cost is c. 30 euros ⁄ test, and this can be reducedfurther if the test is performed routinely. Highplasma concentrations of b-lactams have neuro-logical toxicity [7]. The mental status of patientsreceiving high doses of b-lactams has not beeninvestigated thoroughly, and the effects of thevery high b-lactam concentrations observed in thepresent study could be insidious and remainunsuspected. For example, lethargy, asthenia,depression and anorexia are common symptomsin elderly patients treated for endocarditis and areinvestigated rarely.

The main caveat for the systematic monitor-ing of b-lactam concentrations during treatmentof endocarditis is the lack of data on theoptimal concentrations required. Although thecorrelation between very high b-lactam plasmaconcentrations and toxicity (especially neurolog-ical toxicity) is likely, it has not yet beendemonstrated formally. Likewise, even thoughanimal models suggest that trough plasmab-lactam concentrations of 4· MIC are sufficient,it is not clear whether these data can beextrapolated to humans. Although there is noreason to increase the trough plasma b-lactamconcentrations to > 10· MIC for the treatment ofendocarditis (R. Tulkens, personal communica-tion), the results of the present study suggestthat the use of current guidelines probably leadsto even higher concentrations in most patients,and particularly in the elderly. The conse-quences of these high concentrations remain tobe determined. HPLC could offer an opportun-ity to monitor plasma b-lactam concentrations,and would allow individualised rather thanstandardised treatment. Prospective, random-ised, multicentre studies are required to evalu-ate the impact of such monitoring onendocarditis therapy.

A C K N O W L E D G E M E N T S

This work was presented in part at the 7th InternationalSymposium on Modern Concepts in Endocarditis and Cardio-vascular Infections, Chamonix, France, 2003 and the 14thEuropean Congress of Clinical Microbiology and InfectiousDiseases, Prague, Czech Republic, 2004.

R E F E R E N C E S

1. Cars O. Efficacy of beta-lactam antibiotics: integration ofpharmacokinetics and pharmacodynamics. Diagn Micro-biol Infect Dis 1997; 27: 29–33.

2. Craig W. Pharmacodynamics of antimicrobial agents as abasis for determining dosage regimens. Eur J Clin MicrobiolInfect Dis 1993; 12(suppl 1): S6–S8.

3. Wilson WR, Karchmer AW, Dajani AS et al. Antibiotictreatment of adults with infective endocarditis due tostreptococci, enterococci, staphylococci, and HACEKmicroorganisms. American Heart Association. JAMA 1995;274: 1706–1713.

4. Bayer AS, Bolger AF, Taubert KA et al. Diagnosis andmanagement of infective endocarditis and its complica-tions. Circulation 1998; 98: 2936–2948.

5. Hoen B, Alla F, Selton-Suty C et al. Changing profile ofinfective endocarditis: results of a 1-year survey in France.JAMA 2002; 288: 75–81.

6. Dhawan VK. Infective endocarditis in elderly patients.Clin Infect Dis 2002; 34: 806–812.

7. Bloomer HA, Barton LJ, Maddock RK. Penicillin-inducedencephalopathy in uremic patients. JAMA 1967; 200: 121–123.

8. Durack DT, Lukes AS, Bright DK. New criteria for diag-nosis of infective endocarditis: utilization of specificechocardiographic findings. Duke Endocarditis Service.Am J Med 1994; 96: 200–209.

9. Pehourcq F, Jarry C. Determination of third-generationcephalosporins by high-performance liquid chromatogra-phy in connection with pharmacokinetic studies. J Chro-matogr Anal 1998; 812: 159–178.

10. Cockroft DW, Gault MH. Prediction of creatinine clearancefrom plasma creatinine. Nephron 1976; 16: 31–41.

RESEARCH NOTE

Severe soft tissue infections of theextremities in patients admitted to anintensive care unit

J.-R. Zahar, J. Goveia, P. Lesprit andC. Brun-Buisson

Service de Reanimation Medicale and d’Immu-nologie Clinique, Hopital Henri Mondor, Creteil,France

A B S T R A C T

This report describes a retrospective analysis of 33patients admitted to an intensive care unit withsuspicion of necrotising fasciitis (NF) of the

Corresponding author and reprint requests: J.-R. Zahar,INSERM U 570, Faculte de Medecine Necker–Enfants Malades,Unite de Pathogenie des Infections Systemiques, 156 Avenuede Vaugirard, 75730 Paris cedex 15, FranceE-mail: zahar@necker.fr

Research Note 79

� 2004 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 11, 63–82

extremities. The aim of the study was to clarifythe clinical presentation of NF in order todetermine when early surgery should be consid-ered. Twenty-one patients with surgically con-firmed NF were compared to 12 patients withsuperficial soft tissue infection. At admission,patients with NF were more likely to have skinareas of ischaemia or necrosis, fluid-filled vesicles,and severe sepsis or septic shock.

