Journal of Microbiology, Immunology and Infection (2021) 54, 632e638

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.e-jmii.com

Original Article

Clinical characteristics and relevance of

coagulase-negative Staphylococci other than
S. epidermidis by positive blood culture
Nobuhiro Asai a,b, Daisuke Sakanashi b, Hiroyuki Suematsu b,
Hideo Kato b, Mao Hagihara b, Hiroki Watanabe a,b,
Arufumi Shiota b, Yusuke Koizumi a,b, Yuka Yamagishi a,b,
Hiroshige Mikamo a,b,*

a
Department of Clinical Infectious Diseases, Aichi Medical University Hospital, Aichi, Japan
b
Department of Infection Control and Prevention, Aichi Medical University Hospital, Japan

Received 10 December 2019; received in revised form 21 February 2020; accepted 3 March 2020
Available online 11 March 2020

KEYWORDS Abstract Introduction: Staphylococci is the most commonly isolated bacteria in blood cul-
Coagulase-negative tures (BC), and is mostly regarded as contamination. However, the clinical relevance and char-
staphylococci; acteristics of CNS other than Staphylococcus epidermidis (Non-SE-CNS) is still unknown.
Contamination; Methods: For the purpose of clarifying the epidemiology and clinical significance of Non-Se-
Bacteremia; CNS infections, we retrospectively evaluated BC isolates of Non-Se-CNS at our institute from
Blood culture; May 2013 to March 2017.
Time to positivity Infections were defined as true bacteremia if (1) two or more positive BCs of the same spe-
cies were present (or detection of the same species in another clinically relevant sample), (2)
the patient had clinical symptoms as outlined in the Centers for Disease Control and Preven-
tion’s definition of primary blood stream infection8 and the symptoms or markers of inflamma-
tion (e.g., fever, symptoms, white cell counts, C-reactive protein, or procalcitonin) were
improved after pathogen-directed therapy (e.g. either antibiotic therapy and/or removal of
the foreign body in the case of catheter-related blood stream infections suspected) and (3)
without any other infection.
Results: During the period, 279 patients with BC positive for Non SeeCNS were identified. 44
patients were excluded, either due to missing data or multiple pathogens in the same BC.
Among the remaining 235 patients, 67 (29%) were categorized as true bacteremia (n Z 43)
or possible infections (n Z 24). Resistance rate to methicillin among Non-SE-CNS is about
60%, which is lower than those among S. epidermidis of 78%.
Conclusion: About 30% of Non-Se-CNS isolates were clinically relevant in our study. Antibiotic
susceptibility of Non-SE-CNS is quite different than those of S. epidermidis.

* Corresponding author. Department of Clinical Infectious Diseases, Aichi Medical University School of Medicine, 1-1 Yazakokarimata,
Nagakute, Aichi, 480-1195, Japan.
E-mail address: mikamo@aichi-med-u.ac.jp (H. Mikamo).

https://doi.org/10.1016/j.jmii.2020.03.001
1684-1182/Copyright ª 2020, Taiwan Society of Microbiology. Published by Elsevier Taiwan LLC. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Non-SE CNS by positive blood culture 633

Copyright ª 2020, Taiwan Society of Microbiology. Published by Elsevier Taiwan LLC. This is an
open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).

