A Rapid (--hour) Urine-culture System for Direct Identification and Direct Antimicrobial Susceptibility Testing PAUL A. HEINZE, M.S., LAURI D. THRUPP, M.D., AND CARL R. ANSELMO, PH.D. Heinze, Paul A., Thrupp, Lauri D., and Anselmo, Carl R.: A rapid (--hour) urine-culture system for direct identification and direct antimicrobial susceptibility testing. Am J Clin Pathol : -8, 99. This study evaluates a new direct rapid system for urine cultures, including detection and quantitation of positive specimens by Gram stain, direct identification by --hour incubation of sediment with reagent strips, and antibiotic susceptibility testing by direct (--hour) diskelution methods. Of 98 routine urine specimens, had significant (< r ' colony-forming units/ml) gram-negative bacilluria, of which 89% were detected by the Gram stain. Direct rapid identification was correct in 9%. Results of direct diskelution antimicrobial tests showed overall agreement with results of standard disk diffusion in 9% of tests, and major discrepancies in %. For urine specimens with gram-negative bacilluria, this system permitted detection, quantitation, identification, and antimicrobial susceptibility testing in four to six hours with reasonable, though not complete, accuracy. (Key words: Rapid urine culture; Antimicrobial susceptibility testing.) URINARY-TRACT INFECTIONS are among the most common bacterial infections. _,8,9 As a result, cultures of urine comprise a large portion of the work load in the clinical microbiology laboratory. Furthermore, in hospital practice many patients with suspect urinary tract infection may be acutely ill. Management of such patients may be improved by rapid detection of infected urine specimens and rapid speciation and antibiotic susceptibility testing. Methods for processing urine specimens intended to provide rapid results and to increase laboratory efficiency have been described. These include screening methods to expedite detection of infected urine specimens,, and identification of colonies from primary culture plates,' and rapid direct or indirect antimicrobial susceptibility tests., However, application of such procedures in a routine coordinated system for Received July 9, 9; received revised manuscript and accepted for publication December, 9. Presented in part to the th Interscience Conference on Antimicrobial Agents and Chemotherapy, sponsored by the Infectious Disease Society and the American Society for Microbiology. Address reprint requests to Dr. Thrupp: Professor of Medicine, University of California Irvine, and Consultant to Clinical Microbiology Laboratory Service, University of California Irvine Medical Center, City Drive South, Orange, California 98. California State University, Long Beach, and University of California, Irvine-Orange County Medical Center, Orange, California rapid, direct processing of urine cultures in the hospital microbiology laboratory has not been described. We report herein the development and initial evaluation of a rapid direct urinary culture system with the following components: () rapid detection of positive specimens and () estimated quantitation of positive specimens, both by Gram stain of direct smears; () rapid speciation of gram-negative pathogens by direct testing of urinary sediment with reagent-impregnated filter-paper strips; () direct tests of antimicrobial susceptibility of the gram-negative pathogen in the same fresh specimen of urine using the rapid disk-elution method. This study has demonstrated that for urine specimens with significant gram-negative bacilluria, the system permitted detection of all positive specimens, estimation of colony counts, speciation, and results of antibiotic susceptibility tests to be reported with reasonable accuracy within four to six hours. Urine Specimens Materials and Methods Nine hundred eighty-seven specimens of fresh urine were studied, including specimens from both inpatients and outpatients. To avoid mixed cultures, specimens from infants and from patients with indwelling catheters were excluded. Specimens with volumes of less than approximately ml were also excluded. Detection and Quantitation of Positive Specimens The initial step in processing was screening by Gram stain of a direct smear of uncentrifuged urine in order to select positive specimens (> colony forming units [cfu]/ml). Standardized thin smears, - mm in diameter, eight per slide, were prepared from fresh uncentrifuged specimens using a.-ml calibrated -9/9// $.8 American Society of Clinical Pathologists Downloaded from https://academic.oup.com/ajcp/article-abstract////8 on April 8
