Risk Factors for Mortality in Patients with Nosocomial ... - cgmh.org.tw

infection control and hospital epidemiology

december 2009, vol. 30, no. 12

original article

Risk Factors for Mortality in Patients with Nosocomial Stenotrophomonas maltophilia Pneumonia Chia-Cheng Tseng, MD;a Wen-Feng Fang, MD;a Kuo-Tung Huang, MD; Pei-Wen Chang, MSHI; Mei-Lien Tu, RRT; Yi-Ping Shiang, RPharm; I. S. Douglas, MD; Meng-Chih Lin, MD

objective. The aim of this study was to determine potential risk factors for mortality in patients with nosocomial Stenotrophomonas maltophilia pneumonia. design. setting.

A retrospective, single-center, observational study. A 2400-bed tertiary teaching hospital in southern Taiwan.

patients and methods. This retrospective study evaluated patients (age, at least 18 years) with nosocomial pneumonia (S. maltophilia isolated from respiratory culture) who were seen at Kaohsiung Chang Gung Memorial Hospital over a 3-year period. A total of 406 patients (64% male, mean age Ⳳ standard deviation, 69.6 Ⳳ 14.93 years; mean duration of hospital stay Ⳳ standard deviation, 57.5 Ⳳ 39.47 days) were included. results. Most index isolates (53.9%) were from the first sample cultured. Polymicrobial isolates were cultured from samples from 177 (43.6%) of the 406 study patients. The most common copathogen was Pseudomonas aeruginosa (53.11% of isolates). The all-cause hospital mortality rate was 42.6% (173 deaths among 406 patients). Survivors had a shorter time from admission to a positive index culture result than did nonsurvivors (26.1 vs 31.7 days; P p .04 ). Mortality was significantly higher among patients with malignancy (adjusted odds ratio [AOR], 2.48; 95% confidence interval [CI], 1.52–4.07; P ! .001 ), renal disease (AOR, 2.6; 95% CI, 1.51–4.47; P p .001 ), intensive care unit stay (AOR, 1.72; 95% CI, 1.1–2.7; P p .018 ), and inadequate initial empirical antibiotic therapy (AOR, 2.17; 95% CI, 1.4–3.38; P p .001). conclusions. S. maltophilia pneumonia is associated with a high mortality rate and is commonly associated with concomitant polymicrobial colonization or infection. Underlying comorbidities and inadequate initial empirical antibiotic therapy substantially account for increased mortality rates. Infect Control Hosp Epidemiol 2009; 30:1193-1202

Nosocomial pneumonia, also termed hospital-acquired pneumonia (HAP), is the most common and leading cause of morbidity and mortality in hospitals when caused by multidrug-resistant strains that are unresponsive to initial antibiotic treatment.1 Ventilator-associated pneumonia (VAP) refers to HAP that occurs in patients who require mechanical ventilation. Patients with HAP have a greater risk of mortality, compared with patients who receive mechanical ventilation and do not develop pneumonia.2 The increase in the prevalence of HAP due to antibiotic-resistant bacteria has resulted in more frequent administration of inappropriate antimicrobial treatment and the associated increased risk of hospital mortality and increased medical care costs.3-5 Many studies have investigated risk factors for and consequences of the development of nosocomial infection. However, because

most of these studies did not systematically account for discharge status, such as all-cause mortality, potential for significant bias exists.6 Stenotrophomonas maltophilia often presents as part of a polymicrobial culture and is usually not well established when treatment is indicated.7 The incidence of S. maltophilia infection has increased in recent years, especially in immunocompromised and clinically debilitated patients. Infection due to multidrug-resistant S. maltophilia has emerged as an important nosocomial infection in many hospitals.8-10 Risk factors, such as old age, poor physiological score, and bacteremia, have been found to increase mortality among patients with S. maltophilia infection.11-13 Most of these studies focused only on a particular patient group, such as patients with cancer or patients staying in an intensive care unit (ICU).

1 Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine (C.-C.T., W.-F.F., K.-T.H., M.-C.L.), 2Department of Respiratory Therapy (M.-L.T.), and 3Department of Pharmacy (Y.-P.S.), Chang Gung Memorial Hospital-Kaohsiung Medical Center, Chang Gung University College of Medicine, Kaohsiung, and 4Department of Respiratory Care, Pulmonary Research Center, Chang Gung Institute of Technology, Chiayi (W.-F.F., M.-C.L), Taiwan; and 5Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, University of Colorado, Denver, Colorado (I.S.D.) Received April 17, 2009; accepted July 22, 2009; electronically published October 23, 2009. 䉷 2009 by The Society for Healthcare Epidemiology of America. All rights reserved. 0899-823X/2009/3012-0010$15.00. DOI: 10.1086/648455

1194



table 1. Demographic and Clinical Characteristics of Patients with Nosocomial Stenotrophomonas maltophilia Pneumonia Patients with nosocomial S. maltophilia pneumonia Survivors (n p 233)

Characteristic a

Sex Female Male Age, mean years Ⳳ SD Duration of hospital stay, mean days Ⳳ SD Total Before pneumonia onseta After pneumonia onset ICU staya Antibiotic therapy in previous 14 days Underlying conditions Charlson Comorbidity Index,a mean score Ⳳ SD Malignancya Diabetes mellitusa Chronic respiratory diseaseb COPD Liver diseaseb Renal diseasea Bacteremia Mechanical ventilationa Tracheostomya Polymicrobial infectiona Ventilator-associated pneumoniaa Culture sitea Sputum Endotracheal aspirate Bronchoalveolar lavage fluid Adequacy of initial empirical antibiotic therapya Adequate Inadequate Indeterminate Antibiotic therapy Aminoglycosides Beta-lactam Carbapenems Cephalosporins Glycopeptides Quinolones TMP-SMX Other Status within 7 days after pneumonia onset Improved Not improved Indeterminate

