The impact of COPD on ICU mortality in patients with ventilator-associated pneumonia

The impact of COPD on ICU mortality in patients with ventilator-associated pneumonia

Respiratory Medicine (2011) 105, 1022e1029 available at www.sciencedirect.com journal homepage: www.elsevier.com/locate/rmed The impact of COPD on ...

230KB Sizes 0 Downloads 30 Views

Respiratory Medicine (2011) 105, 1022e1029

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/rmed

The impact of COPD on ICU mortality in patients with ventilator-associated pneumonia Demosthenes Makris a,b,*, Benoit Desrousseaux a, Epaminondas Zakynthinos b, Alain Durocher a, Saad Nseir a a b

Intensive Care Unit, Calmette Hospital, University Hospital of Lille, boulevard du Pr Leclercq, 59037 Lille cedex, France Intensive Care Unit, Critical Care Department, University Hospital of Thessaly, Biopolis, Larisa 41110, Greece

Received 9 November 2010; accepted 1 March 2011 Available online 24 March 2011

KEYWORDS Pneumonia; Mechanical ventilation; Chronic obstructive pulmonary disease; Mortality; Prognosis; Intensive care

Summary Objective: To determine the impact of COPD on intensive care unit (ICU) mortality in patients with VAP. Methods: This prospective observational study was performed in a mixed ICU during a 3-year period. Eligible patients received mechanical ventilation for >48 h and met criteria for microbiologically confirmed VAP. Risk factors for ICU mortality were determined using univariate and multivariable analyses. Results: Two hundred and fifteen patients with microbiologically confirmed VAP were included. Most VAP episodes were late-onset (88%), and Pseudomonas aeruginosa was the most frequently isolated bacterium (39% of VAP episodes). ICU mortality was significantly lower in non-COPD patients (n Z 150) compared to COPD patients (n Z 65) (43.3% vs 60%, p Z 0.027, OR [95% CI] Z 1.96 [1.8e3.54]). Duration (days) of mechanical ventilation and ICU stay median (IQR) in non-COPD patients were 25 (15e42) and 30 (18e48), whereas in COPD patients were 31 (19e45) and 36 (20e48) (p > 0.05). The differences in duration (days) of mechanical ventilation and ICU stay were significant between non-COPD patients and severe COPD (GOLD stage IV) patients (p Z 0.001 and p Z 0.02, respectively). Multivariable analysis identified COPD [OR (95% CI) 2.58 (1.337e5)], SAPS II [1.024 (1.006e1.024)] and presence of shock at VAP diagnosis [3.72 (1.88e7.39)] as independent risk factors for ICU mortality. Conclusion: COPD, SAPS II, and shock at VAP diagnosis are independently associated with ICU mortality in patients who present VAP. ª 2011 Elsevier Ltd. All rights reserved.

Abbreviations: CI, confidence interval; COPD, chronic obstructive pulmonary disease; CPIS, clinical pulmonary infection score; ICU, intensive care unit; LOD, logistic organ dysfunction; MDR, multidrug resistant; MRSA, methicillin-resistant Staphylococcus aureus; OR, odds ratio; SAPS, simplified acute physiology score; VAP, ventilator-associated pneumonia. * Corresponding author. Intensive Care Unit, Critical Care Department, University Hospital of Thessaly, Biopolis, Larisa 41110, Greece. Tel.: þ302413502960; fax: þ302413501280. E-mail address: [email protected] (D. Makris). 0954-6111/$ - see front matter ª 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.rmed.2011.03.001

Mortality in COPD patients with VAP

Introduction Ventilator-associated pneumonia (VAP) is the second most frequent infection in patients hospitalized in the intensive care unit (ICU).1 VAP is associated with prolonged duration of mechanical ventilation and ICU stay and may increase the need of human resources and the burden of cost in ICU.2 In addition, VAP is considered as an important determinant of outcome for critically ill patients.3 Previous studies have shown that patients who experience VAP have a greater risk of death compared to patients without the disease although this was challenged by other investigators.4e6 Prior studies aimed to identify risk factors that may affect VAP prognosis.7e9 The presence of multidrug resistant (MDR) bacteria, the severity of pneumonia itself and the application of inadequate initial antibiotic therapy, have all been identified as determinants of adverse outcome of patients with VAP. Furthermore, it has been suggested that the underlying disease leading to ICU admission might also affect the course of VAP.10 However, the association between specific underlying diseases and VAP outcome is not yet clear. COPD is a known risk factor for VAP occurrence, and is related to risk factors for adverse VAP outcomes such as MDR bacteria.11 Previous studies systematically investigating the association between VAP and COPD are few and COPD populations in those studies were small whereas COPD severity was not known.10,12,13 We have previously investigated whether COPD patients with VAP present increased mortality compared to COPD patients without VAP. In the present prospective observational investigation, we followed a large cohort of critically ill patients who presented VAP and we aimed to evaluate the impact of COPD on ICU mortality in this population. Our hypothesis was that COPD would be associated with increased risk for ICU mortality in VAP patients.

