Professional Documents
Culture Documents
Aaron
Aaron
PII: S0732-8893(18)30196-2
DOI: doi:10.1016/j.diagmicrobio.2018.06.010
Reference: DMB 14623
To appear in: Diagnostic Microbiology & Infectious Disease
Received date: 16 January 2018
Revised date: 21 May 2018
Accepted date: 13 June 2018
Please cite this article as: Aaron Campigotto, Lee Goneau, Larissa M. Matukas , Direct
identification and antimicrobial susceptibility testing of microorganisms from positive
blood cultures following isolation by lysis-centrifugation. Dmb (2018), doi:10.1016/
j.diagmicrobio.2018.06.010
This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.
ACCEPTED MANUSCRIPT
Direct identification and antimicrobial susceptibility testing of microorganisms from positive blood
T
*Both authors contributed equally to this work
IP
CR
a
Division of Microbiology, Department of Pathology and Laboratory Medicine, London Health Sciences
US
Centre, London, ON, Canada; bPublic Health Ontario, Toronto, ON, Canada; cDivision of Infectious
Diseases, Department of Medicine, St. Michael’s Hospital, Toronto, ON, Canada; dDivision of
AN
Microbiology, Department of Laboratory Medicine, St. Michael’s Hospital, Toronto, ON, Canada
M
ED
Present address: Division of Microbiology, St. Michael’s Hospital, 30 Bond Street, Toronto, ON, Canada,
M5B 1W8
CE
AC
Abstract
We evaluated a direct from positive blood culture pelleting procedure that utilizes a lysis-centrifugation
susceptibility testing (AST) and rapid methicillin- and beta-lactam-resistance screening. The
identification evaluation was performed on 125 cultures and resulted in the correct genus level
T
identification in 91.2% of cultures and a species level concordance of 82.4% compared to routine
IP
subcultured growth. For the AST evaluation, susceptibility results from direct pelleting and subcultured
CR
growth for 187 cultures were compared; an average ±twofold dilution agreement of 98.2% (1650/1681)
US
and 98.6% (1375/1394) for Gram-negatives and -positives respectively was found. Major errors fell
below 5% except for MRSA, which was falsely reported as oxacillin sensitive 17.2% (11/66) of the time.
AN
Lastly, the sensitivity and specificity of rapid MRSA screening was 94.7% (36/38) and 90.0% (9/10)
respectively, while the ESBL screening results were 90.3% (65/72) and 100.0% (13/13) respectively.
M
1. Introduction
The accurate identification and antimicrobial susceptibility testing (AST) of the pathogen causing a
bloodstream infection is a one of the most important tasks carried out by the clinical microbiology
laboratory. Sepsis is associated with high mortality rates and the time to effective antibiotic therapy is
an important determinant in reducing both morbidity and mortality (1). While clinical presentation,
T
epidemiologic factors, and Gram stain results can guide empiric antimicrobial therapy, definitive
IP
antimicrobial therapy relies on identification and AST. Importantly, obtaining pathogen identification to
CR
the genus or species level, is essential for those organisms that may not be covered by typical empiric
US
agents or intrinsically resistant organisms that cannot be differentiated from susceptible ones by Gram
staining alone (e.g. Pseudomonas species, Enterobacter species). With traditional microorganism
AN
identification, performed by analyzing biochemical reactions and carbohydrate fermentation, growth of
the organism from solid media is usually required; similarly organism growth on solid media is typically
M
required for AST although methods to expedite both traditional identification and susceptibility testing
ED
Matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) has been
introduced into the clinical microbiology lab for faster and more accurate microorganism identification
CE
(3, 4). Similar to traditional identification tests, MALDI-TOF MS identification is valid for organisms
AC
growing from solid media. Various methods for direct identification of organisms from positive blood
cultures generally rely on the lysis of the human blood components in the blood culture broth followed
by the separation and concentration of the microorganism by centrifugation or filtration (5–8); the
The limitations of these procedures include high cost, large volume of blood culture broth required,
lengthy procedure time, lack of consistently reliable gram positive bacterial identification or lack of
ACCEPTED MANUSCRIPT
viable organism for direct AST of the concentrate (2, 8, 9). The present study evaluated a lysis-
centrifugation method that overcame many of these limitations, in particular the loss of viability for
direct AST (5). Additionally, the use of a centrifugation method allowed for a lower cost per sample and
We compared the accuracy of MALDI-TOF MS identification from positive blood cultures, using a lysis-
T
centrifugation direct pelleting source to the identification from traditional overnight sub-cultured
IP
growth from solid media. We subsequently demonstrated that this lysis-centrifugation method results
CR
in a pellet that may be used in AST leading to a decreased time to a finalised result, thus providing
US
guidance for directed treatment.