Nonsurvivors (n p 173)

All (n p 406)

P

138 95 138 69.99

(59.23) (40.77) (59.23) Ⳳ 14.61

122 51 122 69.15

(70.52) (29.48) (70.52) Ⳳ 15.39

260 146 260 69.63

(64.04) (35.96) (64.04) Ⳳ 14.93

.019

59.73 26.06 33.67 130 228

Ⳳ 36.20 Ⳳ 23.01 Ⳳ 24.47 (55.79) (97.85)

54.45 31.7 22.75 120 168

Ⳳ 43.41 Ⳳ 31.71 Ⳳ 27.63 (69.36) (97.10)

57.48 28.46 29.01 250 396

Ⳳ 39.47 Ⳳ 27.17 Ⳳ 26.93 (61.58) (97.54)

.183 .038 !.001 .005 .75

2.52 44 71 61 49 26 37 46 103 42 109 74

Ⳳ 2.02 (18.88) (30.47) (26.18) (21.03) (11.16) (15.88) (19.74) (44.21) (18.03) (46.78) (31.76)

3.43 61 37 36 29 25 48 43 93 19 68 69

Ⳳ 2.18 (35.26) (21.39) (20.81) (16.76) (14.45) (27.75) (24.86) (53.76) (10.98) (39.31) (39.88)

2.91 105 108 97 78 51 85 89 196 61 177 143

Ⳳ 2.14 (25.86) (26.60) (23.89) (19.21) (12.56) (20.94) (21.92) (48.28) (15.02) (43.60) (35.22)

!.001 !.001

121 (51.93) 90 (38.63) 22 (9.44)

76 (43.93) 84 (48.55) 13 (7.51)

197 (48.52) 174 (42.86) 35 (8.62)

133 (57.08) 72 (30.90) 28 (12.02)

73 (42.20) 82 (47.40) 18 (10.40)

206 (50.74) 154 (37.93) 46 (11.33)

6 39 36 26 24 145 32 6

9 30 39 29 33 81 11 5

15 69 75 55 57 226 43 11

.575

.04 .209 .28 .322 .004 .218 .057 .05 .133 .09 .134

.003

(2.58) (16.74) (15.45) (11.16) (10.30) (62.23) (13.73) (2.58)

203 (87.12) 14 (6.01) 16 (6.87)

(5.20) (17.34) (22.54) (16.76) (19.08) (46.82) (6.36) (2.89)

50 (28.90) 108 (62.43) 15 (8.67)

(3.69) (17.00) (18.47) (13.55) (14.04) (55.67) (10.59) (2.71)

.165 .873 .069 .103 .012 .002 .017 1.99 !.001

253 (62.32) 122 (30.05) 31 (7.64)

note. Data are no. (%) of patients, unless otherwise indicated. COPD, chronic obstructive pulmonary disease; ICU, intensive care unit; SD, standard deviation; TMP-SMX, trimethoprim-sulfamethoxazole. a Variable was entered in the multivariate logistic regression analysis. b Disease was defined in accordance with the Charlson Comorbidity Index.

However, nosocomial pneumonia could occur in any patient and in any ward. Therefore, we conducted this study to review nosocomial pneumonia due to S. maltophilia in our hospital and to identify potential risk factors for mortality in these

patients. An understanding of the epidemiology, possible coinfecting pathogens, and antimicrobial susceptibility profile of nosocomial S. maltophilia pneumonia would provide insight into potential therapeutic approaches.

table 2. Demographic and Clinical Characteristics of Patients with Stenotrophomonas maltophilia Monomicrobial Infection or Polymicrobial Infection

Characteristic

Patients with monomicrobial infection

Patient with polymicrobial infection

Survivors (n p 124)


Survivors (n p 109)


50 (40.32) 74 (59.68) 69.10 Ⳳ 15.82

31 (29.52) 74 (70.48) 70.14 Ⳳ 13.77

.600

45 (41.28) 64 (58.72) 71.00 Ⳳ 13.11

20 (29.41) 48 (70.59) 67.62 Ⳳ 17.60

Ⳳ 30.78 Ⳳ 24.71 Ⳳ 19.61 (73.33) (95.24)

.056 .040 !.001 .004 1.99

a,b

Sex Female Male Age, mean years Ⳳ SDb Duration of hospital stay, mean days Ⳳ SD Total Before pneumonia onsetab After pneumonia onset ICU stay Antibiotic therapy in previous 14 days Underlying condition Charlson Comorbidity Index,a,b mean score Ⳳ SD Malignancya,b Diabetes mellitusb Chronic respiratory diseasec COPDc Liver diseasec Renal diseasea,b,c Bacteremia Mechanical ventilation Tracheostomy Ventilator-associated pneumonia Culture site Sputum Endotracheal aspirate Bronchoalveolar lavage Adequacy of initial empirical antibiotic therapya,b Adequate Inadequate Indeterminate Antibiotic treatment Aminoglycosides Beta-lactam Carbapenems Cephalosporins Glycopeptides Quinolones TMP-SMX Other Status within 7 days after pneumonia onset Improved Not improved Indeterminate

P .089

55.58 22.51 33.07 68 119

Ⳳ 32.69 Ⳳ 20.14 Ⳳ 24.19 (54.84) (95.97)

2.64 Ⳳ 2.54 25 32 28 26 14 19 24 44 16 31

(20.16) (25.81) (22.58) (20.97) (11.29) (15.32) (19.35) (35.48) (12.90) (25.00)

73 (58.87) 39 (31.45) 12 (9.68)

47.48 28.63 18.85 77 100

3.93 Ⳳ 2.75 39 20 18 13 18 28 26 59 8 44

(37.14) (19.05) (17.14) (12.38) (17.14) (26.67) (24.76) (56.19) (7.62) (41.90)