8 HEINZE, THRUPP AND ANSELMO A.J.C.P. February 99 ESTIMATED COUNT FROM MICROSCOPY OF DIRECT GRAM-STAIN (c) CENTRIFUGED TO SEDIMENT BACTERIA- (FPS) INOC. SED. TO FPS HR. DIRECT IDENTIFICATION INOC. DISC-ELUTION CUVETTE FOR k HR. DIRECT (AS) ANTIMICR. SUSCEPTIB. 8+ ORGS/O.I.F. CIO 8 BACTERIA/ML) icy (FPS) CC OF A : DILUTION (AS) - ORGS/O.I.F CIO BACTERIA/ML) <c) KFPS) CC DIRECT INOCULATION "(AS) < ORGS/O.I.F. ( - BACTERIA/ML) -(C) <- -(FPS) CC SALINE SUSPENSION (AS) NO BACTERIA SEEN._ (< BACTERIA/ML) /STANDARD URINE V SCREENING CULTURE FIG.. Preparation and adjustment of inoculum from positive urine specimens and processing scheme for direct identification and direct antimicrobial susceptibility testing. *A sediments resuspended in ml physiologic saline solution to use as inoculum. FPS = filter-paper reagent strips. AS = antibiotic susceptibility tests. platinum loop. Following heat fixation and Gram-staining, a thin layer of immersion oil was applied, and each smear was screened under x magnification. Suggestive positive smears were examined under x, oil immersion, and in specimens showing gramnegative bacilli the number of organisms was estimated quickly according to the logarithmic categories shown in Table. An average time of min per specimen was needed for the entire procedure. Positive specimens were processed by the direct methods outlined below. Quantitative dilution-plate counts were also performed on all specimens tested by the direct methods. Specimens negative or showing apparent mixed flora by Gram stain were excluded from further direct study methods, and were processed only in the clinical microbiology laboratory by standard culture methods using quantitative loops streaked on blood, MacConkey, and polymyxin B- nalidixic-acid-blood agar plates. Direct Identification Specimens positive on smears were processed by the direct rapid methods according to the colony counts Table. Quantitation of Gram-negative Bacteriuria by Direct Gram Stain of Uncentrifuged Specimens Average Number of Gram-negative Bacilli per Oil-immersion Field < - - :8 Estimated Colony Count (cfu/ml) ' 8 estimated from the Gram-stained samples (Fig. ). First, they were centrifuged at, x g for min, and the sediment was resuspended in ml physiologic saline solution for use as inoculum for the reagentimpregnated filter-paper strips (FPS).* Eleven filter-paper-strip reactions were used to identify the gram-negative bacteria in the saline suspension of urinary sediment. Units of. ml of the saline suspension in x -mm test tubes were used as inoculum for the nitrate, phenylalanine deaminase, hydrogen sulfide, indole, urea, Voges-Proskauer, malonate, lysine decarboxylase, ornithine decarboxylase, and o-nitrophenyl- B-D-galactopyranoside (ONPG) test strips. All strips were incubated for hours at C, then read according to the manufacturer's directions. Oxidase tests were performed after hours according to the method of Gaby 8 by use of the twelfth chamber of the disk-elution cuvettet on cultures suggestive of Pseudomonas aeruginosa because of inert filter-paper-strip reactions, slower growth, and characteristic antibiotic susceptibility patterns. Results of direct speciation by filterpaper-strip were compared with results of identification obtained for the same specimen by the clinical microbiology laboratory using the R/B system supplemented by citrate and urease., Direct Antibiotic Susceptibility Tests Disk-elutiont antibiotic susceptibility tests (AS) were performed directly on the specimen in parallel with the * Pathotec, General Diagnostics Division, Warner Lambert Company, Morris Plains, New Jersey. t AutoBac, Pfizer Diagnostics, Groton, Connecticut. Downloaded from https://academic.oup.com/ajcp/article-abstract////8 on April 8
Vol. No. RAPID DIRECT URINARY CULTURE PROCEDURE 9 rapid direct culture for identification by filter-paperstrip reactions. In an effort to standardize the inoculum at approximately cfu/ml, the urinary inoculum was prepared according to the semiquantitative estimate from the results of Gram-staining, as outlined in the flow chart (Fig. ). Thus, specimens estimated by Gram-staining to contain approximately 8 cfu/ml were diluted :, those estimated to contain ml were used directly undiluted, and from those estimated to contain - /ml, the suspension of centrifuged sediment used was the same as that prepared for the direct filter-paper-strip identification procedure. A -ml volume of the adjusted urinary inoculum in 8 ml Eugonic brotht was dispensed to the -chamber AutoBac cuvette. Ten antimicrobial elution disks were added, and the cuvette was incubated and read by light-scattering index at and hours, or until a control-chamber growth index of.9 was reached. 