Fatality of Staphylococcus aureus infections in a Greek university hospital: role of inappropriate empiric treatment, methicillin resistance, and toxin genes’ presence


The aim of the present study was to identify predictors of fatality among patients with S. aureus infections requiring hospitalization. Cases hospitalized with S. aureus infections at the University General Hospital of Patras, Greece, during a 4-year period (2013–2016) were studied. mecA, lukS/lukF-PV (Panton-Valentine leukocidin, PVL), tst (toxic shock syndrome toxin), fnbA (fibronectin-binding protein A), eta, and etb (epidermolytic toxins) genes’ carriage was detected by PCR in 149 selected patients. Among 464 patients, 346 were included (118 with missing data). Primary bacteremia predominated (44.2%), followed by lower respiratory tract infections (13.6%), deep seated infections (9.8%), osteoarticular (9.5%), and catheter-related bloodstream infections (6.1%). Methicillin-resistant S. aureus (MRSA) represented 33.8% of infections and were less likely to receive appropriate empiric treatment (79.5% versus 97.4%; P < 0.001). Thirty-day fatality was 14.5%. Multivariate analysis revealed that development of septic shock, Charlson Comorbidity Index, lower respiratory tract infection, bacteremia (primary or secondary), MRSA, and CRP was significantly associated with fatality. Appropriate empiric treatment was a predictor of good prognosis. Thirty-two out of 149 S. aureus (21.5%) carried lukS/lukF-PV genes, whereas, 14 (9.4%), 133 (78.7%), four (2.7%), and one (0.7%) carried tst, fnbA, eta, and etb genes, respectively. No difference was found among toxin genes’ presence and mortality. PVL was significantly more frequently found among MRSA as compared to MSSA (45.1% versus 9.2%; P < 0.001). MRSA represented one third of the infections requiring hospitalization and were independently associated with fatality, probably since were more likely to receive inappropriate antibiotic treatment as compared to MSSA.


Staphylococcus aureus is one of the most common causes of community and hospital-acquired infections associated with high morbidity and mortality [1]. Even in a setting with a high prevalence of carbapenemase-producing Gram-negative bacteria, such as Greece, it was the fourth most common cause of bacteremias [2]. Methicillin resistance among S. aureus (MRSA) differs among countries, but in a European point-prevalence study in 2011–2012, it represented 41.2% of all S. aureus hospital-acquired infections [1]. During a 3-year period in a university Greek hospital, MRSA accounted for 40.0% of S. aureus bloodstream infections [2]. In Greece, MRSA has also successfully disseminated in the community causing a large proportion of S. aureus infections [3, 4].

The success of S. aureus in disseminating and provoking infections is due to its capacity to produce a wide range of virulence factors contributing to its pathogenicity and invasiveness [4,5,6]. The most important among them is Panton-Valentine leukocidin (PVL) and staphylococcal toxic shock syndrome toxin (TSST-1). The former is commonly associated with MRSA, especially community-acquired clones, such as the ST80-IV in Greece, but can also be isolated form methicillin-susceptible isolates (MSSA) [4]. Strains carrying the tst gene (encoding the TSST-1) were more commonly found in MSSA clones such as the international ST30 clone, and have the capacity to provoke toxic shock syndrome [5, 6]. Epidermolytic toxins A and B (ETA and ETB) are responsible for the staphylococcal scalded skin syndrome of newborn and young infants, whereas, FnbA is one of the most important fibronectin-binding proteins contributing to tissue colonization in various pathological conditions mainly in indwelling medical device-related infections [7, 8].

The importance of appropriate empiric antimicrobial treatment in diminishing mortality among S. aureus infections has been demonstrated [9,10,11]. The high prevalence of MRSA in many European countries poses an additional problem regarding empiric antimicrobial treatment since in case of suspicion of S. aureus infection adding of vancomycin or newer antibiotics with anti-MRSA action, such as linezolid or daptomycin is warranted not only in hospital-acquired infections but also in community ones [3, 4].