AN
2. Materials and Methods
M
St. Michael’s Hospital, a predominantly adult 475-bed academic hospital in Toronto, Canada, which has
PT
three adult ICUs and a coronary care unit, provides microbiology services for itself and to a community
hospital system within the greater Toronto area. Blood cultures in the current study were collected
CE
between February and June 2015 (identification portion) and between July and January 2015 (AST
AC
portion), in BacT/Alert Standard Aerobic (SA), Standard Anaerobic (SN) or (Pediatric Fastidious
Antimicrobial Neutralization) PF plus bottles and incubated in the BacT/Alert 3D (bioMérieux, St.
Laurent, Quebec ) automated detection system at the hospital in which the patient was admitted.
Processing of all positive blood cultures, including those from the outside client was performed at the St.
Michael’s Hospital microbiology laboratory during daytime lab hours; when a positive culture was
ACCEPTED MANUSCRIPT
identified at the outside lab in an offsite automated detection system, the positive blood culture bottle
was also transported to St. Michael’s Hospital for processing during daytime lab hours.
For the identification portion of the study, the first positive blood culture bottle per patient per 24-hour
period was included in the evaluation if Gram-stain demonstrated a single organism. This was designed
to mimic the anticipated use of the pelleting procedure in the laboratory once routinely employed. In
T
addition, this procedure limited replicates of the pelleting procedure on the same microorganism
IP
increasing the diversity of organisms tested for this study.
CR
2.2 Routine Microorganism Identification
US
All blood culture bottles, irrespective of Gram-stain reaction, were subcultured onto 5% sheep blood,
AN
chocolate, MacConkey, and/or colistin-nalidixic acid (CNA) agar (Oxoid, Ontario, Canada). Inoculated
media were incubated at 35⁰C in 5% CO2 with the exception of the MacConkey agar, which was
M
incubated in ambient air. The plates were examined at 24 and 48 hours. If the anaerobic blood culture
ED
bottle was positive, brucella vitamin K agar (BVKA) plate was included and incubated anaerobically for
48 hours. If yeast were seen on Gram-stain, a Sabouraud agar was also inoculated and held aerobically
PT
at 30⁰C.
CE
Following incubation, microorganism growth on solid media was identified by the Bruker Microflex LT
MALDI-TOF MS (Bruker Daltonics). Briefly, isolates were spotted onto a target plate and 1 µl of α-cyano-
AC
4-hydroxycinnamic acid matrix was applied and allowed to dry prior to identification by MALDI-TOF MS.