9, Results of the direct disk-elution antibiotic susceptibility tests were compared with results completed two days later in the clinical laboratory by the routine culture procedures and the standard antibiotic susceptibility procedures (agar-overly modification of the Bauer-Kirby disk-diffusion procedure). Results Detection and Quantitation Of the specimens studied, 9 were positive (> cfu/ml) by standard quantitative culture, and 88% of the 9 were detected by the direct Gram stain. Of the 9 positive specimens, showed pure or predominant growths of gram-negative bacilli; of these, 89% were detected by the Gram stain (Table ). Occasional false-positive results were found in which specimens showed apparent gram-negative bacilli by Gram-stain but actually grew diptheroids. However, each of these specimens, when processed by the rapid direct methods, failed to produce growth in the direct susceptibility tests and failed to produce any reaction with the filter-paper strips; thus, they were ruled out as positive specimens in hours. Of the specimens with mixed cultures, were correctly detected as mixed by direct Gram stain and were eliminated from further direct testing. Seven of the specimens erroneously did not appear mixed by gram-stain, and therefore were not excluded from testing by the rapid direct methods. Of these seven, two were suspected to be mixed cultures because results of filter-paper-strip tests or direct susceptibility tests were aberrant. However, the remaining five of the produced diagnostic errors, as they appeared to contain pure cultures of Escherichia coli by the rapid Table. Reliability of the Direct Gram Stain in Detection of Significant Bacteriuria from 98 Urine Culture Specimens Organisms in Direct Gram-stain Result Positive Cultures* Reliability (a cfu/ml) Positivet Negative Total (% Positive) Gram-negative bacilli Staphylococci Streptococci Yeast Mixed rapidgrowing species 8 9 89 9 * Positive by.-ml loop quantitative culture method. t Direct smear,.-ml loop, any organisms seen in several oil-immersion fields. direct methods, even though mixed organisms were present (Table ). Not all the 8 specimens infected with gram-negative bacilli that were detected by direct smear were tested by the direct methods. Twenty-one of the 8 were eliminated because they came from patients who had indwelling catheters, had been collected from infants, or were of insufficient volume to process by the rapid protocol. Figure illustrates the accuracy of the direct Gram stain both for screening to detect positive cultures and for estimating the quantitative colony counts. Figure also shows the distribution of colony counts per milliliter by pour plate for the 8 positive specimens studied by the direct method. Smear results representing false negatives ( specimens, box B), and false positives (five specimens, box A) are also presented in Figure. The accuracy of the direct Gram stain both for screening to detect positive specimens and for estimating colony counts was greatest with gram negative bacteriuria at levels of 8 cfu/ml and was somewhat less at - cfu/ml. Overall, in 9% of the positive specimens, the semiquantitative smear estimate was correct to within ± log. Rapid Direct Identification Complete identification and susceptibility results by standard methods were available for of the 8 specimens with gram-negative bacteriuria processed by the direct methods. Table compares identification results obtained by the new rapid direct method with the results of standard methods for these specimens. For 8 of the specimens, or 9%, the rapid direct system produced the same speciation results as the standard methods (Table ). Five isolates were misidentified because of discrepancies in key biochemical 88 Downloaded from https://academic.oup.com/ajcp/article-abstract////8 on April 8
8 HEINZE, THRUPP AND ANSELMO A.J.C.P. February 99 INTERPRETATION -» (NEGATIVE) (POSITIVE) >8!! <*9 O U UJ - - li- o O Q - A < I NUMBER OF SPECIMENS * B * «lo^lo 8 9 By Dilution Pour-Plates * By Standard. ml Loops CFU/ml. QUANTITATIVE CULTURE* FIG.. Gram-negative bacteriuria. Detection of positive specimens and estimated quantitation by direct smear and Gram stain compared with the distribution of colony counts by standard quantitative culture, for urine specimens that yielded gram-negative bacilli on culture. reactions, such as a negative indole test for E. coli (Table ). Since the R/B system, supplemented by citrate and urea, constituted the standard routine method in this study, only those biochemical tests that overlapped within the two systems could be compared directly (Table ). For the individual tests, agreements between the methods ranged from 8 to %. The poorest results were obtained with the indole, ornithine decarboxylase, lysine decarboxylase, and ONPG tests. Lower inoculum density was one factor that correlated with discrepant results for these individual tests. It should be noted that the overall results of