Contamination underscores need for lab safeguards
Wisconsin patients get false-positive MTB cultures
Recent reports of laboratory errors in diagnosing tuberculosis, including five incidents in Wisconsin last year, highlight the costly impact of cross contamination and the importance of staff coordination in identifying and preventing it, say public health officials.
"One of the important lessons is that it requires teamwork, the work of the laboratory, the clinicians, and the public health department because you cannot get to the bottom of this unless you pull all the information together," says Mary Jo Trepka, MD, an epidemiologist from the Centers for Disease Control and Prevention, based at the Wisconsin Division of Health in Madison.
Trepka wrote a report, published recently in the Morbidity and Mortality Weekly Report, detailing how false-positive cultures for M. Tuberculosis were identified in five separate instances.1 Four of the incidents resulted from cross-contamination of specimens, while one originated from inoculating a subculture from the wrong medium broth. The report follows other published investigations of cross-contamination of TB that puts the incidence at 1% to 4%, and up to 25% during outbreak situations.
TB specimens, which can survive for several days in a laboratory environment, are prone to cross contamination because of the multiple steps required to make a diagnosis from processing sputum to decontaminating the specimen to growing the culture in solid and liquid media, Trepka says. In recent years the increased volume of testing, combined with more sensitive culture systems, has increased the risk of false-positive tests, she adds.
One of the most definitive studies on the problem of cross contamination with TB specimens was carried out by the CDC in 1995. After pulling specimens from laboratories in Arkansas and performing DNA fingerprinting on them, health officials estimated that 3% to 5% of all positive TB cultures in the state that year were caused by laboratory error.
In Wisconsin last year, the rate was even higher. The five incidents of contamination yielded false-positive cultures for 11 patients. Had these cases been added to reports of TB for the state, it would have represented 9% of cases in 1996, the article notes.
"It was a bad year, and I think we had a rash of problems," Trepka tells TB Monitor. "But from all of these reports there is a rate of false-positive cultures that is not insignificant, so it is a real problem."
So far, reports on false-positive cultures for TB have been done retrospectively. Not until the advent of DNA fingertyping has there been a means of accurately confirming contamination. One of the most comprehensive studies of false-positive tests was undertaken in San Francisco, which has been fingertyping all TB cases since 1990. A review of San Francisco patients in 1992 found that nine out of 500 TB diagnoses were false positives yeilding a rate of about 2%, says Peter Small, MD, assistant professor of medicine in the division of infectious diseases at Stanford University Medical Center in San Francisco.2
Small first began using DNA fingertyping to confirm laboratory contamination.
"I was the person who went back to these laboratories and said I think you have a problem,’" he says. "And I have been impressed by the response of laboratory directors. Their awareness has prevented it prospectively, while the availability of fingerprinting has allowed people to do the testing and say, Hey we are doing something wrong.’"
At the same time, the actual extent of contamination occurring in labs today is hard to quantify. The CDC is conducting several prospective studies to get a better handle on the extend of the problem, says Sarah Valway, MD, chief of the epidemiology section in the CDC’s Division of TB Elimination.
"We have no idea of rates. We have a few studies looking at population-based things, and that is all" she explains. "It will happen, and we actually think lab contamination is not that uncommon. It’s a matter of recognizing it when it happens."
A rash of problems
Delayed recognition of the Wisconsin cases resulted in 10 patients and their families being informed of a TB diagnosis. Eight of these patients received unnecessary medical treatment, including one who was hospitalized and put in respiratory isolation, two who received bronchoscopy, and seven who received anti-TB medication.
Moreover, the public health department made contact investigations on the 10 cases, resulting in tuberculin skin tests for 108 family and social contacts. In addition, 328 hospital employees and patients received skin tests, and nine had chest radiographs. The screening produced no evidence of transmission. The CDC estimates that the case management and contact tracing required 240 person hours from state and local health staff for family and social contacts and 330 hours for employee and patient investigations.
In the first of five incidents in Wisconsin, a state lab received two isolates one from each of two patients both of which demonstrated low-level resistance of isoniazid. Further investigation revealed that three weeks earlier one of the patient’s sputum specimens had grown Streptococcus, and symptoms had stopped. The investigation also found that only one of the same patient’s three specimens was culture positive for TB. DNA fingertyping subsequently revealed that the two isolates were identical, confirming suspicion that the patient’s specimen had been cross contaminated by the other patient’s specimen during initial processing.
In another case at a different laboratory, a proficiency test sample turned up positive for TB. Specimens from three other patients also tested positive later that day. DNA testing found that all three isolates were identical.
A high level of suspicion helped identify cross contamination in another incident in which a hospital reported a cluster of three cases of multidrug-resistant TB, which is rare in Wisconsin. The culture-positive specimens from the three patients had been processed by a lab on two consecutive days, and it was concluded that the specimens from two patients were contaminated by a third patient’s specimen, possibly from a contaminated reagent or from carryover in the BACTEC machine.
Although it is not possible to determine from these retrospective investigations the exact point that contamination took place, Trepka says health officials believe that most, if not all, happened during the decontamination processing steps.
"Each case is different," she says. "In one case, we believe it was one of the reagent bottles, which was very large, and may have caused a splash back while pouring into flasks. In another case, we think the staff got distracted and may have used the same syringe for inoculating a couple of specimens."
Whether most contaminations occur during set-up or from the BACTEC system is a matter of contention, Small says. "I suspect both account for some of it, but a lot of it occurs during setup," he adds.
Proving contamination is not easy and requires several steps. First, one must look at the laboratory information and identify possible red flags for contamination, such as a patient having only one positive culture but one or more negative sputums. Another sign might be a culture that is slow-growing, reflecting the typically small amount of bacteria crossed over from contamination, Trepka says. A second step is to perform DNA fingertyping. If specimens appear to be organically related, that adds weight to evidence of contamination, she adds. But a third, epidemiological step also is required showing that patients involved in the case did not have contact with each other and thus, eliminating the possibility of true transmission, Trepka notes.
"Every lab needs to go through their procedures with their staff continually and say where things could go wrong, what we can do to prevent this," she says.
(Editor’s note: For practical guidelines on how to reduce the risk of cross contamination, see Peter Small’s article published in Journal of Clinical Microbiology, referenced below.)
References
1. Centers for Disease Control and Prevention. Multiple misdiagnoses of tuberculosis resulting from laboratory error Wisconsin, 1996. MMWR 1997; 46:797-801.
2. Small P, McClerry N, Singh S, et al. Molecular strain typing of Mycobacterium tuberculosis to confirm cross resistance in the AFB laboratory and modification of procedures to manage occurrences of false positive cultures. J Clin Micro 1993; 31:1,677-1,682.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content