Identification results were reported on the Biotyper software (3.0). Scores ≥2.000 were accepted as
species level identification, while a Biotyper score between 1.700 and 2.000 was accepted as a genus
level identification consistent with the instrument recommendations and the laboratory’s reporting
structure. If a Biotyper score from the solid media was found to be <1.700, it was discarded from
analysis as we felt this represented either a database or technical limitation of the MALDI-TOF MS. All
ACCEPTED MANUSCRIPT
Candida species and organisms not achieving a Biotyper score ≥2.000 underwent on plate extraction
1.5 mL of blood from positive blood culture bottles was added to a 2 ml microcentrifuge tube to which
T
0.5 ml of lysis buffer [0.6% polyoxyethylene 10 oleoyl ether (Brij 97) in 0.4 M 3-cyclohexylamino-1-
IP
propane sulfonic acid (CAPS)] was added. This mixture was vortexed for 3-5 seconds and then incubated
CR
at room temperature for 2 minutes. The resulting solution was centrifuged at 12,000 rpm for 1 minute
30 seconds and the supernatant was discarded by pipette. The pellet was suspended in 1 ml of wash
US
buffer (20 mM sodium phosphate, 0.05% Brij 97, and 0.45% NaCl). Immediately following suspension,
the solution was again centrifuged at 12,000 rpm for 1 minute 30 seconds. The supernatant was again
AN
discarded by pipetting off the liquid, only leaving the pellet.
M
The pellet was spotted onto the MALDI-TOF target plate with a sterile, autoclaved toothpick. Gram
ED
positive bacteria and Candida species underwent on-plate extraction with 1 µl of formic acid (70% v/v).
This was allowed to air dry and 1 µl of α-cyano-4-hydroxycinnamic acid matrix was applied. For Gram
PT
negative bacteria, only the 1 µl of α-cyano-4-hydroxycinnamic acid matrix was applied. If a score was
CE
<2.000, a one-time refire of the MALDI-TOF target was allowed, as is consistent with our identification
procedure from the solid media. Microorganism identification scores reported on the Biotyper software
AC
were recorded.
187 culture-negative BacT/Alert SA Standard Aerobic blood culture (BC) bottles (bioMérieux) were
spiked with 100 CFU/mL of 24-hour sub-cultured Gram-positive and -negative organisms and incubated
in the BacT/ALERT (bioMérieux) BC system. Organisms were selected based on their resistance profiles,
ACCEPTED MANUSCRIPT
with ~50% of isolates demonstrating a resistance mechanism including methicillin resistant S. aureus
(ESBL), carbapenemase-producing organisms (CPO), and multidrug resistant organisms (MDR; for
example, Pseudomonas and Acinetobacter species). Blood extracted from positive bottles was both
pelleted and subcultured to facilitate method comparison. Susceptibility profiles were generated from
T
sub-cultured and pelleted isolates using the Vitek2 (bioMérieux) AST system and AST-GP67 (Gram-
IP
positive) or AST-N220 (Gram-negative) cards. Acceptance criteria included essential and categorical
CR
agreement ≥90%, with combined major and minor errors of <10% (11, 12). Major errors were
considered as a susceptible result determined by one method but resistant by the other method. Minor
US
errors were considered whenever one method resulted in intermediate susceptibility determination
AN
while the other method determined either susceptible or resistant. The pellet was also inoculated to an
oxacillin screen plate to assess for MRSA resistance. Growth on plates was assessed after 18- to 24-
M
hours.
ED
In addition to susceptibility testing, resistance screening was carried out using both standard subculture
and pellet approaches for both S. aureus and Enterobacteriaceae. In total, 48 S. aureus isolates were
PT
assessed for methicillin resistance (MRSA) using the MRSA Latex Test kit (Denka Seiken), while 85
CE
Enterobacteriaceae were assessed for extended spectrum beta-lactamase (ESBL)-production using the β
LACTA Test kit (Bio-Rad). Manufacturer instructions were followed for both assays, with the exception
AC
3. Results
3.1 Comparison of MALDI-TOF Identification Using Direct Lysis-Centrifugation and Standard Subculture
Approaches
ACCEPTED MANUSCRIPT
125 positive blood cultures were included for study, comprising of 64 Gram-positive bacteria, 57 Gram-
negative bacteria, and 4 yeasts (Table 1). MALDI-TOF MS correctly identified organisms to the genus
level in 117 (93.6%) bacterial pellets. Growth of the organisms from the pellet was demonstrated in
representative Gram-positive and Gram-negative organisms and yeast (data not shown). Furthermore,
AST is dependent on viable organisms, our AST would have failed if the organisms had died through the
T
pelleting process. The eight isolates not identified to the genus level, were mainly Gram-positive
IP
organisms (six of eight isolates) of which three were S. pneumoniae. There were no incorrect MALDI-
CR
TOF MS genus identifications when the source was a blood culture pellet.