Keywords Intensive care unit, necrotising fasciitis,

sepsis, soft tissue infection

Original Submission: 13 February 2004; Revised

Submission: 21 July 2004; Accepted: 3 August 2004

Clin Microbiol Infect 2005; 11: 79–8210.1111/j.1469-0691.2004.01027.x

Severe skin or soft tissue (SST) infections caninvolve fascia planes, thereby constituting necro-tising fasciitis (NF). Such infections are character-ised by extensive necrosis and systemic toxicity[1,2]. Early clinical diagnosis of NF, and differen-tiation between NF and superficial SST infection(erysipela or cellulitis), may be difficult [3,4].However, early surgical debridement of patientswith NF has been associated with improvedsurvival when compared with delayed surgicalexploration [5,6]. The present study describes aretrospective analysis of 33 patients admitted toan intensive care unit (ICU) with severe SSTinfection of the extremities, with the aim ofclarifying the clinical presentation of NF anddetermining when patients should be referred forearly surgery.

Data collected included patient demographics,source of infection, predisposing factors, clinicalpresentation, laboratory parameters, bacteriologi-cal findings, management, duration of ICU andhospital stay, and rate of survival. The severity ofsepsis at presentation was assessed according topublished guidelines and definitions [7], and theseverity of acute illness was assessed with theSimplified Acute Physiology Score II (SAPS II) [8].NF was defined as a soft tissue infection withfascia involvement confirmed at surgery, whilesuperficial SST infection (or ‘medical’ infection)was defined as either (1) soft tissue infectionwithout fascia involvement, as confirmed bysurgical exploration, or (2) soft tissue infectionthat resolved without surgery. In order to exam-ine features associated with NF cases needing

surgical debridement, patients with medical SSTinfections were compared to patients with con-firmed NF. The data were compared by means ofthe Mann–Whitney U-test.

In total, 25 (76%) patients underwent surgicalexploration. The mean delay between diagnosisand surgery was 3.4 days (range 0–23 days); 18patients had surgery within 48 h of diagnosis. Atsurgery, NF was confirmed in 21 (84%) of thesepatients, while four patients were diagnosed withsuperficial SST infection.

Twenty-one patients were referred for primarytherapy, including 16 who failed to improve afterinitial therapy. Four patients presented withsevere sepsis and 11 with septic shock. The meanage of the patients was 59 years (range20–88 years). A medical condition predisposingto soft tissue infection was noted in 21 (64%)patients, namely diabetes mellitus (n = 11), cancer(n = 4; all receiving chemotherapy), corticosteroidtherapy (n = 8; all receiving > 1 mg ⁄ kg ⁄day pred-nisone equivalent) and liver cirrhosis (n = 4); fourpatients had received non-steroidal anti-inflam-matory drugs. A precipitating factor was recordedfor 24 (73%) patients, in that 11 patients had a pre-existing local infection and 13 patients had a recenthistory of trauma. The latter group includedminor or major limb injury (n = 8), intramuscularinjections (n = 4) or surgical amputation (n = 1).Seven cases of infection were hospital-acquiredafter liposuction surgery (n = 1), bedsore infection(n = 2), infected haematoma while receiving anti-coagulant therapy (n = 1), leg ischaemia (n = 1) orsurgical wound infection (n = 2).

Physical findings are detailed in Table 1. NFwas associated significantly with cyanosis, nec-rotic skin areas, and fluid-filled vesicles. Themean SAPS II on admission was 34.5 (range 10–96). Patients with NF presented frequently withsevere sepsis or shock (n = 15; 71%), whereasthere were no such cases among patients withmedical SST infection (p 0.0003). There was nodifference in laboratory test parameters betweenthe two groups (Table 1).

Microbiological findings are detailed inTable 2. Staphylococcus and Streptococus spp. werethe most common isolates (Table 2). No microbialpathogen was isolated from 12 patients. Microor-ganisms were recovered more frequently frompatients with NF (86% vs. 25%; p 0.001). Bloodcultures were positive for 15 patients, of whomten also had positive tissue or fluid culture

80 Clinical Microbiology and Infection, Volume 11 Number 1, January 2005

� 2004 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 11, 63–82

results. For six patients, only tissue or fluidculture results were positive. NF was associatedmore frequently with bacteraemia (57%) than wassuperficial SST infection (25%; p 0.009).

