Non-fermenting Gram-negative bacilli (NF-GNB) colonisation and infection are frequent among critically ill patients 1 2 3 . Antibiotic resistance is common in NF-GNB, resulting in inappropriate initial antibiotic treatment and poor outcomes 4 5 6 . Several studies have investigated incidence, risk factors, and outcome of patients with Pseudomonas aeruginosa and Acinetobacter baumannii colonisation and infection. However, few studies have investigated patients with colonisation and/or infection related to Stenotrophomonas maltophilia.

Isolation rates of S. maltophilia have been increasing since the early 1970s, according to reports from several centres 7 . This pathogen primarily affects patients with co-morbid illness such as cystic fibrosis, immunosuppression, organ transplantation, and malignancies 8 . Infections related to S. maltophilia are associated with high morbidity and mortality rates. Therapy for these infections represents a significant challenge both for the clinician and the microbiologist because of this organism's high level of antibiotic resistance to most of the currently used agents and methodological difficulties in susceptibility testing with this organism 9 10 .

In recent years, several trials have elucidated risk factors for S. maltophilia infection, which include neutropenia, presence of a central venous catheter, prolonged hospitalisation, and previous antibiotic treatment with broad-spectrum antibiotics 11 12 13 14 15 16 . However, only one study was performed in intensive care unit (ICU) patients with ventilator-associated pneumonia (VAP) related to S. maltophilia 16 . In addition, no study has evaluated risk factors for and impact on outcome of ICU-acquired S. maltophilia. Identifying risk factors for S. maltophilia colonisation and/or infection could be useful for future interventional studies aiming at preventing ICU-acquired S. maltophilia. Therefore, we conducted this study to determine incidence of, risk factors for, and outcome of ICU-acquired S. maltophilia.

This prospective case-control study, which was a part of a cohort study, was conducted in a 30-bed medical and surgical ICU from January 1998 to January 2001. Because the study was observational, Institutional Review Board approval was not required; this was in accordance with Institutional Review Board regulation.

All patients who were without severe immunosuppression and who were hospitalised more than 48 hours in the ICU were eligible. Patients with colonisation and/or infection related to NF-GNB at ICU admission and patients without screening for NF-GNB colonisation at ICU admission and more than 48 hours after ICU admission were excluded. Patients without ICU-acquired S. maltophilia who developed an ICU-acquired colonisation and/or infection related to NF-GNB other than S. maltophilia were also excluded. However, patients with ICU-acquired colonisation and/or infection related to S. maltophilia and ICU-acquired colonisation and/or infection related to NF-GNB other than S. maltophilia were studied as cases.

Infection control policy included continuous surveillance of nosocomial infections, isolation techniques, and routine screening of multi-drug-resistant (MDR) bacteria. At ICU admission, isolation techniques were used in all patients until receipt of screening results. Thereafter, these techniques were performed in all patients with colonisation or infection related to MDR bacteria. Isolation techniques included use of protective gowns and gloves and adequate hand-washing with antiseptic soap between patient contacts. No selective digestive decontamination was performed. Antibiotic treatment was based on local antibiotic guidelines, including specific recommendations for each type of infection.

Routine screening of NF-GNB was performed in all patients at ICU admission and on a weekly basis (every Monday) thereafter. This screening included nasal, anal, and axilla swabs. In addition, tracheal aspirate was performed in intubated or tracheotomised patients. Hemocultures, quantitative tracheal aspirates, broncho-alveolar lavage, urine cultures, and other microbiologic cultures were performed according to clinical status.

Infection and colonisation were considered to be ICU-acquired if they were diagnosed more than 48 hours after ICU admission. VAP was defined by the presence of new or progressive radiographic infiltrate associated with two of the following criteria: temperature above 38.5°C or below 36.5°C, leucocyte count above 10,000/μl or below 1,500/μl, purulent tracheal aspirate, and positive (≥10⁶ colony-forming units per ml) tracheal aspirate. Ventilator-associated tracheobronchitis was defined using all of the following criteria: fever (>38°C) with no other recognisable cause, new or increased sputum production, positive (≥10⁶ colony-forming units per ml) endotracheal aspirate culture yielding a new bacteria, and no radiographic evidence of new pneumonia 17 . Other definitions of nosocomial infections were based on criteria of the CDC (Centers for Disease Control and Prevention) 18 . Colonisation was defined as a positive microbiologic culture without clinical signs of infection. Prior antibiotic treatment was defined as any antibiotic treatment during the two weeks preceding ICU admission. Severe immunosuppression was defined by the presence of neutropenia (leucocyte count <1,000/μl or neutrophils <500/μl), active solid or hematology malignancy, long-term corticosteroid therapy (≥1 mg/kg per day for >1 month), or HIV infection (CD4 <50/μl during the previous 6 months). Cases were patients with ICU-acquired S. maltophilia colonisation and/or infection. Controls were patients without NF-GNB colonisation and/or infection. Antibiotics used during the study period included glycopeptides (vancomycin and teicoplanin), extended-spectrum penicillins (amoxicillin/clavulanate, ticarcillin, ticarcillin/clavulanate, piperacillin, piperacillin/tazobactam, and imipenem), fluoroquinolones (ofloxacin and ciprofloxacin), extended-spectrum cephalosporins (cefotaxime, ceftriaxone, ceftazidime, and cefepime), aminoglycosides (gentamicin, tobramycin, and amikacin), and others (erythromycin, metronidazole, fusidic acid, rifampicin, and fosfomycin). Antimicrobial therapy was deemed inappropriate when none of the antibiotics used was active in vitro on S. maltophilia. Outcomes evaluated included ICU mortality, duration of mechanical ventilation, and duration of ICU stay.