The aim of the present study was to identify predictors of fatality among patients with S. aureus infections requiring hospitalization focusing on empiric antimicrobial treatment and pathogen factors such as methicillin resistance and toxin genes’ presence.

Materials and methods

Study design

The retrospective study was carried out during a 4-year period (2013–2016), at the University General Hospital of Patras (UGHP), a tertiary teaching hospital with 700 beds and approximately 100,000 annual admissions. The study was carried out under the Surveillance Programme for S. aureus infections and was approved by the Hospital Ethics Committee that waived the need for informed consent (279/08.06.2017).

According to the database of the Department of Microbiology of the UGHP, a total of 1954 positive cultures for S. aureus was recorded (Fig. 1). Four-hundred and sixty-four patients (464) with 522 S. aureus-positive cultures required hospitalization at UGHP for the treatment of their infection. Analysis was performed among 346 patients for whom clinical records were available.

Fig. 1

Flowchart of study patients. UGHP: University General Hospital of Patras

Patients chart reviews were used in order to collect epidemiologic data. Parameters assessed included demographic characteristics (age, sex), comorbidities, empiric antibiotic treatment (treatment initiated within 48 h after the onset of infection manifestations and the collection of the first blood culture that was detected positive), laboratory results (leucocytes, CRP), development of septic shock, length of stay, and fatality. An empiric antibiotic therapy was considered as appropriate if it included an antibacterial agent with in vitro activity against the infecting isolate and was initiated within 48 h after the onset of symptoms. Infection was characterized as community or healthcare associated according to conventional definitions [12].

S. aureus identification and antibiotic susceptibility testing

Isolates were identified as S. aureus by Gram stain, catalase, and coagulase production (slidex Staph plus test, bioMerieux S.A., Marcy l’ Etoile, France) and verified by molecular methods (PCR for 16S rRNA and nuc genes) [13]. Antibiotic susceptibility testing was performed by the disk diffusion method against anti-staphylococcal agents: penicillin, cefoxitin, gentamicin, rifampicin, fusidic acid, ciprofloxacin, tetracycline, erythromycin, clindamycin, and sulfamethoxazole-trimethoprim (co-trimoxazole). MICs of vancomycin, teicoplanin, linezolid, and daptomycin were determined by Etest (bioMerieux). All results were evaluated according to EUCAST guidelines [14]. Methicillin-resistant S. aureus (MRSA) were defined isolates resistant to cefoxitin [14]. Community-associated MRSA (CA-MRSA) were defined cases among patients without prior hospitalization or antimicrobial therapy the year before sampling, whereas, hospital-associated MRSA (HA-MRSA) those with positive cultures after 48 h of hospitalization or from individuals hospitalized, receiving antimicrobials, residences in a long-term care facility during the last year, as well as, hemodialysis patients and those with an indwelling catheter or a percutaneous device at the time the culture sample was obtained [15].

Polymerase chain reaction procedures

For the molecular studies, the first three isolates from each month were selected (149 isolates). All isolates were tested by PCR for the presence of mecA, lukS/lukF-PV (PVL), tst (TSST-1), fnbA, eta, and etb genes [16, 17].

Statistical analysis

SPSS version 23.0 (SPSS, Chicago, IL) was used for data analysis. Categorical variables were analyzed by using the Fisher exact test and continuous variables with Mann-Whitney U test. Backward stepwise multiple logistic regression analysis used all those variables from the univariate analysis with a P < 0.1. Factors contributing to multicollinearity were excluded from the multivariate analysis. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated to evaluate the strength of any association. P < 0.05 was considered statistically significant.


In total, 1954 positive cultures for S. aureus were identified (Fig. 1). Among them, only 522 represented 464 patients that were hospitalized for the treatment of S. aureus infections. Due to absence of medical records for the 118 patients, only 346 patients were included in the final analysis. The most common types of infections were primary bacteremia (153; 44.2%), lower respiratory tract infections (47; 13.6%), skin and soft tissue infections (34; 9.8%), osteoarticular (33; 9.5%), and catheter-related bloodstream infections (21; 6.1%) (Table 1).

