• Blood culture;
  • clinical impact;
  • matrix-assisted laser desorption/ionization time-of-flight mass spectrometry;
  • rapid diagnosis;
  • sepsis


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. Transparency Declaration
  10. References
  11. Supporting Information

For septic patients, delaying the initiation of antimicrobial therapy or choosing an inappropriate antibiotic can considerably worsen their prognosis. This study evaluated the impact of rapid microbial identification (RMI) from positive blood cultures on the management of patients with suspected sepsis. During a 6-month period, RMI by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) was performed for all new episodes of bacteraemia. For each patient, the infectious disease specialist was contacted and questioned about his therapeutic decisions made based on the Gram staining and the RMI. This information was collected to evaluate the number of RMIs that led to a therapeutic change or to a modification of the patient's general management (e.g. fast removal of infected catheters). During the study period, 277 new episodes of bacteraemia were recorded. In 71.12% of the cases, MALDI-TOF MS resulted in a successful RMI (197/277). For adult and paediatric patients, 13.38% (21/157) and 2.50% (1/40) of the RMIs, respectively, resulted in modification of the treatment regimen, according to the survey. In many other cases, the MALDI-TOF MS was a helpful tool for infectious disease specialists because it confirmed suspected cases of contamination, especially in the paediatric population (15/40 RMIs, 37.50%), or suggested complementary diagnostic testing. This study emphasizes the benefits of RMI from positive blood cultures. Although the use of this technique represents an extra cost for the laboratory, RMI using MALDI-TOF MS has been implemented in our daily practice.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. Transparency Declaration
  10. References
  11. Supporting Information

Sepsis is a major cause of morbidity and mortality in hospitalized patients. In the USA, 750 000 cases of severe sepsis occur annually [1]. In Europe, sepsis occurs in more than 35% of the patients in the intensive care unit. More than 50% of patients who experience septic shock do not survive [2-4].

The management of bacteraemic patients can be improved by the administration of the appropriate treatment without delay [5-7]. Molecular techniques allow for rapid microbial identification (RMI) from blood samples but have limitations, in particular the high cost per analysis and the need for antimicrobial susceptibility testing [8].

Because it allows the identification of microorganisms in a few minutes instead of the hours required by biochemical techniques, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) is a promising alternative diagnostic tool [9]. Since its commercialization at the beginning of the twenty-first century, many new strategies to perform RMI directly from clinical samples have been evaluated [10-13]. Recently, inexpensive strategies that allow RMI within 20 min after the blood culture becomes positive were described [14, 15].

Currently, the usefulness of RMI is still debated. Arguments for RMI state that RMI could lead to the faster adoption of the appropriate antibiotic regimen and help to identify the cause of the sepsis if it is unknown. The still limited information concerning the susceptibility of the microorganisms to antimicrobials is a weakness of RMI [16, 17].

The primary aim of this study was to prospectively evaluate the theoretical impact of RMI from positive blood cultures on the clinical management of bacteraemic patients in our hospitals. The compliance with the recommendations of the infectious disease specialist (IDS) was also retrospectively evaluated to determine the real clinical impact of the RMI technique.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. Transparency Declaration
  10. References
  11. Supporting Information


The Saint-Pierre University Hospital and the Jules Bordet Institute are university-affiliated medical centres located in Brussels, Belgium. Saint-Pierre is a public hospital with vast experience in infectious diseases. Jules Bordet is the only hospital in Belgium that is completely dedicated to cancer. Both institutions are served by the same laboratory, which is open on weekdays from 07.30 until 20.00 h and on Saturday and Sunday from 08.00 until 16.00 h. Positive blood culture bottles are analysed during these time periods. During the night, medical microbiologists and IDSs are on call for emergencies.

Collection of blood cultures and inclusion criteria

From September 2011 to March 2012, the first positive blood culture for each bacteraemic episode in patients from both hospitals was prospectively enrolled in the study. All positive cultures for the same patient obtained within 3 days of each other and presenting the same Gram staining results were considered as belonging to the same episode. When staphylococcal morphology was observed in the Gram staining, RMI was always performed to confirm or rule out contamination.


The positive blood cultures (Bactec Plus Aerobic and Bactec F Lytic Anaerobic; Becton Dickinson, Franklin Lakes, NJ, USA) were prepared and analysed according to a previously described in-house protocol [14]. The spectra were acquired on a Microflex LT system (Bruker Daltonics, Bremen, Germany) and subsequently analysed using MALDI Biotyper Automation Control and Biotyper 3.0 software. At that time, the database (V3.1.2.0) included 3995 spectra. The analyses were performed in batches twice daily. The RMIs were classified as ‘reliable’ or ‘unreliable’ according to previously validated cut-offs of 1.4 and 1.6 for the acceptable identification to the genus and species levels, respectively. A difference of at least 0.3 between the first identification match and the first discrepant match was also required to validate the identification [14]. The last criterion was not required if the identification score value met the manufacturer's instructions (cut-off values of 1.7 and 2 for the acceptable identification to the genus and species levels, respectively). The RMIs that showed a discrepancy with the Gram staining results were also classified as unreliable.

Clinical impact evaluation

The design of the prospective analysis is presented in Fig. 1. For each positive blood culture, the Gram staining result was transmitted to the IDS, who answered in a blinded manner several questions on the clinical presentation and current antibiotic treatment of the patient and explained his therapeutic decision. All data were registered in a standardized case report form by the medical microbiologist (see Supplementary material, Fig. S1). When the RMI was transmitted, the IDS informed the medical microbiologist of any modification of his initial therapeutic decision. All cases in which the RMI provided another benefit were also recorded. Indeed, RMIs from blood cultures showing staphylococcal forms on the Gram staining were expected to exclude or confirm Staphylococcus aureus infections, catheter-related infections and contaminations. The detection of particular organisms could also indicate which additional medical investigations could be conducted to determine the origin of the septic infection if it was unknown.


Figure 1. Design of the prospective analysis.

Download figure to PowerPoint

A retrospective analysis of all cases was then performed. For Saint-Pierre patients, the pharmacy department used billing, medical and nursing files to retrospectively evaluate whether the IDS's therapeutic recommendations were followed. The IDSs of both institutions also retrospectively checked the medical files. In this section, the RMI/conventional identification delay, the IDS contact/adaptation of the treatment delay (when needed) and the influence of the medical unit where blood cultures were collected on the utility of the RMI were also determined.

Statistical analysis

The times needed to obtain the RMI and conventional identification results were evaluated and compared using non-parametric Mann–Whitney–Wilcoxon test. A Fisher's exact test was used to compare the proportion of RMIs that led to an altered treatment at each institution.