Patients and methods This study was conducted in a 30-bed ICU at Albert Calmette University Hospital at Lille, France, from November 2006 to June 2009. The local Institutional Review Board approved the study, and did not require an informed consent because of the non-interventional design of the study. This decision was in accordance with the French law. Eligible patients had received mechanical ventilation for >48 h and met criteria for microbiologically confirmed VAP. Patients who were intubated after non-invasive-ventilation (NIV) failure were included in the study. Patients who received only NIV were not eligible. Patients were prospectively screened on a daily basis for inclusion in the study.

Study population The oropharyngeal cavity was washed four times a day with chlorhexidine solution. Continuous subglottic suctioning was not utilized. The ventilator circuit was not changed routinely. In all patients a heatemoisture exchanger was positioned between the Y piece and the patient, the

1023 heatemoisture exchangers were changed every 48 h or more frequently if visibly soiled. Patients were kept in semirecumbent position during most of their period of mechanical ventilation. Tracheal cuff pressure was maintained around 25 cm H2O. There was no systematic stress ulcer prophylaxis and no selective digestive decontamination. Infection control policy included isolation techniques in patients with MDR bacteria, written antibiotic treatment protocol, and continuous surveillance of nosocomial infections. Sedation and ventilator weaning procedures were standard throughout the whole study period. NIV was used in all COPD patients with severe acute exacerbation if there was no contraindication.

Definitions COPD was defined according to recent ATS/ERS criteria14 and COPD severity was assessed by the Global Initiative for COPD (GOLD) criteria.15 Definition of VAP included the presence of new or progressive radiographic infiltrate associated with two of the three following criteria: 1) temperature >38.5  C or <36.5  C, 2) leukocyte count >10 000/mL or <1500/mL, 3) purulent tracheal aspirate. In addition, a positive tracheal aspirate culture (106 cfu/ mL), or bronchoalveolar lavage culture (104 cfu/mL) was required to confirm the diagnosis of VAP.15 Only first VAP episodes occurring >48 h after starting mechanical ventilation were taken into account. VAP episodes occurring <5 d after starting mechanical ventilation were considered as early-onset. Late-onset VAP was defined as VAP diagnosed 5 d after starting mechanical ventilation. Prior antibiotic use was defined as any antibiotic treatment during the four weeks preceding ICU admission. Antimicrobial therapy was considered inappropriate when none of the antibiotics used was active in vitro on microorganisms causing VAP.15 Shock was defined as hypotension which was not reversed with fluid resuscitation (and required vasopressors) and was associated with organ dysfunction or hypoperfusion abnormalities.16 Immunosuppression was defined as active solid or hematologic malignancy, leucopenia (<1000/mL), and chronic use of immunosuppressive therapy.17 Chronic corticosteroid use was considered as immunosuppressive treatment at 1 mg/kg/day for at least one month during the 3 months preceding ICU admission. Long-term low-dose systemic corticosteroid treatment was defined as corticosteroid treatment at <1 mg/kg/day for at least one month during the 3 months preceding ICU admission. During the study period, no HIV infected patient was admitted to the ICU. MDR bacteria were defined as methicillin-resistant Staphylococcus aureus (MRSA), ceftazidime or imipenem-resistant Pseudomonas aeruginosa, Acinetobacte baumannii, Stenotrophomonas maltophilia, and extended-spectrum b-lactamase producing Gramnegative bacilli. Modified clinical pulmonary infection score (CPIS)18 was used to assess VAP severity. Radiologic score was determined by a modification of the technique of Weinberg et al.19 Anterioreposterior chest roentgenograms were divided into four zones using a horizontal line originating from the hilus; each zone was then graded as

1024 follows: 0, normal; 1, interstitial pulmonary infiltrates; 2, fluffy alveolar infiltrates; 3, dense alveolar infiltrates.

