ARIANE YECHOURON, ANDRE DASCAL,* JANET STEVENSON, AND JACK MENDELSON. Department ofMicrobiology and Infectious Diseases, Sir Mortimer ...
Vol. 29, No. 12
JOURNAL OF CLINICAL MICROBIOLOGY, Dec. 1991, p. 2758-2762 0095-1137/91/122758-05$02.00/0 Copyright © 1991, American Society for Microbiology
Ability of National Committee for Clinical Laboratory StandardsRecommended Quality Control Strains from the American Type Culture Collection To Detect Errors in Disk Diffusion Susceptibility Tests ARIANE YECHOURON, ANDRE DASCAL,* JANET STEVENSON, AND JACK MENDELSON Department of Microbiology and Infectious Diseases, Sir Mortimer B. Davis-Jewish General Hospital, 3755 Cote Ste Catherine Road, and McGill University, Montreal, Quebec, Canada Received 29 July 1991/Accepted 23 September 1991
The National Committee for Clinical Laboratory Standards (NCCLS) recommends, as a quality control for the disk diffusion susceptibility test, the use of three strains from the American Type Culture Collection: Staphylococcus aureus ATCC 25923, Pseudomonas aeruginosa ATCC 27853, and Escherichia coli ATCC 25922. This study assesses the capacity of these strains to detect errors in the overall method. ATCC strains were tested by comparing testing by the standard NCCLS-recommended procedure (ST) with testing under the following conditions: incubation at 25°C, Mueller-Hinton agar depths of 2 mm (AD2) and 8 mm (AD8), agar pHs of 6.5 and 8, inocula with McFarland standards of 0.25 (0.25M) and 4.0 McFarland (4.OM), direct inoculation without preincubation of inoculum (DI), and a 2-h delay between inoculation and disk application (2HR). The frequency of zone measurements outside the NCCLS-recommended control zone limits were as follows: ST, 0%; AD2, 18%; AD8, 9.6%; pH 6.5, 7.9%; pH 8, 5.3%; 0.25M, 3.5%; 4.OM, 24%; DI, 3.4%; 2HR, 1.8%; 25°C (only E. coli and P. aeruginosa were evaluable), 28%. These results suggest that the quality control strains are only partially effective in detecting single extreme laboratory errors and that careful laboratory supervision is necessary even in the setting of properly monitored quality control strains.
To assure that bacterial susceptibility results obtained by disk diffusion of rapidly growing pathogens are both accurate and reproducible, the National Committee for Clinical Laboratory Standards (NCCLS) has recommended performance guidelines (5). Included in these performance standards is the frequent testing of the three American Type Culture Collection (ATCC) quality control strains: Escherichia coli ATCC 25922, Staphylococcus aureus ATCC 25923, and Pseudomonas aeruginosa ATCC 27853. Defined zone diameter limits are recommended against which laboratories can compare their results and thus verify compliance with the recommended method (4). Laboratory errors in the preparation of agar, inoculum, and incubation temperature are known to affect the results obtained by disk diffusion susceptibility tests (3, 7). However, the capacity of these three ATCC strains to detect such errors has not been comprehensively elucidated. We undertook this study to evaluate the ability of these three ATCC strains to detect procedural errors in disk diffusion. To this end, we exposed the three ATCC strains to single variations of the recommended NCCLS protocol. (This material was presented in part at the ASM General Meeting, Dallas, Tex., 5 to 9 May 1991.)
