The content, links, and pdfs are no longer maintained and might be outdated.
Multiple Misdiagnoses of Tuberculosis Resulting from Laboratory Error -- Wisconsin, 1996
A positive culture for Mycobacterium tuberculosis confirms a diagnosis of tuberculosis disease (TB); however, false-positive cultures have resulted from laboratory error (1-8). In 1996, the Division of Health (DOH), Wisconsin Department of Health and Family Services, became aware of five possible incidents of laboratory error associated with the processing of M. tuberculosis cultures; these incidents resulted in the potential misdiagnosis of TB in 11 persons. This report summarizes the findings from the investigation of the five incidents, which suggest that four incidents were associated with cross-contamination of specimens in the laboratory and one with the inadvertent inoculation of a subculture from the incorrect broth medium; these errors resulted in adverse effects on patients and on the medical and public health systems. Incident 1
In March, laboratory A referred mycobacterial isolates to the Wisconsin State Laboratory of Hygiene (SLH). After two of the referred isolates -- one from patient 1 and the other from patient 2 -- demonstrated low-level resistance to isoniazid, SLH staff contacted laboratory A staff to inform them of the results. Three weeks earlier, on notification that M. tuberculosis had been isolated from a sputum sample obtained from patient 1, the patient's physician had informed staff of DOH's TB Program that TB had not been suspected, that a sputum specimen obtained from the patient had grown Streptococcus pneumoniae, and that the patient's symptoms had abated with treatment for pneumococcal pneumonia. Further inquiry revealed that only one of three specimens from patient 1 was culture-positive for M. tuberculosis, that the positive specimen had been processed immediately after that of a smear- and culture-positive specimen obtained from patient 2 who had clinically obvious TB, and that growth was present on both solid and broth media. Restriction fragment length polymorphism (RFLP) patterns of the two isolates were identical, and there was no evidence of contact between the two patients. Based on these findings, DOH concluded that patient 1's culture was false-positive as the result of cross-contamination and that contamination occurred during the initial processing of specimens. Incidents 2 and 3
In March, the TB Program, SLH, and laboratory B staff noted a substantial increase in the number of M. tuberculosis isolates reported by laboratory B. Review of laboratory B logs identified two clusters of positive cultures processed 5 days apart. During the morning shift of the first day, a laboratory proficiency test sample containing M. tuberculosis had been processed. The second, third, and fifth specimens (each from different patients) processed during the afternoon shift were each acid-fast bacilli (AFB) smear-negative but were M. tuberculosis culture-positive. The RFLP patterns for these three isolates were identical to that of the laboratory proficiency isolate. Five days later, the first specimen processed that day was an AFB smear-positive specimen that had been obtained from a patient with clinically obvious TB. The third, fourth, seventh, and 15th specimens (each from different patients) processed were AFB smear-negative but were M. tuberculosis culture-positive. The RFLP patterns for these four isolates were identical to that of the first, and there was no evidence that the patients had had contact with one another.
A total of seven patients had suspected false-positive cultures processed in laboratory B. Record review indicated diagnoses other than TB for six patients. The seventh patient was lost to follow-up after an AFB culture had been requested for the patient as part of routine bronchoscopy procedures rather than because of clinical suspicion of TB. Based on these findings, DOH concluded that cross-contamination had occurred during incidents 2 and 3. Because laboratory B used the BACTEC 9000 * system, which measures mycobacterial growth noninvasively, cross-contamination probably occurred during initial processing of the specimens rather than during mycobacterial growth measurements. Incident 4
In July, laboratory C and TB Program staff noted a cluster of three cases of multidrug-resistant TB (MDR-TB) reported from one hospital. MDR-TB is rare in Wisconsin: during 1991-1995, only four cases were reported. The culture-positive specimens from these three patients had been processed in laboratory C on 2 consecutive days. The third and fifth specimens processed on the first day had been obtained from patient 1, who had AFB smear-positive pulmonary TB. The sixth (and last) specimen of the day was AFB smear-negative and the only M. tuberculosis culture-positive specimen among five specimens obtained from patient 2. The first specimen of the second day was AFB smear-negative and the only M. tuberculosis culture-positive specimen among four specimens obtained from patient 3. The RFLP patterns were identical for isolates from patients 1, 2, and 3. Patient 2 had cryptococcal pneumonia diagnosed, and no evidence of TB was detected in patient 3 during postmortem examination. In addition, laboratory C used the BACTEC 460 system, which has needles that enter vials during sampling. Based on these findings, DOH concluded that the specimens from patients 2 and 3 had been cross-contaminated by the specimen from patient 1, possibly from contaminated reagent or from carryover in the BACTEC 460 machine. Incident 5
In May, a physician called laboratory D about the report of a positive M. tuberculosis subculture from a bronchoalveolar lavage (BAL) specimen obtained during bronchoscopy to evaluate a mass present on chest radiograph. A biopsy specimen of the mass showed lung cancer. The BAL specimen was AFB smear-negative with no growth on solid media or in the original broth medium. At the time the subculture was inoculated, the original broth medium was in an incubator next to a broth culture that was growing M. tuberculosis and was from a patient with clinically obvious TB. RFLP analysis of the subculture was not possible. Based on these findings, DOH concluded that the patient's subculture had probably been inoculated with the broth medium growing M. tuberculosis from the patient with active TB.
