Evaluation of two bioluminescence-measuring instruments, the Turner Design and Lumac systems, for the rapid screening of urine specimens

Two bioluminescence-measuring instruments, the Turner Design and Lumac systems, were compared with a standard plate culture method for their ability to rapidly screen 400 urine specimens. For cultures with less than 1,000 CFU/ml the Turner Design, with old and new evaluation formulas, gave 6.5 and 50.6% false-positive results, respectively, versus 17.6% at greater than or equal to 500 relative light units with the Lumac. For cultures which had greater than 10(5) CFU/ml the Turner Design gave 39% (old formula) and 14% (new formula) false-negative results compared with 4% at less than 200 relative units with the Lumac. The microorganisms most frequently isolated in the false-negative cultures from either system were gram-positive cocci. Predictive values for a positive test at greater than 10(5) CFU/ml were 77.4% (old formula) and 35.7% (new formula) for the Turner Design versus only 50% for the Lumac at greater than or equal to 500 relative light units. Predictive values for a negative test for both instruments were greater than 88% at greater than 10(5) CFU/ml. The Turner Design and Lumac systems were 4.0 and 3.7 times as expensive, respectively, as the plate culture method. Although both systems greatly reduce the time required to process urine specimens, their high costs as compared with that of plate culture, their failure to detect many specimens having greater than 10(5) CFU of gram-positive cocci per ml, and the numerous false-positives reported by both instruments suggest that additional improvements in the systems are warranted.     

Use of rapid screening tests in processing urine specimens by conventional culture and the AutoMicrobic system

Two rapid urine screening tests, the Chemstrip LN (BioDynamics, Indianapolis, Ind.) and the Bac-T-Screen urine screening device (Marion Laboratories, Inc., Kansas City, Mo.), were evaluated as techniques to predict bacteriuria as quantitated by either conventional culture or the AutoMicrobic system (Vitek Systems, Inc., Hazelwood, Mo.). A total of 666 urine specimens were analyzed by both screening tests as well as the AutoMicrobic system and quantitative culture. The sensitivities of both Chemstrip LN and Bac-T-Screen for the detection of low levels of bacteriuria (greater than or equal to 10(3) CFU/ml) were comparable (73.3 and 74.4%, respectively) and were too low to recommend their use as a primary urine screen. Their excellent predictive value of a negative result at the 10(5) CFU/ml level (96 and 97.5%, respectively) makes them potentially useful in predicting urine specimens with less than 10(5) CFU/ml. The use of either of these tests in combination with the AutoMicrobic system markedly decreased the time required to classify urine specimens. Their low cost relative to the AutoMicrobic system urine card makes the use of either test cost effective as a screen for the AutoMicrobic system.     

Detection of bacteriuria and pyuria within two minutes.

A study was performed to evaluate two rapid urine screening methods, Bac-T-Screen (Marion Laboratories, Inc., Kansas City, Mo.) and Chemstrip LN (Boehringer Mannheim Diagnostics, BioDynamics, Indianapolis, Ind.), for their ability to screen for bacteriuria and pyuria within 2 min. A total of 1,000 urine specimens were tested with the Bac-T-Screen and the Chemstrip LN and compared with a semiquantitative plate culture method. Of the 1,000 specimens tested, 249 had colony counts of greater than or equal to 10(5) CFU/ml by the culture method. Of these, the Bac-T-Screen detected 94.8% (236 of 249) and the Chemstrip LN detected 84.7% (210 of 249). There were 120 pure cultures of probable pathogens of which the Bac-T-Screen detected 97.5% (117 of 120) and the Chemstrip LN detected 91.7% (110 of 120). Leukocyte counts were performed on all specimens, and both methods have the ability to detect greater than 10 leukocytes per mm3 in a majority (greater than 93%) of the specimens. The cost per test for a negative screen is approximately $1.30 for the Bac-T-Screen and $0.40 for the Chemstrip LN. Overall there is a similar negative predictive value with both methods for bacteriuria and pyuria.     

Bioluminescence screening for bacteriuria.