We also used Fisher's exact test to determine whether the RMI had a higher or lower impact on particular patient populations; a lower impact could be expected in medical units where clinicians routinely monitor infectious diseases (e.g. patients from intensive care units and sterile units).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. Transparency Declaration
  10. References
  11. Supporting Information

Rapid microbial identifications

During the study period, positive blood cultures from 243 patients were included, and 277 RMIs were performed (see Supplementary material, Table S1).

Based on the cut-off criteria, 61.01% and 38.99% of RMIs (169/277, 108/277) were classified as ‘reliable’ and ‘unreliable’, respectively [14]. The medical microbiologist transmitted 71.12% of all RMIs (197/277) to the IDS. All of the bacterial RMIs transmitted to the IDS are presented in Tables 1, 2 and 3 and in Fig. 2. An RMI suggesting a Candida krusei fungaemia was also transmitted to the IDS, with caution because of a poor score value (see Table S1, patient no. 174).


Figure 2. Description of the rapid microbial identification (RMI) classification.

Download figure to PowerPoint

Table 1. Rapid microbial identification (RMI) transmitted to the infectious disease specialist for cultures that showed Gram-negative bacteria (n = 71)
Patient no.Gram stainingRMIScoreHLogReliable RMI?T?CITRMI/CIT discrepancy?From incubator removal to RMI delay (min)From incubator removal to CIT delay (min)
  1. CIT: conventional identification technique; GNR: Gram-negative rods; GNRcc: Gram-negative rods coccobacilli; H: Homogeneity; T?: Transmission of the RMI to the IDS?; Y: yes, N: no; WC: with caution; NA: not applicable; MD: missing or incomplete data; A. genomospecies, Acinetobacter genomospecies; A. johnsonii, Acinetobacter johnsonii; A. lowfii, Acinetobacter lowfii; A. faecalis, Alcaligenes faecalis; C. fetus, Campylobacter fetus; E. cloacae, Enterobacter cloacae; E. coli, Escherichia coli; H. influenzae, Haemophilus influenzae; K. sedentarius, Kytococcus sedentarius; K. oxytoca, Klebsiella oxytoca; K. pneumoniae, Klebsiella pneumoniae; P. agglomerans, Pantoea agglomerans; P. buccae, Prevotella buccae; P. mirabilis, Proteus mirabilis; P. aeruginosa, Pseudomonas aeruginosa; P. veronii, Pseudomonas veronii; R. ornithocolytica, Raoultella ornithocolytica; R. planticola, Raoultella planticola; S. maltophilia, Steonotrophomonas maltophilia; S. paratyphi, Salmonella paratyphi; S. typhimurium, Salmonella typhimurium.

  2. Homogeneity: number of successive matches identical to the first one; log: difference between the score of the first match and the first discrepancy.

  3. a

    Known limitations of the MALDI-TOF MS.

  4. b

    The medical microbiologist told the IDS that an anaerobic bacterium was suspected (bad MALDI-TOF MS result and anaerobic bottles).

204GNR A. genomospecies 1.47310.157NWC A. faecalis Y502950
111GNR A. johnsonii 1.95760.774YY A. johnsonii N733064
90GNR A. genomospecies 2.03530.651YY A. lowfii Ya502975
241GNR A. lowfii 2.00370.818YYAcinetobacter sp.N1511479
56GNR K. sedentarius b 1.11810.113NWC Anaerobic GNR YMDMD
97GNR Campylobacter suspected C. fetus 1.92470.645YYCampylobacter sp.N1013336
27GNR E. cloacae 1.93840.339YY E. cloacae complex N931559
125GNR E. cloacae 1.88810NAYY E. cloacae complex N541437
232GNR E. cloacae 2.02640.21YY E. cloacae complex N221150
262GNR E. cloacae 2.26660.374YY E. cloacae complex N1041579
10GNR E. coli 2.24110NAYY E. coli N651683
14GNR E. coli 280.275YY E. coli N211369
25GNR E. coli 2.25410NAYY E. coli N921558
30GNR E. coli 1.80660.273NWC E. coli N681395
38GNR E. coli 1.97970.492YY E. coli N471098
45GNR E. coli 2.27980.272YY E. coli N751061
57GNR E. coli 2.3310NAYY E. coli N611497
70GNR E. coli 1.26180.26YY E. coli N1481555
95GNR E. coli 2.07390.574YY E. coli N582970
131GNR E. coli 2.19690.605YY E. coli N801180
144GNR E. coli 1.85110NAYY E. coli N711676
154GNR E. coli 2.27680.423YY E. coli N2031639
155GNR E. coli 2.29190.578YY E. coli N1671544
175GNR E. coli 1.95410NAYY E. coli N971555
181GNR E. coli 230790.522YY E. coli N3251472
213GNR E. coli 2.08510NAYY E. coli N371473
214GNR E. coli 2.13990.548YY E. coli N341161
224GNR E. coli 234210NAYY E. coli N861625
233GNR E. coli 1.83710NAYY E. coli N4521639
239GNR E. coli 2.18760.278YY E. coli N1681552
243GNR E. coli 1.88890.493YY E. coli N151788
250GNR E. coli 2.36790.46YY E. coli N861583
258GNR E. coli 2.16710NAYY E. coli NMDMD
263GNR E. coli 2.19210NAYY E. coli N1051516
264GNR E. coli 2.29210NAYY E. coli N1061517
273GNR E. coli 2.15380.413YY E. coli N2061595
281GNR E. coli 1.79890.636YY E. coli N2731595
282GNR E. coli 1.39160.28NWC E. coli N2741596
286GNR E. coli 2.31390.451YY E. coli N1871529
288GNR E. coli 2.14880.303YY E. coli N3361456
296GNR E. coli 2.31490.361YY E. coli N2571516
299GNR E. coli 2.20980.547YY E. coli N601101
300GNR E. coli 2.09180.465YY E. coli N381463
255GNRcc H. influenzae 2.12510NAYY H. influenzae N953062
259GNRcc H. influenzae 1.63510NAYY H. influenzae NMDMD
109GNR K. oxytoca 2.2730.365YY K. oxytoca N731553
218GNR K. oxytoca 2.23420.283YY K. oxytoca N951516
13GNR K. pneumoniae 1.9240.4YY K. pneumoniae N651691
41GNR K. pneumoniae 1.90360.435YY K. pneumoniae N1101569
58GNR K. pneumoniae 2.15340.306YY K. pneumoniae N1801621
77GNR K. pneumoniae 1.82930.267NWC K. pneumoniae N562199
78GNR K. pneumoniae 1.32570.426YY K. pneumoniae N551648
99GNR K. pneumoniae 2.23970.411YY K. pneumoniae N1001646
112GNR K. pneumoniae 2.2260.369YY K. pneumoniae N741572
134GNR K. pneumoniae 2.08810.258YY K. pneumoniae N1362075
135GNR K. pneumoniae 1.78670.362YY K. pneumoniae N1361515
168GNR K. pneumoniae 2.19950.223YY K. pneumoniae N621654
211GNR K. pneumoniae 2.07190.402YY K. pneumoniae N431209
297GNR K. pneumoniae 1.93210.237NWC K. pneumoniae N391030
26GNR P. aeruginosa 2.17861.06YY P. aeruginosa N931559
85GNR P. aeruginosa 2.26560.68YY P. aeruginosa N951548
145GNR P. aeruginosa 2.25360.757YY P. aeruginosa N1811923
289GNR P. agglomerans 1.80730.24NWC P. agglomerans N834561
153GNR P. buccae 2.1561.084YY P. buccae N10410499
119GNR P. mirabilis 2.33590.8YY P. mirabilis N4681523
67GNR P. veronii 1.61310.109NWCPseudomonas sp.N2001323
268GNR R. ornitholytica 2.50810.39YWC R. planticola Ya1541507
265GNR S. maltophilia 1.41910.186NWC S. maltophilia N2591384
55GNRSalmonella sp.2.24290.36YY S. paratyphi N1181519
190GNRSalmonella sp.1.91910NAYY S. typhimurium N2171512
150GNRSalmonella sp.2.32110NAYYSalmonella sp.N2241614
Table 2. Rapid microbial identifications (RMIs) transmitted to the infectious disease specialist (IDS) for cultures that showed Gram positive bacteria (n = 107)
Patient no.Gram stainingRMIScoreHLogReliable RMI?T?CITRMI/CIT discrepancyFrom incubator removal to RMI delay (min)From incubator removal to CIT delay (min)
  1. CIT, conventional identification technique; GPR, Gram-positive rods; GPCc, Gram-positive cocci in clusters; GPCpch, Gram-positive cocci in pairs and chains; Y: yes, N: no, WC: with caution; NA: not applicable; MD, missing or incomplete data; B. cereus, Bacillus cereus; B. licheniformis, Bacillus licheniformis; C. amycolatum, Corynebacterium amycolatum; S. pettenkoferi, Staphylococcus pettenkoferi; E. canis, Enterococcus canis; E. durans, Enterococcus durans; E. faecalis, Enterococcus faecalis; E. faecium, Enterococcus faecium; L. crispatus, Lactobacillus crispatus; M. luteus, Micrococcus luteus; K. sedentarius, Kytococcus sedentarius; S. agalactiae, Streptococcus agalactiae; S. anginosus, Streptococcus anginosus; S. aureus, Staphylococcus aureus; S. capitis, Staphylococcus capitis; S. dysgalactiae, Streptococcus dysgalactiae; S. epidermidis, Staphylococcus epidermidis; S. hominis, Staphylococcus hominis; S. gallolyticus, Streptococcus gallolyticus; S. haemolyticus, Staphylococcus haemolyticus; S. pneumoniae, Streptococcus pneumoniae; S. mitis, Streptococcus mitis; S. salivarius, Streptococcus salivarius; S. pyogenes, Streptococcus pyogenes; S. saprophyticus, Staphylococcus saprophyticus; S. thermophilus, Streptococcus thermophilus.