Outcomes and study variables ICU mortality was assessed as a primary outcome in this study. Duration of mechanical ventilation and ICU stay were assessed as secondary outcomes. For all study patients, the following characteristics were prospectively recorded at admission: age; gender, severity of illness based on simplified acute physiology score II (SAPS II), McCabe score,20 logistic organ dysfunction (LOD) score,21 surgery during the last two weeks, presence of co-morbidities, transfer from other wards, infection, shock, long-term low-dose systemic use of corticosteroids, prior use of antibiotics, and findings of the last spirometry in COPD patients. The following parameters were recorded at the day of VAP diagnosis: duration of mechanical ventilation prior to VAP, late-onset VAP, type of causative pathogen, VAP related to MDR bacteria, polymicrobial VAP, radiographic score, CPIS, temperature, leukocyte count, albumin level, presence of shock, use of systemic corticosteroids, and the presence of acute renal failure as evaluated by the RIFLE scale.22 In addition, information on ICU mortality, total duration of mechanical ventilation, and length of ICU stay was collected.

Statistical analysis Continuous variables were compared using the Student’s t test for normally distributed variables or the ManneWhitney U test for non-normally distributed variables. The c2or Fisher’s exact test was used to compare categorical variables, as appropriate. To determine the relationship between ICU mortality (dependent variable) and several independent variables, a multivarable regression analysis was performed. All variables with a p value <0.1 by univariate analysis were entered in the forward stepwise multiple logistic regression model. Results of the logistic regression analyses are reported as adjusted odds ratios (ORs) with 95% confidence intervals (CIs). COPD patients were compared with those without COPD with regards to characteristics at ICU admission, and at the day of VAP diagnosis. Results are expressed as mean  SD or median (25th, 75th interquartile range) for normally and non-normally distributed continuous variables, respectively. Quantitative variables are presented as numbers (percentages). All p values were two-tailed and p values of <0.05 were considered to indicate statistical significance.

Results During the study period, 215 patients met criteria for VAP and were all prospectively followed. Characteristics of study patients at ICU admission and at the day of VAP diagnosis are presented in Tables 1 and 2. One-hundred and fifty seven patients (76%) were admitted due to medical causes (COPD exacerbation 17.1%, Community-acquired pneumonia 19.4%, Health-care-associated pneumonia 25.1%, ARDS 2.8%, acute intoxications 10.8%, severe

D. Makris et al. cellulites 8%, other causes including cardiac arrest, embolism, pancreatitis 16.6%). The mean mechanical ventilation duration (days) prior to VAP diagnosis was 18  12. A total of 235 microrganisms were isolated at significant threshold. The most frequently isolated microrganisms were P. aeruginosa (39%), and S. aureus (17% of VAP episodes). A total of 136 MDR bacteria were isolated in 125 (57%) patients, including 56 (26%) ceftazidime or imipenemresistant P. aeruginosa, 30 (13%) A. baumannii, 27 (12%) ESBL producing Gram-negative bacilli, 16 (7%) MRSA, 7 (3%) S. maltophilia (Table 3). Sixty-five out of 215 (30%) patients with VAP had COPD. Spirometric findings prior to admission were available for 51 out of 65 (78.5%) COPD patients. Forced expiratory volume in 1 s (FEV1) %predicted was 54  16%. According to the GOLD stages of COPD severity 4(7.8%) of them were stage I, 18(35.2%) stage II, 14(27.4%) stage III and 15(29.5%) stage IV. Twenty-two out of 65 (33%) COPD patients had long-term oxygen therapy, 64 (98%) were treated with inhaled bronchodilators, 23 (35%) had received corticosteroids during the last three months, and 4 (6%) had chronic non-invasiveventilation at home (Table 3). P. aeruginosa, S. aureus, Enterobacter species and A. baumannii accounted for 74% of isolated microrganisms in COPD patients. Table 4 depicts the distribution of pathogens in COPD according to the GOLD staging of severity. Age, rate of patients with ultimately or rapidly fatal disease, rate of patients with infection at ICU admission, rate of patients with long-term systemic corticosteroid use, rate of patients with prior antibiotic treatment, and CPIS were significantly higher in COPD patients compared to patients without COPD. However, SAPS II and LOD score were significantly lower in COPD patients compared to those without COPD. No significant difference was found in other patient characteristic or microbiological data between the two groups (Tables 1e3).