!.001

.004 .224 .306 .085 .203 .034 .324 .002 .193 .007 .014

44 (41.90) 53 (50.48) 8 (7.62)

.111

64.44 30.10 34.34 62 109

Ⳳ 39.44 Ⳳ 25.38 Ⳳ 24.87 (56.88) (100)

2.74 Ⳳ 2.28 19 39 33 23 12 18 22 59 26 43

(17.43) (35.78) (30.28) (21.10) (11.01) (16.51) (20.18) (54.13) (23.85) (39.45)

48 (44.04) 51 (46.79) 10 (9.17)

65.21 36.44 28.76 43 68

Ⳳ 56.34 Ⳳ 39.96 Ⳳ 36.08 (63.24) (100)

3.59 Ⳳ 2.79 22 17 18 16 7 20 17 34 11 25

(32.35) (25.00) (26.47) (23.53) (10.29) (29.41) (25.00) (50.00) (16.18) (36.76)

54 (51.43) 42 (40.00) 9 (8.57)

4 18 23 15 11 69 18 2

2 13 25 17 18 54 6 2

(3.23) (14.52) (18.55) (12.10) (8.87) (55.65) (14.52) (1.61)

100 (80.65) 9 (7.26) 15 (12.10)

(1.90) (12.38) (23.81) (16.19) (17.14) (51.43) (5.71) (1.90)

31 (29.52) 60 (57.14) 14 (13.33)

.693 .638 .330 .373 .061 .524 .003 1.99 !.001

.175 .916 .198 .226 .403

.029 .022 .134 .587 .705 1.99 .042 .452 .593 .222 .721 .877

32 (47.06) 31 (45.59) 5 (7.35)

.172 79 (63.71) 37 (29.84) 8 (6.45)

P

.002 54 (49.54) 35 (32.11) 20 (18.35)

19 (27.94) 40 (58.82) 9 (13.24)

2 21 13 11 13 76 14 4

7 17 14 12 15 27 5 3

(1.83) (19.27) (11.93) (10.09) (11.93) (69.72) (12.84) (3.67)

103 (94.50) 5 (4.59) 1 (0.92)

(10.29) (25.00) (20.59) (17.65) (22.06) (39.71) (7.35) (4.41)

.028 .366 .119 .146 .072 !.001 .251 1.99 !.001

19 (27.94) 48 (70.59) 1 (1.47)

note. Data are no. (%) of patients, unless otherwise indicated. COPD, chronic obstructive pulmonary disease; ICU, intensive care unit; SD, standard deviation; TMP-SMX, trimethoprim-sulfamethoxazole. a Variable was entered into the multivariate logistic regression analysis for the monomicrobial group. b Variable was entered into the multivariate logistic regression analysis for the polymicrobial group. c Disease was defined in accordance with the Charlson Comorbidity Index.

1196



table 3. Copathogens with Stenotrophomonas maltophilia

Pathogen

No. (%) of isolates (n p 177)

Acinetobacter baumannii Corynebacterium species Enterobacter species Klebsiella pneumoniae Pseudomonas aeruginosa Staphylococcus aureus Other

17 21 5 4 94 18 24

(9.60) (11.86) (2.82) (2.26) (53.11) (10.17) (13.56)

methods Setting and Study Design This was a retrospective study conducted at the Kaohsiung Chang Gung Memorial Hospital, a 2400-bed, tertiary teaching hospital in southern Taiwan. During a 3-year period (January 2005–December 2007), all hospitalized patients aged 18 years or older who had nosocomial S. maltophilia pneumonia microbiologically confirmed by sputum culture were recruited. Patients with known tuberculosis or severe immunosuppression, such as human immunodeficiency virus infection or solid organ or bone marrow transplantation, were excluded. In patients who had multiple episodes of pneumonia, only the first episode and its associated first sputum culture positive for S. maltophilia were used for the analysis. This study was approved by the Institutional Review Board of the Chang Gung Memorial Hospital, and the requirement for patient consent was waived. Definitions Pneumonia was defined in accordance with modified criteria proposed by the US Centers for Disease Control and Prevention. This definition requires 2 of the following criteria to be met: fever (increase in temperature of at least 1⬚C or body temperature greater than 38.3⬚C), leukocytosis (25% increase in leukocyte count and a leukocyte count of at least 10,000 cells/mm3) or leukopenia (25% decrease in leukocyte count and a leukocyte count of no more than 5000 cells/ mm3), and purulent tracheal secretions (more than 25 neutrophils per high-power field). The definition also requires one of the following criteria to be met: new and persistent infiltrates apparent on a chest radiograph, same microorganism isolated from pleural fluid and tracheal secretions, radiographic cavitation or histological evidence of pneumonia, or positive results of culture of bronchoalveolar lavage samples (greater than or equal to 1 # 10 4 colony-forming units/ mL). Pneumonia was considered to be hospital acquired if it occurred within 48 hours after hospital admission or if the patient was readmitted within 1 week after the last discharge from the hospital; VAP was defined as pneumonia that occurred within 48–72 hours after mechanical ventilation. Patients were classified as having an ICU stay if they were in