A rapid screening test (45 min) for bacteriuria was evaluated in 1,000 clinical urine specimens. The test procedure is based upon firefly luciferase analysis of bacterial ATP and uses the Lumac kit and Lumac M2010 Biocounter (3M Co., St. Paul, Minn.). The procedure allows for removal and destruction of nonbacterial ATP and subsequent analysis of bacterial ATP by firefly luciferase with a single photon counter. Results, expressed in relative light units, were compared with actual CFU by the calibrated loop technique. Sensitivities and specificities were calculated separately for clean-catch midstream specimens and for urines obtained by catheterization. The sensitivity for 719 clean-catch specimens with a Lumac cutoff of greater than or equal to 500 relative light units, representing greater than or equal to 10(5) CFU/ml, was 93%. The sensitivity for 281 catheterized specimens with a Lumac cutoff of greater than or equal to 200 relative light units, representing greater than or equal to 10(4) CFU/ml, was 95%. There were 19 false-negative results in the 1,000 specimens tested; more than 50% of these were contaminated cultures and were not considered significant in determining bacteriuria. In conclusion, the Lumac bioluminescence assay is a reliable, rapid bacteriuria screening technique with the potential of reducing the laboratory cost and for reducing the turnaround time in processing negative urine cultures.     

Evaluation of three bacteriuria screening methods in a clinical research hospital.

In a study conducted to compare three screening methods for their ability to detect significant bacteriuria, 2,815 urine specimens were screened by Chemstrip LN (BioDynamics, Division of Boehringer Mannheim Chemicals, Indianapolis, Ind.), 1,000 were screened by Bac-T-Screen (Marion Scientific Laboratory, Kansas City, Mo.), and 289 were screened by ATP assay (Turner Designs, Mountain View, Calif.). Results were compared with those obtained by quantitative culture plate method. The ATP assay showed the highest sensitivity (91%) compared with the Bac-T-Screen (67%) and Chemstrip LN (50%) tests but had the lowest specificity (64%) compared with the Bac-T-Screen (83%) and Chemstrip LN (91%). In 101 leukopenic patients with significant bacteriuria, the Bac-T-Screen test showed a higher sensitivity (33% at 10(4) to 10(5) CFU/ml and 80% at greater than or equal to 10(5) CFU/ml). It is concluded from this study that none of the three methods are sufficiently sensitive for the clinical research patients in this institution.     

Evaluation of the modified Bac-T-Screen and FiltraCheck-UTI urine screening systems for detection of clinically significant bacteriuria

Previous evaluations of the Bac-T-Screen system (Vitek Systems, Inc., Hazelwood, Mo.) demonstrated excellent sensitivity with specimens from patients with clinically significant bacteriuria (including infections with small numbers of uropathogens) but poor specificity with specimens from noninfected patients. In the study reported here, the sensitivity and specificity of the Bac-T-Screen system with a modified decolorizing reagent were evaluated. A manual filtration system, FiltraCheck-UTI (Vitek Systems), for screening urine specimens, Gram stains of mixed urine specimens, and quantitative cultures were also evaluated. The test sensitivity for clinically significant bacteriuria was greater than 96% with the original Bac-T-Screen system as well as the modified system and the manual system. In comparison, the sensitivities of the Gram stains and quantitative cultures (greater than or equal to 10(5) CFU/ml) were 82 and 77%, respectively. Of the 375 patients classified as noninfected by clinical parameters, 34% had positive screening tests with the original Bac-T-Screen system, as compared with 13 and 11% with the modified Bac-T-Screen and FiltraCheck-UTI systems, respectively. Thus, the modified Bac-T-Screen system and the manual FiltraCheck-UTI system have sensitives comparable to that of the original Bac-T-Screen system and markedly improved test specificities.     