Introduction Microbiology

Coagulase-negative staphylococci (CNS) inhabit the skin and Blood culture sets were obtained according to a standard
mucous membranes, and have been commonly isolated in protocol. Blood cultures were processed in the microbi-
blood cultures. CNS are also the most common cause of foreign ology laboratory using BACTEC system (BD, Tokyo, Japan).
body infections such as those with central venous catheter A blood culture set consists of an anaerobic and an aerobic
(CVC), pace makers, prosthetic joints, and others.1e4 Most bottle. TTP was calculated according to laboratory stan-
episodes of positive CNS blood culture were considered as dards. Bacterial species were identified by matrix-assisted
contamination, previously accounting for 80e85%. With the laser desorption ionization-time-of-flight mass spectrom-
emergence of biologics, chemotherapy agents in the treat- etry (MALDI-TOF MS) (Bruker Daltonik, Bremen, Germany).
ment of rheumatic disease, and malignancy, clinical charac- Antimicrobial susceptibility testing was performed for the
teristics and relevance of positive CNS blood culture could strain using the broth microdilution method (Dry Plate R,
possibly vary from the results of previous reports. Eiken Chemical co., Ltd, Tokyo, Japan) according to the
Staphylococcus epidermidis is the most commonly iso- Clinical and Laboratory Standards Institute guidelines.7 As
lated in blood culture, and is mostly considered as for Methicillin-resistance, MIC interpretive criteria (mg/ml)7
contamination. Among 41 CNS species including S. epi- was used in Non-SE CNS except for Staphylococcus lugdu-
dermidis, only a few have been associated with human in- nensis by oxacillin (S0.5), and in S. lugdunensis by oxacillin
fections.5 While CNS is regarded as one group to distinguish (S8) and/or cefoxitin (S4).
them from Staphylococcus aureus, CNS in fact comprises a
heterogeneous group, ranging from true nonpathogenic to
facultative pathogenic species with low, medium or even Definition as infection and contamination
high virulence potential. Species distribution also depends
on the examined clinical specimen, since different staph- Infections were defined as true bacteremia if (1) two or
ylococci species show niche preferences, e.g., S. capitis is more positive BCs of the same species were present (or
found on the human head, S. saprophyticus is mainly iso- detection of the same species in another clinically relevant
lated in bladder urine (cystitis in young females), S. cohnii sample), (2) the patient had clinical symptoms as outlined
is isolated from human feet, and Staphylococcus auricularis in the Centers for Disease Control and Prevention’s defini-
is isolated from the external auditory meatus.2,4 S. lugdu- tion of primary blood stream infection,8 and the symptoms
nensis shows a virulence and is an important cause of in- or markers of inflammation (e.g., fever, symptoms, white
fectious endocarditis.5,6 CNS other than S. epidermidis cell counts, C-reactive protein, or procalcitonin) were
(Non-SE) is commonly isolated, however, the clinical rele- improved after pathogen-directed therapy (e.g. either
vance and characteristics of Non-SE are still unknown. We antibiotic therapy and/or removal of the foreign body in
conducted a retrospective study evaluating the clinical the case of catheter-related blood stream infections sus-
characteristics and relevance of positive CNS blood culture pected) and (3) without any other infection. Infection was
at Aichi Medical University hospital in Japan. This is the first defined as possible if a peripheral venous catheter (PV) or
study on Non-SE-CNS bacteremia since 2010 worldwide, as central venous catheter (CV) was present with one set of
far as we could search. positive blood cultures and the above (2) and (3) were
fulfilled. Several studies could demonstrate the relevance
of a single positive BC in Non-SE-CNS infection.4,9,10 All the
Patients and methods other cases were considered contaminations.
Isolates (species and antibiotics susceptibility), patients’
Study design and patients characteristics (age, sex, comorbidity, and antibiotics
used), time to positivity (TTP) in BC and outcomes were
For the purpose of clarifying the epidemiology and clinical evaluated. Comorbidity was evaluated by the Charlson
significance of coagulase-negative staphylococci other than Comorbidity Index (CCI).11
S. epidermidis (Non-Se-CNS) infection, we retrospectively
evaluated blood culture (BC) isolates of Non-Se-CNS at Aichi
Medical University Hospital in Japan from May 2013 to Definition of variables
March 2017. Our institute is a 900-bed tertiary care center
located in the countryside at Aichi prefecture in central Disseminated intravascular coagulation (DIC) was diagnosed
Japan. Patients under 16 of age were excluded in this study. according to the disseminated intravascular coagulation
This study was approved by the Institutional Review Board diagnostic criteria established by the Japanese Association
of Aichi Medical University Hospital. for Acute Medicine (JAAM DIC diagnostic criteria).12
634 N. Asai et al.