the ICU when S. maltophilia was isolated from a sputum sample. Polymicrobial infection was defined as S. maltophilia isolated with a mixture of other microorganisms from the same respiratory sample. The susceptibilities of S. maltophilia isolates to antimicrobial agents were determined using the disk-diffusion methods, as recommended by the National Committee for Clinical Laboratory Standards.14 Indeterminate susceptibility to antibiotics was considered to be drug resistance. To evaluate the initial empirical antibiotic response for treating underlying pneumonia, we defined an antibiotic as adequate if it was able to cover pathogens isolated from culture, with clinically improving condition in the patient assessed at 48 hours after the start of treatment. Inadequate initial antibiotic treatment was defined as inability to cover any 1 bacterium and lack of improvement of the patient’s clinical condition within 48 hours after the start of antibiotic treatment. If clinical status after initial antibiotic treatment could not be identified as adequate or inadequate, it was designated as indeterminate. After use of antibiotics to treat pneumonia for 7 days, clinical status was re-evaluated and classified as improved if fever had subsided, sputum production had decreased, and there was regression of pneumonia infiltration on chest radiograph and improvement of laboratory data; otherwise, it was classified as not improved. If the clinical status could not be identified as improved or not improved, it was designated as indeterminate. Data Collection Clinical data were retrieved from medical and microbiology records and included age, sex, underlying disease or condition, duration of hospital stay, initial empirical antibiotic treatment, culture site, susceptibilities of antimicrobial agent, and discharge status. Comorbidity was quantified using the Charlson Comorbidity Index, as described elsewhere.15 Bacteremia; stay in an ICU; invasive procedures, including those required for tracheostomy and mechanical ventilation; and any antibiotic use in the previous 14 days were considered as possible risk factors if performed or existing before onset of S. maltophilia pneumonia. The primary aim of this study was to determine the potential risk factors of in-hospital mortality. The risk factors for polymicrobial infection during hospitalization were also examined. Statistical Analysis Categorical variables were analyzed using the x2 test or Fisher’s exact test when appropriate, and continuous variables were compared using Student’s t test or the Mann-Whitney U test. Multivariate logistic regression analyses were performed to identify risk factors for mortality and polymicrobial infection. Each variable considered to be a risk factor and with a P value !.20 in the univariate analysis was entered into the multivariate model. A backward elimination procedure was used to define

clinical outcomes for s. maltophilia pneumonia

1197

table 4. Risk Factors for Polymicrobial Infection with Stenotrophomonas maltophilia Patients with S. maltophilia infection Risk factor

Monomicrobial (n p229)

Duration of hospital stay, mean days Ⳳ SD Total Before pneumonia onseta After pneumonia onset Antibiotic therapy in previous 14 daysa Resistance to ciprofloxacina Resistance to TMP-SMXa Diabetesa Chronic pulmonary diseasea Mechanical ventilationa Tracheostomya

51.86 25.31 26.55 219 53 31 52 46 103 23

Ⳳ 32.02 Ⳳ 22.51 Ⳳ 23.27 (95.6) (23.14) (13.54) (22.71) (20.09) (44.98) (10.04)

Polymicrobial (n p 177) 64.73 32.54 32.2 177 61 35 56 51 93 38

Ⳳ 46.51 Ⳳ 31.82 Ⳳ 29.71 (100) (34.46) (19.77) (31.64) (28.81) (52.54) (21.47)

P .001 .008 .032 .006 .012 .091 .043 .041 .130 .001

note. Data are no. (%) of patients, unless otherwise indicated. SD, standard deviation; TMPSMX, trimethoprim-sulfamethoxazole. a Variable was used in the multivariate logistic regression analysis.

the final independent risk factors. The Hosmer-Lemeshow test was used to assess goodness-of-fit. Cumulative mortality was estimated by the Kaplan-Meier analysis comparing patients with adequate and inadequate initial empirical antibiotic therapy. Difference in survival was calculated according to log-rank statistics. Results are presented as absolute number (percentage) or mean (Ⳳ standard deviation). Adjusted odd ratios and 95% confidence intervals are reported for logistic regression analyses. A 2-tailed P value !.05 was considered to be statistically significant. All statistical analyses were performed using SPSS, version 14.0 (SPSS).

results Patient Characteristics Of the total 4084 patients with positive respiratory culture results during hospitalization who were initially evaluated during the 3-year study period, 406 met the inclusion criteria. Characteristics of the study patients are presented in Table 1. Study patients tended to be older (59% were at least 70 years of age; mean age, 69.63 years) and male (64.0%). Patients had a mean Charlson Comorbidity Index score of 2.91. Approximately half of the patients (196 [48%]) received mechanical ventilation before isolation of S. maltophilia. The incidence of tracheostomy (15%), diabetes (27%), and microbiologically confirmed bacteremia (22%) was high among our patients. One hundred ninety-seven cultures (48.52%) were of sputum samples, 174 (42.86%) were of endotracheal aspirate samples, and 35 (8.62%) were of bronchoalveolar lavage samples. Adequate initial empirical antibiotic treatment was administered to 50.74% of the study population. Most conditions (62.32%) were evaluated as improved in clinical status after 7 days of antibiotic treatment. Quinolone was the most frequently administered initial empirical

antibiotic for the treatment of pneumonia (to 226 patients [55.67%]) (Table 1). Approximately one-third of cases (143 [35.22%]) were VAP. Patient characteristics and demographic data, by monomicrobial and polymicrobial infection, are presented in Table 2. Microbiology and Polymicrobial Infection Respiratory cultures positive for S. maltophilia were obtained on day 28 after admission. Most S. maltophilia isolates (54%) were isolated from the first positive sample. Polymicrobial infections were found in 177 (43.6%) patients. P. aeruginosa was the most frequently found copathogen (in 94 samples [53.1%]), followed by Corynebacterium species, Staphylococcus aureus, and Acinetobacter baumannii (Table 3). In our study, polymicrobial infections were not associated with hospital mortality but were more common in patients with antibiotic use in the previous 14 days, resistance to ciprofloxacin, diabetes mellitus, chronic pulmonary disease, and tracheostomy (Table 4). Furthermore, patients with polymicrobial infection had a longer hospital stay and tended to be infected for a longer period, compared with patients with monomicrobial infection. According to multivariate analysis, chronic pulmonary disease, resistance to ciprofloxacin, and tracheostomy were independent risk factors for polymicrobial infection with S. maltophilia (Table 5). Antimicrobial Susceptibility Table 6 shows the antibiotic susceptibility results for S. maltophilia isolates. The most effective antibiotics for S. maltophilia were levofloxacin and moxifloxacin. Among all isolates, the rates of resistance to flumarin, ciprofloxacin, and trimethoprim-sulfamethoxazole were 87.2%, 28.1%, and 16.3%, respectively. A slightly increasing trend in the rate of resistance to trimethoprim-sulfamethoxazole was observed.