Evaluation of a two-minute test for urine screening

A study was conducted to evaluate the ability of a urine filtration system (Bac-T-Screen, Marion Laboratories, Inc., Kansas City, Mo.) to detect negative urine cultures within 2 min. A total of 1,000 urine specimens were tested with the Bac-T-Screen and compared with a standard semiquantitative culture plate method and the Autobac system (General Diagnostics, Warner-Lambert Co., Morris Plains, N.J.). Of the 1,000 clean voided urine specimens tested, 246 specimens had colony counts greater than or equal to 10(5) CFU/ml by the culture plate method. Of these, the Bac-T-Screen detected 65.4% (161 of 246), and the Autobac detected 63.0% (155 of 246). When pure cultures of diphtheroids, lactobacilli, and viridans streptococci other than group D and cultures containing multiple organisms were considered to be contaminants and, therefore, were excluded, there were 106 pure cultures of probable pathogens of which the Bac-T-Screen detected 76.4% (81 of 106) and the Autobac detected 90.6% (96 of 106). Some 133 specimens were uninterpretable with the Bac-T-Screen because 36 clogged the filter and 97 left a residual pigment on the filter. A majority of those clogging the filter (69.4%) had positive plate counts, whereas the majority of the pigmented urines had negative plate counts. Of those urine specimens tested. 754 were negative by the culture plate method. The false-positive rates for Bac-T-Screen and Autobac were 16.2 and 5.8%, respectively. As a urine screen, the Bac-T-Screen has a negative predictive value comparable to the Autobac system and has the advantage of being a 2-min test.     

Clinical evaluation of the Lumac bioluminescence method for screening urine specimens.

A bioluminescence method for screening urine cultures to provide rapid reporting of negative specimens and to select appropriate urine cultures for direct application of automated identification methods was evaluated. A total of 2,000 specimens were processed in the Lumac Biocounter (3M, St. Paul, Minn.), and the results were compared with quantitative culture techniques by using a 0.001-ml inoculating loop. A total of 841 specimens were positive by the bioluminescence method; 291 specimens were culture positive (greater than or equal to 50,000 CFU of one or two organisms per ml). Positive cultures represented more than 20 different organisms. Approximately two-thirds of the false-positive results represented mixed flora or pure cultures of less than 5 X 10(4) organisms per ml. The predictive value of a negative result was 98.4%, reflecting a false-negative rate of only 0.7%. No advantages in cost or technician time were noted, but the Lumac method appears to be a useful technique in decreasing reporting time, especially for negative urine cultures.     

Clinical laboratory evaluation of a urine screening device

A study was conducted to compare the Bac-T-Screen Bacterial Detection Device for Urines (BDD; Marion Laboratories, Kansas City, Mo.) with urine Gram stain as a screen for bacteriuria. We analyzed 631 urine samples with the BDD and compared the results to urine Gram stains and quantitative cultures. A total of 90 (14%) specimens could not be analyzed with the BDD due to interfering pigments (67 specimens) or clogging of the filter (23 specimens). Of the 541 specimens that were analyzed, the BDD correctly identified 67 (88.2%) of the 76 specimens with greater than or equal to 10(5) CFU/ml but only 294 (63.2%) of the 465 specimens with less than 10(5) CFU/ml. The majority of the false negative specimens had either gram-positive organisms or yeasts. The predictive value of a negative BDD reading was 97.0%. The urine Gram stain correctly identified 92.1% of all positive cultures and 77.8% of all negative cultures. The predictive value of a negative urine Gram stain was 98.4%. In summary, the BDD compares favorably with the urine Gram stain as a screen for bacteriologically negative urine specimens.     

Use of the Bac-T-Screen to predict bacteriuria from urine specimens held at room temperature.

Results from the Bac-T-Screen (BTS) of fresh urine specimens were compared with the BTS results obtained when the same urine specimens had been held at room temperature for 24 h. Of the 246 specimens studied, 43 were initially BTS positive, 11 were false-negative, and 39 had greater than or equal to 10(5) CFU/ml. After 24 h at room temperature an additional 60 specimens had greater than or equal to 10(5) CFU/ml, of which only 16 were BTS positive; 10 specimens still gave false-negative results, and the number of false-positive specimens increased by only 6.5% of all specimens. For significant specimens (containing greater than or equal to 10(5) CFU of probable pathogens per ml), the predictive value of a negative test changed by only 0.1% (99.5 to 99.4%), whereas the sensitivity of the test remained at 96.4% for incubated specimens. Of those specimens that developed greater than or equal to 10(5) CFU/ml in vitro, 85% contained gram-negative bacilli. Neither bacteria grown in vitro nor urine specimens from normal females containing greater than or equal to 10(5) CFU/ml were positive with the BTS. For reasons not entirely understood, the BTS system may be unique in its ability to discriminate between bacteria which represent true bacteriuria and those which are present because of contamination, possibly due to other cellular elements present in infection-related bacteriuria, namely leukocytes and sloughed bladder epithelial cells.     