Table 1 Univariate analysis of 30-day fatality among S. aureus infections

Thirty-day fatality was 14.5% (50 out of 364 patients). Univariate analysis of S. aureus infections’ fatality is shown in Table 1. Multivariate analysis revealed that development of septic shock (P < 0.001; OR 17.1, CI 6.1–47.8), Charlson Comorbidity Index (P 0.001; OR 1.4, CI 1.2–1.8), lower respiratory tract infection (P 0.015; OR 4.0, CI 1.3–12.1), bacteremia (primary or secondary) (P 0.007; OR 7.4, CI 1.7–31.8), MRSA (P 0.008; OR 6.6, CI 1.7–26.4), and CRP (P 0.013; OR 1.0, CI 1.0–1.1) were significantly associated with fatality, whereas appropriate empiric treatment was identified as predictor of good prognosis (P 0.017; OR 0.173, CI 0.041–0.727).

MRSA represented 33.8% of all infections (117 out of 346). Their percentage was significantly higher among healthcare-associated infections (42.2%; 62 out of 161 infections) as compared to community associated (26.5%; 49 out of 185). Among MRSA-infected patients, 49 (41.9%) were community associated (CA-MRSA). Univariate analysis of differences among MSSA and MRSA infections is depicted in Table 2. MRSA infections were more common among patients hospitalized in ICU (12.8% versus 5.7%; P 0.035). Lower respiratory tract was the site of infection most commonly associated with MRSA as compared to MSSA (20.5% versus 10.0%; P 0.012). Patients with MRSA infections were less likely to receive appropriate empiric treatment (79.5% versus 97.4%; P < 0.001) that can partially explain the trend towards higher fatality among patients with MRSA infections (19.7% versus 11.8%; P 0.054).

Table 2 Univariate analysis of differences among MSSA and MRSA infections

Molecular studies for the presence of toxin genes were conducted in 149 selected isolates, the first three recovered each month. In total, 32 (21.5%) S. aureus carried the lukS/lukF-PV gene, whereas, 14 (9.4%) the tst gene (no patient developed toxic shock syndrome). No isolate carried simultaneously both tested toxin genes. fnba, eta, and etb genes were present in 133 (78.7%), four (2.7%), and one (0.7%) isolates, respectively. No difference was found among toxin genes’ presence and fatality (Table 2). On the contrary, PVL was significantly more frequently detected among MRSA as compared to MSSA (45.1% versus 9.2%; P < 0.001) (Table 2). Carriage of other toxin genes was evenly distributed among MSSA and MRSA.


Staphylococcus aureus has been characterized as the 4th etiologic agent of BSIs (8.9%), after S. epidermidis (19.6%), Klebsiella pneumoniae (18.2%), and Escherichia coli (9.2%), in a previous study conducted in the same Hospital setting, with 40% MRSA [2]. In the present study, MRSA represented a high percentage of infections (33.8%) which was in accordance also to the rate reported by the European Centre for Disease Prevention and Control (ECDC) for the same period (39.0%) [1]. The high percentage of MRSA among community-associated infections reflects the successful dissemination of the MRSA ST80-IV clone carrying PVL, as it was detected in a 12-year multi-center study in Greece where CA-MRSA were 68.6% and ST80-IV constituted 73.3% of total MRSA infections [4]. These previous data explain why patients even those with community-associated infections received an empiric treatment that comprised of glycopeptides or newer anti-staphylococcal agents such as linezolid and daptomycin [3, 4]. Administration of such newer anti-staphylococcal antibiotics rose in Greek hospitals as it was observed in a study, with an increase of linezolid (r 0.481, P < 0.001) and daptomycin consumption (r 0.995, P < 0.001) [18]. Our finding that MRSA infection as compared to MSSA was independently associated with fatality is in accordance with other studies [10] and can partially be explained by the fact that patients with MRSA infection were less likely to receive appropriate empiric antibiotic therapy.

