False-Positive Culture Diagnosis Of Tuberculosis

Abstract & Commentary

Synopsis: Clinician beware: Tuberculosis cultures were falsely positive in 4% of patients.

Source: Burman WJ, et al. The incidence of false-positive cultures for Mycobacterium tuberculosis. Am J Respir Crit Care Med 1997;155:321-326.

"It is one thing to show a man that he is in error, and another to put him in possession of truth."

—John Locke

An Essay Concerning Human Understanding

Book IV, Chapter 7

The "gold standard" for diagnosis of infection with Mycobacterium tuberculosis (MTB) is recovery of the organism in culture. Burman et al report that this standard is sometimes not as golden as we would like.

Between 1988 and 1994, MTB was recovered at the Mycobacteriology Laboratory of Denver Health and Hospitals from 696 specimens obtained from 199 patients. Two hundred twenty isolates from 198 patients were studied. IS6110 DNA fingerprinting of isolates was performed at the Little Rock, Arkansas, Regional Mycobacteriology Genotyping Laboratory with confirmation of identity by probing with pTBN12. At the time of diagnosis, 15 patients had multiple isolates obtained from more than one body site; all such isolates from an individual patient were identical. Forty-four patients had only a single positive culture. Based on apparent genotypic identity with isolates processed within a 42-day window, eight (3.6%) of these 220 were definite (5) or probable (3) false-positives.

If one assumes that the 476 positive cultures not fingerprinted were all true positives, 1.1% of culture results were false positives. After eliminating multiple isolates from the same patient, a false-positive result was obtained in eight (4%) of 199 patients with a positive culture. An additional five false-positive cultures from another laboratory were studied, giving a total of 13 in order to determine the manner of contamination. All 13 had been submitted to the laboratory within 17 days of the presumed source specimen. Nine (62%) were processed or reprocessed because of microbial contamination on the same day as the presumed source of contamination.

The 13 false-positive specimens had been contaminated from nine source specimens. Nine false-positive cultures occurred in laboratories using the BACTEC culture system. However, needle carry-over, a recognized cause of contamination while using this system, was a possible mechanism in only one case. The mechanism of contamination of three false positives in a laboratory that did not use the BACTEC system was most likely via multiple-use containers of decontamination fluid and phosphate buffer.

The culture results were recognized as false-positive by clinicians in only three cases. Of the remaining 10 cases, one patient died, and two were lost to follow-up before treatment could be prescribed. Seven patients with false-positive cultures, however, received multidrug antituberculous therapy; one of these developed rifampin-induced hepatitis.

COMMENT BY STAN DERESINSKI, MD, FACP

If we can’t rely on TB cultures, what can we rely on? The problem of false-positive cultures for this organism has previously been identified, and the rates of occurrence reported in these prior studies were similar to that reported by Burman et al. In most instances, however, the results were not recognized as being falsely positive by clinicians. As a result, many patients are unnecessarily treated for tuberculosis, sometimes with adverse consequences.

In some settings, the frequency of false-positive MTB cultures may be a great deal higher than those reviewed here. The authors point out that in two previous reports of nosocomial outbreaks of TB, the rates of false-positive cultures were, astoundingly, 16.4% and 26% (Fischl MA, et al. Ann Intern Med 1992;117:177-183; Valway S, et al. Tubercle Lung Dis 1994;75:S42).

Among the means by which false-positive culture results may occur are mislabeling of specimens at any of multiple processing steps and cross-contamination during initial or subsequent processing. As listed by the authors, contamination may occur via aerosolization, splashing, or by carryover from contaminated reagents, pipettes, microbiologic loops, containers or container lids, or, in the case of the BACTEC system, needle carryover.

The only means of clearly demonstrating cross-contamination of MTB cultures is by comparison of genotypes, a methodology not readily available to most clinicians. Furthermore, in the instance of a single source outbreak, a single genotype would also be expected. The key to the recognition of false-positive MTB cultures is an alert, properly skeptical clinician.