Zdrav Vestn 2007; 76: 145–9 

Research article/Raziskovalni prispevek 

CHROMOGENIC MEDIA FOR URINE CULTURES CAN BE
COST-EFFECTIVE


KROMOGENA GOJIŠČA ZA BAKTERIOLOŠKO PREISKAVO URINA SO LAHKO
STROŠKOVNO SMOTRNA


Matjaž J. Retelj, Tatjana Harlander 

Mikrobiološki laboratorij, Zavod za zdravstveno varstvo Novo mesto, Mej vrti 5, SI-8000 Novo mesto 

Abstract 
Background Chromogenic media for diagnostic urinary bacteriology have several advantages over 
traditional media, such as cysteine-lactose-electrolyte deficient (CLED) medium. Chromogenic media allow for easier recognition of mixed growth, save time, reduce workload and 
provide higher detection rates. However, the cost of chromogenic media is significantly 
higher compared to CLED and performance of chromogenic media varies depending on 
the manufacturer. In the present study, performance, turn-around time and cost of Uriselect4 chromogenic medium was compared to CLED. 
Methods For performance analysis, 351 midstream urine (MSU) samples from September 2005 to 
December 2005 were directly plated in parallel on Uriselect4 and CLED agar using the 
calibrated loop technique. Isolates on Uriselect4 were presumptively identified according 
to the product insert. 
For cost-effectiveness analysis, we included 1,972 consecutive MSU samples from May 2005 
to July 2006. We compared the cost of required materials as well as technologists’ or specialists’ time for each medium examined. 
Results No significant differences were found between the isolation rates of urinary pathogens 
on the studied media. The procedure using chromogenic media for uropathogens is 
slightly cheaper than the procedure using CLED, considering the proportion of bacteriuria 
positive samples (50.5 %) and the distribution of taxa among isolates (namely Escherichia 
coli with 59.6 %) observed in our laboratory. At the current isolation proportion in MSU 
samples processed in our laboratory, the average time to reporting results could be 
decreased by 0.3 days. 
Conclusions Use of chromogenic media for urine investigations offers multiple advantages without 
increasing costs compared to procedures using CLED. 
Key words laboratory techniques and procedures; urinalysis; urinary tract infections; costs and cost 
analysis; comparative study; chromogenic compounds; culture media 

Izvleček 

Izhodišča 
Kromogena gojišča za bakteriološko preiskavo urina imajo številne prednosti pred tradicionalnimi gojišči, kot je gojišče s cisteinom in laktozo ter brez elektrolitov (CLED). Kromogena gojišča omogočajo lažje odkrivanje mešane rasti povzročiteljev okužb urinskih poti 
ter skrajšajo čas preiskave, zmanjšajo delovno intenzivnost ter imajo večjo stopnjo detekcije mikroorganizmov. Največja ovira pri uvedbi teh gojišč v rutinske postopke je njihova cena, saj je od 2- do 13-krat višja od cene gojišča CLED. Poleg tega se kromogena gojišča 
različnih izdelovalcev razlikujejo v zmogljivosti. Stroškovna smotrnost zamenjave tradi-

Corresponding author / Avtor za dopisovanje: 

Matjaž Retelj, Zavod za zdravstveno varstvo Novo mesto, Mej vrti 5, SI-8000 Novo mesto, tel.: 07 / 39 34 121, faks: 07 / 39 34 101, 
e-pošta: matjaz.retelj@zzv-nm.si 