identification by the direct filter-paper-strip tests (9% correct) were skewed by the fact that 8% of the isolates were E. coli. Further evaluation of this direct system with strains other than E. coli is needed. Rapid Direct Antibiotic Susceptibility Tests Antimicrobial susceptibility results were recorded for the direct rapid disk-elution system following - and -hour incubation periods. Forty-nine (%) of the specimens with gram-negative bacteriuria achieved an acceptable growth index (>.9) in hours, while 8 (9%) achieved adequate growth by hours. Six specimens needed incubation for an additional hour, or hours total. Results for the -hour rapid test readings, including the six specimens that needed additional incubation, are compared with the standard 8-hour disk-diffusion results in Table. A very major discrepancy (sensitive by the rapiddirect method but resistant by the standard disk-diffusion test) for a single drug occurred in six (%) of the individual specimens tested. A major discrepancy for any single drug occurred in an additional ten individual specimens (%), and major discrepancies for two or three drugs occurred in seven more (8%) of the individual specimens. From an overall total of antimicrobic-organism test pairs, there were.% very major,.% major (sensitive by the standard disk-diffusion but resistant by the rapid-direct test), and.% minor (sensitive or resistant by one test, and intermediate by the other) discrepancies. For those specimens that Downloaded from https://academic.oup.com/ajcp/article-abstract////8 on April 8
Vol.. No. RAPID DIRECT URINARY CULTURE PROCEDURE 8 Table. Rapid (-Hour) Direct Identification of Gram-negative Pathogens in Urinary Sediment with the Use of Reagent-impregnated Filter-paper Strips (Pathotec) Organism E. coli Klebsiella Proteus mirabilis Other* Correct Identification Total Number of Urine Specimens Number (%) 9 9 U 9 * Includes correct identification of one Enterobacter, two Providencia, and one Citrobacter, and incorrect identification of one Pseudomonas and one Enlerobacter. had achieved adequate growth by three hours, the proportion of discrepancies was similarly low. Overall, the error rate for the rapid-direct test compared with the standard indirect disk-diffusion test was approximately equal to the rate of nonreproducibility of the disk-diffusion test. Discussion Rapid reporting of relevant, accurate results is an important goal for improving clinical microbiology data in the management of infectious disease. A comprehensive combined system for reporting culture results for urine specimens in a few hours has not been reported previously. We have described initial evaluation of a new system of rapid, simple procedures for detection of positive specimens and for quantitation, identification, and antimicrobial susceptibility testing of the direct specimen with reasonable, though not complete, accuracy in - hours. The Gram stain of the direct urine smear proved reliable for screening to detect positive specimens (Table ) and to eliminate specimens with high-count mixed flora. The Gram-stain was also useful for estimating cfu/ml, permitting standardization of the inoculum for the rapid direct antimicrobial susceptibility test. Prior studies have shown the -hour identification procedure using Pathotec strips to be quite accurate as a standard rapid method for identifying strains already grown on primary or subculture plates. In our study with gram-negative bacteriuria, mostly due to E. coli, application of the same method directly on suspensions of urinary sediment proved 9% accurate. The few identification errors usually were the result of equivocal or false-negative biochemical tests. Eleven filter-paperstrips were used in this system, but fewer strips could be selected, which would be more economic and still produce sufficient data for presumptive identification of gram-negative urinary pathogens. Table. Rapid (-Hour) Direct Identification of Pathogens in Urinary Sediment with the Use of Reagent-impregnanted Filter-paper Strips (Pathotec) Errors in Identification of Gram-negative Bacilli Organism (Standard Methods) Escherichia coli Escherichia coli Escherichia coli Enterobacter cloacae Pseudomonas aeruginosa Rapid Identification Unable to identify Pseudomonas Mixed culture Klebsiella pneumoniae Mixed culture Reason for Misidentification Indole-negative Indole-negative ONPG-negative Oxidase ± Sensitive only to gentamicin + polymyxin B "False" resistance to polymyxin B Lysine decarboxylasenegative Ornithine decarboxylasenegative Nonmotile "False" resistance to polymyxin B Inoculum (cfu/ml) x 8 x " x 8 x x The results of the --hour antimicrobial susceptibility tests performed directly on the urine specimens correlated well with results of the standard