isms. The first of each strain (Sj) was purchased from Difco (Detroit, Mich.) in lyophilized form in February 1988, subcultured twice, and frozen at -70°C in individual aliquots. A fresh aliquot, subcultured consecutively three times on 5% sheep blood agar, was used for each experiment. The second of each strain (S2) was generously provided by the Laboratoire de Sante Publique du Quebec (Ste-Anne-de-Bellevue, Quebec, Canada). These strains were acquired directly from ATCC in 1982 or 1983 and frozen at -70°C. They were subsequently subcultured and lyophilized, the form in which they were received at our institution. These were rehydrated with trypticase soy broth and subcultured consecutively three times on 5% sheep blood agar before use. Media. The media used consisted of two lots of MuellerHinton Il agar (BBL, Mississauga, Canada) and one lot of trypticase soy broth (TSB; Difco). Mueller-Hinton agar plates were hand poured to depths of 4 to 6 mm (standard depth), 2 mm, and 8 mm. Plates with pH 7.2 to 7.4 (recommended pH), pH 6.5, and pH 8.0 were prepared by acidification (with 1 N HCI) or alkalinization (with 1 N NaOH) of media after sterilization. Measurements of pH were reverified after pouring and solidification with a surface electrode. Media were used within 1 week of preparation. Antibiotics. Antibiotic disks used were those of Organon Teknika (Scarborough, Ontario, Canada). Disk potencies (in micrograms) were as follows: amikacin, 30; ampicillin, 10; carbenicillin, 100; cefazolin, 30; ceftazidime, 30; ceftriaxone, 30; chloramphenicol, 30; clindamycin, 2; gentamicin, 10; imipenem, 10; norfloxacin, 10; oxacillin, 1; tetracycline, 30; tobramycin, 10; trimethoprim-sulfamethoxazole, 1.25/ 23.75; and vancomycin, 30. The disk potency of penicillin was 10 IU. Methods. Disk diffusion by the standard method (ST) was
MATERIALS AND METHODS Bacteria. The three strains, E. coli ATCC 25922, S. aureus ATCC 25923, and P. aeruginosa ATCC 27853, were each obtained from two different sources for a total of six organ-
*
Corresponding author. 2758
VOL. 29, 1991 o~fij-e g _ c
O
cn
~
m
c:
~~
_,
iD
U:cn
O
_
k>
k>
m O
t'
N C O
It»
IJ
gU
U>
Oc
k>
_
-&
k> k> > -_E
k>
oO
'O OC JA
OC
k>
J
k>
f»
S
a' aO'
kt
e-*
k>k>k>k>J
k> k>
en
c
s
O
oc
ON
_. -k
Oc
oc
loc
k>
oo
k>0
00 ON
tnc
OJ 1
Oc
O
O
>
CA >
1_
k>n tz> _-
k> k>
o
rD
k0>
k>0
c
O
4à- ->
o rs
k>
k>o
k> O
n
oc
k> _
J
"
en
J
ooc
m
ItnCA 3
k>o
_> CD _.
k>
Ot c>
r_ 0
00
e
r_.O
J'
.
CkF
q k>k-
-J
tl
--J
O
O w
o
o
k>
k>
0
Oc
c k>
CtO
"
s oc
4àJ
-
_
ON
> oc
eA
-J
>tA
c
Oc
,O
O
(i
_ --
1oc
o
S Ic
ok "OCO c
ccc>
_
>»0 lkCD
-Ju
k>
CD
w
k>C
fi
" I'-
10 ,k
Ik
CO
k>-
Ik
5 0
O
n
1
o
o _.
k>
>
k>
k>
Oc
o -
cc>X c
O
,O
O',0O
N-J -J
ui u
tà 4
Oc
Oc Ot
toi k>
k
5-e cccb
ucN
k>
E
k>-
X -
i
k>
>
k>
tJ
Oc
" -
,
k>
k>
k>
k>
k>
J
p-
ON,
k>
k>
k> UJ
tk
Un
_
Oo
tested as recommended in the NCCLS M2A4 document (5). Experimental errors introduced into the standard NCCLS protocol were as follows: room temperature (23 to 25°C) incubation (RT), agar depths of 2 mm (AD2) and 8 mm (AD8), agar pHs of 6.5 and 8, and a 2-h delay between inoculation of organism and application of disks (2HR). The same tube of TSB used in the standard method was used for the inoculation of these plates. This tube of TSB was used again to prepare an inoculum of a 0.25 McFarland standard (0.25M); 10 drops of a 0.5 McFarland standard was diluted one in two and immediately inoculated. A 4.0 McFarland standard (4.OM) was prepared by inoculating 15 colonies into TSB and incubating it for 3 h at 35°C. At the end of this period, it was compared to a McFarland nephelometer standard of 4.0 and adjusted. A second McFarland standard of 0.5 was prepared directly from the blood plate and inoculated immediately on MuellerHinton agar (DI). The antibiotics against which each organism was tested are listed with their acceptable zone measurements in the first columns of Tables 1, 2, and 3. Each strain was tested twice on separate days. Zone sizes were read by a single observer (A.Y.) at the end of 18 h of incubation by using a caliper, and results were rounded to the nearest whole number.