146 Zdrav Vestn 2007; 76 
cionalnih gojišč s kromogenimi gojišči še ni bila podrobno raziskana. V tej raziskavi smo 
primerjali zmogljivost, trajanje in stroške postopka s kromogenim gojiščem Uriselect4 in 
postopka z gojiščem CLED. 
Metode Za analizo zmogljivosti gojišča smo 351 zaporednih vzorcev srednjega curka urina od 
septembra 2005 do decembra 2005 neposredno cepili na gojišči Uriselect4 in CLED s tehniko kalibrirane zanke. Izolate na gojišču Uriselect4 smo identificirali skladno z navodilom 
za uporabo. 
V analizo stroškovne smotrnosti smo vključili 1972 zaporednih vzorcev srednjega curka 
urina od maja 2005 do julija 2006. Primerjali smo stroške potrebnega materiala ter potrebne delovne ure laboratorijskih tehnikov in mikrobiologov. 
Rezultati Med gojiščema nismo odkrili statistično pomembnih razlik v stopnji izolacije povzročiteljev okužb urinskih poti. Postopek z uporabo kromogenega gojišča je nekoliko cenejši od 
postopka z uporabo gojišča CLED, če upoštevamo delež za bakteriurijo pozitivnih vzorcev 
srednjega curka urina (50,5 %) in porazdelitev bakterijskih skupin med izolati (zlasti 
bakterije Escherichia coli z 59,6 % vseh izolatov) v našem laboratoriju. Pri tem deležu za 
bakteriurijo pozitivnih vzorcev srednjega curka urina se povprečno trajanje preiskave 
urina z uporabo kromogenega gojišča skrajša za 0,3 dneva. 
Zaključki Uporaba kromogenih gojišč pri bakteriološki preiskavi urina nudi številne prednosti pred 
postopki z uporabo gojišča CLED brez dodatnih stroškov. 
Ključne besede laboratorijske tehnike in postopki; analiza urina; okužbe sečil; analiza stroškov; primerjalna raziskava; kromogene spojine; gojišča 

Introduction 
All samples were transported, refrigerated and processed within 24 hours of sampling.

Traditionally, isolation of urinary pathogens is performed on cysteine-lactose-electrolyte deficient Media and reagents(CLED) medium. Recently, chromogenic media for 
diagnostic urinary bacteriology were developed that We used Uriselect4 agar (#64694) from Bio-Rad (Marprovide presumptive identification of common uri-nes-La-Coquette, France) at € 276 for 500 g and CLED 
nary tract pathogens. Their performance was con-agar (#40129012) from Biolife (Milano, Italy) at € 41 
firmed in several studies.1-7 The greatest advantages for 500 g. Both media were supplied in dehydrated 
of chromogenic media over CLED are easier recogni-form and prepared in the laboratory according to 
tion of mixed growth, shorter analysis duration, work-manufacturers’ instructions. The preparation process 
load reduction and higher detection rates. However, and the prepared plates were routinely checked in 
chromogenic media from different manufacturers our ISO 9001:2000 certified quality assurance system. 
vary in performance and cost. In particular, the cost The plates were used within 14 days of preparation. 
of chromogenic media presents a major obstacle for Kovac’s indole reagent (Fluka, Buchs, Switzerland) 
their use in routine practice, since it ranges from 2 to was used for indole production testing on Uriselect4. 
13-times the cost of CLED. The cost-effectiveness of 
switching from CLED to a chromogenic medium has Inoculation of media and incubation 
not been fully established.1-5, 8 In the present study, 
the performance, turn-around time and cost for Urise-MSU samples were directly plated in parallel on CLED 

lect4 chromogenic medium were compared to CLED. 
and Uriselect4 using the calibrated loop technique.9 
Samples were inoculated with a 1 µL calibrated plastic disposable loop (Copan, Brescia, Italy) using one 

Materials and methods 
plate per sample. All plates were incubated in the air 
at (35 ± 1) °C for 18–20 hours. 

Specimens 

For performance analysis, 351 consecutive midstream Interpretation and identification 
urine (MSU) samples from September 2005 to Decem-Uriselect4 and CLED plates were read independently 
ber 2005 were included in the study. Of these, 147 of each other by separate qualified specialists. The 
samples were referred from the local hospital and 204 number of colonies on each medium was recorded. 
samples were from general practitioners. All plates that produced 50 colonies or more (i. e. . 5 
For cost-effectiveness analysis, 1,972 consecutive MSU × 104 colony forming units/ml of urine) of any mor-
samples from May 2005 to July 2006 were included in phological type and contained . 2 isolates were con-
the study. Of these, 897 samples were referred from sidered indicative for infection and subjected to fur-
the local hospital and 1,075 samples were from gene-ther identification. Isolates on Uriselect4 were preral practitioners. sumptively identified according to the product insert. 