1198



table 5. Independent Risk Factors in Multivariate Analyses Patients with Stenotrophomonas maltophilia infection All Risk factor For polymicrobial infection Chronic pulmonary disease Resistance to ciprofloxacin Tracheostomy For hospital mortality ICU stay Malignancy Diabetes Renal disease Mechanical ventilation Inadequate initial empirical antibiotic therapy note.

Monomicrobial P

1.75 (1.08–2.82) 1.61 (1.03–2.54) 2.12 (1.18–3.80)

.023 .037 .012

… … …

.018

2.03 (1.11–3.7) 3.01 (1.57–5.77)

.021 .001

2.52 (1.24–5.12) 2.37 (1.33–4.22) …

.011 .003

1.72 2.48 0.58 2.6

(1.1–2.7) (1.52–4.07) (0.35–0.96) (1.51–4.47) … 2.17 (1.4–3.38)

!0.001

.035 .001 .001

AOR (95% CI)

Polymicrobial

AOR (95% CI)

P

AOR (95% CI)

P

… … … … … … 2.83 (1.11–7.19) … 2.64 (1.36–5.11)

.029 .004

AOR, adjusted odds ratio; CI, confidence interval; ICU, intensive care unit.

Of the 406 study patients, 7 (1.7%) had multidrug-resistant S. maltophilia infection. There were no statistically significant differences in susceptibility patterns between the monomicrobial and polymicrobial groups. Because our results revealed that a significant proportion of coinfection was attributable to P. aeruginosa, we examined the antibiotic susceptibility for P. aeruginosa isolates and found that the most effective antibiotic for both S. maltophilia and P. aeruginosa was levofloxacin, followed by ciprofloxacin. Mortality The all-cause hospital mortality rate was 43% (173 deaths), and 90 patients died within 14 days after the onset of S. maltophilia pneumonia. There was no statistically significant difference in mean overall duration of hospital stay between survivors and nonsurvivors, but survivors had a shorter duration of hospital stay before pneumonia onset and a longer duration of hospital stay after pneumonia onset. Mortality was significantly higher among male patients, patients who stayed in an ICU, patients with a greater Charlson Comorbidity Index score, and patients with malignancy or renal disease. Adequacy of initial empirical antibiotic therapy and improvement of pneumonia status within 7 days after initiation of treatment resulted in a good outcome. Multivariate analysis revealed that ICU stay, malignancy, renal disease, and inadequate initial empirical antibiotic therapy were independent risk factors for hospital mortality (Table 5), and diabetes mellitus seemed to provide some survival benefit protection. Adequate initial empirical antibiotic therapy also resulted in better outcome than did inadequate antibiotic therapy during the hospital stay, as assessed by the Kaplan-Meier method (Figure 1). Mortality was slightly higher in the monomicrobial group than in the polymicrobial group (45.85% vs 38.42%; P p .133). There was no difference in the rate of VAP between the groups (P p .236); however, VAP was associated with higher mortality in the monomicrobial group but not in the

polymicrobial group. Adequate initial empirical antibiotic therapy affected patient outcome in the polymicrobial group but not in the monomicrobial group. Improved pneumonia status within 7 days after initiation of treatment resulted in better survival rates in both groups. The use of trimethoprimsulfamethoxazole in the monomicrobial group and quinolones in the polymicrobial group as initial empirical antibiotic therapy contributed to good outcomes (Table 2). Multivariate analysis revealed that ICU stay, malignancy, and renal disease were independent risk factors for hospital mortality in the monomicrobial group and that only malignancy and renal disease were mortality risk factors in the polymicrobial group.

discussion Our study has 3 main findings. First, HAP due to S. maltophilia is associated with high mortality. Second, P. aeruginosa was commonly seen in polymicrobial cultures positive for S. maltophilia. Third, mortality was significantly higher among male patients and patients who required an ICU stay, had a high Charlson Comorbidity Index score, had malignancy or renal disease, or received inadequate initial empirical antibiotic treatment. S. maltophilia is an important nosocomial pathogen that has been responsible for an increasing frequency of infection in recent years, especially in immunocompromised and clinically debilitated patients. As was shown in recent studies, severe infection caused by S. maltophilia is associated with a high mortality rate (41%–62%).16-18 S. maltophilia pneumonia–associated mortality was previously reported to be 23%– 44%12,19 and was 42.6% in the present study. Unlike most of the previous studies, our study not only included patients from both the general ward and the ICU but also did not exclude patients with polymicrobial infection. In our study, we found that polymicrobial infection was common in nosocomial pneumonia caused by S. maltophilia, and P. aeruginosa was the most frequently found copathogen. Similar to

clinical outcomes for s. maltophilia pneumonia

other study findings,20 our results showed that polymicrobial infection was not associated with hospital mortality. Strains of S. maltophilia have been reported to transfer antibiotic resistance to other bacteria, such as P. aeruginosa, Enterobacteriaceae species, and Proteus mirabilis.21-23 Therefore, the presence of S. maltophilia in polymicrobial infection deserves attention, and we suggest that antibiotics chosen for the treatment of S. maltophilia pneumonia should also cover P. aeruginosa. Some researchers may consider studies excluding polymicrobial infections to be cleaner; however, we chose to include polymicrobial infections because we encounter many such cases in clinical practice. Especially in our study, we found that inadequate initial antibiotic treatment (unable to cover any bacterium isolated from culture) may result in higher mortality in the group with polymicrobial infection than in the group with monomicrobial infection. This prompted us to consider all pathogens. As suggested in most of the previous studies, malignancy, severe septic shock, and organ failure were independent risk factors for mortality associated with S. maltophilia infection.24,25 Del Toro et al directly indicated that mortality is highly correlated with the severity of the baseline situation.26 Therefore, it was not surprising that our study found that malignancy, renal disease, and ICU stay were the independent risk factors for mortality among patients with nosocomial S. maltophilia pneumonia. However, an interesting finding in our study was that diabetes mellitus was associated with a lower mortality risk, which is in discordance with other reports.27,28 We cannot find a better explanation than chance for this finding. This result could have also been obscured by use of the diagnosis of diabetes from the chart abstract data in the analysis, which failed to accurately reflect the blood