  2. Homogeneity: number of successive matches identical to the first one; log: difference between the score of the first match and the first discrepancy.

  3. a

    The medical microbiologist suggested an ‘Enterococcus sp.’ Identification.

  4. b

    RMI reliable to the genus level.

  5. c

    Contamination by cutaneous flora was not excluded.

  6. d

    Known limitation of the MALDI-TOF MS.

257GPR B. cereus 1.68920.109NWC B. cereus NMDMD
157GPR B. licheniformis 1.89150.802YYBacillus sp.N1651572
141GPR C. amycolatum 1.38740.234NWC C. amycolatum N2641236
106GPCc S. pettenkoferi 1.71620.568YY CNS NMDMD
164GPCc E. canisa 1.20410.126NWC E. durans Y1833034
140GPCpch E. faecalis 1.91280.763YY E. faecalis N351524
180GPCpch E. faecalis 173280.584YY E. faecalis N541519
197GPCpch E. faecalis 1.67380.565YY E. faecalis N881582
287GPCpch E. faecalis 2.03480.811YY E. faecalis N411022
156GPCpch E. faecium 1.95590.892YY E. faecium N1671770
272GPR L. crispatus 1.32110.267NWCLactobacillus sp.N764459
88GPCc M. luteus 1.52620.513YbYMicrococcus sp.N481167
183GPCc M. luteus 1.39120.512NWCMicrococcus sp.N471315
184GPCc K. sedentarius 1.6310.526YYMicrococcus sp.N1443006
194GPCc M. luteus 1.42510.297NWCMicrococcus sp.N901707
292GPCc M. luteus 1.79830.499YYMicrococcus sp.N1521617
102GPRPropionibacterium sp.1.68470.607YYPopionibacterium sp.N964659
35GPCpch S. agalactiae 2.14790.704YY S. agalactiae N461641
215GPCpch S. anginosus 1.44420.241NWC S. anginosus N941540
71GPCc S. aureus 1.62480.59YY S. aureus N1701506
80GPCc S. aureus 2.09910NAYY S. aureus N781515
98GPCc S. aureus 1.85410NAYY S. aureus N2331518
151GPCc S. aureus 1.4570.352YbWC S. aureus N521382
167GPCc S. aureus 1.81710NAYY S. aureus N2811564
285GPCc S. aureus 169290.475YY S. aureus N2741596
75GPCc S. capitis 1.58940.466YbWC S. capitis N5961507
83GPCc S. capitis 1.98750.743YY S. capitis N1251192
114GPCc S. capitis 1.81350.59YY S. capitis NMDMD
128GPCc S. capitis 1.63720.604YY S. capitis N851579
147GPCc S. capitis 1.59740.474YY S. capitis N2261611
148GPCc S. capitis 1.45750.393YbWC S. capitis N2251611
277GPCc S. capitis 1.73230.35YY S. capitis N771457
115GPCpch S. dysgalactiae 1.82440.188NWC S. dysgalactiae N871152
3GPCc S. epidermidis 280.616YY S. epidermidis N1311554
32GPCc S. epidermidis 1.9970.63YY S. epidermidis N691397
36GPCc S. epidermidis 1.28720.18NWC S. epidermidis N471560
59GPCc S. epidermidis 1.72870.65YY S. epidermidis N871527
64GPCc S. epidermidis 2.00780.735YY S. epidermidis N931542
76GPCc S. epidermidis 1.42240.255NWC S. epidermidis N1721555
82GPCc S. epidermidis 1.88560.719YY S. epidermidis NMDMD
87GPCc S. epidermidis 1.39530.201NWC S. epidermidis N961526
89GPCc S. epidermidis 1.22340.384NWC S. epidermidis N5341482
96GPCc S. epidermidis 1.65360.651YY S. epidermidis N651492
103GPCc S. epidermidis 1.8890.844YY S. epidermidis N951601
107GPCc S. hominisc 1.97860.934YY S. epidermidis YMDMD
110GPCc S. epidermidis 1.7960.566YY S. epidermidis N731553
121GPCc S. epidermidis 1.9180.615YY S. epidermidis N1301511
123GPCc S. epidermidis 1.70860.65YY S. epidermidis N991509
126GPCc S. epidermidis 1.7150.351YY S. epidermidis N411155
129GPCc S. epidermidis 1.26220.162NWC S. epidermidis N1901335
133GPCc S. epidermidis 1.46310.318YbWC S. epidermidis NMDMD
136GPCc S. epidermidis 2.08670.753YY S. epidermidis NMDMD
139GPCc S. epidermidis 1.77970.618YY S. epidermidis N361525
146GPCc S. epidermidis 1.86560.658YY S. epidermidis N1501387
158GPCc S. epidermidis 1.60120.505YY S. epidermidis N1431176
160GPCc S. epidermidis 1.7260.457YY S. epidermidis N2441324
161GPCc S. epidermidis 2.18780.966YY S. epidermidis N1271545
162GPCc S. epidermidis 1.35720.059NWC S. epidermidis N571139
166GPCc S. epidermidis 2.0560.502YY S. epidermidis NMDMD
186GPCc S. epidermidis 1.50920.352YbWC S. epidermidis N931594
201GPCc S. epidermidis 1.62380.316YY S. epidermidis N761537