Retelj MJ, Harlander T. Chromogenic media for urine cultures can be cost-effective 

All suspected Escherichia coli and Proteus spp. colonies on Uriselect4 were tested for indole production 
on the plate as recommended by the manufacturer. A 
positive indole reaction for E. coli presumptives and 
negative indole reaction for Proteus mirabilis presumptives on Uriselect4 were taken as definitive identification. Further identification of other presumptive 
colonies from Uriselect4, and all identifications from 
CLED, were performed using conventional morphological and biochemical methods, namely the indole-
methyl red-Voges Proskauer-citrate-double sugar agar-
urea agar (IMVC) panel for Enterobacteriaceae. 

Statistical methods 

Differences between media were compared using 
McNemar’s test. 

Calculation of cost 

For cost-effectiveness analysis, we compared three 
different protocols for the isolation of urinary pathogens: (i) CLED as the primary medium, followed by 
the IMVC panel for confirmation of presumptive Enterobacteriaceae; (ii) Uriselect4 as the primary medium, followed by the IMVC panel for confirmation 
of presumptive Enterobacteriaceae not already identified on the plate; (iii) Uriselect4 as the primary medium, followed by the ID32GN panel (bioMerieux, 
Marcy l’Etoile, France, € 2.84 per panel) for confirmation of all presumptive gram-negative bacteria not already identified on the plate. 
We included the costs of required materials and technologists’ or specialists’ time (€ 15/h and € 26/h, respectively). The cost of materials included isolation 
media, Petri dishes, sterile inoculating loops, confirmation reagents needed for presumptive colonies of 

Pseudomonas, enterococci and 
,,
Enterobacteriaceae

aeruginosa. The cost of work-up included the tech-

excluded from the performance analysis. From 151 
bacteriuria positive MSU samples, 144 isolates were 
recovered on Uriselect4 and 145 isolates were recovered on CLED. On each medium, two of these 
isolates were yeasts. Three MSU samples exhibiting 
mixed growth were detected. Table 1 shows the distribution of isolates between both media. Statistical 
analysis indicated no significant difference between 
the proportion of isolations from CLED vs. Uriselect4 
(p > 0.05). 

Table 1. Results of plating mid-stream urine samples 
on cysteine-lactose-electrolyte deficient (CLED) medium and Uriselect4 chromogenic medium. 

Razpr. 1. Rezultati nacepljanja vzorcev urina iz srednjega curka na gojišče s cisteinom in laktozo ter brez 
elektrolitov (CLED) in kromogeno gojišče Uriselect4. 

Uriselect4 p value 
. 1 isolate Non-significant 
CLED 
. 1 isolate 134 9 > 0.05 
Non-significant 8 181 

Isolates that were cultured from Uriselect4, but not 
CLED, were enterococci (5 samples), P. mirabilis (2 
samples), Klebsiella spp. (1 sample) and group B streptococci (1 sample). Some isolates of E. coli (4 samples), P. aeruginosa (2 samples), enterococci (2 samples), coagulase-negative staphylococci (2 samples) 
were cultured only from CLED and not Uriselect4. 
Sensitivity of the media were calculated based on the 
proportion of strains recovered from each medium. 
One-hundred-fifty-four isolates were recovered from 
either CLED or Uriselect4. The media sensitivity was 

93.5 % and 94.2 % for Uriselect4 and CLED, respectively. 
All but one (98.9 %) of the 87 E. coli isolates from Uri
nologists’ time needed for preparation of media, plat

ing of samples, inoculating of confirmation media, 

preparing gram stains, administration work at receipt 

and reporting; as well as the specialists’ time needed 

for reading of plates, colony counting, reading of con

firmation media and reading gram stains. 

Costs incurred for confirmation of presumptives were 

included in proportion to the isolation frequency of 

particular organisms within samples included in the 

cost-effectiveness analysis. 
We also determined the theoretical turn-around for 
each medium by applying durations of laboratory 
operations to the above mentioned isolation frequencies. The laboratory operations included in the turnaround analysis were: plating of samples, plate analysis, counting, inoculating of confirmation media, preparing gram stains, administration work at receipt and 
reporting, and the incubation periods required. 

