In the aftermath of the September 2001 terrorist attacks, the US federal government wasted no time in allocating large amounts of resources to build facilities for research into dangerous pathogens that might be used by terrorists as bioweapons. But it is now emerging that there are problems with the way some of these facilities operate, which suggests that the overall process has been poorly managed. These issues must be addressed before any further expansion goes ahead.

As construction of the first facilities got under way, some specialists were warning — at least privately — that many of the labs would lack people properly experienced in handling 'select agents', as microorganisms that could be used in bioweapons are now euphemistically called. Critics said that many of the institutions selected to host the labs lacked the capacity to manage the task.

Two weeks ago, these predictions seemed to come true. On 30 June, the Centers for Disease Control and Prevention (CDC) in Atlanta, Georgia, suspended research with select agents in five labs at Texas A&M University in College Station. The university is home to a National Center for Foreign Animal and Zoonotic Disease Defense, one of several such facilities established by the US Department of Homeland Security.

The shutdown was the result of Texas A&M's failure to report in a timely manner that four members of staff had been infected by bacteria in early 2006. One worker fell ill from Brucella in February last year and two months later routine blood tests results revealed that three other staff members had been exposed to Coxiella burnetii. Both bacteria come from cattle; both are fever-causing pathogens and could be used in bioweapons.

Fortunately, no one was seriously injured, nor does any disease seem to have been spread to the wider population. Texas A&M insists that it is now improving its reporting procedures. And the federal government is treating the incident as it has others: a bump on the road to better biodefence through more laboratory research.

But a closer look at the case raises serious concerns. Texas A&M only got around to reporting the four exposures to the CDC on 9 April this year. This was just one month before officials from the Department of Homeland Security were due to arrive on the campus to discuss a major new biodefence contract: the $450-million National Bio and Agro-Defense Facility, for which Texas A&M and its partner institutions are competing. The Department of Homeland Security, which will fund the complex, will this week narrow the 18 bidders down to a short list of three to five, with a final decision due in 2009. The new facility will replace the ageing Plum Island Animal Disease Center, which has operated for half a century off Long Island in New York state.

The federal government should not have to be prompted by activists into telling the American public the truth.

But even when Texas A&M made its report, neither the university nor the federal agencies found it necessarily to share news of the infraction with the general population. It was only on 26 June that the problem became public, when the Sunshine Project, a small watchdog group based in Austin, Texas, revealed details of the infections at Texas A&M. A week later, the watchdog disclosed associated problems at nine other laboratories nationally.

Back in April 2006, an audit from the Office of Inspector General at the US Department of Health and Human Services reported that 11 of the 15 biodefence labs it funds at universities had identifiable deficiencies in training, security and, most disturbingly, accountability. In fact, eight of the labs had accountability issues, including lax inventories for pathogens and inadequate controls on who can enter the laboratory.

If biodefence labs are to be run safely and successfully, the regulatory process needs to instil public confidence. The federal government should not have to be prompted by activists into telling the American public the truth about its workings. Many communities are inherently suspicious of these facilities and their mode of operation should be made as transparent as is realistically possible. The need to keep some of the technical details of the work secret should not be used to cover up salient facts about management and operations that ought to be squarely in the public domain.