US
One hundred and fourteen isolates from subculture growth met the more stringent species
identification acceptance criteria compared to 103 (82.4%) from the direct blood culture pellet source.
AN
This resulted in a concordance rate to the species level, using a Biotyper score of ≥2.000 by both
methods, of 90.4% (Table 2). Organisms that failed to identify to species level were predominantly
M
Gram-positive bacteria, regardless of which source was used. Of the 64 Gram-positive isolates identified
ED
to the species level from the subcultured solid media, 52 (81.3%) were also identified from the direct
pellet source. Fewer coagulase negative staphylococci species (CNSS) were identified to the species
PT
level from either subcultured growth (20/23, 87%) or the direct pellet source (16/23, 70%), whereas, all
CE
S. aureus (n=13) were identified correctly from either source method (Table 1).
AC
Gram-negative bacteria were identified to the species level in 100% and 93% from subcultured growth
and the direct pelleting source respectively. The four fungemias due to C. albicans, C. parapsilosis, and C.
tropicalis, were all adequately named to the species level from either the subcultured growth or from
3.2 Evaluation of Antimicrobial Susceptibility Testing Results Using Direct Lysis-Centrifugation and
A comparison between using standard subcultured growth versus blood culture pellets as the source for
AST testing was performed on 187 isolates (95 were Gram-negative, 92 were Gram-positive); these
isolates were unique from those used for the bacterial identification study.
A total of 3075 antibiotic comparisons were evaluated, with an average ±twofold dilution agreement of
98.2% (1650/1681) and 98.6% (1375/1394), and categorical agreement of 98.6% (1628/1651) and 98.5%
T
(1373/1394) for Gram-negative and -positive cards respectively (Table 3ab). Major error rates fell below
IP
5% for all antibiotics tested except for MRSA, which was falsely reported as oxacillin sensitive 17.2%
CR
(11/66) of the time resulting in a very major error. Minor error rates were high (12.6%) for the
US
antimicrobial agents amoxicillin-clavulanic acid and nitrofurantoin in Gram-negatives tested using the
GN220 AST card. However, this could be attributed to intermediate MIC values, which fall only one two-
AN
fold dilution away from resistance interpretation. Thus, these errors could be attributed to normal test-
to-test variation using the Vitek2 instrument. High essential agreement values (97.9%) for these agents
M
support this hypothesis. Data logged in the laboratory information system facilitated comparison of AST
ED
resulting time for both methods. Pelleting reduced the average time from Gram-stain result to Vitek2
AST report from 41.6 to 11.6 hours compared to the sub-culture method (Table 4). The presence of
PT
methicillin resistance, detected through the use of plating to oxacillin screen plates, was correctly
CE
Forty-eight Staphylococcus aureus isolates were assessed for methicillin resistance (MRSA), while 85
sensitivity and specificity of the MRSA Latex Test was 94.7% (36/38) and 90.0% (9/10), while the ESBL β
LACTA Test was 90.3% (65/72) and 100.0% (13/13) respectively (Table 5). The ESBL β LACTA test was
reliably interpreted without difficulty. However, MRSA Latex Test results were much less strongly
ACCEPTED MANUSCRIPT
reactive when conducted with pelleted S. aureus compared to those done using 24-hour sub-cultured
4. Discussion
T
With the ability of the MALDI-TOF MS to rapidly and reliably identify microorganisms, its use in the
IP
clinical microbiology laboratory has become routine and innovative methods to utilize this technology
CR
are evolving. Direct MALDI-TOF MS identification and AST from sterile sites, like blood cultures studied
here, has the potential to decrease organism identification by hours if not days. In this study, we were
US
able to compare the use of a direct pelleting procedure, which is a modification of published methods,
AN
to routine MALDI-TOF MS identification from subcultured growth on solid media; compared to the
previously published methods, the modifications, specifically the use of a pelleting procedure as