Results 

Performance of Uriselect4 

Nineteen (5.4 %) of the 351 MSU samples had incomplete data about their processing and, as a result, were 

select4 produced colonies of the expected color. The 

atypical isolate produced a cream colored colony. 

There were no other cases in which the final identifi

cation of isolates did not conform to the colony 

appearance stated in the Uriselect4 product insert. 

Only 9 MSU samples gave a result above bacteriuria 

threshold and below the upper detection limit of the 

agar plates (200 CFU/plate), so no conclusions could 

be confirmed regarding the correlation of CFU levels 

between CLED and Uriselect4. 

Costs 

Uriselect4 plates cost significantly more than CLED 
plates; however Uriselect4 has the ability to satisfactorily identify E. coli and P. mirabilis on the plate, thereby avoiding the cost of the identification procedure 
for these organisms. Therefore, the cost of the Uriselect4 procedure for urinalysis depends greatly on 
the frequency distribution of major urinary pathogens. MSU samples used for cost-effectiveness analysis included 50.5 % samples with . 1 isolate. The most 
frequently isolated taxa from these samples were included in the cost-effectiveness analysis and had the 
following isolation frequencies: E. coli (59.6 %), enterococci (15.3 %), P. aeruginosa (6.3 %). P. mirabilis 


Zdrav Vestn 2007; 76 

"
. 


. 


. 

*		
	*+,(./01*
2	%34*+,(./01*

#
-
!./01*
-
!./("56

	
		
	






. 


. 

. 

. 
"
. 
!
. 
 
. 

. 

. 

. 

. 

. 

*+,(
-
!

. 


. 

. 

. 
"
. 
!
. 
 
. 

. 

. 

. 

. 

. 

$%&	%	%
'

	
. 

$%&	%	%
'

	
. 
(
)	

	
. 

Figure 1. Cost ratio of procedure using Uriselect4 
followed by IMVC panel for confirmation of presumptive Enterobacteriaceae or ID32GN panel for confirmation of presumptive gram-negative bacteria against 
the procedure using cysteine-lactose-electrolyte deficient (CLED) medium followed by IMVC panel for confirmation of presumptive Enterobacteriaceae. 

Sl. 1. Razmerje stroškov postopka z gojiščem Uriselect4, 
kateremu sledi potrjevanje izolatov, sumljivih za 
enterobakterije na IMVC, ali potrjevanje izolatov, sumljivih za gramnegativne bakterije na ID32GN testu 
proti postopku z gojiščem s cisteinom in laktozo ter 
brez elektrolitov (CLED), kateremu sledi potrjevanje 
izolatov, sumljivih za enterobakterije na IMVC. 

(5.6 %), Klebsiella spp. (4.4 %), Proteus vulgaris 
(1.9 %), Morganella spp. (1.4 %), Enterobacter spp. 
(1.0 %). 
We constructed a model for urine analysis costs 
that takes this frequency distribution of isolates into 
account, and allows for a variable proportion of MSU 
samples with . 1 isolate. The cost ratio of two different Uriselect4 procedures vs. the CLED procedure is 
shown in Figure 1. The CLED and Uriselect4 procedures depicted are described in the Materials and 
Methods section. 
Our model shows that the procedure using Uriselect4 
followed by IMVC panel for confirmation of presumptive Enterobacteriaceae not already identified on the 
plate, costs the same as the procedure using CLED 
followed by IMVC panel for confirmation of all presumptive Enterobacteriaceae, when the proportion 
of MSU samples with . 1 isolate is higher that 48.5 %. 
The theoretical turn-around time differs between the 
two media, since E. coli and P. mirabilis are identified 
on Uriselect4 at the same time the plate is read. An 
additional 24 h for identification of these organisms 
is necessary on CLED. Figure 2 shows the theoretical 
average turn-around time for Uriselect4 and CLED 
procedures. 
Discussion 

Presumptive identification of urinary pathogens and 
detection of mixed cultures on CLED medium as well 
as other traditional media is time-consuming and requires extensive experience in clinical microbiology. 