disk-diffusion procedure, which were not available until 8 hours later. Major limitations of direct susceptibility testing are misleading results from mixed flora and variability of inoculum size; direct susceptibility test- Table. Direct Identification of Gram-negative Organisms in Urinary Sediment Results of Rapid (-Hour) Reagent Strip Tests (Pathotec) Compared with Standard (8--Hour) Media Inoculated from Isolated Colonies Test Phenylalanine deaminase Urease Indole H S Lysine decarboxylase Ornithine decarboxylase o-nitrophenolgalactoside (ONPG) Number of Tests 8 Rapid (+)/ Standard (-) Rapid (-)/ Standard ( + ) % Agreement 9 9 8 9 9 Downloaded from https://academic.oup.com/ajcp/article-abstract////8 on April 8
8 HEINZE, THRUPP AND ANSELMO A.J.C.P. February 99 Table. Rapid (-Hour*) Direct Antimicrobial Susceptibility by Disk-elution (Autobac) of Gram-negative Bacteria in Urinary Sediment Results Compared with Standard (8-Hour) Disk-diffusion Testst No. Discrepancies % Discrepancies Antimicrobial Number of Tests Very Majort Major Minor Very Major Major Minor Tetracycline Chloramphenicol Ampicillin Cephalothin Kanamycin Gentamicin Polymyxin B Carbenicillin Nitrofurantoin Nalidixic acid... * Includes three strains that needed - hours of incubation and three strains that needed hours of incubation. t Standardized agar-overlay modification of the Bauer-Kirby method using NCCLS interpretive standards. t Sensitive by the rapid direct test but resistant by standard disk-diffusion test. Sensitive by standard disk-diffusion test but resistant by rapid direct test. "' Shift from sensitive or resistant to intermediate, or from intermediate to sensitive ing, therefore, has not been recommended as a routine procedure., The few exceptions, when direct susceptibility testing has been justified, were urgent specimens that could be predicted by experience and by Gram-stain to contain one organism alone or a clearly predominant organism present in large numbers. In the proposed rapid system most mixed cultures were successfully screened out and the inoculum was standardized. The system was adequately reliable for most clinical purposes, with overall accuracy of more than 9%. In fact, the "error" rate of discrepant results in our direct-rapid method and the standard indirect diskdiffusion test was approximately equal to the rate of nonreproducibility of the standard disk-diffusion test, and the rate of errors from overnight incubation of direct disk tests after elimination of mixed cultures. For the few individual antimicrobial-agent rapid tests that were discrepant compared with the standard tests, the majority of the discrepant results showed falseresistance ("major" discrepancy), which clinically is less critical than false-sensitive results ("very major" discrepancy). Success of both the rapid identification method and the rapid antimicrobial susceptibility test procedures depended on the following characteristics of fresh clean-voided or catheterized urine specimens: () the large majority of positive urines have high counts of a single organism (most contained more than /ml); () any minority species present from urethral, meatal, or vulval flora did not interfere, because they were usually present in small numbers (< /ml); and () such low-count commensual organisms were usually relatively slow-growing species. The presence of two rapid-growing gram-negative species as contaminants, both in high counts, resulted in occasional misleading identification and antimicrobial susceptibility test results, as anticipated. The chance of inadvertent failure to detect such specimens was minimized by three procedures: () excluding specimens from patients with indwelling catheters and from infants; () detecting mixed flora by Gram stain, () observing mixed, inconsistent results with the rapididentification filter-paper-strip tests or the rapid antimicrobial susceptibility, or both tests. It should be noted that this study was performed on direct fresh clinical urine specimens, not on stored urines or on specimens simulated by inoculation from stock cultures. As expected, therefore, the majority of the positive urine specimens were infected with E. coli. The overall accuracy of the direct identification and antimicrobial susceptibility testing thus reflected the true clinical distribution of urinary pathogens and included relatively few specimens with non-. coli gram-negative bacilluria. Further experience is needed to establish the accuracy of these procedures for larger numbers of specimens with non-e. coli gram-negative bacilluria and also for gram-positive pathogens. Not all urine cultures are sufficiently urgent clinically to necessitate rapid tests. However, in the clinical microbiology laboratory serving a hospital population, enough specimens are received from patients with acute symptomatic infection that rapid culture methods often could facilitate diagnosis and treatment. A laboratory system could efficiently process all urine specimens from hospital inpatients (with indicated exceptions) by the proposed rapid system. All specimens positive by the first step of the method (Gram-staining) could then be processed by the rapid direct identifica- Downloaded from https://academic.oup.com/ajcp/article-abstract////8 on April 8