216GPCc S. epidermidis 1.85880.635YY S. epidermidis N951516
249GPCc S. epidermidis 1.77480.481YY S. epidermidis N861583
253GPCc S. epidermidis 1.79380.636YY S. epidermidis N941824
256GPCc S. epidermidis 1.71740.606YY S. epidermidis NMDMD
260GPCc S. epidermidis 1.23810.186NWC S. epidermidis NMDMD
267GPCc S. epidermidis 1.86670.606YY S. epidermidis N1541507
269GPCc S. epidermidis 182880.698YY S. epidermidis N1531506
271GPCc S. epidermidis 1.37840.297NWC S. epidermidis N771559
275GPCc S. epidermidis 1.60620.17NWC S. epidermidis N1331558
278GPCc S. epidermidis 1.30350.349NWC S. epidermidis N1191581
306GPCc S. epidermidis 1.69850.622YY S. epidermidis NMDMD
234GPCpch S. gallolyticus 1.84930.434YY S. gallolyticus N1742066
93GPCc S. haemolyticus 1.92170.62YY S. haemolyticus N391434
219GPCc S. capitis 1.50960.341YbWC S. haemolyticus YMDMD
7GPCc S. hominis 1.88260.436YY S. hominis N921324
11GPCc S. hominis 1.89340.739YY S. hominis N651681
21GPCc S. hominis 1.70660.621YY S. hominis N2731583
39GPCc S. hominis 2.09561.053YY S. hominis N1341462
72GPCc S. hominis 1.7850.78YY S. hominis N1481440
73GPCc S. hominis 1.8460.833YY S. hominis N1771439
113GPCc S. hominis 2.2760.85YY S. hominis N1611485
122GPCc S. hominis 2.11960.81YY S. hominis N1001510
143GPCc S. hominis 2.12850.8YY S. hominis N711596
207GPCc S. hominis 2.21160.868YY S. hominis N2511346
212GPCc S. hominis 1.83830.682YY S. hominis N371473
248GPCc S. hominis 2.22961.004YY S. hominis N541563
293GPCc S. hominis 2.19260.854YY S. hominis N751171
305GPCc S. hominis 2.06960.99YY S. hominis N571138
124GPCpch S. pneumoniae 1.0920.09NWC S. mitis/oralis Yd513549
235GPCpch S. salivarius 1.38640.095NWC S. mitis/oralis N1751804
1GPCpch S. pneumoniae 1.92470.729YWC S. pneumoniae N301396
18GPCpch S. pneumoniae 1.73560.513YWC S. pneumoniae N942933
42GPCpch S. pneumoniae 1.45330.358YbY S. pneumoniae N951021
50GPCpch S. pneumoniae 2.1370.79YWC S. pneumoniae NMDMD
105GPCpch S. pneumoniae 1.82240.346YWC S. pneumoniae NMDMD
120GPCpch S. pneumoniae 1.71960.477YWC S. pneumoniae N731571
185GPCpch S. pneumoniae 1.34130.193NWC S. pneumoniae N1861771
198GPCpch S. pneumoniae 1.36440.165NWC S. pneumoniae N871627
242GPCpch S. pneumoniae 1.98660.602YWC S. pneumoniae N1501528
65GPCpch S. pyogenes 2.1580.571YY S. pyogenes N661532
68GPCpch S. pyogenes 1.89680.635YY S. pyogenes N221146
165GPCpch S. pyogenes 1.84770.613YY S. pyogenes N711597
195GPCpch S. pyogenes 2.00660.637YY S. pyogenes N311157
246GPCpch S. pyogenes 2.2780.678YY S. pyogenes N701523
60GPCc S. saprophyticus 1.76160.631YY S. saprophyticus N1971608
159GPCpch S. thermophilus 2.09760.67YY S. thermophilus NMDMD
Table 3. Rapid microbial identifications (RMIs) transmitted to the infectious disease specialist (IDS) for mixed cultures (n = 18)
Patient nrGram stainingRMIScoreHLogReliable RMI?T?CITRMI/CIT discrepancy?From incubator removal to RMI delay (min)From incubator removal to CIT delay (min)
  1. CIT, conventional identification technique; GNR, Gram-negative rods; GNRcc, Gram-negative rods coccobacilli; GPR, Gram-positive rods; GPCc, Gram-positive cocci in clusters; GPCpch, Gram-positive cocci in pairs and chains; Y, yes; N, no; WC, with caution; NA, not applicable; IR, incomplete result.

  2. C. brakii, Citrobacter brakii; C. freundi, Citrobacter freundi; C. perfringens, Clostridium perfringens; E. coli, Escherichia coli; E. faecalis, Enterococcus faecalis; H. influenzae, Haemophilus influenzae; K. pneumoniae, Klebsiella pneumoniae; L. lactis, Lactococcus lactis; M. morganii, Morganella morganii; P. aeruginosa, Pseudomonas aeruginosa; S. aureus, Staphylococcus aureus; S. capitis, Staphylococcus capitis; S. epidermidis, Staphylococcus epidermidis; S. haemolyticus, Staphylococcus haemolyticus; S. hominis, Staphylococcus hominis; S. maltophila, Stenotrophomonas altophilia; S. mitis, Streptococcus mitis.

