Canadian study disputes health care worker risk

Conversion rates linked to ventilation levels?

Several recent U.S. studies of TB exposure and transmission among health care workers say risks are down — so far it’s hard to tell whether workers were infected at work or at home. In one large Midwestern hospital, for example, health care workers from a high-prevalence zip code had higher rates of TB infection than those from other zip codes, suggesting, in accordance with other data, that hospitals are among the best places to be if you’re hoping to avoid TB infection.

That’s why a Canadian study has TB experts in the United States scratching their heads. Presented at an American Thoracic Society-sponsored con ference in Arlington, VA, that looked at TB control issues in low-prevalence settings, the Canadian study found conversion rates in hospitals were closely linked to ventilation levels. That evidence shows that exposures probably did take place at the job site — not in the community — because job sites with poor ventilation showed a positive correlation with higher rates of worker skin-test conversion.

Study surveys for skin-test conversions

The study, unpublished as of TB Monitor’s press time, took a retrospective look at 17 Canadian hospitals, where researchers surveyed workers for skin-test conversions, says Richard Menzies, MD, director of the Montreal Chest Clinic and principal investigator for the study. Of the thousands of workers surveyed, 1,300 were located whose records contained documentation of a negative TB skin-test. Then, researchers used a measurement technique common in oncology by looking at "survival times" for the group; that is, the time it took for the 1,300 workers to progress from a negative to a positive TB skin test.

Those survival times were correlated with ventilation levels at the facilities where the workers were employed. After adjusting for confounding factors, including country of origin and history of BCG vaccination, the two sets of data were correlated. Interestingly, the analysis shows a strong correlation between ventilation levels and risk for TB conversions among health care workers in non-isolation patient rooms but only a weak correlation for isolation rooms, Menzies notes. "My take is that once people get into isolation, even if the ventilation’s not so great, there are other factors at work [to prevent worker exposure]," he says. "At that point, whether ventilation’s good or bad isn’t as important as the fact that the patient’s in treatment, and people are wearing masks and taking other precautions."

Undiagnosed case is the culprit

That brings up a point about which Americans and Canadians agree: The undiagnosed case, the one that’s not already in the isolation room, is causing most of the exposures to health care workers.

Menzies says the difference in study design may explain why more of those undiagnosed patients were in the Canadian hospitals, driving up delays to diagnosis and resulting in more patient days without needed TB treatment. A look-back study, unlike a prospective study, doesn’t draw attention to itself, he says; swarms of researchers don’t inadvertently prompt clinicians to "think TB." Perhaps if the American researchers looked back in time, as their friends across the border did, they’d see things differently, he adds.

HCW transmission still occurs

"It’s easy to say that now that the big outbreaks of the past are past, the problem of exposure among health care workers is a thing of the past," he concludes. "I think it may be truer to say that there still is transmission." Granted, rates of conversion in the Canadian study were small — about 1% to 2% a year — but over a lifetime of employment in the health care setting, the percentage adds up, he notes.

That doesn’t mean hospitals practicing excellent infection control can’t bring conversion rates down to community levels, he adds. Obviously, such places can, and do. The real question, he says, is this: What’s happening in the 99% of places that aren’t practicing excellent infection control? "I’d suggest that what’s happening is transmission."