(
)	

	
. 

Figure 2. Theoretical average time to reporting of 
results against the proportion of mid-stream urine 
samples with . 1 isolate for the procedure using Uriselect4 or cysteine-lactose-electrolyte deficient (CLED) 
medium. 

Sl. 2. Teoretično povprečno trajanje preiskave do rezultata v odvisnosti od deleža vzorcev urina srednjega curka z .1 izolatom za gojišče Uriselect4 in gojišče 
s cisteinom in laktozo ter brez elektrolitov (CLED). 

The introduction of novel chromogenic compounds 
into clinical bacterial diagnostics promised an easier 
and faster identification of isolates, as well as enhanced detection of mixed cultures. The chromogenic compounds used in media for urinary pathogens 
act as substrates for specific bacterial enzymes, allowing for easier differentiation of taxa and better visualization of mixed cultures. 
Several researchers have previously demonstrated 
equal or superior performance of various chromogenic media over traditional media for identification of 
urinary tract pathogens.1-3, 5, 6, 8 Their findings suggest 
that the chromogenic media studied can be used as a 
single primary isolation medium for urine samples. 
We did not find any significant differences between 
isolations of urinary pathogens on the studied media. 
We agree that Uriselect4 can be used as a single primary isolation medium for MSU samples. However, 
we isolated only two yeasts and, thus, are unable to 
conclude whether Uriselect4 can be successfully used 
for isolation of these organisms. Our results do not 
agree with the findings of a previous author, who 
found that chromogenic media for urinary pathogens 
have poor growth support for gram-positive bacteria.3 
Only one E. coli isolate on Uriselect4 had a colony 
appearance that was discordant with the product insert. The proportion of E. coli not exhibiting ß-galactosidase activity in our study was comparable to results obtained in other studies.1 Failure to detect group 
B streptococci on CLED in one sample confirms the 
idea that chromogenic urinary media better support 
the growth of more fastidious strains of Streptococcus agalactiae as well as low numbers of this organism compared to traditional media.1 
In previous studies, chromogenic media detected significantly more mixed growth than CLED agar.6 Some 
authors warn when the concurrent presence of mul


Retelj MJ, Harlander T. Chromogenic media for urine cultures can be cost-effective 

tiple isolates is found on a plate, Kovac’s indole 
reagent should be avoided because tryptophanase 
metabolites can spread in the medium and reach 
indole-negative colonies.2, 5 In our study, only three 
MSU samples with detected mixed growth did not 
allow media comparison from this viewpoint. 
Uriselect4, like most other chromogenic media for 
urinary pathogens, contains chromogenic substrates 
for ß-galactosidase and ß-glucosidase. In addition, the 
presence of tryptophan allows for the detection of 
tryptophan deaminase activity (TDA) as well as 
tryptophanase activity (production of indole) after 
adding a drop of Kovac’s indole reagent to a colony 
on the plate. The medium’s greatest contribution to 
the speed of identification is the ability to identify E. 
coli and P. mirabilis immediately after reading the 
plate. As a result, the need for subculturing and 
performing multiple biochemical tests to identify 
these organisms is eliminated. E. coli is identified by 
the presence of ß-galactosidase activity (pink colonies), absence of ß-glucosidase activity and a positive 
indole reaction. P. mirabilis is identified by the absence of ß-galactosidase and ß-glucosidase activity, 
presence of TDA (brown precipitate in agar surrounding the colony) and a negative indole reaction. 
Based on this information, and the fact that E. coli is 
the predominant species in MSU samples followed 
closely by P. mirabilis isolates, using chromogenic 
media is potentially cheaper than using traditional 
media. However, cost-effectiveness depends on the 
actual proportion of MSU samples with . 1 isolate and 
the proportion of E. coli and P. mirabilis isolated in 
each laboratory. Other researchers speculated that the 
speed and reliability of identification on chromogenic media may render them cost-effective, although 
the price of chromogenic media is much higher than 
traditional media.1-5, 8 Nevertheless, a detailed cost-
benefit study was not performed. 
In contrast to the general opinion that chromogenic 
urinary media are costly, we have shown that the procedure using chromogenic media for uropathogens 
is cheaper than the procedure using CLED, considering the proportion of bacteriuria positive samples and 
the distribution of taxa among isolates observed in 
our laboratory. In our laboratory, the proportion of 
MSU samples with . 1 isolate is 50.5 % (61.9 % for 
samples from the local hospital and 38.6 % for samples from general practitioners). In this context, the 
procedure incorporating Uriselect4 costs practically 
the same as the procedure using CLED. Even if we 
upgraded the identification of all presumptive gram-
negative bacteria not already identified on chromogenic agar to ID32GN panel, the additional overall cost 
would be 8.1 %. We consider this to be acceptable, 
considering the wide coverage of taxa identified by 
the ID32GN panel. At the current isolation propor