Our results suggest that S. maltophilia colonisation and/or infection is not common in this population of immunocompetent ICU patients. COPD and duration of antibiotic treatment are independently associated with ICU-acquired S. maltophilia. ICU-acquired S. maltophilia is associated with high ICU mortality and morbidity rates. In addition, ICU-acquired infection related to S. maltophilia is an independent risk factor for ICU mortality.

In the SENTRY Antimicrobial Surveillance Program, S. maltophilia represented 0.6% to 0.9% of all isolates collected during a three year period 7 . In that study, the respiratory tract was the most commonly reported S. maltophilia site of infection in all geographic regions. Incidence of ICU-acquired S. maltophilia found by our study (2%) is higher than previously reported. This is probably related to the fact that screening of patients with S. maltophilia was not performed in previous studies. Our study differs from previous studies in its methodology. In previous studies, patients with infection related to S. maltophilia were compared with patients without S. maltophilia infection. However, in our study immunocompetent patients with ICU-acquired S. maltophilia were compared with immunocompetent patients without NF-GNB colonisation and/or infection. Risk factors for S. maltophilia and other NF-GNB are probably similar. Therefore, exclusion of patients with colonisation and/or infection related to NF-GNB other than S. maltophilia probably allowed a more accurate evaluation of risk factors for ICU-acquired S. maltophilia.

COPD was identified as an independent risk factor for ICU-acquired S. maltophilia. Previous studies identified COPD as a risk factor for VAP and for respiratory tract colonisation by GNB 20 21 . Factors predisposing to respiratory tract colonisation include impairment of mucosal clearance and loss of mucosal integrity 22 . A multicentre study evaluated the relationship between bacterial flora in sputum and functional impairment in patients with acute exacerbation of COPD hospitalised in pneumology units 23 . P. aeruginosa was isolated more frequently in patients with low FEV₁ (forced expiratory volume in one second). Unfortunately, information on severity of COPD was not available in our study. Another potential explanation for the association between COPD and ICU-acquired S. maltophilia is the frequent antibiotic use in patients with COPD. Patients with severe COPD are prone to frequent acute exacerbations requiring antimicrobial therapy 24 . In our study, rate of prior antibiotic use was higher in patients with COPD than in patients without COPD. However, this difference did not reach statistical significance.

Prolonged duration of antibiotic treatment is a well-known risk factor for emergence of MDR bacteria 25 26 . Shortening duration of antibiotic treatment could be useful in preventing ICU-acquired S. maltophilia. Two recent randomised trials demonstrated that shorter duration of antibiotic treatment is effective and safe in ICU patients with VAP 26 27 . In addition, serial measurements of clinical pulmonary infection score can define the clinical course of VAP resolution 28 . Based on this score, identifying patients with good outcome as early as day three could be of help to define strategies to shorten the duration of antibiotic treatment.

Recent reports documented an increasing incidence of NF-GNB, including S. maltophilia, in patients with cystic fibrosis 29 30 . In our study, cystic fibrosis was not significantly associated with S. maltophilia acquisition. However, few patients with cystic fibrosis were included in this study. Marchac et al. 31 reported that prevalence of S. maltophilia, in respiratory tract of patients with cystic fibrosis, increased from 3.3% to 15% during the last decade. Antibiotic treatment, isolation of Aspergillus fumigatus in sputum, and oral corticosteroid use were significantly associated with S. maltophilia 31 . Another recent study found no association between lung function decline and S. maltophilia acquisition in patients with cystic fibrosis, suggesting that S. maltophilia is rather a marker of worse lung function in these patients 32 .