  3. Homogeneity: number of successive matches identical to the first one; log: difference between the score of the first match and the first discrepancy.

  4. a

    RMI was performed on two positive blood culture bottles for this patient.

  5. b

    Both identifications appeared alternatively in the ten matches, with acceptable score values.

302GPCc S. hominis 1.25110.008NWC

S. hominis

Streptococcus sp.

Acinetobacter sp.

Y, IR1351176
19GPCpch GNR C. brakii 2.02110.035YY

E. faecalis

C. brakii

P. aeruginosa

Y, IR2702902
20GNR GPCpch E. faecalis 2.25780.897YY

E. coli

E. faecalis

S. aureus

Y, IR2731599
40GPCc S. hominis 1.96960.783YY

S. epidermidis

S. hominis

Y, IR461629
44GPCpch GNR

E. faecalis

C. freundii

2.22 2.138 50.693 0.292YY

E. faecalis

C. freundii

46GPCc S. epidermidis 1.76760.639YY

S. epidermidis

S. epidermidis

101GNR GPCpch E. coli 2.16280.366YY

E. coli

S. mitis/oralis

Y, IR943115
137GPCc S. capitis 1.72740.586YY

S. capitis

S. haemolyticus

S. epidermidis

138GNRcc H. influenzae 2.07110NAYY

H. influenzae

K. pneumoniae

Y, IR1365853
152GNR K. pneumoniae 2.07830.325YY

K. pneumoniae

E. coli

Y, IR591206
176GNR E. coli 2.10910NAYY

E. coli

K. pneumoniae

Y, IR981556
177GNR GPR K. pneumoniae 2.27760.407YY

K. pneumoniae

L. lactis

Y, IR971555
188GNR E. coli 2.02710NAYY

E. coli

E. coli

189GPCc GPCpch S. aureus 2.18510NAYY

S. aureus

Streptococcus sp.

Y, IR2173146
193GPCc GPCpch S. haemolyticus 1.91970.502YY

S. haemolyticus

E. avium

Y, IR4871498

C. perfringens b

E. coli



Alternate IDNAYY

C. perfringens

E. coli

S. salivarius

Y, IR1751578
261GNR S. maltophilia 2.06520.298YY

S. maltophilia

Ochrobactrum sp.

Y, IR1151415
279GNR GPCpch M. morganii 2.49210NAYY

M. morganii

E. faecalis

Y, IR1231583

The RMIs that were not transmitted (28.88%, 80/277) included five reliable RMIs that were erroneously not transmitted by the medical microbiologist (oversight) and 75 unreliable RMIs.

Among the unreliable RMIs, 30.56% were nevertheless transmitted to the IDS, with caution (33/108; Fig. 2). These RMIs showed high levels of homogeneity in the ten identification matches, had good score values but failed to meet the 0.3 log difference criterion (a difference of at least 0.3 between the best match and the first discrepant match was required) or were partially informative (genus information or suspicion of ‘coagulase-negative’ Staphylococcus). The decision to transmit data based on unreliable RMIs was made by the medical microbiologist according to his experience.

Among the 197 transmitted RMIs (164 + 33), 88.32% were confirmed by the culture-based identification methods (174/197), 4.06% (8/197) showed erroneous identifications (six RMIs transmitted with caution, one Acinetobacter sp.—RMI no. 90—and one Staphylococcus epidermidis identified as Staphylococcus hominis that most likely resulted from contamination with cutaneous flora). The remaining 7.61% (15/197) were RMIs based on mixed cultures; for eight of them, the Gram staining suggested the presence of several organisms, and for all of them, at least one microorganism remained undetected using the RMI technique. The RMIs for the mixed cultures are presented in Table 3.

Prospective analysis

The transmitted RMIs corresponded to 40 paediatric and 90 adult patients from Saint-Pierre and 67 patients from Jules Bordet. The survey showed that the Gram staining results led to a modification of the patient's treatment in 17.26% of cases (34/197, see Table 4).

Table 4. Clinical context and description of cases for whom the Gram result led to a modification of the empirical antimicrobial treatment (n = 34)
N°Gram stainingRMI resultCITCurrent antibiotic treatmentClinical contextTreatment modification according to the IDS surveyComment
  1. CIT, conventional identification technique; CNS, central nervous system; PTZ, piperacillin/azobactam; AMC, amoxicillin/clavulanic acid; IDS, infectious disease specialist.

Saint-Pierre University Hospital
119GNR Proteus mirabilis P. mirabilis 0Septic shock <UTIAMC additionCommunity acquired UTI
98GPCc Staphylococcus aureus S. aureus 0Dyspnoea, dialysisVancomycin additionSuspicion of CNS infection in a patient with multiple or permanent catheters
38GNR Escherichia coli E. coli AMCPyelonephritisShift to temocillinHospital acquired pyelonephritis

