Abstract: Evaluation and introduction of a chromogenic urinary tract infection (UTI) medium for the primary culture of routine urine specimens in a very large clinical microbiology laboratory is described. A modified scheme for the direct identification of the main groups of organisms causing UTI has led to quality improvements, with enhanced discrimination of mixed cultures and in efficiencies in staffing and working time that counterbalance increases in media costs. These improvements, together with the concurrent introduction of boric acid preservative for specimens from general practice, significantly reduced the level of positive results by 4.6% (2383 specimens) over an eight-month period (P
Key words: Boric acids. Chromogenic compounds. Culture media. Urinary tract infections.
Microbiological examination of urine is the most frequently requested investigation in diagnostic clinical microbiology laboratories. Development of methods for the routine culture of urine for the presence of urinary tract infection (UTI) in laboratories with very high routine specimen workloads has sought to combine both rapid, efficient techniques with the level of accuracy and reliability that clinical needs demand. The pursuit of such techniques often leads to the introduction of compromises in identification procedures, resulting inevitably in reduced accuracy of organism identification.
One solution to the problem is to re-evaluate reporting procedures to coincide more closely with clinical need. This may be achieved by reaching agreement on the level of identification necessary for effective clinical management and dividing isolates into well-recognised types rather than individual species. Development of commercial chromogenic media provides the opportunity to investigate the potential of this approach by direct isolate identification.
At University Hospital, Nottingham, with an annual urine workload of 140 000 specimens, it was decided that, of the chromogenic media currently available, Chromogenic Urinary Tract Medium CM949 (Oxoid Ltd., Basingstoke, Hampshire, UK) - a complex agar based on electrolyte deficient CLED medium (Oxoid) - most closely matched local clinical criteria.
Presumptive, direct differential identification of the main pathogens that cause UTI is derived from colour changes in either the bacterial colony or the medium itself. A small number of simple additional tests are necessary to complete or confirm initial findings on some isolates, and organisms outside the typical range of morphological types described in the manufacturer's supporting literature may be recognised reliably with experience.
It is well established that bacterial counts of >= 10^sup 5^ colony-forming units (cfu)/mL are indicative of infection, and that counts of less than this are likely to indicate contamination.1,2 Subsequently, this narrow approach has been modified to take account of the many patient categories and the variety of specimen types submitted. Routine culture methods within the Public Health Laboratory Service are designed to consider bacterial counts of >= 10^sup 4^ cfu/mL as potentially significant, depending on other factors which include the method of specimen collection and number of organism types present.3
In the current study, this level of sensitivity is accepted as suitable for the routine diagnosis of most UTIs using the calibrated loop technique4 as the method of bacterial quantification. Factors contributing to an increase in the bacterial count, such as delay in transporting specimens to the laboratory, increase the potential for misdiagnosis owing to overgrowth, and while it is reasonable to expect specimen delay to be minimal where laboratories are situated within the hospital they serve, those submitted from general practice often are subjected to extended transit times. At Nottingham, it was decided to support the introduction of the chromogenic medium with the use of boric acids as a preservative in the containers supplied for general practice specimens.
Materials and methods
Evaluation comprised a series of studies designed to investigate the relative discrimination and sensitivity of culture results obtained using Oxoid chromogenic medium compared with those obtained using Oxoid CLED medium. Following introduction as the routine urine culture method, and the addition of boric acid as a specimen preservative, a retrospective evaluation of general practice specimens was undertaken that compared results obtained during the subsequent eight months with those of the corresponding eight months of the previous year. All evaluations were made using quarter plates and 1 (mu)L inocula (standard loop), which is the routine culture method used in the laboratory.
Discrimination of organisms
A total of 87 urine specimens known to contain one or two organism when tested using the existing CLED culture method were replated onto CLED and the chromogenic media using a standard 1 (mu)L loop. The culture plates were incubated in air overnight (18 h) at 37 deg C and then examined for growth. Organisms were recognised on the basis of colonial appearance and ascribed to agreed types.