The timely administration of appropriate antimicrobial therapy is a well described predictor of survival among staphylococcal infections [9, 11, 19]. In total, 215 patients (62.1%, including 174 as appropriate empiric and 41 as inappropriate) received in addition to a broad-spectrum beta-lactam, an antibiotic with an action against MRSA (glycopeptides, linezolid, or daptomycin). This resulted in the appropriate empiric treatment of 79.5% of MRSA infections, which is higher than that reported in a study from Israel, (32.9%) with a higher mortality (43.9%) [9]. On the other hand, 125 of the 229 MSSA infected patients received empiric treatment by anti-MRSA antibiotics, which highlights the difficulty to stratify the risk for MRSA infections in settings where MRSA prevalence among community onset infections was high [3, 4]. Even though glycopeptides and especially vancomycin were mostly used for anti-MRSA coverage, a significant percentage of patients received as empiric antibiotic treatment linezolid and daptomycin. These agents were used in specific situations including linezolid in nosocomial pneumonia or skin and soft tissue infection, daptomycin for skin and soft tissue infection especially in renal insufficiency where glycopeptide usage might have been deleterious as compared to aforementioned options [20, 21]. Linezolid and daptomycin usage in Greek hospitals increased during the last decade [18]; this increase was motivated by the fact that a large percentage of MRSA isolates had an MIC to vancomycin > 1 mg l−1 (47 out of 117), for which the efficacy of vancomycin was known to be suboptimal [19, 22].

PVL was higher among MRSA infections, but its presence did not influence fatality which was in accordance to the literature [23], even though some studies showed a deleterious effect of PVL on outcome [24]. Contrary to the literature, in our cohort, patients with lower respiratory tract infection were more likely to be infected with PVL-positive isolates as compared to those with skin and soft tissue disease [23]. As previously shown, lower respiratory tract infection was independently associated with worst outcome [10, 11]. In our study, fatality of lower respiratory tract infection was 31.9% (15 out of 47 patients), comparable to previously published studies (39–67%) [12].

The presence of tst remained low among included infections and did not have any effect on outcome which contrasted with a previous study form Eastern China where tst carriage was independently associated with fatality [25]. The presence of tst gene is not synonym to toxic shock syndrome since none of the patients infected by such an isolate developed the aforementioned syndrome. Among 15 infections due to tst-positive strains, only one was from a patient hospitalized in the adult ICU. A previous study (2010–2012) from that ICU found a silent dissemination of strains belonging to ST30, especially MSSA, carrying the tst gene but no infection was caused then by that clone [6].

The present study has several limitations. First, this was conducted in a setting of high MRSA prevalence not only in healthcare environment but also at the community meaning that results cannot be extrapolated to settings with much lower prevalence. Our cohort represents the most severe cases, since we collected data from patients requiring hospitalization for the treatment of staphylococcal infection and patients treated in ambulatory basis were excluded due to lack of detailed medical records. Even though vancomycin’s trough levels were available in our hospital, it was not routinely measured in all patients receiving vancomycin; therefore, no such data were included in our analysis.

MRSA represented one third of the studied infections, highlighting the dissemination of MRSA clones not only in the hospital but also at the community. PVL was more frequent among MRSA strains, but its presence did not show an impact on fatality. Infections due to MRSA were independently associated with fatality, probably because were more likely to receive inappropriate antibiotic treatment as compared to MSSA.


  1. 1.

    European Centre for Disease Prevention and Control (ECDC) (2013) Point prevalence survey of healthcare associated infections and antimicrobial use in European acute care hospitals. ECDC, Stockholm

    Google Scholar 

  2. 2.