M
opposed to a filtration method, means that there is both less equipment to purchase and less
ED
Multiple methods of direct identification from positive blood cultures have been proposed with correct
identification to the species level ranging from 67% to 93%. The performance of the direct pellet
CE
method described here was within the upper ranges reported (5, 7). Commonly found with these
published procedures is the ability to correctly identify Gram-negative bacteria more frequently than
AC
Gram-positive bacteria, a finding we also encountered. To partially overcome this limitation, formic acid
was applied to all Gram-positive bacteria. Many (7/12, 58%) of the Gram-positive bacteria that were
unable to be identified to the species level were CNSS; identification to the genus level, however, was
successful. This identification only to the genus level is unlikely to be significant in many clinical
situations since the isolation of a CNSS may represent contamination. It is important to note that S.
The overall concordance of the pelleting technique to identify isolates to the genus level was higher
than most other described methods at 93.6%. For instance, the use of a short incubation plate may only
reach a correct genus level identification greater than 80% after 4 hour (13) or even up to 8 hours of
incubation (14). No isolate was identified to the genus level through the direct pelleting procedure
incorrectly.
T
The major limitations for Gram-positive bacterial identification to the species level occurred most often
IP
with CNSS, where the species differed between the identification from solid media and the direct
CR
pelleting method, and S. pneumoniae, where no identification was able to be obtained from the direct
US
method while identification from subcultured media was correct. In each of the three cases of S.
pneumoniae bacteremia, the blood culture bottle remained in the automated culture system for an
AN
additional eight to twelve hours after flagging positive, as a result of not having a 24-hours laboratory
for processing cultures. In each of these instances, only a few colonies were recovered on the subculture
M
plate indicating a very low viable bacterial load in the bottle. The minimum bacterial density required to
ED
produce a pellet for MALDI-TOF MS identification has been shown to be 106 CFU/ml, and with this
density a minimum of 6 ml of blood was required to produce a pellet; in the above instance, this
PT
threshold may not have been achieved due to the autolysis of S. pneumoniae prior to its processing (15).
CE
Automated AST conducted using this lysis-centrifugation approach was a reliable method for rapidly
AC
determining resistance profiles of bacteria. Removing the subculture step has the potential to allow
same-day susceptibility results, while antimicrobial resistance screening from pellets permitted
resistance determination in approximately one hour post blood culture positivity. However, we did
observe low discrimination of MRSA identification from blood culture pellets tested using the Vitek2
automated system, with only 66.6% of MRSA correctly identified (20/30) compared to the standard sub-
culture approach. Importantly, this was not of significant concern, as subculturing pellets of MALDI-TOF
ACCEPTED MANUSCRIPT
MS identified S. aureus directly to oxacillin screening plates correctly identified the presence of
methicillin resistance. Thus, oxacillin screening plates can be used in place of the Vitek2 for the
detection of MRSA while still providing results in less than 24-hours. Furthermore, as latex agglutination
was able to detect 9/10 MRSA isolates, we believe that additional studies examining this technique are
warranted to determine if this method may supplement or replace oxacillin screen plates.
T
5. Conclusion
IP
CR
The ability of this procedure to generate a reliable identification in almost 90% of positive blood
cultures, and permit rapid AST resulting and resistance screening, when compared to identification from
US
solid media, demonstrates that this method may have an impact on changing microbiology practices.