Carbapenem use was not significantly associated with ICU-acquired S. maltophilia. However, previous studies identified carbapenem use as an independent risk factor for S. maltophilia acquisition 33 34 . These different results could be explained by the small number of patients treated with carbapenem in our study. However, other studies found no relationship between carbapenem use and S. maltophilia acquisition, whereas agents other than carbapenems were identified as risk factors for the acquisition of S. maltophilia 35 36 . These findings suggest that exposure to broad-spectrum antibiotics might be more important than exposure to any single agent. Immunosuppression is another well-known risk factor for S. maltophilia acquisition 14 37 . Although patients with severe immunosuppression were excluded from our study, ICU patients may be at risk for immunoparalysis. A temporary immunoparalysis may occur in critically ill patients, resulting in higher risk for nosocomial infections 38 .

Although risk factors for colonisation and infection related to S. maltophilia were similar, outcomes of patients with colonisation or infection related to S. maltophilia were clearly different. A recent meta-analysis demonstrated that risk factors for colonisation and infection related to MDR bacteria were similar 39 .

In our study, S. maltophilia was associated with significantly increased ICU mortality and longer duration of mechanical ventilation and ICU stay. However, S. maltophilia was previously considered to have limited pathogenicity. In a recent retrospective study, S. matophilia was isolated in the respiratory tract specimens of 64 patients without pneumonia 11 . No significant difference was found in mortality rate between treated and untreated patients. However, few patients (29%) were mechanically ventilated. In another retrospective study, patients with S. maltophilia VAP were compared with patients with late-onset VAP related to other GNB 16 . Although mortality rate was similar in patients with S. maltophilia VAP and patients with late-onset VAP related to other GNB, S. maltophilia was associated with increased patient morbidity. Recent studies indicated that infection with this organism was associated with significant morbidity and mortality, particularly in severely compromised patients 12 13 15 . A retrospective case study, including 69 patients with infection or colonisation related to multi-resistant S. maltophilia, has evaluated risk factors for mortality 40 . McCabe score and organ dysfunction were associated with increased risk for mortality in these patients. These results suggest that infection related to S. maltophilia has an indirect impact on mortality. In contrast, our results suggest a direct impact of S. maltophilia infection on ICU mortality given that this infection was found to be an independent risk factor for ICU mortality. In cases with ICU-acquired infection related to S. maltophilia, inappropriate initial antibiotic treatment was associated with increased ICU mortality rate. Inappropriate initial antibiotic treatment is a well-known risk factor for mortality in patients with infection related to NF-GNB 5 6 .

Our study has several limitations. First, it was performed in a single ICU. Therefore, our results may not be applicable to other ICU patients. Similarly, only immunocompetent patients were included. This precludes application of our results to patients with immunosuppression. Second, during the study period, no molecular typing was performed on S. maltophilia isolates. Therefore, the impact of patient-to-patient transmission could not be determined. Previous studies using molecular typing in newborn and pediatric ICUs have demonstrated that cross-transmission of S. maltophilia could be identified using molecular typing 41 42 . However, no S. maltophilia outbreak occurred during the study period. Finally, the number of patients with ICU-acquired S. maltophilia was small. Therefore, some of the trends observed in this study could have reached statistical significance if the study sample had been larger.

S. maltophilia colonisation and/or infection is not frequent in this cohort of immunocompetent ICU patients. COPD and duration of antibiotic treatment are independently associated with ICU-acquired S. maltophilia. ICU-acquired S. maltophilia is associated with high mortality and morbidity rates. In addition, ICU-acquired infection related to S. maltophilia is an independent risk factor for ICU mortality.

• ICU-acquired S. maltophilia is not common among this cohort of immunocompetent ICU patients.

• COPD and duration of antibiotic treatment are independently associated with ICU-acquired S. maltophilia and with ICU-acquired infection related to S. maltophilia.

• ICU-acquired S. maltophilia is associated with high mortality and morbidity rates. In addition, ICU-acquired infection related to S. maltophilia is an independent risk factor for ICU mortality.

CI = confidence interval; COPD = chronic obstructive pulmonary disease; ICU = intensive care unit; MDR = multi-drug-resistant; NF-GNB = non-fermenting Gram-negative bacilli; OR = odds ratio; SAPS = Simplified Acute Physiology Score; VAP = ventilator-associated pneumonia.

The authors declare that they have no competing interests.

SN and AD designed the study. All authors contributed to data collection. CDP performed statistical analyses. SN wrote the manuscript, all authors participated in its critical revision. SN had full access to all data in the study and had final response ability for the decision to submit for publication. All authors read and approved the final manuscript.