Empirical antibiotics were administered to allpatients, of whom nine received a single antibi-otic, 14 received a combination of two antibiotics,and ten received a combination of three or more

antibiotics. The antibiotics used were b-lactams(26 patients), aminoglycosides (18), third-genera-tion cephalosporins (seven), metronidazole (ten),clindamycin (two), and other miscellaneous anti-biotics (five). The overall mortality rate was 33%.The mean length of stay in the ICU was 10.6 days(range 1–49 days). There were no deaths in the‘medical’ group, whereas 52% of patients withNF died (Table 1).

Controversies persist over the classification ofsevere SST infections [9,10]. However, a prag-matic clinical–anatomical classification [3] is usedmost commonly, based on the different anatom-ical planes involved and the causative disease. Oncomparison of patients with NF confirmed atsurgery to those who progressed favourably withmedical therapy only, it was found that theformer were much more likely to have severesepsis or shock. None of the patients with medicalsevere SST infection presented with shock,despite a similar SAPS II assessment (Table 2).Therefore, the results suggest that patients pre-senting with septic shock should undergo rapidsurgical exploration, together with resuscitation.Indeed, systemic symptoms are an indication thata necrotising infection has already spread across alarge surface area and has disseminated toxins[11,12]. The present study also confirmed previ-ous observations [13–15] that patients with NFpresent more frequently with cyanosis, skinnecrosis and fluid-filled vesicles. However, thespecificity of these findings was not 100%, and afew patients with severe medical SST infectionpresented with such signs (Table 2).

Microbiological findings also differed betweenpatients with NF and those with medical SST.Bacteraemia was much more common in patientswith NF (57% vs. 25%), often in association withsevere sepsis and septic shock. Polymicrobialinfection was recorded rarely in this series ofpatients, which included only SST infection of theextremities, and was not more frequent in patientswith NF. Indeed, severe SST infection of the limbsis more often monomicrobial, with the involve-ment of skin flora [16].

NF is a rare but serious bacterial disease that isoften underestimated and poorly recognised [17].Of the 21 cases in this study, only nine weresuspected by the referring or admitting physician,and diagnosis was delayed by 2–10 days follow-ing the onset of symptoms. A high index ofsuspicion is mandatory for an early diagnosis,

Table 1. Clinical characteristics of 33 patients with necro-tising fasciitis and severe skin or soft tissue infection

Variables All patients

Necrotisingfasciitis

n = 21

SuperficialSST infection

n = 12 p

Age, years (mean) 58.8 ± 16.9 61.5 ± 15.4 54.6 ± 19.1 0.4Temperature (mean) 38.1 ± 1.2 38.2 ± 1.3 37.8 ± 1.3 0.42SAPS II (mean) 36.5 ± 22 36.9 ± 21 31.2 ± 16 0.5DIC (no. of patients) 6 5 1 0.3Clinical presentation

Crepitus 9 7 2 0.5Necrosis 18 16 2 0.003Ischaemia 11 10 1 0.1Cyanosis 14 12 2 0.02Fluid-filled vesicles 16 14 2 0.01Pain 27 17 10 0.6Erythema 26 17 9 0.6

Immunodepression 21 14 7 0.4Biological findings

Leukocyte count 12.5 ± 7.9 12.9 ± 7.7 11.9 ± 8.5 0.7Fibrin 6.4 ± 3 5.5 ± 2.7 6.7 ± 3.1 0.3CPK 2008 ± 5241 2819 ± 6243 304.5 ± 439.6 0.2C-reactive protein 252 ± 147 291 ± 101 240 ± 161 0.6Platelets 226 ± 160 203 ± 154 238 ± 165 0.6

Severe sepsis or shockat admission(no. of patients)

15 15 0 < 0.001

Mechanical ventilation 18 16 2 0.003Renal replacementtherapy

7 7 0 0.03

Vasopressive drugs 12 11 1 0.02Adequate antibiotictherapy within 48 h

30 19 11 1

Survivors 11 12 0.005

CPK, creatine phosphokinase; DIC, disseminated intravascular coagulation; SAPSII, Simplified Acute Physiology Score II; SST, skin or soft tissue.