Risk factors for ICU mortality Risk factors for ICU mortality by univariate analysis are presented in Tables 1 and 2. Multivariable logistic regression analysis model (HosmereLemshow goodness-of-fit, p Z 0.578) showed that COPD, SAPS II, and shock at VAP diagnosis were independently associated with ICU mortality (Table 5).

Impact of COPD on outcome ICU mortality was significantly higher in COPD patients compared to non-COPD patients (60% vs 43% respectively, p Z 0.027). ICU mortality in the different GOLD stages of COPD severity were 25%, 82%, 77.5% and 66.5% for stage I, II, III and IV respectively. No significant difference was found in mean FEV1 %predicted between survivors and nonsurvivors (48  20 vs 45  17%, p Z 0.674). Overall, there was no significant difference in duration of mechanical ventilation and length of ICU stay between the COPD and non-COPD. Durations (days) of mechanical ventilation median (25th, 75th interquartile range) in survivors were 31 (16e67) and 26 (15e41) for COPD and non-COPD patients respectively (p Z 0.2). Lengths of ICU stay (days) in survivors were 39 (21e71) and 34 (28e58) for COPD and non-

Mortality in COPD patients with VAP Table 1

1025

Characteristics of study patients at ICU admission. Survivors, n Z 111 Non-survivors, n Z 104 p value COPD

p value

Yes, n Z 65 No, n Z 150 Age 57  16 Male gender 82 (73) SAPS II score 48  17 McCabe score 2 53 (47) LOD score 6.2  4 Category of admission Medical 84 (75) Surgical 27 (24) Co-morbidities Immunosuppression 17 (15) Liver cirrhosis 3 (2) Chronic renal failure 4 (3) Diabetes 18 (16) Cancer 13 (10.8) 23 (20.7) Cardiac diseaseb COPD 26 (23) Transfer from other wards 71 (63) Infection 72 (64) Shock 30 (27) Long-term systemic corticosteroid use 6 (5) Prior antimicrobial treatment 39 (35)

65  15 78 (75) 55  17 73 (70) 6.6  4

<0.001 0.877 0.006 0.001a 0.125 0.442

73 (70) 31 (29) 24 4 4 14 24 29 39 74 80 44 10 59

(23) (3) (3) (13) (23) (27.8) (37) (71) (76) (42) (9) (56)

0.167 0.714 >0.999 0.702 0.031 0.265 0.027a 0.308 0.050 0.022a 0.302 0.002a

68  11 53 (81) 47  16 49 (75) 53

58  17 107 (71) 53  18 77 (51) 74

54 (83) 11 (16)

103 (68) 47 (31)

8 (12) 3 (4) 4 (6) 7 (10) 10 (15.4) 21 (32.3) NA 45 (69) 53 (81) 19 (29) 17 (26) 37 (56)

33 4 4 25 28 31 NA 100 99 39 6 61

(22) (2) (2) (16) (18.6) (20.6) (66) (66) (26) (4) (40)

<0.001 0.128 0.027 0.001a 0.013 0.013a

0.130 0.434 0.247 0.304 0.841 0.086 NA 0.753 0.033a 0.738 <0.001a 0.037a

Data are N (%) or mean  SD. ICU, intensive care unit; SAPS, simplified acute physiology score; LOD, logistic organ dysfunction; COPD, chronic obstructive pulmonary disease; NA, not applicable. a Odds ratio (95% CI) 2.6 (1.5e4.5), 2.9 (1.5e5.5), 0.44 (0.21e0.93), 1.9 (1.1e3.5), 2.2 (1.1e4.5), 1.9 (1.1e3.5), 11 (4e29), 2.4 (1.3e4.1), 1.9 (1e3.4); respectively. b Chronic heart failure or ischemic disease.