Antibiotic treatment was classified as appropriate or as years (range 20e93 years). Forty-three patients (64%) were
inappropriate when the identified pathogens were sensitive males. For infection sites, catheter-related blood stream
and resistant respectively to the initially prescribed infection was the most commonly seen in 35 patients (52%),
antibiotics. followed by unknown in 23 (34%), and others in 9 (14%).
Others include 3 infectious endocarditis, 1 necrotizing
fasciitis, 1 pyogenic arthritis, 1 lead infection of implanted
Results pacemaker and 3 deep tissue abscesses.
In terms of patients’ conditions, shock and DIC were
During the period, 279 patients with BC positive for Non-Se- confirmed in 8 (12%) and 5 (7%), respectively. As for the
CNS were identified. Forty-four patients were excluded, underlying diseases, the most common underlying disease
either due to missing data or multiple pathogens in the was malignancies in 27 (40%), followed by cerebrovascular
same BC. Of the remaining 235 patients, 67 (29%) were diseases (CVD) in 19 (28%). Mean CCI was 3.1  2.4
categorized as true bacteremia (n Z 43) or possible in- [mean  standard deviation (SD)].
fections (n Z 24). The patients’ characteristics and clinical As for the clinical outcomes, 56 (84%) survived and 11
outcomes were shown in Table 1. The mean age was 77 (16%) died. The methicillin susceptibility test of BC isolates
revealed resistance in 130 of 279 (47%). Anti-methicillin-
resistant S. aureus (MRSA) antibiotics were used empirically
Table 1 Patients’ characteristics and clinical outcomes. in 55 patients (82%).
Variables (n Z 67) Number (%) Fig. 1 shows the proportions of positive blood cultures
for Non-SE CNS. The most important point here is that
Mean age (years, mean  SD) 73.8  14.7 Staphylococcus capitis, caprae and hominis were mostly
Median age (years, range) 77 (20e93) contaminations, however, about 60% of Staphylococcus
Male sex 43 (64) haemolyticus and 80% of S. lugdunensis were responsible
External device for bacteremia.
Central venous catheter 17 (25)
Central venous port 7 (10)
Peripheral venous catheter 37 (55)
Infection site Staphylococcal species isolated by blood culture
CRBSI 35 (52) and the antibiotic susceptibility
Othersa 9 (14)
Unknown 23 (34) A total of 301 strains were isolated from 279 patients as
Underlying diseases shown in Table 2. The most common isolates were Staph-
Diabetes mellitus 11 (16) ylococcus hominis in 114 (38%), followed by S. capitis in 72
Chronic kidney disease 11 (16) (24%) and Staphylococcus caprae in 47 (16%) patients.
Hemodialysis 4 (6) Antibiotic susceptibility rates for these strains are shown
Cardiac diseases 14 (21) in Table 2. S. epidermidis isolated by blood culture during
Pulmonary diseases 5 (7) the study period are set as a control group. Susceptibility
Cerebrovascular disease 19 (28) rates of MPIPC, gentamycin, clindamycin, levofloxacin and
Collagen diseases 6 (9) sulfamethoxazole-trimethoprim among Non-SE-CNS are
Malignancy 27 (40) higher than those among S. epidermidis. As for species,
Charlson comorbidity index (mean  SD) 3.1  2.4 resistance rates to methicillin of S. haemolyticus and of S.
Methicillin resistantb 130/279 (47) lugdunensis are much higher than those of other species
Patients’ conditions (88%, v. v. 39%, p < 0.001) as shown in Table 3.
Shock 8 (12)
Disseminated intravascular coagulation 5 (7)
Performance status (ECOG)
0-1 12 (18) Relationship between TTP and bacteremia
2 8 (12)
3 18 (27) TTP of BC for true or possible infections was much shorter
4 29 (43) than that for contaminations (mean TTPSD 29.3  15.7 vs.
Outcome 46.2  24.4 h, p < 0.001). With respect to the diagnostic
Survival 56 (84) value of TTP for bacteremia or contamination, the area
Death 11 (16) under the receiver-operating characteristic curve (AUROC)
Treatment for TTP was 0.735 (p < 0.001, 95% confidential interval
Anti-MRSA drug use 55 (82) 0.662e0.808) as shown in Fig. 2. The TTP cutoff value was
a
Resistance to methicillin was examined in all species 36 h and had a sensitivity of 65%, a specificity of 76%, a
isolated. positive predictive value of 87%, and a negative predictive
b
Others were 3 infectious endocarditis, 1 necrotizing fascii- value of 47%. The cut-off was set based on the Youden
tis, 1 pyogenic arthritis, 1 lead infection of implanted pace- Index (Table 4).
maker and 3 deep tissue abscesses.
Twenty-eight of 29 (97%) Non-Se-CNS isolates that were
ECOG: Eastern Cooperative Oncology Group; MRSA: methicillin-
considered contaminations grew only in the anaerobic BC
resistant Staphylococcus aureus; SD: standard deviation.
bottle.
Non-SE CNS by positive blood culture 635