1199

figure 1. Kaplan-Meier survival curves for patients who received adequate empirical antibiotic therapy and patients who received inadequate empirical antibiotic therapy for nosocomial Stenotrophomonas maltophilia pneumonia. P ! .001, by log-rank test.

glucose level at the time of infection. Further investigation of glycemic control status might be necessary to clarify the significance of this factor. Mechanical ventilation is used for patients with acute respiratory failure, and those patients were at high risk of infection. Thus, pneumonia occurring after intubation with

table 6. In Vitro Susceptibilities of Stenotrophomonas maltophilia Isolates Susceptibility, % (proportion of isolates)

2007

Pseudomonas aeruginosa isolates

… … … … … … (117/163) (21/163) … … (157/163) … (159/163) … …

97.8 (91/93) … 64.5 (60/93) 91.4 (85/93) 80.6 (75/93) … 92.5 (86/93) … 91.4 (85/93) 64.5 (60/93) 92.9 (13/14) 71.0 (49/69) … 71.0 (66/93) 71.0 (45/57)

S. maltophilia isolates Antimicrobial agent Amikacin Amoxicillin-clavulanic acid Aztreonam Cefepime Ceftazidime Ceftriaxone Ciprofloxacin Flumarin Gentamicin Imipenem Levofloxacin Meropenem Moxifloxacin Piperacillin Piperacillin Trimethoprimsulfamethoxazole

Monomicrobial Polymicrobial Overall

2005

2006 0 (0/18) 5.6 (1/18) … 26.3 (5/19) 33.3 (6/18) 0 (1/18) 72.7 (88/121) 14.9 (18/121) 0 (0/18) … 89.3 (92/103) … 96.1 (99/103) 5.6 (1/18) …

5.1 0.0 … 11.4 30.4 0 76.9 11.4 3.8 … 96.0 … 97.3 1.3 …

4.9 6.6 … 14.5 41.0 0 65.5 14.7 3.3 … 90.5 … 96.6 1.6 …

5.0 2.9 … 12.8 35.0 0 71.9 12.8 3.6 … 93.6 … 97.0 1.4 …

5.7 (7/122) 2.5 (3/122) … 10.7 (13/122) 35.2 (43/122) 0 (0/122) 71.3 (87/122) 10.7 (13/122) 4.1 (5/122) … … … … 0.8 (1/122) …

86.5

80.2

83.7

86.1 (105/122) 84.3 (102/121) 81.6 (133/163)

71.8 12.9

96.3 97.5

…

1200



mechanical ventilation (VAP) may have a worse outcome. We identified patients with VAP in the cohort of patients with S. maltophilia pneumonia and assessed whether this factor affected the outcome. We found that the mortality rate was higher among patients with S. maltophilia VAP than among with patients with HAP who did not receive mechanical ventilation; however, the difference did not reach statistical significance (P p .09). When patients were grouped in monomicrobial and polymicrobial groups, we found that there was a statistically significant difference in mortality between the group with VAP and the group with HAP not associated with mechanical ventilation in the monomicrobial group but not in the polymicrobial group. Tracheostomy can enable collection of a lower respiratory tract specimen for culture without contamination by normal flora of the oral cavity, but its effect on pneumonia outcomes remains controversial. Several authors have demonstrated that tracheostomy may decrease the risk of VAP,29,30 although others have reported the opposite finding.12 In our study, tracheostomy was shown to have a potential protective effect with borderline statistical significance in univariate analysis but no significance in multivariate analysis. A possible explanation for this finding could be the compounding effects of other factors. Early and adequate antibiotic therapy is important to optimize the treatment of patients with HAP.5,14,31-33 In other words, the appropriateness of initial antibiotic therapy is very important in patients with pneumonia, especially those with HAP caused by an antibiotic-resistant pathogen. Kollef et al indicated that inappropriate initial antibiotic therapy of microbiologically confirmed VAP was associated with greater 30-day mortality.5 In our study, the adequacy of initial empirical antibiotic therapy was correlated with hospital survival among all patients with S. maltophilia pneumonia, and inadequate initial empirical antibiotic therapy was the independent factor for predicting hospital mortality by multivariate analysis. We also found that better outcomes may occur in patients with S. maltophilia pneumonia when sputum specimens are cultured as early as possible. Sputum culture may not only identify the pathogen causing pneumonia but may also guide clinicians to administer the most effective antibiotic treatment. Trimethoprim-sulfamethoxazole is a therapeutic option for S. maltophilia infection and has frequently been prescribed, as suggested by several reports.34-37 However, the rate of resistance to this antibiotic has increased in recent years.38-41 Trimethoprim-sulfamethoxazole is not a favorable drug to treat infection with nosocomial pathogens other than S. maltophilia. If patients with S. maltophilia pneumonia are coinfected with other pathogens, this drug may fail to treat nosocomial pneumonia. However, trimethoprim-sulfamethoxazole may contribute to treatment benefit and result in a good survival rate for monomicrobial (S. maltophilia only) infections. Quinolone was the most reliable antibiotic in the polymicrobial group. Therefore, we suggest that trimethoprim-