tion in MSU samples processed in our laboratory, the 
average time to reporting results could be decreased 
by 0.3 days. 
In conclusion, we believe that the use of chromogenic media, like Uriselect4, in urine investigations, 
offers multiple advantages without increasing costs 
in comparison to procedures using CLED. Chromogenic media have equal or greater sensitivity of 
detection, superior detection of mixed cultures as a 
result of better visualization of different species, and 
enable shorter times to reporting results. Chromogenic media also provide much more useful information regarding appropriate antibiotic therapy after 
only 24 hours compared to CLED. Use of these media 
also allow for substantial upgrading of biochemical 
identification procedures without a significant rise in 
total costs. 

Acknowledgements 

We are grateful to the staff of our laboratory for their 
generous cooperation and excellent technical assistance, 
and Prof. dr. Mario Poljak for critical reading of the manuscript. 

References 

1. 
Hengstler KA, Hammann R, Fahr AM. Evaluation of BBL CHROM-
agar orientation medium for detection and presumptive identification of urinary tract pathogens. J Clin Microbiol 1997; 35: 
2773–7. 
2. 
Carricajo A, Boiste S, Thore J, Aubert G, Gille Y, Freydiere AM. 
Comparative evaluation of five chromogenic media for detection, enumeration and identification of urinary tract pathogens. 
Eur J Clin Microbiol Infect Dis 1999; 18: 796–803. 
3. 
Aspevall O, Osterman B, Dittmer R, Sten L, Lindback E, Forsum U. 
Performance of four chromogenic urine culture media after one 
or two days of incubation compared with reference media. J Clin 
Microbiol 2002; 40: 1500–3. 
4. 
Fallon D, Andrews N, Frodsham D, Gee B, Howe S, Iliffe A, et al. 
A comparison of the performance of cystine lactose electrolyte 
deficient (CLED) agar with Oxoid chromogenic urinary tract infection (CUTI) medium for the isolation and presumptive identification of organisms from urine. J Clin Pathol 2002; 55: 524–9. 
5. 
Scarparo C, Piccoli P, Ricordi P, Scagnelli M. Comparative evaluation of two commercial chromogenic media for detection and 
presumptive identification of urinary tract pathogens. Eur J Clin 
Microbiol Infect Dis 2002; 21: 283–9. 
6. 
Fallon D, Ackland G, Andrews N, Frodsham D, Howe S, Howells 
K, et al. A comparison of the performance of commercially available chromogenic agars for the isolation and presumptive identification of organisms from urine. J Clin Pathol 2003; 56: 608–12. 
7. 
Perry JD, Butterworth LA, Nicholson A, Appleby MR, Orr KE. Evaluation of a new chromogenic medium, Uriselect4, for the isolation and identification of urinary tract pathogens. J Clin Pathol 
2003; 56: 528–31. 
8. 
Samra Z, Heifetz M, Talmor J, Bain E, Bahar J. Evaluation of use of 
a new chromogenic agar in detection of urinary tract pathogens. 
J Clin Microbiol 1998; 36: 990–4. 
9. 
Pezzlo M, York MK. Aerobic bacteriology. In: Isenberg HD, eds. 
Clinical microbiology procedures manual. 2nd ed. Washington: 
ASM Press; 2004. p. 3.12.1–3.12.19. 
Arrived 2006-10-26, accepted 2007-02-07