Table 2. Microbiological data for 21 cases of soft tissueinfection

Pathogens

Medical SST

infection NF

Total no. of

isolates

Microbiological dataOrganism(s) isolated 3 18Bacteraemia 3 12Monomicrobial ⁄polymicrobial 3 ⁄ 0 15 ⁄ 3Gram-negative bacilli 0 8Gram-positive cocci 3 12

Gram-negative bacilli 8Escherichia coli 2 2Klebsiella pneumoniae 2 2Citrobacter freundii 1 1Pseudomonas aeruginosa 1 1Proteus mirabilis 1 1Morganellla morganii 1 1

Gram-positive cocci 15Staphylococcus aureus 2 7 9Streptococcus pyogenes 0 3 3Streptococcus spp. 1 2 3

NF, necrotising fasciitis; SST, skin or soft tissue.

Research Note 81

� 2004 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 11, 63–82

since early and adequate surgical debridementand fasciotomy has been associated with im-proved survival compared to delayed surgicalintervention [13,18].

The present study had several limitations,including its retrospective design and the smallnumber of patients; however, it was possible toidentify four clinical factors and one microbiolo-gical factor that were associated significantly withNF in patients admitted to an ICU with severeSST infection. The presence of these signs shouldprompt early surgical exploration.

R E F E R E N C E S

1. Weinbren MJ, Perinpanayagam RM. Streptococcal necro-tising fasciitis. J Infect 1992; 25: 299–302.

2. Donaldson PM, Naylor B, Lowe JW, Gouldesbrough DR.Rapidly fatal necrotising fasciitis caused by Streptococcuspyogenes. J Clin Pathol 1993; 46: 617–620.

3. Bisno AL, Stevens DL. Streptococcal infections of skin andsoft tissues. N Engl J Med 1996; 334: 240–245.

4. Barker FG, Leppard BJ, Seal DV. Streptococcal necrotisingfasciitis: comparison between histological and clinicalfeatures. J Clin Pathol 1987; 40: 335–341.

5. McHenry CR, Piotrowski JJ, Petrinic D, Malangoni MA.Determinants of mortality for necrotizing soft-tissueinfections. Ann Surg 1995; 221: 558–563.

6. Majeski JA, Alexander JW. Early diagnosis, nutritionalsupport, and immediate extensive debridement improvesurvival in necrotizing fasciitis. Am J Surg 1983; 145: 784–787.

7. Anonymous. ACCP-SCCM Consensus Conference defini-tions for sepsis and organ failure and guidelines for the

use of innovative therapies in sepsis. Chest 1992; 101: 1644–1655.

8. Le Gall JR, Lemeshow S, Saulnier F. A new simplifiedacute physiology score (SAPS II) based on Euro-pean ⁄North American multicentre study. JAMA 1993; 270:2957–2963.

9. Simmons RL, Ahrenholz DH. Infections of the skin andsoft tissues. In: Howard RJ, Simmons RL, eds. Surgicalinfectious diseases. Norwalk: Appleton & Lange, 1988; 404–408.

10. Umbert IJ, Winkelmann RK, Oliver GL et al. Necrotizingfasciitis: a clinical, microbiologic, and histopathologicstudy of 14 patients. J Am Acad Dermatol 1989; 20: 774–781.

11. Stevens DL. Invasive group A streptococcal infections. ClinInfect Dis 1992; 14: 2–13.

12. Stevens DL. Streptococcal toxic shock syndrome: spectrumof diseases, pathogenesis, and new concept in treatment.Emerg Infect Dis 1995; 1: 69–78.

13. Elliott DC, Kufera JA, Myers RA. Necrotizing soft tissueinfections. Risk factors for mortality and strategies formanagement. Ann Surg 1996; 224: 672–683.

14. Ward RG, Walsh MD. Necrotizing fasciitis: 10 years’experience in a district general hospital. Br J Surg 1991; 78:488–489.

15. Callahan TE, Schecter WP, Horn JK. Necrotizing soft tissueinfection masquerading as cutaneous abscess followingillicit drug injection. Arch Surg 1998; 133: 812–817.

16. Green RJ, Dafoe DC, Raffin TA. Necrotizing fasciitis. Chest1996; 110: 219–229.

17. Rangaswamy M. Necrotizing fasciitis: a 10-year retro-spective study of cases in a single university hospital inOman. Acta Trop 2001; 80: 169–175.

18. McHenry CR, Piotrowski JJ, Petrinic D et al. Determinantsof mortality for necrotizing soft-tissue infections. Ann Surg1995; 221: 558–565.

82 Clinical Microbiology and Infection, Volume 11 Number 1, January 2005

� 2004 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 11, 63–82