may have inhaled the substance, did not fall ill.

In all, Texas A&M, UT Southwestern, the University of Texas Medical Branch in Galveston and the UT Health Science Center in Houston reported a combined 58 infectious disease exposures in and outside of lab settings since 2002, none of which posed serious public health threats.

These sound bad, but they are really fairly typical. These incidents happen in labs all over the country. The News story said that responses to inquiries to Texas Tech and the UT Health Science Center in San Antonio brought back no reports of incidents. Frankly, I don't believe it for a moment. Someone ought to examine the record keeping and notification compliance practices of these institutions.

And the Dallas Morning News has found still other documents at A&M:

A mouse reported missing in December after being infected with Q fever was never found, though investigators believe it must have been thrown out with other biohazard waste. Two outside health care officials crossed paths with a Brucella researcher in April before the researcher had decontaminated after an experiment. That same month, another lab worker reported high levels of Q-fever antibodies, though officials can't say whether the researcher was exposed at A&M or another lab.

The bottom line here is that the more biodefense work that is funded to more likely there is to have a nasty accident. If the cost benefit equation for this work were so overwhelmingly favorable on the benefit side, perhaps we could just say, "Well, it's a trade-off."

Unfortunately the balance isn't in our favor.