Enterococcus faecalis

Citrobacter freundii

E. faecalis

C. freundii

PTZShock of unknown originVancomycin and metronidazole additionTo cover E. faecium and Bacteroides sp.
258GNR E. coli E. coli 0UTITemocillin additionE. coli previously found in urine sample
273GNR E. coli E. coli Levofloxacin + oxacillinMediastinitis post cardiac surgery + UTIShift to PTZHospital acquired infection, broader antimicrobial spectrum
214GNR E. coli E. coli AMCSeptic shock, angiocholitisShift to PTZSevere infection, broader antimicrobial spectrum
97GNR Campylobacter suspected Campylobacter fetus Campylobacter sp.PTZDiarrhoeaShift to clarithromycinAccording to guidelines
106GPCc CNS CNS AMCPulmonary infectionShift to PTZTreatment modification probably explained by the clinical context more than the Gram result
211GNR Klebsiella pneumoniae K. pneumoniae 0Septic shock, brain tumourPTZ additionHospital acquired infection, broad antimicrobial spectrum
42GPCpch Streptococcus pneumoniae S. pneumoniae 0Pulmonary infectionAMC additionS. pneumoniae highly suspected
55GNRSalmonella sp. S. paratyphi 0SalmonellosisLevofloxacin additionSalmonella highly suspected, according to guidelines
150GNRSalmonella sp.Salmonella sp.0Diarrhoea, fever and nauseaCeftriaxon additionSalmonella highly suspected, according to guidelines (for paediatric patient)
159GPCpch Streptococcus thermophilus S. thermophilus Ampicillin + cefotaximeBronchiolitis, feverStop ampicillinPulmonary origin suspected
246GPCpch Streptococcus pyogenes S. pyogenes TemocillinUTIShift to cefotaximPulmonary origin suspected
45GNR E. coli E. coli Ampicillin + cefotaximeUTIStop ampicillinBecause of the high ampicillin resistance in GNR
120GPCpch S. pneumoniae S. pneumoniae AMCSuspicion of ethmoiditisShift to intravenous administrationAccording to guidelines
Jules Bordet Institute
96GPCc Staphylococcus epidermidis S. epidermidis 0LymphomaVancomycin additionSuspicion of CNS infection in a patient with a permanent catheter
180GPCpch E. faecalis E. faecalis TemocillinEndometrial cancer, pyelonephritis, feverPTZ additionGenital origin suspected, broader antimicrobial spectrum
197GPCpch E. faecalis E. faecalis 0Abdominal surgeryPTZ and vancomycin additionAbdominal origin suspected, broad antimicrobial spectrum
161GPCc S. epidermidis S. epidermidis CefepimAcute leukaemia, neutropeniaVancomycin additionSuspicion of CNS infection in a patient with multiple or permanent catheters
267GPCc S. epidermidis S. epidermidis 0Medullary graft, Systematic blood culture collection for patients receiving high doses of corticosteroidsVancomycin additionSuspicion of CNS infection in a patient with multiple or permanent catheters
261GNR Stenotrophomonas maltophilia S. maltophilia + Ochrobactrum sp.0Intestinal diseasePTZ additionAbdominal origin suspected, broad antimicrobial spectrum
262GNR Enterobacter cloacae E. cloacae 0CholangiocarcinomaPTZ additionAbdominal origin suspected, broad antimicrobial spectrum
250GNR E. coli E. coli MeropenemSeptic shock <UTIShift to PTZE. coli susceptible to PTZ previously found in clinical samples
215GPCpch Streptococcus anginosus S. anginosus OxacillinEarn, Nose and Throat neoplasyShift to moxifloxacinBetter cellular penetration
184GPCc Kytococcus sedentarius Micrococcus sp.MeropenemAcute leukaemia, neutropenia, feverVancomycin additionSuspicion of CNS infection in a patient with multiple or permanent catheters
189GPCc+GPCpch S. aureus S. aureus Streptococcus sp.TZPNeutropenia, feverVancomycin additionSuspicion of CNS infection in a patient with multiple or permanent catheters
41GNR K. pneumoniae K. pneumoniae 0Ear, Nose and Throat neoplasyPTZ and amikacin additionP. aeruginosa was suspected in an infected wound
146GPCc S. epidermidis S. epidermidis 0Systematic blood culture collection for patients receiving high doses of corticosteroidsVancomycin additionSuspicion of CNS infection in a patient with multiple or permanent catheters
101GNR+GPCc E. coli

E. coli

Streptococcus mitis/oralis

CefepimAcute leukaemia, neutropeniaVancomycin additionSuspicion of CNS infection in a patient with multiple or permanent catheters
90GNR Acinetobacter genomospecies Acinetobacter lowfii AMCHepatic disease, feverPTZ+amikacinBecause of the high AMC resistance in GNR
70GNR E. coli E. coli 0Acute leukaemia, septic shockMeropenem additionHospital acquired infection in a patient in sterile unit
73GPCc Staphylococcus hominis S. hominis 0MDVancomycin additionSuspicion of CNS infection in a patient with multiple or permanent catheters

All clinical data for patients who benefited from the RMIs according to our survey (confirmation of contamination excluded, n = 29) are presented in Table 5.

Table 5. Clinical context and description of cases for whom rapid microbial identification (RMI) showed a benefit either in terms of antimicrobial treatment or in terms of general management of the patient (n = 29)Thumbnail image of

Of the adult patients, the RMIs led to a modification of the empirical treatment in 11.11% (10/90) and 14.93% (10/67) of cases according to the IDSs of Saint-Pierre and Jules Bordet, respectively (p 0.48).

For adult patients at Saint-Pierre, the modification of the empirical treatment consisted of the addition of a new drug in 80% (8/10). At Jules Bordet, the RMIs led to the addition of a new drug in only 20% of cases (2/10), the cessation of treatment in 20% of cases (2/10) and changes in the antimicrobial treatment regimen in 30% of cases (3/10); in addition, the RMIs prevented the initiation of useless treatments in 30% of cases at this institution (3/10).

In the paediatric population, the IDSs reported a modification of the empirical treatment (escalation) in only 2.50% of cases (1/40).

According to the survey, 37.5% of the RMIs (15/40) for the paediatric population allowed for the rapid confirmation of contamination. In 73.3% of those RMIs (11/15), the RMI highlighted the presence of a coagulase-negative Staphylococcus but a blood infection was excluded in all cases because blood cultures were collected at the paediatric emergency department and none of these patients had either catheters or other medical devices. The RMI confirmed contamination for only 11.11% (10/90) and 5.97% (4/67) of adult patients from Saint-Pierre and Jules Bordet, respectively (see Supplementary material, Table S2). The confirmation of contaminated samples never led to the modification of the antimicrobial treatment.

For adult patients, the IDSs reported other benefits of RMI for 4.44% and 7.46% of patients at Saint-Pierre and Jules Bordet, respectively. Such benefits included requests for new blood cultures (n = 2), removal or control of catheters (n = 4), additional medical investigations (n = 3) and resolving confusion over the samples of two patients (n = 1); as presented in Table 5.

Retrospective analysis

Compliance with IDS recommendations, according to the IDS and pharmacy data, is presented in Table 5.

For recommendations regarding the patients at Saint-Pierre (n = 15), the pharmacy data suggested that the recommended treatment adaptations were implemented in 9/10 cases before the results of the conventional methods were available. Several discrepancies were observed between the IDS and pharmacy data that may be partially explained by the lack of information in the billing, medical and nursing files.

The IDSs also highlighted an eleventh case not detected by the survey in which the RMI led to the rapid adaptation of the treatment regimen.

For recommendations regarding the patients at Jules Bordet (n = 14), with the exception of one patient, the IDS confirmed that all recommendations were respected and that all treatment modifications were implemented the same day as the RMI results became available.

The median times required to obtain an identification result from both the RMI and the conventional methods were evaluated using 178 of the 197 transmitted RMIs (19 were excluded because of missing data) and found to be 1 h 35 min (95 min; minimum–maximum: 15–596 min) and 25 h 43 min (1543 min, 1021–10499 min), respectively. This difference was statistically significant (p <0.001) (see Supplementary material, Table S3).

The delay between the transmission of the RMI to the IDS and the administration of the modified treatment was >4 h in 50% of the cases (4/8, two incomplete or missing data, Table 5).

The Fisher's exact test showed no significant difference in the impact of the RMI according to the medical unit where the samples were collected (see Supplementary material, Fig. S2).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. Transparency Declaration
  10. References
  11. Supporting Information

MALDI-TOF MS has emerged as a promising tool for the rapid identification of organisms from positive blood cultures [13-15]. Numerous strategies were evaluated to optimize RMI from positive blood cultures. A previous evaluation showed that our RMI protocol was able to correctly identify 73.7% of the blood culture bacteria at the species level in <1 h [14]. The inclusion of a higher proportion of Gram-positive bacteria that are usually less successfully identified by MALDI-TOF MS RMI techniques may explain the lower rate of successful identifications observed in the present study (61.01% vs 73.7%).