Figure 1. Shows proportions of positive blood cultures in each staphylococcal species.

Discussion addition, our results clarified the clinical relevance of

non-SE CNS blood culture. These could help physicians
CNS possess a sufficient armamentarium of factors and make a reasonable judgment for patients with a positive
strategies to act as opportunistic pathogens, which are blood culture of Non-SE CNS as to whether it is a true
determined by specific host conditions as well as by spe- bacteremia, and thus whether anti-MRSA agents are
cific species. The most common pathogen isolated by necessary or not. It could contribute to avoiding an un-
blood culture is CNS in 28.1%, followed by Escherichia coli necessary empirical therapy and reduce an additional
in 13.4%, S. aureus in 12.8%, and Candida species in 10.1% medical cost.
in our institute during the study period. This result seems We hypothesized that time to positivity (TTP) in BC
similar to the one worldwide.13,14 Progress in current might be a diagnostic tool for bacteremia as previous
medicine relies on the use of indwelling catheters or other studies had reported.17 As shown in Fig. 2, TTP could be a
foreign bodies and is often associated with an opportu- useful diagnostic tool for bacteremia. Combination of using
nistic Non-SE-CNS infections. As mentioned above, CNS both MALDI-TOF-MS and TTP, most cases with Non-SE CNS
shows a variety of pathogenesis from nonpathogenic to positive blood culture could easily be determined as con-
medially or severely pathogenic. CNS-associated infection taminations or as true bacteremia. As for pattern of posi-
could cause a skin infection as well as disseminated in- tive BC bottle, 97% non-SE-CNS isolates that were
fections like infective endocarditics or vertebritis. Thus, considered contaminations grew in the anaerobic BC bottle
physicians are likely to empirically prescribe anti-MRSA only. This tendency is similar to the result of previous
drugs when patients had a set of positive blood culture studies and might be one of the characteristics in
of CNS. Empiric use of anti-MRSA drugs could lead to be contaminations.4
socioeconomic loss as well as dysbiosis in patients. Clos- Depending on Non-SE Staphylococcal species, pro-
tridioides difficile infection could occur in immunocom- portions of true bacteremia varies ranging about 20e80% as
promised hosts with imbalance of microbiota, resulting in shown in Fig. 1. Isolations of S. capitis, S. caprae and S
an expansion of social medical costs. Availability of hominis in BC were mostly contaminations, while about
medical equipment like MALDI-TOF-MS15 and Verigene,16 60e80% of S. lugdunensis and S. haemolyticus were in-
physicians can rapidly identify a pathogen by blood cul- fections. Previous report demonstrated that bacteremia by
ture such as nonpathogenic or a virulent Non-SE CNS. In S. haemolyticus was associated with serious clinical
636 N. Asai et al.

Table 2 Resistance to methicillin of non-S. epidermidis coagulase negative staphylococci isolated.