Rapid bioluminescence method for bacteriuria screening.

A study was performed to evaluate the UTIscreen (Los Alamos Diagnostics, Los Alamos, N. Mex.), a rapid bioluminescence bacteriuria screen. The UTIscreen was compared with three other rapid bacteriuria screens: the Bac-T-Screen (Vitek Systems, Hazelwood, Mo.), an automated filtration device; the Chemstrip LN (Boehringer Mannheim Diagnostics, BioDynamics, Indianapolis, Ind.), an enzyme dipstick; and the Gram stain. A semiquantitative plate culture was used as the reference method. Of the 1,000 specimens tested, 276 had colony counts of greater than 10(5) CFU/ml by the culture method. Of these, the UTIscreen detected 96% (265 of 276) using greater than or equal to 5% of the integrated light output of the standard reading as a positive interpretive breakpoint, the Bac-T-Screen detected 96% (266 of 276), the Chemstrip LN detected 90% (249 of 276), and the Gram stain detected 96% (264 of 276). Of the 214 probable pathogens isolated at greater than 10(5) CFU/ml, the UTIscreen detected 95% (204 of 214), the Bac-T-Screen detected 98% (210 of 214), the Chemstrip LN detected 92% (198 of 214), and the Gram stain detected 98% (209 of 214). The predictive values of negative test results at greater than 10(5) CFU/ml for the UTIscreen, the Bac-T-Screen, the Chemstrip LN, and the Gram stain were 98, 97, 93, and 98%, respectively. The overall specificities at greater than 10(5) CFU/ml for the UTIscreen, the Bac-T-Screen, the Chemstrip LN, and the Gram stain were 70, 48, 51, and 69%, respectively. There were 532 specimens with colony counts of >10(3) CFU/ml, and of these, the UTIscreen, the Bac-T-Screen, the Chemstrip LN, and the Gram stain detected 72, 81, 76, and 73%, respectively. Of the 249 probable pathogens isolated at >10(3) CFU/ml, the UTIscreen, the Bac-T-Screen, the Chemstrip LN, and the Gram stain detected 91, 95, 89, and 93%, respectively. The overall specificities at > 10(3) CFU/ml for these methods were 79, 55, 57, and 78%, respectively. The cost per test for detection was approximately $0.50 for the Chemstrip LN. Overall, the UTIscreen is rapid and easy to perform; its sensitivity compared favorably with those of the other screening methods; it had higher specificity than the Bac-T-Screen and Chemstrip LN; and it allowed for bathing of specimen.     

Detection of bacteriuria and pyuria by URISCREEN a rapid enzymatic screening test.