Further, these approaches have the potential to greatly impact patient management through rapid and
AN
judicious use of appropriate antibiotics, while improving upon infection prevention and control
M
initiatives through timely isolation of patients infected with antimicrobial resistant organisms. We are
currently in the process of implementation within our laboratory with further investigations required to
ED
Funding
AC
This research did not receive any specific grant from funding agencies in the public, commercial, or not-
for-profit sectors.
ACCEPTED MANUSCRIPT
T
IP
CR
References
1. Kumar A, Roberts D, Wood KE, Light B, Parrillo JE, Sharma S, Suppes R, Feinstein D, Zanotti S,
US
Taiberg L, Gurka D, Cheang M. 2006. Duration of hypotension before initiation of effective
AN
antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care
Med2006/04/21. 34:1589–1596.
M
2. Lupetti A, Barnini S, Castagna B, Capria A-LL, Nibbering PH. 2010. Rapid identification and
ED
antimicrobial susceptibility profiling of Gram-positive cocci in blood cultures with the Vitek 2
PT
3. Sauer S, Kliem M. 2010. Mass spectrometry tools for the classification and identification of
CE
4. Tan KE, Ellis BC, Lee R, Stamper PD, Zhang SX, Carroll KC. 2012. Prospective evaluation of a
clinical microbiology laboratory for identification of bacteria and yeasts: a bench-by-bench study
5. Fothergill A, Kasinathan V, Hyman J, Walsh J, Drake T, Wang YF. 2013. Rapid identification of
ACCEPTED MANUSCRIPT
bacteria and yeasts from positive-blood-culture bottles by using a lysis-filtration method and
matrix-assisted laser desorption ionization-time of flight mass spectrum analysis with the
6. Foster AGW. 2013. Rapid identification of microbes in positive blood cultures by use of the vitek
T
Microbiol 51:3717–3719.
IP
CR
7. Ferroni A, Suarez S, Beretti JL, Dauphin B, Bille E, Meyer J, Bougnoux ME, Alanio A, Berche P,
Nassif X. 2010. Real-time identification of bacteria and Candida species in positive blood culture
US
broths by matrix-assisted laser desorption ionization-time of flight mass spectrometry. J Clin
Microbiol 48:1542–1548.
AN
8. Meex C, Neuville F, Descy J, Huynen P, Hayette MP, De Mol P, Melin P. 2012. Direct identification
M
of bacteria from BacT/ALERT anaerobic positive blood cultures by MALDI-TOF MS: MALDI
ED
Sepsityper kit versus an in-house saponin method for bacterial extraction. J Med Microbiol
61:1511–1516.
PT
9. Klein S, Zimmermann S, Köhler C, Mischnik A, Alle W, Bode KA. 2012. Integration of matrix-
CE
10. Machen A, Drake T, Wang YF (Wayne), Anderson A, Melendez A. 2014. Same Day Identification
and Full Panel Antimicrobial Susceptibility Testing of Bacteria from Positive Blood Culture Bottles
Made Possible by a Combined Lysis-Filtration Method with MALDI-TOF VITEK Mass Spectrometry
11. Clark R, Lewinski M, Loeffelholz M, Tibbetts R. 1997. Cumitech 31A, Verification and Validation of
ACCEPTED MANUSCRIPT
12. CLSI. 2015. Verification of Commercial Microbial Identification and Antimicrobial Susceptibility
Testing Systems,1st ed. CLSI guideline M52. Wayne, PA: Clinical and Laboratory Standard
Institute.