Species Number of isolates which are resistant Resistant rate to methicillin (%)
to methicillin/total species (n Z 301)
Staphylococcus capitis 39/72 54
Staphylococcus caprae 30/47 64
Staphylococcus cohnii 1/4 25
Staphylococcus haemolyiticuss 27/30 90
Staphylococcus hominis 59/114 52
Staphylococcus intermedius 2/4 50
Staphylococcus lugdunensis 9/11 89
Staphylococcus saccharolyticus 3/3 100
Staphylococcus simulans 8/8 100
S. warneri 0/5 9
Othersa 1/3 33
Total species of 180/301 60
resistance to methicillin
a
S. saprophyticus, S. xylosus and S. sciuri are seen as one for each.

Table 3 Comparison of antibiotic susceptibility between non-SE CNS and S. epidermidis.

Antibiotics susceptibility (%) Non-SE-CNS (n Z 301) S. epidermidis (n Z 293) p-value
Oxacillin 43 (128/301) 22 (65/293) <0.001
Gentamycin 63 (191/301) 47 (137/293) <0.001
Clindamycin 86 (260/301) 77 (226/293) 0.004
Levofloxacin 48 (144/301) 28 (81/293) <0.001
Minocycline 93 (281/301) 94 (276/293) 0.736
Vancomycin 100 (300/300) 100 (292/293) 0.494
Teicoplanin 96 (290/301) 98 (288/293) 0.205
Linezolid 99 (298/301) 100 (293/293) 0.249
Sulfamethoxazole-Trimethoprim 93 (281/301) 85 (248/293) <0.001
Non-SE CNS, non-S. epidermidis coagulase negative Staphylococci.

complications.9 However, there was no difference of CCI lugdunensis is higher than those among other species. The
score among each species (data not shown). mechanism of biofilm accumulation and maturation and
For resistance against methicillin, 60% of total isolates factors involved in biofilm detachment and dissemination
were resistant to methicillin. During the study period, the have been clarified.5,9,20e22 There is evidence that S. epi-
most common CNS species isolated by BC was S. epi- dermidis biofilm formation interferes with phagocytic up-
dermidis, accounting for 49% (293/594) in all isolates. The take and with pro-inflammatory activation of macro-
resistance rate to methicillin of S. epidermidis was 78%, phages.23 This effect was irrespective of the intercellular
which was higher than those of non-SE-CNS (Table 3). As for adhesin used.24 Although there were very little studies
a resistance to methicillin among non-SE-CNS species, the about the difference between CNS and S. lugdunensis, it
resistance rates of both S. haemolyticus and S. lugdunensis has been known that the mechanisms of biofilm formation
were seen in 90 and 89%, respectively in our study (Table are quite different from those of Staphylococcus epi-
2). Due to these results, empirical anti-MRSA agents dermids.25 Because clinical S. lugdunensis strains form PIA-
should be considered for patients with bacteremia of these independent and extracellular DNA-dependent biofilms, the
pathogens. Previous studies revealed that antimicrobial S. lugdunensis nuclease activity identified may play a role
susceptibility of S. lugdunensis is lower than the one of in biofilm detachment.5,25 These differences might
other CNS.18 However, in the recent report published in contribute to a high clinical relevance of S. lugdunensis in
Japan,19 methicillin resistance was observed in 55% of all positive BC as above mentioned. Additionally, S. lugdu-
isolates in children. Of these, the resistance rate was nensis could cause a severe infection such as infective
confirmed to be 62% of hospital acquired infections. In our endocarditis.
study, methicillin resistance rate of S. lugdunensis was 34% Twenty-four of the 67 patients (36%) exhibited bacter-
of all isolates (data not shown), and was 75% of the in- emia were diagnosed as having bacteremia by one positive
fections (75% of all the infection cases). BC alone. The rate was lower than we expected, since
Antimicrobial resistance rate of S. lugdunensis may be previous studies had shown higher rates (38e55%).4,15e17 It
increasing and it is necessary to continue the surveillance could be due to a strict criterion as bacteremia of this
of CNS isolates. The rate of true bacteremia in S. study. As for patients’ general condition, 70% of the
Non-SE CNS by positive blood culture 637

Figure 2. Shows ROC curves of TTP as a diagnostic value for bacteremia and possible infections among patients who have positive
BCs.