kn k> en
n Un
RESULTS Details of results are given in Tables 1, 2, and 3. When method ST was used, all measurements were within the acceptable zone limits. After 18 h of RT incubation, S. aureus exhibited minimal growth. After 36 h, zones were definable but very large. Of 28 P. aeruginosa measurements at RT, 16 (57%) were outside acceptable limits, compared with 4 of 44 (9%) E. coli measurements, for an overall total of 20 of 72 (28%) unacceptable measurements for these two organisms. At AD2, 21 of 114 (18%) measurements were outside acceptable zone limits (S. aureus, 11 of 42; E. col, 1 of 44; and P. aeruginosa, 9 of 28). At AD8, 11 of 114 (9.6%) measurements were outside the control limits (S. aureus, 2 of 42; E. coli, 6 of 44; and P. aeruginosa, 3 of 28). At pH 6.5, 9 of 114 (7.9%) results were out of limits (S. aureus, 6 of 42; E. coli, 3 of 44; and P. aeruginosa, 0 of 28). At pH 8, 6 of 112 (5.3%) results were out of range (S. aureus, 6 of 41; E. coli, 0 of 44; and P. aeruginosa, 0 of 27). At 0.25M, 4 of 114 (3.5%) measurements were out of range (S. aureus, 3 of 42; E. coli, 0 of 44; and P. aeruginosa, 1 of 28). At 4.OM, 27 of 114 (24%) results were out of limits (S. aureus, 11 of 42; E. coli, 14 of 44; and P. aeruginosa, 2 of 28). DI as well as 2HR resulted in only rare deviations: 4 of 114 (3.4%) (S. aureus, 2 of 42; E. coli, 1 of 44; and P. aeruginosa, 1 of 28) and 2 of 114 (1.8%) (S. aureus, 0 of 42; E. coli, 2 of 44; and P. aeruginosa, 0 of 28), respectively. Among the two strains of each individual organism, we noted six discrepant measurements: E. coli S1 and S2 tested at 4.OM demonstrated discrepant results against carbenicillin, and S. aureus S, and S2, when tested against clindamycin and rifampin, gave different results with a large inoculum or AD8.
UJ
k> Oc
DISCUSSION
Oc
-
cz o
C
"O
k> O
2759
CD
k>
sC
k> k>
m
ONJ
-J
wO
k>
r
OC
j
-a w
kt>
.
k>
Oc
UJ
-k~
--J
"
cO
" k>
J
t»J
Oc
O
"
-4~ l "J It w ->- tA
:
B :A
O
oJ
_
S r 1< 0 -J
o. c -c
z
k>
s> c.. O C Br 2nc ' æt
O - O
:
B~ S 3< °n _
11
CND
~r
O
N 00 a,,
rr- .E OC
'I . -. e,.-
N
..c>o
N
1
0N Z c
rq ON O
Oq
c> c>
OC o
cj00 0. N
N
N 0-
N
E. E O 000. 010 0 .0
Z 00-
OOc>
eo
c.
o . .
cE oo Z-o -o
Z =
.
- c> Oc o
4.,
O z O X a: Es F > UU
F
VOL. 29, 1991
ERROR DETECTION ABILITY OF QUALITY CONTROL STRAINS
O3
Z> Il
g. a O o 3 o