T
13. Chen Y, Porter V, Mubareka S, Kotowich L, Simor AE. 2015. Rapid Identification of Bacteria
IP
Directly from Positive Blood Cultures by Use of a Serum Separator Tube, Smudge Plate
CR
Preparation, and Matrix-Assisted Laser Desorption Ionization-Time of Flight Mass Spectrometry. J
US
14. Idelevich EA, Schüle I, Grünastel B, Wüllenweber J, Peters G, Becker K. 2014. Rapid identification
AN
of microorganisms from positive blood cultures by MALDI-TOF mass spectrometry subsequent to
15. Stevenson LG, Drake SK, Murray PR. 2010. Rapid identification of bacteria in positive blood
ED
T
staphylococci
S. auricularis 2 1 1 0 2 0
IP
S. epidermidis 13 11 2 9 4 0
S. hominis 6 6 0 6 0 0
CR
S. lugdunesis 1 1 0 1 0 0
S. warneri 1 1 0 0 1 0
Streptococci
US
S. agalactiae 1 1 0 1 0 0
S. pneumoniae 3 3 0 0 0 3
S. gordonii 1 0 1 0 1 0
AN
S. milleri group1 2 2 0 1 0 1
2
AHS 2 1 1 1 0 1
Other
M
E. faecalis 4 4 0 4 0 0
E. faecium 1 1 0 1 0 0
ED
M. luteus 4 3 1 4 0 0
3
ASB 4 3 1 2 1 1
L. monocytogenes 1 1 0 1 0 0
PT
4
Corynebacterium spp. 2 1 1 0 2 0
B. longum 1 0 1 0 1 0
P. acnes 1 1 0 1 0 0
CE
A. odontolyticus 1 0 1 1 0 0
Gram positive total 64 54 10 46 12 6
AC
Gram-negative bacteria
Enterobacteriaceae
E. coli 36 36 0 34 2 0
K. pneumoniae 8 8 0 8 0 0
Enterobacter spp5 3 3 0 3 0 0
S. marcescens 1 1 0 1 0 0
C. koseri 1 1 0 1 0 0
P. stuartii 1 1 0 1 0 0
Salmonella spp 1 1 0 1 0 0
Other
P. aeruginosa 2 2 0 1 0 1
ACCEPTED MANUSCRIPT
P. multocida 1 1 0 1 0 0
M. catarrhalis 1 1 0 1 0 0
B. fragilis 1 1 0 1 0 0
P. nigrescens 1 0 1 0 0 1
Gram negative total 57 56 1 53 2 2
Yeast 4 4 0 4 0 0
Total 125 114 11 103 14 8
1
S. intermedius, S. anginosus; 2S. sanguinis, S. mitis; 3B. cereusx2, B. pumilus, P.
T
odorifir; 4C. afermentens, C. Striatum; 5E. cloacae complex, E. aerogenes x2
IP
CR
US
AN
M
ED
PT
CE
AC
ACCEPTED MANUSCRIPT
Adjusted identification to
T
species level2
IP
Total (n=114) - 103 (90.4)
Gram-positive (n=54) - 46 (85.2)
CR
Gram-negative (n=56) - 53 (94.6)
Candida (n=4) - 4 (100)
1
Species level identification; 2Only isolates with a Biotyper score ≥2.000 on
US
routine identification included
AN
M
ED
PT
CE
AC
ACCEPTED MANUSCRIPT
T
Cefoxitin 95 100.0 96.8 98.4 0.0 3.2 3.2
IP
Cefixime 95 100.0 96.8 98.4 0.0 3.2 3.2
Cefpodoxime 95 98.9 100.0 99.5 0.0 0.0 0.0
Cefotaxime 95 96.8 95.8 96.3 0.0 4.2 4.2
CR
Ceftazidime 95 95.8 98.9 97.4 0.0 1.1 1.1
Imipenem 95 93.7 95.8 94.7 5.0 4.2 9.2
Amikacin 95 97.9 100.0 98.9 0.0 0.0 0.0
US
Gentamicin 95 98.9 97.9 98.4 0.0 2.1 2.1
Tobramycin 95 100.0 98.9 99.5 0.0 1.1 1.1
Ciprofloxacin 95 98.9 98.9 98.9 1.3 1.1 2.3
AN
Tetracycline 95 97.9 97.9 97.9 1.8 0.0 1.8
Nitrofurantoin 95 97.9 87.4 92.6 0.0 12.6 12.6
Cotrimoxazole 95 97.9 98.9 98.4 0.8 0.0 0.8
M
1
Essential Agreement; 2Categorical Agreement; 3Major Error; 4Minor Error
ED
PT
CE
AC
ACCEPTED MANUSCRIPT
TABLE 3b Automated antibiotic susceptibility test results for Gram-positive organisms compared using both pellet and
routine methods.