    Kolonitsiou F, Papadimitriou-Olivgeris M, Spiliopoulou A, Stamouli V, Papakostas V, Apostolopoulou E et al (2017) Trends of bloodstream infections in a University Greek Hospital during a three-year period: incidence of multidrug-resistant bacteria and seasonality in gram-negative predominance. Pol J Microbiol 66:171–180

    Article  Google Scholar 

  3. 3.

    Bouchiat C, Curtis S, Spiliopoulou I, Bes M, Cocuzza C, Codita I et al (2017) MRSA infections among patients in the emergency department: a European multicentre study. J Antimicrob Chemother 72:372–375

    CAS  Article  Google Scholar 

  4. 4.

    Drougka E, Foka A, Liakopoulos A, Doudoulakakis A, Jelastopulu E, Chini V et al (2014) A 12-year survey of methicillin-resistant Staphylococcus aureus infections in Greece: ST80-IV epidemic? Clin Microbiol Infect 20:O796–O803

    CAS  Article  Google Scholar 

  5. 5.

    Argudin MA, Argumosa V, Mendoza MC, Guerra B, Rodicio MR (2013) Population structure and exotoxin gene content of methicillin-susceptible Staphylococcus aureus from Spanish healthy carriers. Microb Pathog 54:26–33

    CAS  Article  Google Scholar 

  6. 6.

    Papadimitriou-Olivgeris M, Drougka E, Fligou F, Dodou V, Kolonitsiou F, Filos KS et al (2017) Spread of tst-positive Staphylococcus aureus strains belonging to ST30 clone among patients and healthcare workers in two intensive care units. Toxins (Basel) 9:E270

    Article  Google Scholar 

  7. 7.

    Krakauer T (2019) Staphylococcal superantigens: pyrogenic toxins induce toxic shock. Toxins (Basel) 11

  8. 8.

    Josse J, Laurent F, Diot A (2017) Staphylococcal adhesion and host cell invasion: fibronectin-binding and other mechanisms. Front Microbiol 8:2433

    Article  Google Scholar 

  9. 9.

    Paul M, Kariv G, Goldberg E, Raskin M, Shaked H, Hazzan R et al (2010) Importance of appropriate empirical antibiotic therapy for methicillin-resistant Staphylococcus aureus bacteraemia. J Antimicrob Chemother 65:2658–2665

    CAS  Article  Google Scholar 

  10. 10.

    Soriano A, Martinez JA, Mensa J, Marco F, Almela M, Moreno-Martínez A et al (2000) Pathogenic significance of methicillin resistance for patients with Staphylococcus aureus bacteremia. Clin Infect Dis 30:368–373

    CAS  Article  Google Scholar 

  11. 11.

    van Hal SJ, Jensen SO, Vaska VL, Espedido BA, Paterson DL, Gosbell IB (2012) Predictors of mortality in Staphylococcus aureus Bacteremia. Clin Microbiol Rev 25:362–386

    Article  Google Scholar 

  12. 12.

    Horan TC, Andrus M, Dudeck MA (2008) CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control 36:309–332

    Article  Google Scholar 

  13. 13.

    Zhang K, Sparling J, Chow BL, Elsayed S, Hussain Z, Church DL et al (2004) New quadriplex PCR assay for detection of methicillin and mupirocin resistance and simultaneous discrimination of Staphylococcus aureus from coagulase-negative staphylococci. J Clin Microbiol 42:4947–4955

    CAS  Article  Google Scholar 

  14. 14.

    The European Committee on Antimicrobial Susceptibility Testing (EUCAST) (2017) Breakpoint tables for interpretation of MICs and zone diameters. Version 7.1. http://www.eucast.org

  15. 15.

    Morrison MA, Hageman JC, Klevens RM (2006) Case definition for community-associated methicillin-resistant Staphylococcus aureus. J Hosp Infect 62:241

    CAS  Article  Google Scholar 

  16. 16.

    Jarraud S, Mougel C, Thioulouse J, Lina G, Meugnier H, Forey F et al (2002) Relationships between Staphylococcus aureus genetic background, virulence factors, agr groups (alleles), and human disease. Infect Immun 70:631–641

    CAS  Article  Google Scholar 

  17. 17.