Reported by: S Foldy, MD, J Rach, City of Milwaukee Health Dept; J Tjaden, S Muesegades, G Tyszka, A Vatland, Sheboygan County Div of Public Health, Sheboygan; G Madlom, K Bleau, D Ubbink, Ozaukee Health Dept, Ozaukee; N Kreuser, PhD, G Sigmann, M Swanson, Wauwatosa Health Dept, Wauwatosa; L Kubista, D Fett, Wisconsin State Laboratory of Hygiene; TO Beyer, M Proctor, PhD, JP Davis, MD, Chief Medical Officer and State Epidemiologist for Communicable Disease, Div of Health, Wisconsin Dept of Health and Family Svcs. Div of Tuberculosis Elimination, National Center for HIV, STD, and TB Prevention; State Br, Div of Applied Public Health Training (proposed), Epidemiology Program Office, CDC.
Editorial Note: Incidents 1-5 resulted in false-positive cultures for M. tuberculosis for 11 patients. ** Before recognition that these cultures were false-positive, 10 patients or their families had been informed of the diagnosis of TB, and eight patients received unnecessary medical treatments including hospitalization in respiratory isolation (one), bronchoscopy (two), and anti-TB medication (seven). Ten of these false-positive cases were reported to local health departments. As a result, 108 family and social contacts received tuberculin skin tests (TSTs); all were negative. The case management and contact investigations of these cases accounted for approximately 240 person-hours of labor by local and state health department staff. In addition, 328 hospital employees and patients received TSTs, and nine had chest radiographs; no evidence of transmission was found. Hospital infection-control and employee health staff expended an additional estimated 330 person-hours as a result of these episodes.
These findings in Wisconsin are similar to those in other recent reports that have documented the occurrence of false-positive M. tuberculosis cultures. The percentage of false-positive cultures in these reports ranged from 1.2% to 4.0% (1-3,5,8). False-positive results may be even more common in outbreak situations: based on a review of records for 223 MDR-TB culture-positive patients in outbreaks in five states, the clinical course was inconsistent with TB in 26% of patients (7).
False-positive cultures can adversely affect patients, their contacts, hospitals, and the public health system. Examples of these effects include psychological stress, social stigmatization of patients and their families (3), unnecessary and costly medical treatment (e.g., additional physician visits, radiographic studies, and additional specimen collection and culturing), and adverse side effects resulting from unnecessary anti-TB treatment (1-4,6).
Potential mechanisms resulting in contamination and laboratory error include mislabeling or switching specimens during handling (4), and instrument or reagent contamination resulting in carryover of mycobacteria from one sample to another during initial processing, processing for susceptibility testing, or sampling of sequential vials by the BACTEC 460 system (2,3,6,7). Primary prevention of laboratory error requires use of standardized laboratory procedures (6) that minimize the potential for errors.
The process of culturing mycobacteria is inherently prone to errors because of factors including the ability of some culture systems to identify mycobacteria when few organisms are present, the multiple steps involved in processing mycobacterial cultures, the viability of M. tuberculosis for long periods in laboratory environments, and the large number of mycobacteria present in some specimens (2). In addition, some laboratories may process a volume of specimens that exceeds their capabilities (6). The potential for errors underscores the need for prompt recognition of false-positives. Indicators of potential false-positive M. tuberculosis cultures are that:
The DNA fingerprint pattern of the suspected isolate is identical to that of the putative source isolate;
cultures of M. tuberculosis requires the cooperation of and communication between clinicians, public health and private laboratories, and local and state health departments. When culture results are inconsistent with the patient's signs and symptoms or clinical course, clinicians should discuss M. tuberculosis results with the laboratory and the responsible local or state health department (1). Laboratory staff should record the date and order of processing to enable easy identification of clusters of positive cultures and should prospectively track positivity rates and establish a threshold which, when exceeded, will prompt an investigation (6). If false-positive cultures and/or contamination are suspected, laboratory staff should notify their local or state health department and the patient's physician and should consider obtaining DNA fingerprinting of the putative source and the potentially contaminated specimens (1). Finally, TB-control program staff should routinely analyze surveillance data for clusters of positive cultures from a laboratory and for case-patients associated with predictors for false-positive cultures.
* Use of trade names and commercial sources is for identification only and does not imply endorsement by the Public Health Service or the U.S. Department of Health and Human Services.
** If these 11 cases had been included with the 114 true incident cases of TB in Wisconsin in 1996, they would have represented 9% of statewide reported cases.
Disclaimer All MMWR HTML versions of articles are electronic conversions from ASCII text into HTML. This conversion may have resulted in character translation or format errors in the HTML version. Users should not rely on this HTML document, but are referred to the electronic PDF version and/or the original MMWR paper copy for the official text, figures, and tables. An original paper copy of this issue can be obtained from the Superintendent of Documents, U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800. Contact GPO for current prices.**Questions or messages regarding errors in formatting should be addressed to firstname.lastname@example.org.
Page converted: 09/19/98
This page last reviewed 5/2/01