A multicenter study was performed to evaluate the ability of the URISCREEN (Analytab Products, Plainview, N.Y.), a 2-min catalase tube test, to detect bacteriuria and pyuria. This test was compared with the Chemstrip LN (BioDynamics, Division of Boehringer Mannheim Diagnostics, Indianapolis, Ind.), a 2-min enzyme dipstick test; a semiquantitative plate culture method was used as the reference test for bacteriuria, and the Gram stain or a quantitative chamber count method was used as the reference test for pyuria. Each test was evaluated for its ability to detect probable pathogens at greater than or equal to 10(2) CFU/ml and/or greater than or equal to 1 leukocyte per oil immersion field, as determined by the Gram stain method, or greater than 10 leukocytes per microliter, as determined by the quantitative count method. A total of 1,500 urine specimens were included in this evaluation. There were 298 specimens with greater than or equal 10(2) CFU/ml and 451 specimens with pyuria. Of the 298 specimens with probable pathogens isolated at various colony counts, 219 specimens had colony counts of greater than or equal to 10(5) CFU/ml, 51 specimens had between 10(4) and 10(5) CFU/ml, and 28 specimens had between 10(2) and less than 10(4) CFU/ml. Both the URISCREEN and the Chemstrip LN detected 93% (204 of 219) of the specimens with probable pathogens at greater than or equal to 10(5) CFU/ml. For the specimens with probable pathogens at greater than or equal to 10(2) CFU/ml, the sensitivities of the URISCREEN and the Chemstrip LN were 86% (256 of 298) and 81% (241 of 298), respectively. Of the 451 specimens with pyuria, the URISCREEN detected 88% (398 of 451) and Chemstrip LN detected 78% (350 if 451). There were 204 specimens with both greater than or equal to 10(2) CFU/ml and pyuria; the sensitivities of both methods were 95% (193 of 204) for these specimens. Overall, there were 545 specimens with probable pathogens at greater than or equal to 10(2) CFU/ml and/or pyuria. The URISCREEN detected 85% (461 of 545), and the Chemstrip LN detected 73% (398 of 545). A majority (76%) of the false-negative results obtained with either method were for specimens without leukocytes in the urine. There were 955 specimens with no probable pathogens or leukocytes. Of these, 28% (270 of 955) were found positive by the URISCREEN and 13% (122 of 955) were found positive by the Chemstrip LN. A majority of the false-positive results were probably due, in part, to the detection of enzymes present in both bacterial and somatic cells by each of the test systems. Overall, the URISCREEN is rapid, manual, easy-to-perform enzymatic test that yields findings similar to those yielded by the Chemstrip LN for specimens with both greater than or equal to 10(2) CFU/ml and pyuria or for specimens with greater than or equal to 10(5) CFU/ml and with or without pyuria. However, when the data were analyzed for either probable pathogens at less 10(5) CFU/ml or pyuria, the sensitivity of the URISCREEN was higher (P less than 0.05).     

Clinical evaluation of three urine screening tests

Evaluations of screening tests for bacteriuria have traditionally compared the test results with those of quantitative urine cultures. However, many patients with symptomatic urinary tract infections can have less than 10(5) CFU/ml in their urine. Therefore, the results of urine culture and three screening tests (Bac-T-Screen, Chemstrip LN [which tests for leukocyte esterase and nitrate reductase], and Gram stain) were correlated with the clinical classification of urinary tract infection. The Bac-T-Screen test detected 98, 93, and 100% of the infections classified as probable, possible, and asymptomatic, respectively. In contrast, the Gram stain, leukocyte esterase, and nitrate reductase tests were all insensitive screening tests for infection. Additionally, only 45% of the patients with probable infections had greater than or equal to 10(5) CFU/ml. Thus, the majority of infected patients would not have been detected if quantitative urine cultures were used alone.     

Rapid detection of insignificant bacteriuria by concomitant use of Lumac system and Gram's stain

The Lumac system, which assays bacterial ATP by bioluminescence, is a rapid method (less than 1 hour) for detection of bacteriuria. Conventional culture by calibrated loop technic and the Lumac system were compared using 2,000 urine specimens. Interpretation of Gram's stains of uncentrifuged specimens in addition to results of the Lumac system provided a second comparison with culture. Using a criterion of greater than or equal to 10(4) CFU/mL, conventional culture yielded 17% of the 2,000 specimens positive for bacteriuria. By Lumac + smear 27% were positive opposed to 41% positive by the Lumac system alone. The Lumac + smear method produced sensitivity (97%), specificity (88%), positive predictive value (62%), and negative predictive value (99.3%). False negative rates by the Lumac alone and Lumac + smear were 0.65% and 0.5%, respectively.     

Bacteriuria screening by use of acridine orange-stained smears.