    Peacock SJ, Moore CE, Justice A, Kantzanou M, Story L, Mackie K et al (2002) Virulent combinations of adhesin and toxin genes in natural populations of Staphylococcus aureus. Infect Immun 70:4987–4996

    CAS  Article  Google Scholar 

  18. 18.

    Papadimitriou-Olivgeris M, Kolonitsiou F, Zerva L, Lebessi E, Koutsia C, Drougka E et al (2015) Activity of vancomycin, linezolid, and daptomycin against staphylococci and enterococci isolated in 5 Greek hospitals during a 5-year period (2008-2012). Diagn Microbiol Infect Dis 83:386–388

    CAS  Article  Google Scholar 

  19. 19.

    Soriano A, Marco F, Martinez JA, Pisos E, Almela M, Dimova VP et al (2008) Influence of vancomycin minimum inhibitory concentration on the treatment of methicillin-resistant Staphylococcus aureus bacteremia. Clin Infect Dis 46:193–200

    CAS  Article  Google Scholar 

  20. 20.

    Wang SZ, Hu JT, Zhang C, Zhou W, Chen XF, Jiang LY et al (2014) The safety and efficacy of daptomycin versus other antibiotics for skin and soft-tissue infections: a meta-analysis of randomised controlled trials. BMJ Open 4:e004744

    Article  Google Scholar 

  21. 21.

    Jiang H, Tang RN, Wang J (2013) Linezolid versus vancomycin or teicoplanin for nosocomial pneumonia: meta-analysis of randomised controlled trials. Eur J Clin Microbiol Infect Dis 32:1121–1128

    CAS  Article  Google Scholar 

  22. 22.

    Lodise TP, Graves J, Evans A, Graffunder E, Helmecke M, Lomaestro BM et al (2008) Relationship between vancomycin MIC and failure among patients with methicillin-resistant Staphylococcus aureus bacteremia treated with vancomycin. Antimicrob Agents Chemother 52:3315–3320

    CAS  Article  Google Scholar 

  23. 23.

    Shallcross LJ, Fragaszy E, Johnson AM, Hayward AC (2013) The role of the Panton-Valentine leucocidin toxin in staphylococcal disease: a systematic review and meta-analysis. Lancet Infect Dis 13:43–54

    CAS  Article  Google Scholar 

  24. 24.

    Zhang C, Guo L, Chu X, Shen L, Guo Y, Dong H et al (2016) Presence of the Panton-Valentine leukocidin genes in methicillin-resistant Staphylococcus aureus is associated with severity and clinical outcome of hospital-acquired pneumonia in a single center study in China. PLoS One 11:e0156704

    Article  Google Scholar 

  25. 25.

    Wang M, Zheng Y, Mediavilla JR, Chen L, Kreiswirth BN, Song Y et al (2017) Hospital dissemination of tst-1-positive clonal complex 5 (CC5) methicillin-resistant Staphylococcus aureus. Front Cell Infect Microbiol 7:101

    CAS  PubMed  PubMed Central  Google Scholar 

Download references


This work received no specific grant from any funding agency and was supported by internal funding.

Author information



Corresponding author

Correspondence to Iris Spiliopoulou.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of Ethics Committee of the University General Hospital of Patras and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Katsarou, I., Paraskevopoulou, NM., Papadimitriou-Olivgeris, M. et al. Fatality of Staphylococcus aureus infections in a Greek university hospital: role of inappropriate empiric treatment, methicillin resistance, and toxin genes’ presence. Eur J Clin Microbiol Infect Dis 39, 443–450 (2020). https://doi.org/10.1007/s10096-019-03742-5

Download citation


  • Staphylococcus aureus
  • MRSA
  • Bacteremia
  • Pneumonia
  • PVL
  • Toxic shock syndrome toxin (TSST-1)
  • Fatality