Table 2

Characteristics of study patients at the day of VAP diagnosis. Survivors, n Z 111 Non-survivors, n Z 104 p value

COPD

p value

Yes, n Z 65 No, n Z 150 Duration of MV prior to VAP Late-onset VAP VAP related to MDR bacteria Polymicrobial VAP Radiographic score CPIS PO2/FiO2 Temperature,  C Leucocytes, cells/mL  103 Albumin level, mg/dL Shock Use of systemic corticosteroids Renal failure Hemodialysis RIFLE Risk Injury Failure Loss End stage kidney disease Inappropriate antibiotic treatment

18  11 99 (89) 65 (58) 8 (7) 3.7  1.8 5.3  1.7 250  106 38.1  0.9 13.5  7.3 30.7  4.5 16 (14) 30 (27)

18  14 92 (88) 60 (57) 12 (11) 4.6  2.3 5.5  1.7 194  100 37.6  1.4 14.8  8.8 28.9  5 42 (40) 45 (43)

22 (19)

31 (29)

11 8 16 6 2 38

13 10 22 11 4 42

(9) (7) (14) (5) (1) (34)

(12) (9) (21) (10) (3) (40)

0.731 >0.999 0.889 0.349 0.012 0.420 0.004 0.011 0.304 0.010 0.001a 0.021a 0.113 0.165

0.471

19  14 57 (87) 39 (60) 4 (6) 4.6  2.3 5.8  1.6 207  109 38  1 15  8 30  4 19 (25) 24 (36)

18  13 134 (89) 86 (57) 16 (10) 3.7  1.8 5.2  1.7 227  105 38  1.3 14  8 30  5 39 (26) 51 (34)

16 (24)

37 (24)

9 5 14 5 2 20

15 13 24 12 4 60

(13) (7) (21) (7) (3) (30)

(10) (8) (16) (8) (2) (40)

0.419 0.814 0.880 0.443 0.109 0.023 0.247 0.796 0.557 0.870 0.738 0.876 >0.999 0.842

0.274

MV, mechanical ventilation; VAP, ventilator-associated pneumonia; MDR, multidrug resistant; CPIS, clinical pulmonary infection score. Data are N (%) or mean  SD. a Odds ratio (95% CI) Z 3.98 (2.05e7.7), 2.2 (1.1e3.5), respectively.

1026 Table 3

D. Makris et al. Microrganisms isolated in 215 patients with VAP. Survivors, n Z 111

Non-survivors, n Z 104

COPD Yes, n Z 65

No, n Z 150

Gram-positive Staphylococcus aureus Methicillin sensitive Methicillin resistant Streptococous pneumoniae

18 12 6 1

(16) (10) (5) (0.9)

19 9 10 0

(18) (8) (9) (0)

10 3 7 0

(15) (4) (10) (0)

27 18 9 1

(18) (12) (6) (0.6)

Gram-negative Enterobacter species Haemophilus influenzae Pseudomonas aeruginosa Serratia marcescensa Proteus mirabilis Acinetobacter baumannii Escherichia coli Citrobacter freundi Stenotrophomonas maltophilia Klebsiella species

13 1 40 7 2 17 7 2 5 3

(11) (0.9) (36) (6) (1) (15) (6) (1) (4) (2)

10 0 44 1 1 13 9 2 5 6

(9) (0) (42) (0.9) (0.9) (12) (8) (1) (4) (5)

8 1 22 0 1 10 6 1 3 5

(12) (1) (33) (0) (1) (15) (9) (1) (4) (7)

15 0 62 8 2 20 10 3 7 4

(10) (0) (41) (5) (1) (13) (6) (2) (4) (2)

Data are N (%). a p Z 0.066 for survivors vs non-survivors, p > 0.2 for all other comparisons (survivors vs non-survivors and COPD vs non-COPD patients).

COPD patients respectively (p Z 0.38). However, when COPD population was stratified according to GOLD stages, we found that patients with advanced disease (stage IV) had significantly longer ICU stay and mechanical ventilation duration compared to patients without COPD (Fig. 1). COPD stage IV survivors had marginally longer mechanical ventilation duration and ICU stay compared to non-COPD survivors [69 (24e103) vs 26 (15e41) respectively, p Z 0.05)] and [74 (37e113) and 34 (28e58) respectively, p Z 0.07].