With CLED medium, additional identification of strains was limited to the recognition of the presence or absence of lactose fermentation, whereas with the chromogenic medium a wider range of presumptive identifications was possible owing to the range of colour reactions possible with the chromogens present in the medium. However, for clinical purposes it was decided to limit the identification level of coliform bacilli to a description of type only (Table 1).
In some instances, a small number of additional qualifying tests were necessary to complete the identifications. It was found helpful to use oblique lighting to examine the chromogenic medium plates for small clear colonies typical of streptococci (other than enterococci) and Candida spp. that might otherwise have been missed. Examination of a wet film easily identified Candida spp., and a positive or negative catalase test distinguished staphylococci and streptococci, respectively. Presumptive identification of Staphylococcus aureus was made on the basis of positive rapid latex agglutination (Pastorex, Bio-Rad Laboratories Ltd., Hemel Hempstead, Herts, UK) and confirmed by production of deoxyribonuclease (DNase). Many strains of S. aureus produce a yellow pigmentation on the chromogenic medium. Staphylococci that were negative by rapid latex agglutination were further distinguished into S. epidermidis and S. saprophyticus on the basis of their sensitivity to novobiocin. Streptococci (other than enterococci) were assigned to groups on the basis of their specific carbohydrate antigens (Streptococcal Grouping Kit; Oxoid).
Effect of bacterial overgrowth on organism recognition
Overnight broth cultures of common urinary tract pathogens in various combinations were inoculated onto quarter plates of CLED and chromogenic media using a 1 (mu)L standard loop. In addition, standard tenfold dilutions were prepared from the overnight broth cultures, following the Miles and Misra surface-- viable counting technique,6 and inoculated similarly. Cultures were incubated in air overnight at 37 deg C and examined for growth. The relative visible growth of each organism was recorded in a semiquantitative manner for each dilution.
Preliminary comparative study
The first 124 daily urine specimens were taken on two separate occasions and cultured separately on the CLED and chromogenic media using the routine laboratory method. Following overnight incubation in air at 37 deg C, the cultures were examined independently and the results recorded. Where discrepancies occurred, cultures were re-examined jointly and a consensus reached to agree results. Cultures were designated as infected (positive), no significant growth (negative), or as either contaminated/colonised (equivocal) according to the specimen type, and the viable count and number of organism types present (Table 2).
Comparative discrimination studies
Of the 87 known positive specimens examined, 69 produced identical results on both media and the other 18 showed one additional organism present on the chromogenic medium only. No additional organisms were isolated solely on the CLED medium. This improved discrimination of organisms in mixed culture on the chromogenic medium was demonstrated further in the second study (Table 3), but was not true for all organism combinations.
Comparable discrimination was obtained with two-- organism mixtures, the notable exception being that Proteus spp. were not discernible in combination with Pseudomonas spp. on the chromogenic medium. With three-organism combinations, enterococci were discriminated better in mixed culture with coliform bacilli on the chromogenic medium but, once again, Proteus spp. could not be recognised. The reason appears to be that Proteus spp. are identified by the formation of brown pigment produced by a reaction with the medium itself, rather than a distinctive colony colouration, and this pigmentation cannot easily be seen when the medium is masked by an overgrowth of other organisms.
In-use comparative studies
Results of the preliminary study indicated a relatively small reduction in positive and equivocal results, and a corresponding increase in negative results (Table 4). Although appearing to indicate a trend, the sample number was small and the findings not statistically significant. However, when combined with the results of the other comparative studies, the chromogenic medium proved comparable to CLED medium for routine urine culture, and had the advantages of improved discrimination of mixed growths and an extended range of direct isolate identification. The introduction of the chromogenic medium into routine urine culture practice, concurrently with the use of boric acid in urine specimen containers submitted from general practitioners (GPs), permitted a retrospective eight-month evaluation (of GP specimens) and comparison with the corresponding eight months from the previous year. This showed that the trend indicated in the preliminary study was highly significant (Table 5).