Antibiotic Comparisons EA%1 CA%2 (EA+CA)% VME%3 ME%4 MinE%5 (VME+ME+MinE)%
Benzylpenicillin 92 100.0 100.0 100.0 0.0 0.0 0.0 0.0
Ampicillin 26 100.0 100.0 100.0 0.0 0.0 0.0 0.0
Oxacillin 66 83.3 84.8 84.1 17.2 0.0 0.0 17.2
Gentamicin 66 98.5 98.5 98.5 0.0 0.0 1.1 1.1
Ciprofloxacin 92 100.0 98.9 99.5 0.0 0.0 1.1 1.1
T
Levofloxacin 92 98.9 100.0 99.5 0.0 0.0 0.0 0.0
Moxifloxacin 92 100.0 98.9 99.5 0.0 0.0 1.1 1.1
IP
Erythromycin 92 94.6 93.5 94.0 0.0 0.0 6.5 6.5
Clindamycin 92 100.0 100.0 100.0 0.0 0.0 0.0 0.0
CR
Synercid 92 100.0 100.0 100.0 0.0 0.0 0.0 0.0
Linezolid 92 100.0 100.0 100.0 0.0 0.0 0.0 0.0
Vancomycin 92 100.0 100.0 100.0 0.0 0.0 0.0 0.0
US
Tetracycline 92 100.0 100.0 100.0 0.0 0.0 0.0 0.0
Tigecycline 92 100.0 100.0 100.0 0.0 0.0 0.0 0.0
Nitrofurantoin 92 100.0 97.8 98.9 0.0 0.0 2.2 2.2
AN
Rifampicin 66 100.0 100.0 100.0 0.0 0.0 0.0 0.0
Cotrimoxazole 66 98.5 100.0 99.2 0.0 0.0 0.0 0.0
1
Essential Agreement; 2Categorical Agreement; 3Very Major Error; 4Major Error; 5Minor Error
M
ED
PT
CE
AC
ACCEPTED MANUSCRIPT
TABLE 4 Time from blood culture bottle positivity to MALDI identification and Vitek 2 susceptibility
test result.
Sub-culture Pellet Sub-culture Pellet
Organism identification identification susceptibility susceptibility
(mean, SD) (mean, SD) (mean, SD) (mean, SD)
Gram Negative 26.1 hrs (1.61) 1.0 hrs (0.24) 38.6 hrs (3.32) 12.0 hrs (1.21)
Gram Positive 29.9 hrs (1.75) 0.9 hrs (0.23) 44.5 hrs (3.77) 11.2 hrs (1.23)
T
IP
CR
US
AN
M
ED
PT
CE
AC
ACCEPTED MANUSCRIPT
TABLE 5 Performance characteristics of rapid ESBL (β LACTA) and MRSA (Denka) screening from
bacterial pellet.
True True False False Sensitivity Specificity
Positive Negative Positive Negative (%) (%)
MRSA
(Denka) 36 9 1 0 94.7 90
ESBL (β
LACTA) 65 13 0 7 90.3 100
T
IP
CR
US
AN
M
ED
PT
CE
AC
ACCEPTED MANUSCRIPT
Highlights
T
pelleted organisms
IP
Rapid MRSA and ESBL screening from the pellet demonstrated a >90% sensitivity and
specificity
CR
US
AN
M
ED
PT
CE
AC