Vol.. No. RAPID DIRECT URINARY CULTURE PROCEDURE 8 tion and rapid direct antimicrobial susceptibility test procedures. All urines negative or mixed by Gram stain could be examined by a standard urine-culture procedure. Any positive culture missed by the Gram stain (as well as any exceptions, such as specimens from patients with indwelling catheters or from infants) would thus be processed by the standard methods. In some laboratories, other alternatives might be more practical for limited implementation of the proposed rapid system. For example, for laboratories in which trained microbiology technologists are not available in the evening hours, the rapid system could be implemented only for eligible specimens received in the morning, to allow - hours for processing by the dayshift microbiology staff. Specimens received in the evening could be processed by standard plating methods. Another alternative could be to apply the rapid system only to clinically urgent specimens upon request by the clinician. However, application of such a system to only a few specimens might limit the opportunity for achievement of proficiency by the technologists, and would fail to take advantage of the potential for increased laboratory efficiency. Acknowledgments. The staff of the Clinical Microbiology Laboratory, Orange County-University of California Irvine Medical Center, M. Pezzlo, M.A., microbiology supervisor, and P. Valter, M.D., director of clinical microbiology and clinical chemistry, provided cooperation and excellent clinical microbiology service data. References. Barry AL, Garcia F, Thrupp LD: An improved single disc method for testing the antibiotic susceptibility of rapidly growing pathogens. Am J Clin Pathol :9-8, 9. Barry AL, Joyce LJ, Adams AP, et al: Rapid determination of antimicrobial susceptibility for urgent clinical situations. Am J Clin Pathol 9:9-99, 9. Bartlett RC, Carrington GO, Mielert C: Quality Control in Clinical Microbiology. American Society of Clinical Pathologists. Chicago, III., 98, p 8-. Bauer AW, Kirby WMM, Sherris JC, et al: Antibiotic susceptibility testing by a standardized single disc method. Am J Clin Pathol :9-9, 9. Binford JS, Binford LF, Adler P: A semiautomated microcalorimeter method of antibiotic sensitivity testing. Am J Clin Pathol 9:-9, 9. Cobbs CG: Presumptive tests for urinary tract infection. Urinary Tract Infections and Its Management. Edited by Kqye D. St. Louis, C. V. Mosby, 9, pp -. Craig WA, Kunin CM, DeGroot J: Evaluation of new urinary tract infection screening devices. Appl Microbiol :9-, 9 8. Gaby WL, Hadley C: Practical laboratory test for the identification of Pseudomonas aeruginosa. J Bact :-8, 9 9. Heden C, Illeni T: Automation in Microbiology and Immunology. New York, John Wiley and Sons, 9, pp 9-. Hollick GE, Washington JA: Comparison of direct and standardized disc diffusion susceptibility testing of urine cultures. Antimicrob Agents Chemother 9:8-89, 9. Kass EH: Asymptomatic infections of the urinary tract. Trans Assoc Am Phys 9:-, 9. Kass EH: Chemotherapeutic and antibiotic drugs in the management of infections of the urinary tract. Am J Med 8:-8, 9. Kunin CJ: Detection, Prevention and Management of Urinary Tract Infections. Philadelphia, Lea and Febiger, 9, pp -. Kunin CM: The Quantitative significance of bacteria visualized in the unstained urinary sediment. N Engl J Med :89-9, 9. Mcllroy GT, Uy P, Martin NJ, et al: Evaluation of modified R-B system for identification of members of the family Enterobacleriaceae. Appl Microbiol :8-, 9. Perez JR, GillenwaterJY: Clinical evaluation of testing immediate antibiotic disc sensitivities in bacteriuria. J. Urol : -, 9. Rosner R: Evaluation of the PathoTec '"Rapid I-D System" and two additional experimental reagent impregnated paper strips. Appl Microbiol :89-89, 9 8. Seneca H, Avakian S: Rapid bacteriological identification of uropathogens. J Urol :-8, 9 9. Strauss MB, Welt LG: Disease of the Kidney. Boston, Little, Brown, 9, p -. Thornsberry C, Gavan T, Sherris J, et al: Laboratory evaluation of a rapid automated susceptibility testing system: Report of a collaborative study. Antimicrob Agents Chemother :- 8, 9 Downloaded from https://academic.oup.com/ajcp/article-abstract////8 on April 8