In conclusion, about 30% of Non SeeCNS isolates were

Table 4 Relation between time to positivity in blood clinically relevant in our study. Only the growth of Non-Se-
culture and diagnostic accuracy. CNS in anaerobic BC bottle and a TTP S36 h were associ-
Sensitivity (%) Specificity (%) PPV (%) NPV (%) YI ated with contamination.
S24 h 80 59 87 46 39
S36 h 65 76 87 47 41
S48 h 91 36 31 93 27
Acknowledgment
PPV, positive predictive value; NPV, negative predictive value;
We are grateful for the diligent and thorough critical
YI, youden index.
reading of our manuscript by Dr. Yoshihiro Ohkuni, Chief
Physician, Taiyo and Mr. John Wocher, Executive Vice
President and Director, International Affairs/International
patients had poor PSs of 3e4. Patients with poor condition Patient Services, Kameda Medical Center (Japan).
should be considered as having Non-SE-CNS infections.
There are some limitations in our study. First, this is a
retrospective study in a small population. The classification References
of bacteremia or contamination was based on electric re-
cords, which might have been not entirely comprehensive 1. von Eiff C, Peters G, Heilmann C. Pathogenesis of infections
in some cases. Second, we set a strict criterion of true due to coagulase-negative staphylococci. Lancet Infect Dis
infection in this study as mentioned above. Also, only cases 2002;2:677e85.
who were admitted were analyzed and those who were not 2. Becker K, Heilmann C, Peters G. Coagulase-negative staphy-
admitted were excluded. It is possible that some cases lococci. Clin Microbiol Rev 2014;27:870e926.
were underdiagnosed as contamination. Third, the methi- 3. Weinstein MP, Towns ML, Quartey SM, Mirrett S, Reimer LG,
cillin resistance rate of S. lugdunensis was confirmed by Parmigiani G, et al. The clinical significance of positive blood
cultures in the 1990s: a prospective comprehensive evaluation
cefoxitin/oxacillin disc according to CLSI guideline. How-
of the microbiology, epidemiology, and outcome of bacteremia
ever, the SCCmecA gene was not checked in all isolates. and fungemia in adults. Clin Infect Dis 1997;24:584e602.
This might influence the resistance rate. Fourth, a number 4. Hitzenbichler F, Simon M, Salzberger B, Hanses F. Clinical sig-
of several Non-SE- CNS species were so small that analyses nificance of coagulase-negative staphylococci other than S.
of those were not large enough. More cases should be epidermidis blood stream isolates at a tertiary care hospital.
collected to be analyzed. Infection 2017;45:179e86.
638 N. Asai et al.