sulfamethoxazole would still be beneficial in treating monomicrobial S. maltophilia pneumonia. However, when taking polymicrobial S. maltophilia pneumonia into consideration, we suggest that all pathogens should be carefully identified to exclude colonized isolates and to assess the clinical condition before antibiotic administration. As shown in our study, patients with S. maltophilia pneumonia may be coinfected with P. aeruginosa, and treatment with levofloxacin or ciprofloxacin may result in good outcomes in patients with this polymicrobial infection. The strengths of the present study are its relatively large sample of unselected patients and detailed collection of clinical data; furthermore, its retrospective design was primarily aimed at generating hypotheses for future studies. However, this study has several limitations. First, we did not investigate the precise cause of death in our patients but used hospital mortality instead. Indeed, it is difficult to differentiate deaths directly due to S. maltophilia infection from those that are the result of comorbidities or presence of other pathogens. However, improvement of pneumonia status within 7 days after the initiation of treatment may attenuate the hospital mortality rate, which suggests that S. maltophilia infection may have a large influence on hospital mortality. Second, we may have missed a certain percentage of infections detected using other methods, including protected specimen brush and bronchoalveolar lavage fluid analysis, for the diagnosis of S. maltophilia pneumonia. However, in our hospital, culture of sputum and endotracheal aspirate specimens are the most frequently used diagnostic methods for patients with suspected pulmonary infection. Third, polymicrobial infections may be mixed with colonized isolates, and we lacked direct evidence to prove that all bacteria were true pathogens, although adequate antibiotics to manage all isolated bacteria may result in good outcomes. In conclusion, we suggest that S. maltophilia HAP may result in high mortality. Underlying conditions of malignancy or renal disease, ICU stay, and inadequate initial empirical antibiotic treatment may result in poor outcomes. Polymicrobial infections are common, especially in patients with chronic pulmonary disease and tracheostomy. S. maltophilia pneumonia frequently co-occurs with P. aeruginosa, and antibiotic treatment to cover both pathogens may be needed.

acknowledgments We thank all of the staff and clinicians of the medical intensive care unit who participated in the study for their support and participation. Potential conflicts of interest. All authors report no conflicts of interest relevant to this article. Financial support. This study was supported in part by the Taiwan National Science Council (grant NSC 97-2314-B-182A-082-) and the Chang Gung Memorial Hospital (grant CMRPG870191 to W.-F.F.). Address reprint requests to Meng-Chih Lin, MD, 123 Ta-Pei Road, NiaoSung Hsiang, Kaohsiung Hsien, Taiwan 83301 ([email protected] .tw).

clinical outcomes for s. maltophilia pneumonia a

C.-C.T. and W.-F.F. contributed equally to the article.

references 1. Leroy O, d’Escrivan T, Devos P, Dubreuil L, Kipnis E, Georges H. Hospitalacquired pneumonia in critically ill patients: factors associated with episodes due to imipenem-resistant organisms. Infection 2005; 33:129–135. 2. Iregui MG, Kollef MH. Ventilator-associated pneumonia complicating the acute respiratory distress syndrome. Semin Respir Crit Care Med 2001; 22:317–326. 3. Trouillet JL, Chastre J, Vuagnat A, et al. Ventilator-associated pneumonia caused by potentially drug-resistant bacteria. Am J Resp Crit Care Med 1998; 157:531–539. 4. Valles J, Pobo A, Garcia-Esquirol O, Mariscal D, Real J, Fernandez R. Excess ICU mortality attributable to ventilator-associated pneumonia: the role of early vs late onset. Intensive Care Med 2007; 33:1363–1368. 5. Kollef KE, Schramm GE, Wills AR, Reichley RM, Micek ST, Kollef MH. Predictors of 30-day mortality and hospital costs in patients with ventilator-associated pneumonia attributed to potentially antibiotic-resistant gram-negative bacteria. Chest 2008; 134:281–287. 6. Wolkewitz M, Vonberg RP, Grundmann H, et al. Risk factors for the development of nosocomial pneumonia and mortality on intensive care units: application of competing risks models. Crit Care 2008; 12:R44. 7. Kwa AL, Low JG, Lim TP, Leow PC, Kurup A, Tam VH. Independent predictors for mortality in patients with positive Stenotrophomonas maltophilia cultures. Ann Acad Med Singapore 2008; 37:826–830. 8. Ferrara AM. Potentially multidrug-resistant non-fermentative Gram-negative pathogens causing nosocomial pneumonia. Int J Antimicrob Agents 2006; 27:183–195. 9. Spencer RC. The emergence of epidemic, multiple-antibiotic-resistant Stenotrophomonas (Xanthomonas) maltophilia and Burkholderia (Pseudomonas) cepacia. J Hosp Infect 1995;30(Suppl):453–464. 10. Tan CK, Liaw SJ, Yu CJ, Teng LJ, Hsueh PR. Extensively drug-resistant Stenotrophomonas maltophilia in a tertiary care hospital in Taiwan: microbiologic characteristics, clinical features, and outcomes. Diagn Microbiol Infect Dis 2008; 60:205–210. 11. Aisenberg G, Rolston KV, Dickey BF, Kontoyiannis DP, Raad II, Safdar A. Stenotrophomonas maltophilia pneumonia in cancer patients without traditional risk factors for infection, 1997–2004. Eur J Clin Microbiol Infect Dis 2007; 26:13–20. 12. Hanes SD, Demirkan K, Tolley E, et al. Risk factors for late-onset nosocomial pneumonia caused by Stenotrophomonas maltophilia in critically ill trauma patients. Clin Infect Dis 2002; 35:228–235. 13. Nseir S, Di Pompeo C, Brisson H, et al. Intensive care unit-acquired Stenotrophomonas maltophilia: incidence, risk factors, and outcome. Crit Care 2006; 10:R143. 14. Pines JM. Timing of antibiotics for acute, severe infections. Emerg Med Clin North Am 2008; 26:245–257, vii. 15. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40:373–383. 16. Wang WS, Liu CP, Lee CM, Huang FY. Stenotrophomonas maltophilia bacteremia in adults: four years’ experience in a medical center in northern Taiwan. Wei Mian Yu Gan Ran Za Zhi 2004; 37:359–365. 17. Morrison AJ Jr, Hoffmann KK, Wenzel RP. Associated mortality and clinical characteristics of nosocomial Pseudomonas maltophilia in a university hospital. J Clin Microbiol 1986; 24:52–55. 18. Gopalakrishnan R, Hawley HB, Czachor JS, Markert RJ, Bernstein JM. Stenotrophomonas maltophilia infection and colonization in the intensive care units of two community hospitals: a study of 143 patients. Heart Lung 1999; 28:134–141. 19. Maningo E, Watanakunakorn C. Xanthomonas maltophilia and Pseudomonas cepacia in lower respiratory tracts of patients in critical care units. J Infect 1995; 31:89–92.