Acridine orange (AO)-stained smears of 1,042 urine specimens were examined for the presence of bacteria and compared with quantitative culture results. The detection of one or more organisms per three AO fields at X200 magnification was noted in 161 of 162 and 193 of 195 urine specimens that grew greater than 10(5) and 10(4) CFU/ml, respectively, of a clinically relevant organism. However, a high number of false-positive AO smears (356 and 324, respectively) was observed among urines that failed to grow organisms at 10(5) and 10(4) CFU/ml. Sensitivity, specificity, and positive and negative predictive values of AO smears were 99, 58, 26, and 99%, respectively, for cultures of greater than or equal to 10(5) CFU/ml and 98, 59, 32, and 99%, respectively for cultures of greater than or equal to 10(4) CFU/ml. Despite the poor specificity of the AO smear, the very high negative predictive value would allow for the ruling out of bacteriuria defined at 10(4) CFU/ml and would eliminate the need for culture of ca. 50% of the urine specimens in this study. Employment of the rapid, inexpensive AO procedure only as a means to eliminate specimens for culture would allow significant cost savings and permit the laboratory to send out a large number of potentially negative or low-count urine results within a short time of specimen receipt.     

Comparison of the automicrobic system, acridine orange-stained smears, and gram-stained smears in detecting bacteriuria

We compared the accuracy of the Gram-stained smear, the acridine orange-stained smear, and the AutoMicrobic system (AMS; Vitek Systems, Inc., Hazelwood, Mo.) in screening for bacteriuria, as detected by conventional cultures. For 1,024 clinical specimens, results with the acridine orange-stained smear and the Gram-stained smear were very similar. When read for the presence of one or more microorganisms or leukocytes per 20 oil immersion fields, both smears were highly sensitive (92.1 and 93.3%, respectively) and moderately specific (70.0 and 61.7%, respectively). Sensitivity was greater for specimens yielding greater than or equal to 10(5) CFU/ml (96.1 and 98.9%, respectively) than for those with 10(3) to 10(4) CFU/ml (81.4 and 78.0%, respectively). Preliminary classification based upon the tinctorial and morphological characteristics of the Gram-stained smear was compatible with culture results in nearly all cases. The accuracy of the Gram-stained smears was not influenced by special cleaning of the microscopic slides, or the level of expertise of the microscopist. For 715 specimens, the sensitivity of the AMS in detecting bacteriuria (91.5%) was very similar to that of the stained smears (92.1 and 95.7%, respectively), but the specificity was significantly higher (83.2% versus 42.6 and 70.0%). Detection of microorganisms by the AMS took an average of 6.3 +/- 3.0 h. These data suggest that the Gram-stained smear is easily interpreted, very sensitive, acceptably specific, and still the optimal rapid method for screening for bacteriuria in most clinical microbiology laboratories.     

Visual and clinical analysis of Bac-T-Screen urine screen results.

Of 337 urine specimens evaluated, 75 of the 113 that showed positive readings on the Bac-T-Screen urine-screening instrument were found by subsequent semiquantitative culture to yield either less than 10,000 CFU of mixed bacteria per ml or no growth (less than 100 CFU/ml by our criteria). We tried to determine what factors contributed to the positive Bac-T-Screen results by examining the 75 Bac-T-Screen-positive urine specimens with three visual methods: Gram staining, hemacytometer chamber counting, and filtering through a 5.0-microns-pore-size nitrocellulose filter with subsequent microscopic examination of the stained filter. Somatic cells and other particles present in those urine specimens that yielded positive readings by the Bac-T-Screen included epithelial cells in 43%, crystals and amorphous material in 33%, and leukocytes in 17% of the specimens. There was no relationship between the numbers of particles seen in urine and the magnitude of the relative absorbance reading obtained with the Bac-T-Screen. A retrospective chart review was conducted for patients with positive Bac-T-Screen results and negative cultures. Of the 75 patients, 6 were thought to have urinary tract infections on the basis of clinical criteria; the majority of the remaining 69 patients had clinical histories revealing systemic or urogenital conditions consistent with shedding of particles in the urine. A positive reading by the Bac-T-Screen system seemed to be related to the presence of somatic cells and other particles in urine; bacteriuria was not always detectable in these cases.     

Evaluation of the B-D urine culture kit.