To date, the clinical impact of this type of RMI has been poorly investigated [16].

The major strength of our study is the prospective collection of data that led to similar observations for two medical institutions treating different patient populations.

Our results suggest that MALDI-TOF MS may hasten the modification of empirical treatment regimens in 13.38% of cases in the adult population (21/157). The same proportion of treatment regimens were altered (p 0.64) at both Saint-Pierre (11/90, 12.22%, including the eleventh case, which was highlighted during the retrospective analysis) and Jules Bordet (10/67, 14.93%). A recent study by Stoneking et al. [18] that retrospectively evaluated the effect of more rapid microorganism identification in bacteraemic emergency department patients suggests that the empirical therapy would remain unchanged in only 23% of cases; the remaining 21.3% and 55.7% of patients would receive treatment with an additional antibiotic for organisms not covered by the initial regimen or the adjustment of the regimen to reduce the spectrum of the antibiotics, respectively. Applying the same analysis (excluding contaminated blood cultures and missing data) to adult patients at Saint-Pierre, 78.95% of treatments would remain unchanged. In our study, the proportion of patients who were infected with pathogens not affected by the initial antibiotic regimen was in agreement with the number reported by Heenen et al. [19] in their study of de-escalation in a medico-surgical intensive care unit (16%). We first thought that the high rate of unchanged treatments in our study could be explained by the fact that two-thirds of the patients were hospitalized and received treatment that was based on previous positive samples other than blood cultures. However, the medical unit where the blood cultures were collected did not seem to affect the proportion of treatment modifications in our study. The difference may therefore be explained by the implementation of antimicrobial stewardship programmes in both the Saint-Pierre and Jules Bordet medical centres.

In the Jules Bordet patient population, most modifications of the antimicrobial treatment regimens due to the RMI involved de-escalation. In contrast, most modifications at Saint-Pierre involved treatment escalation. This difference can most likely be explained by the specific population of the Jules Bordet facility, which manages only cancer patients. For such patients, IDSs are confronted with known clinical presentations and must follow rational schemes. Broad-spectrum antibiotics are also most likely to be used in this population.

In both medical centres, the RMI of Enterococcus species and β-hemolytic streptococci was responsible for 30% of the modifications of the antimicrobial treatment, which suggests that performing RMI on positive blood cultures showing Gram-positive cocci in pairs or chains may be of interest despite the known limitations of the MALDI-TOF MS technique (no discrimination of Streptococcus mitis/oralis from Streptococcus pneumoniae).

In the oncological population, the confirmation/exclusion of non-fermenting Gram-negative rod involvement was also responsible for 30% of the observed modifications in the antimicrobial treatment, confirming the previous results of Clerc et al. [20], who highlighted the great benefit of the RMI on the clinical management of patients infected with Gram-negative bacteria.

For paediatric patients, MALDI-TOF MS was especially useful for the rapid confirmation of contaminated blood cultures (37.5%). However, these observations never led to a de-escalation of the antimicrobial treatment. As previously suggested, failure to follow the IDSs’ recommendations may be a result of physicians’ reluctance to modify treatment in patients who are improving [3].

In our study, most changes to the treatment regimen were made before the genus and/or species identification results provided by the conventional methods were available. This high level of compliance certainly results from the design of the study; telephone calls, which allow for a two-way exchange, considerably improved the communication between microbiologists and IDSs [3]. Additionally, it is well known that the implementation of multidisciplinary teams is of major importance for the optimization of antibiotic therapy in clinical settings [21]. However, the delay for the administration of a modified treatment was high (>4 h in 50% of cases), and the communication between other health professionals involved in the antimicrobial treatment administration process (clinicians, nurses, pharmacists) should therefore be improved in our institution. Major improvements of our electronic prescribing system and of the delivery of urgent antimicrobial agents are also awaited. This point will be addressed to our antimicrobial stewardship team members who are particularly skilful at improving such processes.

With a decrease of 26.85 h in the time required for identification and a 13.38% increase in the proportion of patients receiving an appropriate antimicrobial treatment 24 h after the positive blood culture, our results are in perfect agreement with those presented in the recent publication of Vlek et al. [22] (28.8 h, 11.3%). The major difference between their study and the present study is the study design: Vlek et al. compared a ‘standard care’ group with an ‘intervention’ group, whereas we opted for a blinded prospective analysis of a single patient group.

Although it may not always lead to the modification of the treatment regimen, RMI plays an important—and unfortunately difficult to quantify—role in the global management of the patient. The confirmation of a contaminant is of particular interest for non-hospitalized patients and can help to avoid the administration of unnecessary antibiotics. On the other hand, the confirmation of a catheter-related infection—usually made by ‘time-to-positivity’ determination—will allow for the rapid removal of infected devices and can help to prevent further infections. The detection of particular organisms may also allow the clinician to identify the origin of the sepsis and therefore contribute to cost-saving measures.

From a financial point of view, performing RMI requires an additional cost, primarily because of the need for additional staff and an adapted workflow. In addition, the use of conventional identification methods is still needed to resolve cases of mixed cultures or unreliable RMIs.

Because of these practical considerations and because a case-by-case selection of which RMI to perform in collaboration with the IDS would probably slow down the analytical workflow, we decided to perform the analysis only twice daily, which seems to be the most efficient option for our laboratory. The optimization of the technique, including automation, should lessen the financial impact of the technique in the future.

Our study has several limitations. First, the analysis of the pharmacy data highlighted the difficulty in identifying accurate indicators and revealed discrepancies between the billing, medical and nursing files. At our institution, improvements in computerization in the near future will most likely improve data handling [23]. Second, the observations presented in this study are only valid for similar organizations. Indeed, a higher clinical impact might be observed in institutions where the laboratory and pharmacy are open 24 h per day and an IDS is on site at all times. Both Saint-Pierre and Jules Bordet have antimicrobial stewardship programmes and numerous IDSs. These factors might explain a high rate of appropriate empirical treatment and the rapid adaptation of antimicrobial therapies because of frequent re-evaluations. Medical institutions without IDSs would most likely receive a greater benefit from RMI. The benefit of the RMI could also be different in institutions dealing with a high proportion of antimicrobial-resistant bacteria (e.g. methicillin-resistant S. aureus or extended spectrum β-lactamase producers), which is currently not the case in our hospitals.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. Transparency Declaration
  10. References
  11. Supporting Information

In an adult population, 13.38% of the MALDI-TOF MS RMIs from positive blood cultures resulted in the faster adaptation of the antimicrobial treatment regimen. The technique is also able to rapidly confirm contamination, especially in the paediatric population (37.5%), and is able to hasten the removal of infected catheters and suggest complementary diagnostic investigations. Despite the increased cost for the laboratory, RMI analyses are routinely performed twice daily in our laboratory. However, the use of RMI should not be considered unless there is efficient communication between health professionals.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. Transparency Declaration
  10. References
  11. Supporting Information

We thank the team of technicians of the Laboratoire de la Porte de Hal (Bacteriology Department) for their technical assistance and Prudence Mitangala from the School of Public Health, Université Libre de Bruxelles, for his help in the statistical analysis process. The input of Joanna Rodriguez Avelaneda and Salva Guadiola Bagan regarding this work is also greatly appreciated.