1201

20. Combes A, Figliolini C, Trouillet JL, et al. Incidence and outcome of polymicrobial ventilator-associated pneumonia. Chest 2002; 121:1618–1623. 21. Blahova´ J, Kra´likova´ K, Krcme´ry V Sr, Chmelarova´ E, Torsova´ V. Two nosocomial strains of Stenotrophomonas maltophilia transferring antibiotic resistance to Proteus mirabilis P-38 recipient strain. J Chemother 1998; 10:22–24. 22. Hupkova M, Blahova J, Kralikova J, Krcmery V. Transferable resistance to beta-lactams in a nosocomial strain of Xanthomonas maltophilia. Antimicrob Agents Chemother 1995; 39:1011–1012. 23. Vartivarian S, Anaissie E, Bodey G, Sprigg H, Rolston K. A changing pattern of susceptibility of Xanthomonas maltophilia to antimicrobial agents: implications for therapy. Antimicrob Agents Chemother 1994; 38: 624–627. 24. Paez JI, Costa SF. Risk factors associated with mortality of infections caused by Stenotrophomonas maltophilia: a systematic review. J Hosp Infect 2008; 70:101–108. 25. Garcia Paez JI, Tengan FM, Barone AA, Levin AS, Costa SF. Factors associated with mortality in patients with bloodstream infection and pneumonia due to Stenotrophomonas maltophilia. Eur J Clin Microbiol Infect Dis 2008; 27:901–906. 26. del Toro MD, Rodriguez-Bano J, Martinez-Martinez L, et al. Epidemiology, clinical features and prognosis of infections due to Stenotrophomonas maltophilia. Enfermedades Infecciosas y Microbiologia Clinica 2006; 24:4–9. 27. Falguera M, Pifarre R, Martin A, Sheikh A, Moreno A. Etiology and outcome of community-acquired pneumonia in patients with diabetes mellitus. Chest 2005; 128:3233–3239. 28. Kornum JB, Thomsen RW, Riis A, Lervang HH, Schonheyder HC, Sorensen HT. Type 2 diabetes and pneumonia outcomes: a population-based cohort study. Diabetes Care 2007; 30:2251–2257. 29. Nseir S, Di Pompeo C, Jozefowicz E, et al. Relationship between tracheotomy and ventilator-associated pneumonia: a case control study. Eur Respir J 2007; 30:314–320. 30. Moller MG, Slaikeu JD, Bonelli P, Davis AT, Hoogeboom JE, Bonnell BW. Early tracheostomy versus late tracheostomy in the surgical intensive care unit. Am J Surg 2005; 189:293–296. 31. Leone M, Garcin F, Bouvenot J, et al. Ventilator-associated pneumonia: breaking the vicious circle of antibiotic overuse. Crit Care Med 2007; 35: 379–385. 32. Luna CM, Vujacich P, Niederman MS, et al. Impact of BAL data on the therapy and outcome of ventilator-associated pneumonia. Chest 1997; 111: 676–685. 33. Dembinski R, Rossaint R. Ventilator-associated pneumonia. Der Anaesthesist 2008; 57:825–842. 34. Ansari SR, Hanna H, Hachem R, Jiang Y, Rolston K, Raad I. Risk factors for infections with multidrug-resistant Stenotrophomonas maltophilia in patients with cancer. Cancer 2007; 109:2615–2622. 35. Falagas ME, Valkimadi PE, Huang YT, Matthaiou DK, Hsueh PR. Therapeutic options for Stenotrophomonas maltophilia infections beyond co-trimoxazole: a systematic review. J Antimicrob Chemother 2008; 62: 889–894. 36. Safdar A, Rolston KV. Stenotrophomonas maltophilia: changing spectrum of a serious bacterial pathogen in patients with cancer. Clin Infect Dis 2007; 45:1602–1609. 37. Yilmaz M, Celik AF, Mert A. Successfully treated nosocomial Stenotrophomonas maltophilia bacteremia following desensitization to trimethoprim-sulfamethoxazole. J Infect Chemother 2007; 13:122–123. 38. Toleman MA, Bennett PM, Bennett DM, Jones RN, Walsh TR. Global emergence of trimethoprim/sulfamethoxazole resistance in Stenotrophomonas maltophilia mediated by acquisition of sul genes. Emerg Infect Dis 2007; 13:559–565.

1202



39. Chang LL, Lin HH, Chang CY, Lu PL. Increased incidence of class 1 integrons in trimethoprim/sulfamethoxazole-resistant clinical isolates of Stenotrophomonas maltophilia. J Antimicrob Chemother 2007; 59:1038–1039. 40. Al-Jasser AM. Stenotrophomonas maltophilia resistant to trimethoprim-

sulfamethoxazole: an increasing problem. Ann Clin Microbiol Antimicrob 2006; 5:23. 41. Looney WJ, Narita M, Muhlemann K. Stenotrophomonas maltophilia: an emerging opportunist human pathogen. Lancet Infect Dis 2009; 9:312–323.