Creating the environment for process improvement.

The incorporation of quality improvement methods into the practice of laboratory medicine has become widespread in recent years. However, a significant number of laboratory professionals feel that quality improvement efforts take an excessive amount of time to achieve results. Using quality improvement methods and tools will have limited success without first creating the proper environment. Steps that we have found to be beneficial in creating the environment for successful improvement activities using the FOCUS-PDSA model include having senior leaders perform specific activities, aiming the team, and strategic contracting.

Identification of gram-negative bacilli using the Autobac IDX.

The Autobac IDX is a new system for the rapid identification of clinically significant members of the Enterobacteriaceae and Aeromonas, Acinetobacter, Alcaligenes, Flavobacterium, Moraxella, and Pseudomonas species. The use of 18 differentially inhibitory compounds such as dyes and antibiotics along with a computerized algorithm based on a multivariate analysis provides the basis for the identification of 30 different groups of gram-negative bacilli. Required preliminary tests include observations on the presence or absence of swarming on a sheep blood agar plate and noting the following: growth, lactose fermentation, and bile precipitation from a MacConkey plate. Spot indole and spot oxidase tests must be performed as well. Identification by the Autobac IDX System takes 3-6 hr after completion of the preliminary tests. From a total of 403 isolates tested, the Autobac system agreed with the MicroID AND N/F systems on 382 identifications (94.8%). Four isolates, two Acinetobacter anitratus, one Serratia marcescens and one Moraxella osloensis could not be identified by IDX. Additional testing was required on 35 (8.7%) of the isolates.

Acridine orange staining and radiometric detection of microorganisms in blood cultures.

To determine whether acridine orange (AO) staining of blood cultures could be used as a substitute for blind subculture when used in conjunction with the BACTEC system (Johnston Laboratories, Inc., Towson, Md.), the two methods were compared on all BACTEC-negative specimens. Since blind subcultures were routinely performed in our laboratory on days 2 and 6 of incubation, AO staining was also performed on these days. Cultures which were BACTEC positive on day 1 of incubation were not included in the study. Of the 2,395 bottles tested after 2 days of incubation, 106 were subculture positive. Of these, 96 (90.6%) were also AO positive and BACTEC positive, 3 (2.8%) were AO positive and BACTEC negative, and 7 (6.6%) were AO negative and BACTEC positive. Of the 3,487 bottles tested on day 6 of incubation, 14 were subculture positive; 7 (50%) of these were AO positive and BACTEC positive, and seven were AO positive and BACTEC negative. Of the total of 10 culture-positive bottles missed by BACTEC, all were positive, and all 10 companion aerobic bottles were BACTEC positive. In both phases of the experiment, there was a total of only four false-positive AO stains. As a result of this investigation, we have substituted AO staining for blind subculturing of BACTEC-negative bottles.

Rapid detection of Haemophilus influenzae type b in cerebrospinal fluid by commercial coagglutination and latex agglutination kits.

The ability of two commercial kits to detect Haemophilus influenzae type b antigen in cerebrospinal fluid was evaluated. Results obtained by Bactogen, a latex agglutination test, and Phadebact, a coagglutination test, were compared to counterimmunoelectrophoresis, Gram stain and culture results. One hundred and seven specimens of cerebrospinal fluid were tested. Thirty were found to contain bacteria, 20 of which were Haemophilus influenzae type b. All 20 were positive by Bactogen and Phadebact testing, 19 were culture positive, 18 were positive by counterimmunoelectrophoresis and 15 had gram-negative bacilli seen on Gram stain. The culture negative specimen contained microscopically visible gram-negative bacilli and was from a patient on antimicrobial therapy who was previously Haemophilus influenzae type b culture positive. No false positives from other genera or the 77 culture negative specimens occurred with Phadebact, Bactogen or counterimmunoelectrophoresis.