Split samples of urine transported to the laboratory at 5 degrees C and in a boric acid-glycerol-sodium formate preservative (B-D Urine Culture Kit; Becton, Dickinson & Co.) were cultured immediately and, in the case of preserved urine, after 24 and 48 h of storage at 25 degrees C. Agreement between the results for cultures of specimens originally yielding greater than 10(5) colony-forming units (CFU) per ml and the results for urine preserved for 24 and 48 h was 85 and 71%, respectively. One-third of the specimens originally yielding 10(4) to 10(5) CFU per ml yielded less than 10(4) CFU per ml after 24 h of storage in preservative. Provided greater than 10(4) CFU per ml in specimens preserved for up to 24 h is regarded as equivalent to greater than 10(5) CFU per ml in original urine specimens, agreement of results was greater than 90%.     

Screening of Urine Cultures by Three Automated Systems

A study was conducted to compare three automated systems and the Gram stain for their ability to detect significant bacteriuria. A total of 1,000 urine specimens were evaluated by Autobac MTS (General Diagnostics), Auto Microbic system (AMS; Vitek Systems, Inc.), and MS-2 (Abbott Laboratories) and compared with a semiquantitative culture plate method. Two hundred thirty-nine (23.9%) specimens had colony counts of >10⁵ colony-forming units (CFU)/ml by the culture plate method (group I). Of these, 204 (85.3%) were positive by Autobac, 198 (82.8%) were positive by AMS, and 179 (74.9%) were positive by MS-2. When pure cultures of diphtheroids, lactobacilli, and viridans streptococci not group D were considered contaminants and therefore excluded, there were 118 specimens containing pure cultures of probable pathogens. The percentage of significant isolates detected was 97.4% (115 of 118) by the Gram stain, 96.6% (114 of 118) by Autobac, and 95.8% (113 of 118) by AMS and MS-2. The average detection time for all organisms was 2.2 h by Autobac, 6.1 h by AMS, and 1.8 h by MS-2; therefore, all three methods were more rapid than the 18- to 24-h standard plate culture method. One hundred sixty-one (16.1%) specimens had colony counts of 10⁴ to 10⁵ CFU/ml (group II). The probable pathogens not detected in this group were two (1.2%) by Autobac and MS-2 and three by AMS (1.9%). The average detection time for group II was 4.2 h by Autobac, 8.9 h by AMS, and 3.8 h by MS-2. Six hundred specimens had colony counts of <10⁴ CFU/ml. Of these, 188 had colony counts equal to 10³ and <10⁴ CFU/ml (group III), and 412 cultures were below detectable limits by the standard plate method (group IV). Less than 37 and 15% of groups III and IV, respectively, were detected by instrumentation. Average detection times for groups III and IV were 4.6 and 4.8 h by Autobac, 10 and 11 h by AMS, and 4.2 and 4.4 h by MS-2. The cost of supplies and technical time with Gram stain, Autobac, and MS-2, when used as screening methods, were comparable and considerably less expensive than for the reference method. The AMS was the least expensive system when the cost for identifying probable pathogens was included.     

Direct identification and susceptibility testing by the AutoMicrobic system of gram-negative bacilli from urine specimens.

A total of 1,800 urine specimens were screened by Gram stain to detect bacteriuria. Pellets of bacteria were obtained by centrifuging specimens containing greater than or equal to 1 gram-negative bacillus of a single morphological type per oil immersion field. Direct susceptibility tests and identifications were performed from pellets by using the AutoMicrobic system (AMS). Results were compared with culture results by routine AMS methods. Of the 145 specimens showing only gram-negative bacilli on Gram stain, 113 grew greater than or equal to 10(5) CFU of a single species per ml. Compared with routine AMS identifications, the direct method correctly identified 105 (92.9%) of the isolates. Identifications were available within 8 h for 77% of the isolates. When compared by MICs, 93.2% of the direct susceptibility test results agreed with routine AMS results within one twofold dilution. Comparisons by category call indicated that overall complete and essential agreements were 89.9 and 97.8%, respectively, with 1.0% very major, 1.0% major, and 8.1% minor errors. Cefamandole and cephalothin had the lowest correlations by both comparisons. Within 8 h, susceptibility results were available for 94.3% of the isolates. This method offers the advantage of rapid detection, prompt processing, and earlier reporting of complete results for positive urine specimens.