5. Heilmann C, Ziebuhr W, Becker K. Are coagulase-negative conventional medical microbiology laboratories. J Clin Micro-
staphylococci virulent? Clin Microbiol Infect 2019;25: biol 2010;48:900e7.
1071e80. 16. Hayakawa K, Mezaki K, Kobayakawa M, Yamamoto K, Mutoh Y,
6. Non LR, Santos CA. The occurrence of infective endocarditis Tsuboi M, et al. Impact of rapid identification of positive blood
with Staphylococcus lugdunensis bacteremia: a retrospective cultures using the Verigene system on antibiotic prescriptions:
cohort study and systematic review. J Infect 2017;74:179e86. a prospective study of community-onset bacteremia in a ter-
7. Clinical and Laboratory Standards Institute. Performance tiary hospital in Japan. PloS One 2017;12:e0181548.
standards for antimicrobial susceptibility testing; twenty-fifth 17. Herwaldt LA, Geiss M, Kao C, Pfaller MA. The positive predic-
informational supplement. M100-S25. Wayne, PA: Clinical and tive value of isolating coagulase-negative staphylococci from
Laboratory Standards Institute; 2015. blood cultures. Clin Infect Dis 1996;22:14e20.
8. Garner JS, Jarvis WR, Emori TG, Horan TC, Hughes JM. CDC 18. Higaki S, Kitagawa T, Morohashi M, Yamagishi T. Distribution
definitions for nosocomial infections, 1988. Am J Infect Contr and antimicrobial susceptibility of coagulase-negative staphy-
1988;16:128e40. lococci from skin lesions. J Int Med Res 1999;27:191e5.
9. Ruhe J, Menon A, Mushatt D, Dejace P, Hasbun R. Non- epi- 19. Sakurai H, Horikoshi Y. Clinical and microbiological charac-
dermidis coagulase-negative staphylococcal bacteremia: clin- teristics of community-acquired and hospital-acquired in-
ical predictors of true bacteremia. Eur J Clin Microbiol Infect fections with Staphylococcus lugdunensis in children.
Dis 2004;23:495e8. Syonikansenmeneki 2019;1:21e5. in Japanese.
10. Kim SD, McDonald LC, Jarvis WR, McAllister SK, Jerris R, 20. Rice KC, Mann EE, Endres JL, Weiss EC, Cassat JE, Smeltzer MS,
Carson LA, et al. Determining the significance of coagulase- et al. The cidA murein hydrolase regulator contributes to DNA
negative staphylococci isolated from blood cultures at a com- release and biofilm development in Staphylococcus aureus.
munity hospital: a role for species and strain identification. Proc Natl Acad Sci U S A 2007;104:8113e8.
Infect Control Hosp Epidemiol 2000;21:213e7. 21. Izano EA, Amarante MA, Kher WB, Kaplan JB. Differential roles
11. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of poly-N-acetylglucosamine surface polysaccharide and
of classifying prognostic comorbidity in longitudinal studies: extracellular DNA in Staphylococcus aureus and Staphylococcus
development and validation. J Chron Dis 1987;40:373e83. epidermidis biofilms. Appl Environ Microbiol 2008;74:470e6.
12. Gando S, Iba T, Eguchi Y, Ohtomo Y, Okamoto K, Koseki K, et al. 22. Qin Z, Ou Y, Yang L, Zhu Y, Tolker-Nielsen T, Molin S, et al. Role
A multicenter, prospective validation of disseminated intra- of autolys in-mediated DNA release in biofilm formation of
vascular coagulation diagnostic criteria for critically ill pa- Staphylococcus epidermidis. Microbiology 2007;153:2083e92.
tients: comparing current criteria. Crit Care Med 2006;34: 23. Büttner H, Mack D, Rohde H. Structural basis of Staphylococcus
625e31. epidermidis biofilm formation: mechanisms and molecular in-
13. Wisplinghoff H, Bischoff T, Tallent SM, Seifert H, Wenzel RP, teractions. Front Cell Infect Microbiol 2015;5:14.
Edmond MB. Nosocomial bloodstream infections in US hospi- 24. Schommer NN, Christner M, Hentschke M, Ruckdeschel K,
tals: analysis of 24,179 cases from a prospective nationwide Aepfelbacher M, Rohde H. Staphylococcus epidermidis uses
surveillance study. Clin Infect Dis 2004;39:309e17. distinct mechanisms of biofilm formation to interfere with
14. Lamagni TL, Evans BG, Shigematsu M, Johnson EM. Emerging phagocytosis and activation of mouse macrophage-like cells
trends in the epidemiology of invasive mycoses in England and 774A.1. Infect Immun 2011;79:2267e76.
Wales (1990-9). Epidemiol Infect 2001;126:397e414. 25. Rajendran NB, Eikmeier J, Becker K, Hussain M, Peters G,
15. van Veen SQ, Claas EC, Kuijper EJ. High-throughput identifi- Heilmann C. Important contribution of the novel locus comEB
cation of bacteria and yeast by matrix-assisted laser desorp- to extracellular DNA-dependent Staphylococcus lugdunensis
tion ionization-time of flight mass spectrometry in biofilm formation. Infect Immun 2015;83:4682e92.