Screening Employees for Tuberculosis

In my experience, few infection control topics confuse people as much as tuberculosis screening. Except maybe the chickenpox-shingles relationship, but mostly it’s TB.  Whether it’s who gets tested after known exposure, dealing with new employees with vague histories, or BCG vaccine, it’s always something.

I’ve gone in circles for hours with staff who’ve had BCG while they insist that not only should they not be screened for TB, the BCG has given them a force field that will magically repel TB for the rest of their days on earth.  There’s little point in trying to convince them otherwise, as they will, without exception, grab another foreign-born staff member who will vehemently confirm their beliefs, and say, “See, I told you so.  This is what we were told 30 years ago in {insert name of country here}.”  For added emphasis, I’ve had doctors confirm such beliefs for them. Topic for another day.

Another side of TB screening that I’ve seen pop up on infection control message boards lately is the frequency of testing. This is determined by the TB assessment done on your facility each year.  Like any assessment, you’ll consider the population you serve, the incidence of disease in the facility and the surrounding community, protections you have in place, and anything unique. Use it to classify your facility into a risk category according to that assessment-- low-, medium-, or ongoing risk, which will determine your testing frequency.  

TB screening involves a skin test or blood test, and a risk and symptom questionnaire.  If you are a facility at medium-risk, meaning you have cases of TB regularly or serve a high-risk population, then you need to be testing your staff.  Let’s first be clear on why we screen staff for TB.

1. We test new staff on hire to identify anyone with latent or active TB coming to work at your facility  
2. We offer treatment to anyone with latent TB before it becomes active
3. We remove any staff with active TB from the premises, and get them treated so they don’t infect staff or patients
4. We identify any current staff who may have been exposed to TB during their employment, so that you can do #2 and #3 above
5. We use #4 to identify lapses in infection control

I see so many variations on testing, that it seems we’ve forgotten what it’s for.  It is to identify exposed or sick staff, get them treated, and find out where your infection control practices went wrong. If your facility sees TB patients or those at high-risk, you need to make sure you are identifying them and protecting your staff.  This is done through annual TB testing (or more often in settings with ongoing risk, like TB hospitals or clinics). Not every 2 years, not every 3.  Not with a health questionnaire on odd years, and a skin test on even years.  And certainly not by testing everyone at the same time.  

If you have a low-risk facility, where you rarely or never have or anticipate TB patients, you don’t need annual testing.  So don’t test every 2 or 3 years just to make yourself feel better, or because it seems like a good idea.  Your staff are either at risk or at no/low risk.  Testing staff who are at low risk is expensive in the time it takes, the false positives it produces, and the lack of benefit it has.  This is not how diagnostic testing is meant to be used.  It’s the opposite.

If you have determined through your assessment that staff need regular testing, you must do it an interval that is useful.  You cannot pro-rate it, depending on what Human Resources finds convenient or financially acceptable.   You cannot test your contracted or agency staff less frequently than employed staff who are doing the same jobs with the same exposure risk.  You should not do all employees during the same week. And you cannot do one-step testing when 2-step is indicated.    

If you test your staff throughout the year (either on their date of hire or birthday), and some have a conversion, you will use that valuable information.  You will see what they had in common, where they worked.  Perhaps you’ll find no relationship.  But you must arrange your system so that you have every opportunity to find the issues if they exist.

Let’s say an unknown TB patient is present on unit/ward 4West in January, and he is discharged without ever getting a correct TB diagnosis.  If I perform annual testing on all of the 4West staff in February (because we’ve elected to follow a unit-based schedule), many staff will have been exposed, unlikely any will yet test positive in February, and we won’t know about the exposure until they are tested next February, at which point it’ll be quite late to determine which patient had TB.  

Similar caveat for testing staff based on job title.  Nurses will have the highest likelihood of exposure, as they often spend the most time with the patients.  So if I test all of the hospital’s nurses in January or February, I’ll miss the exposure yet again.

However, if I test staff more randomly, by date of hire or alphabetically, I’ll be testing the 4West staff all year long, and as they start popping up positive, I may identify the issue in just 2 months time, rather than 12 months.  I’ll figure out which patient it was perhaps, identify how he went unnoticed, test the rest of the 4West staff, get them treatment if they want it, and put measures into place to prevent the next occurrence.

If I test every 2 years, I’ll never find that patient, some of my 4West staff will have moved to other jobs in parts unknown, and some may unfortunately develop active tuberculosis.  If I wait 3 years, well, what’s the point at all?

TB screening is not like insulin, where you adjust the dose depending on the blood glucose.   There isn’t “more” or “less” screening by spreading out the interval; there is one way to do it properly: testing at a useful interval, and using the health screening to improve the sensitivity and specificity of the diagnostic test.  Making unfounded adjustments alters the effectiveness significantly.