This work was partially presented at the 52nd Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC), San Francisco, USA, in September 2012 (Poster D1413).

Medical records of patients were reviewed retrospectively by linking microbiology test results with the respective medical records. Research ethics approval was granted by the Ethics Review Boards of the hospitals.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. Transparency Declaration
  10. References
  11. Supporting Information
  • 1
    Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med 2001; 29: 13031310.
  • 2
    Alberti C, Brun-Buisson C, Burchardi H et al. Epidemiology of sepsis and infection in ICU patients from an international multicentre cohort study. Intensive Care Med 2002; 28: 108121.
  • 3
    Farrington M. Therapy: the microbiologist's view. J Antimicrob Chemother 1990; 25 (suppl C): 8198.
  • 4
    Vincent JL, Sakr Y, Sprung CL et al. Sepsis in European intensive care units: results of the SOAP study. Crit Care Med 2006; 34: 344353.
  • 5
    Retamar P, Portillo MM, López-Prieto MD et al. SAEI/SAMPAC Bacteremia Group. Impact of inadequate empirical therapy on the mortality of patients with bloodstream infections: a propensity score-based analysis. Antimicrob Agents Chemother 2012; 56: 472478.
  • 6
    Lever A, Mackenzie I. Sepsis: definition, epidemiology, and diagnosis. BMJ 2007; 335: 879883.
  • 7
    Zubert S, Funk DJ, Kumar A. Antibiotics in sepsis and septic shock: like everything else in life, timing is everything. Crit Care Med 2010; 38: 12111212.
  • 8
    Paolucci M, Landini MP, Sambri V. Conventional and molecular techniques for the early diagnosis of bacteraemia. Int J Antimicrob Agents 2010; 36 (suppl 2): S6S16.
  • 9
    Bizzini A, Greub G. Matrix-assisted laser desorption ionization time-of-flight mass spectrometry, a revolution in clinical microbial identification. Clin Microbiol Infect 2010; 16: 16141619.
  • 10
    Gaibani P, Rossini G, Ambretti S et al. Blood culture systems: rapid detection—how and why? Int J Antimicrob Agents 2009; 34 (suppl 4): S13S15.
  • 11
    Emonet S, Shah HN, Cherkaoui A, Schrenzel J. Application and use of various mass spectrometry methods in clinical microbiology. Clin Microbiol Infect 2010; 16: 16041613.
  • 12
    Ferreira L, Sánchez-Juanes F, González-Avila M et al. Direct identification of urinary tract pathogens from urine samples by matrix-assisted laser desorption ionization-time of flight mass spectrometry. J Clin Microbiol 2010; 48: 21102115.
  • 13
    Croxatto A, Prod'hom G, Greub G. Applications of MALDI-TOF mass spectrometry in clinical diagnostic microbiology. FEMS Microbiol Rev 2012; 36: 380407.
  • 14
    Martiny D, Dediste A, Vandenberg O. Comparison of an in-house method and the commercial Sepsityper™ kit for bacterial identification directly from positive blood culture broths by matrix-assisted laser desorption-ionisation time-of-flight mass spectrometry. Eur J Clin Microbiol Infect Dis 2012; 31: 22692281.
  • 15
    Drancourt M. Detection of microorganisms in blood specimens using matrix-assisted laser desorption ionization time-of-flight mass spectrometry: a review. Clin Microbiol Infect 2010; 16: 16201625.
  • 16
    Emonet S, Schrenzel J. How could rapid bacterial identification improve the management of septic patients? Expert Rev Anti Infect Ther 2011; 9: 707709.
  • 17
    Hrabák J, Chudáčková E, Walková R. Matrix-assisted laser desorption ionization–time of flight (MALDI-TOF) mass spectrometry for detection of antibiotic resistance mechanisms: from research to routine diagnosis. Clin Microbiol Rev 2013; 26: 103114.
  • 18
    Stoneking LR, Patanwala AE, Winkler JP et al. Would earlier microbe identification alter antibiotic therapy in bacteremic emergency department patients? J Emerg Med 2013; 44: 18.
  • 19
    Heenen S, Jacobs F, Vincent JL. Antibiotic strategies in severe nosocomial sepsis: why do we not de-escalate more often? Crit Care Med 2012; 40: 14041409.
  • 20
    Clerc O, Prod'hom G, Vogne C, Bizzini A, Calandra T, Greub G. Impact of matrix-assisted laser desorption ionization time-of-flight mass spectrometry on the clinical management of patients with gram-negative bacteremia: a prospective observational study. Clin Infect Dis 2013; 56: 11011107.
  • 21
    Deresinski S. Principles of antibiotic therapy in severe infections: optimizing the therapeutic approach by use of laboratory and clinical data. Clin Infect Dis 2007; 3: S177S183.
  • 22
    Vlek AL, Bonten MJ, Boel CH. Direct matrix-assisted laser desorption ionization time-of-flight mass spectrometry improves appropriateness of antibiotic treatment of bacteremia. PLoS ONE 2012; 7: e32589.
  • 23
    Smith K, Smith V, Krugman M, Oman K. Evaluating the impact of computerized clinical documentation. Comput Inform Nurs 2005; 23: 132138.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. Transparency Declaration
  10. References
  11. Supporting Information
clm12282-sup-0001-FigureS1.pdfapplication/PDF5KFigure S1. Case report form.
clm12282-sup-0002-FigureS2.pdfapplication/PDF18KFigure S2. Impact of the rapid microbial identification according to the medical unit where the blood cultures were collected.
clm12282-sup-0003-TableS1.pdfapplication/PDF91KTable S1. Results of all rapid microbial identifications (RMI; n = 277)
clm12282-sup-0004-TableS2.pdfapplication/PDF8KTable S2. Rapid microbial identification (RMI) that allowed a fast confirmation of contamination (n = 29).
clm12282-sup-0005-TableS3.pdfapplication/PDF50KTable S3. Data used to determine the identification delay by both conventional techniques and rapid microbial identification (RMI; n = 178).

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.