US hospital mycobacteriology laboratories: Status and comparison with state public health department laboratories

In response to the resurgence of tuberculosis, the Centers for Disease Control and Prevention recommended the use of certain mycobacteriology laboratory methods to improve the accuracy of diagnosis and/or minimize times to complete specimen processing, A study to determine the extent to which these recommended methods were being used in hospital laboratories was needed, In 1992, a survey was mailed to infection control and laboratory personnel at 1,076 hospitals with greater than or equal to 100 beds to determine the mycobacterial laboratory services being performed, the methods being used, the number of specimens being processed, and the times to completion during 1991, In 1995, a 20% sample of hospital laboratories that responded to the initial questionnaire was resurveyed, Responses to the 1992 survey were received from personnel at 756 (70%) hospitals representing 750 laboratories. Among laboratories performing the services, the use of recommended methods was as follows: fluorochrome stain for acid-fast bacillus microscopy (47%); radiometric methods for primary culture (29%); rapid (radiometric methods, use of nucleic acid probes, high-performance liquid chromatography, or gas-liquid chromatography) methods for identification of Mycobacterium tuberculosis (59%); and radiometric methods for drug susceptibility testing (55%), Reported times to complete specimen processing were shortest for laboratories that used recommended methods and longest for hospitals that referred specimens to outside laboratories. Only 46% of surveyed laboratories performed at least the minimal number of mycobacterial cultures (20/week) deemed necessary to maintain competence, Among 145 laboratories that performed the services and were resurveyed in 1995, use of recommended techniques increased from 44 to 73% for acid-fast bacillus microscopy, from 27 to 37% for primary culture, from 59 to 88% for M. tuberculosis identification, and from 55 to 75% for drug susceptibility testing. These changes were associated with reductions in reported specimen turnaround times. Use of the methods recommended by the Centers for Disease Control and Prevention increased at the resurveyed hospital mycobacteriology laboratories between 1991 and 1995, However, continued efforts are needed to increase the use of recommended methods at moderate- and high-volume laboratories, encourage referral of specimens from low-volume laboratories, and transmit results rapidly from all laboratories.