Joint introduction of the Oxoid Chromogenic Urinary Tract medium and boric acid as a preservative has brought the positive infection rate of specimens from general practice into line with those from hospital sources. Positive results from general practice specimens have reduced by 4.6% (2383 specimens) and negatives increased by an identical figure. The equivocals remained constant. Application of 99% confidence intervals to the positive results indicates that the comparative reduction is highly significant (P
The laboratory serves two very large teaching hospitals where the ratio of in-patients to out-patients, age spectrum, proportion catheterised and specialty mix provided a relatively stable population for the duration of the study. Additionally, the number of specimens from general practice during the eight-month comparative periods was almost identical. Reduction in positive results and corresponding increase in negatives can be attributed to the improved discrimination of mixed cultures by the chromogenic medium used and prevention of bacterial overgrowth by the boric acid preservative; however, it has not been possible to quantify the specific effects of each method change because they were introduced simultaneously.
Benefits to the laboratory include a smaller number of organisms to process for identification and sensitivities, amounting to around 70 isolates per week. Introduction of the chromogenic medium has removed the need for the additional formal identification that is necessary when CLED medium is used alone, and reduced the number of repeat and follow-up specimens. This has resulted in staffing reductions at the bench and improved specimen turnaround times - benefits that offset the increased cost of the chromogenic medium. Improvements for the patients include reduced false-- positive reporting, with the obvious benefit of savings in unnecessary antibiotic treatments.
Accuracy of Chromogenic Urinary Tract medium in bacterial identification has been investigated by other groups7 but was not critical with the system adopted for our study, in which reporting was more structured to practical clinical management. The limited identification of coliform bacilli was balanced with the reliable direct recognition of organisms beyond the manufacturers stated range. Oxoid chromogenic UTI medium provides a reliable and costeffective alternative to CLED medium for the routine culture of urine specimens.
We would like to thank Dr V. Weston for her help, advice and support.
1 Kass EH. Bacteriuria and the diagnosis of infections of the urinary tract. Arch Intern Med 1957; 100: 709-14.
2 Kass EH. Asymptomatic infections of the urinary tract. Trans Assoc Am Physicians 1956; 69: 56-63.
3 PHLS Standard Operating Procedure: Investigation of Urine. BSOP 41, Issue 1, 1998.
4 Stevens M. Screening urines for bacteriuria. Med Lab Sci 1989; 46: 194-206
5 Johnson HH, Moss MV, Guthrie GA. The use of boric acid for the preservation of clinical urine specimens. The bacterial examination of urine: report of a workshop on needs and methods. Public Health Laboratory Service Monogram Series. London: HMSO, 1978; No.10: 22-8.
6 Miles AA, Misra SS. The estimation of the bactericidal power of the blood. J Hyg Camb 1938; 38: 732.
7 Iliffe AC, Fallon DA. Comparison of the performance of CLED agar with chromogenic UTI medium for the isolation and presumptive identification of isolates from urine: Part 1. PHLS 25th Annual Scientific Conference, University of Warwick, September 2000.
8 Edwards PR, Kauffmann F. A simplification of the Kauffmann-White schema. Am J Clin Pathol 1952; 22: 692.
9 Rowe B, Hall MLM. Kaufmann-White Scheme 1989. Laboratory of Enteric Pathogens, Central Public Health Laboratory, 61 Colindale Avenue, London NW9 SDF.
10 Elston HR, Fitch DM. Determination of potential pathogenicity of staphylococci. Am J Clin Pathol 1964; 42: 346.
11 Esserts I, Radebold K. Rapid and reliable identification of Staphylococcus aureus by a latex agglutination test. J Clin Microbiol 1980; 12(5): 641-83
12 Schleifer KH, Kloos WE. A simple test system for the separation of staphylococci from micrococci. J Clin Microbiol 1975; 1: 337
13 Steel KJ. The oxidase reaction as a taxonomic tool. J Gen Microbiol 1961; 25: 297.
M. BAKER, P. CHEETHAM, J. SHING and N. YAU
Public Health Laboratory, University Hospital, Queen's Medical Centre, Nottingham NG7 2UH, UK
(Accepted 1 June 2001)
Copyright Royal Society of Medicine Press Ltd. 2001
Provided by ProQuest Information and Learning Company. All rights Reserved