At. Nitta et al., MISDIAGNOSIS OF MULTIDRUG-RESISTANT TUBERCULOSIS POSSIBLY DUE TO LABORATORY-RELATED ERRORS, JAMA, the journal of the American Medical Association, 276(24), 1996, pp. 1980-1983
Objective.-To describe 9 cases where a misdiagnosis of multidrug-resis
tant tuberculosis (MDR TB) was made, possibly due to laboratory-relate
d errors. Design.-Case series. Setting.-Public and private hospitals,
outpatient clinics, and mycobacteriology laboratories serving those in
stitutions in Los Angeles County, Calif. Patients.-Consecutive sample
of 70 patients diagnosed with MDR TB who were identified between Augus
t 1993 and August 1994 by the Multidrug-Resistant Unit within TB Contr
ol in Los Angeles County. Outcome Measure.-Detection of laboratory-rel
ated diagnostic errors. Results.-Pulmonary MDR TB was misdiagnosed in
9 (13%) of 70 patients. Reasons why the diagnoses appeared to be erron
eous are as follows. growth of MDR TB from an old tuberculous lesion i
n a patient who was never treated for TB and whose diagnosis predated
anti-TB drugs (1 case), documented contamination with Mycobacterium av
ium complex (1 case), suspected cross-contamination (1 case), suspecte
d specimen mislabeling (1 case), successful treatment using drugs to w
hich the isolate was reportedly resistant (4 cases), discrepant suscep
tibility test results on additional sputum specimens submitted by the
patient (2 cases), and no clinical evidence of TB (3 cases). Conclusio
ns.-These cases emphasize the diagnostic errors that can occur if myco
bacterial susceptibility results are not correlated with all clinical
data including other laboratory results for a given patient. We conclu
de that susceptibility results alone are not enough to dictate treatme
nt, and that careful clinical correlation is necessary in making the d
iagnosis of MDR TB.