Journal of Hepatology 1999; 30:249-253 Printed in Denmark • All rights reserved Munksgaard. Copenhagen Copyright © EuropeanAssociation for the Study of the Liver 1999 Journal of Hepatology ISSN 0168-8278 Staphylococcus aureus nasal carriage in 104 cirrhotic and control patients A prospective study Christian Chapoutot 1, Georges-Philippe Pageaux 1, Pierre-Francois Perrigault 2, Zouberr Joomaye 3, Pascal Perney 3, Helene Jean-Pierre 4, Olivier Jonquet 5, Pierre Blanc 1 and Dominique Larrey 1 1Department of Hepato-Gastro-Enterology, 21ntensive Care Unit B, 3Department of lnternal Medicine E, 4Department of Microbiology and 5Department of lnfectious Diseases, School of Medicine of Montpellier, France Background~Aims: Bacterial infections, specially Staphylococcus aureus (S. aureus) septicemia, remain a leading cause of death following liver transplan- tation. It has been demonstrated that nasal carriage of S. aureus is associated with invasive infections in patients undergoing hemodialysis and could be de- creased by use of antibiotic nasal ointment. However, in cirrhotic patients, the frequency of nasal carriage is unknown. The aims of this study were to determine the prevalence of S. aureus nasal carriage in cirrhotic patients and to assess nosocomial contamination. Methods: One hundred and four patients were in- cluded in a prospective study, 52 cirrhotic and 52 con- trol (hospitalized patients without cirrhosis or disease which might increase the rate of nasal carriage of S. aureus). On admission and after a few days of hospi- talization, nasal specimens from each anterior naris were obtained for culture. S. aureus was identified by the gram strain, positive catalase and coagulase reac- tions; antibiotic susceptibility was determined using a disk-diffusion test. Results: Both groups were similar with regard to age and sex. The prevalence of nasal colonization on hos- pital admission was 56% in cirrhotic patients and 13% in control patients (p=0.001). After an average of 4 days, 42% of cirrhoties and 8% of control pa- tients were colonized (p=0.001), without any noso- comial contamination. Three strains out of 29 were oxacillin-resistant in cirrhotic patients, and none in controls (p>0.05). There was no statistical difference in carriage rate according to sex, age, cause of cir- rhosis and Child-Pugh score. Previous hospitalization (OR, 6.3; 95% CI, 2.3 to 19.9; p=0.0006) and cir- rhosis (OR, 4.4; 95% CI, 1.5 to 13.4; p=0.0048) were independent predictors of colonization. Conclusion: Cirrhotic patients had a higher S. aureus nasal carriage rate than control subjects. Previous hospitalization and cirrhosis diagnosis were correlated to nasal colonization. Further studies are necessary to determine if nasal decontamination could reduce S. aureus infections after liver transplantation. Key words: Cirrhotic patients; Nasal carriage; Staphylococcus aureus. D ESPITE ADVANCES in prevention, diagnosis and treatment, bacterial infections remain a major cause of morbidity and mortality in the weeks follow- ing a liver transplantation. Infections are the second most frequent complication after acute rejection (1). Bacterial infections occur in the first 8 weeks posttrans- plantation in 30 to 55% of patients (2-4). They are the main cause of death in 14% of patients who die within Received 23 February; revised 1 September; accepted 8 September 1998 Correspondence: Georges-Philippe Pageaux, D6partement d'H6pato-Gastro-Ent6rologie, H6pital St Eloi, Avenue Bertin Sans 34295 Montpellier Cedex 5, France. Tel: +33 4 67 33 70 62. Fax: +33 4 67 52 38 97. the 2 months after surgery, or a contributory factor to death in 29% of patients (2). About 60% of infections are caused by gram-positive microorganisms, and Staphylococcus aureus (S. aureus) represents 20% of all these bacterial infections (2,4). S. aureus is a common inhabitant of the skin, and nasal carriage is the princi- pal endogenous reservoir for infection (5-8). The pos- tulated sequence which leads to auto-infection is in- itiated with S. aureus nasal carriage. The organisms are then disseminated via hand carriage to other sites of the body where infection might occur if there is a break in the dermal surfaces by surgical incision or vascular catheterization (9). Auto-infection is common in pa- tients undergoing hemodialysis, continuous ambulat- 249