Discussion Our findings suggest that the presence of COPD in VAP patients is an independent risk factor for ICU mortality. In

Table 4

addition, increased SAPS II at ICU admission and the presence of shock at the day of VAP diagnosis, are also independently associated with ICU mortality in VAP patients. Previous studies reported that COPD was associated with increased mortality in VAP patients.12,13 However, COPD was not independently associated with mortality, suggesting that the higher mortality rate observed in these patients was related to other factors such as various comorbidities. In a cohort of 129 VAP patients, including 24 COPD patients, mortality was significantly higher in COPD patients compared to patients without COPD (33% vs 9%, p Z 0.002, respectively).13 However, the association between COPD and VAP mortality did not remain significant after adjustment for coexisting factors by multivariable analysis. Similar results were reported by Torres et al.12

Microrganisms isolated in COPD patients according to GOLD staging of COPD severity. COPD stages IeII, n Z 22

III, n Z 14

IV, n Z 15

Gram-positive Staphylococcus aureus Methicillin sensitive Methicillin resistant

3 (13) 1 (4) 2 (9)

3 (21) 1 (7) 2 (14)

4 (26) 1 (6) 3 (20)

Gram-negative Enterobacter species Pseudomonas aeruginosa Proteus mirabilis Acinetobacter baumannii Escherichia coli Stenotrophomonas maltophilia Klebsiella species

3 8 0 4 3 0 1

3 6 1 1 1 0 0

2 5 0 1 1 1 1

Data are N (%) p > 0.05 for all other comparisons.

(13) (36) (0) (18) (13) (0) (4)

(21) (43) (7) (7) (7) (0) (0)

(13) (33) (0) (6) (6) (6) (6)

Mortality in COPD patients with VAP Table 5

1027

Risk factors for ICU mortality by multivariable analysis.

COPD SAPS II Shock at the day of VAP diagnosis

Odds ratio

95% CI

p value

2.58 1.02 3.72

1.33e5.02 1.01e1.04 1.88e7.39

0.005 0.010 <0.001

COPD, chronic obstructive pulmonary disease; SAPS, simplified acute physiology score.

who aimed to identify prognostic factors of nosocomial pneumonia in mechanically ventilated patients. Nevertheless, COPD population was relatively small in those previous studies and they might have been underpowered to depict significant differences in ICU mortality in a multivariable analysis. The present study included a larger COPD population and aimed to investigate as well whether GOLD staging of COPD severity was associated with the outcomes of these patients. Our findings provide clear evidence that COPD is associated with poor prognosis in VAP patients. Previous studies performed in patients with communityacquired pneumonia found similar results.23e25 Several factors might explain the higher rates of ICU mortality in COPD patients with VAP. COPD has an adverse impact on respiratory muscle function.28 Autopsy findings in diaphragms of patients with COPD have revealed indices of increased injury.29 These alterations may be amplified during critical illness. In our study we have not assessed changes in respiratory muscle structure or contraction properties. However, we found no evidence that could support indirectly the above hypothesis, such as differences in the mechanical ventilation duration in patients with COPD or not. In addition, we, and other investigators, have pointed out an association between COPD and severe acute exacerbations or VAP secondary to MDR bacteria.11,30,31 Patients with VAP related to these bacteria are at higher risk for receiving inappropriate initial antibiotic treatment which has been identified as a risk factor for mortality in ICU patients.30,32,33 Nevertheless, no significant difference was found in microbiologic findings and inappropriate initial antibiotic treatment between COPD and

non-COPD patients. In this respect we assume that increased mortality in critically ill COPD might be related to other components of the disease. COPD is an inflammatory disease with complex pathobiology where various underlying mechanisms are implicated.26,27 COPD has also a significant systemic component and patients may be characterized by nutritional depletion e cachexia28 that may affect adversely the immune response.34 Certainly, our investigation has not assessed this hypothesis but this might be the subject of a future study. In addition in the present study we investigated whether COPD is related to longer ICU stay and mechanical ventilation. We found no significant association between the COPD population overall and these indices. When we analyzed further our data according to GOLD staging of COPD severity we found that ICU stay and mechanical ventilation duration were longer in the group with advanced COPD (GOLD stage IV) compared to non-COPD patients (Fig. 1). COPD patients with less severe disease (GOLD stages IeIII) were not different with non-COPD patients in terms of ICU stay and MV duration. This underlines a potential relationship between ICU morbidity and disease severity. On the other hand, advances or innovations in COPD management in the ICU (i.e. recognition and management of dynamic hyperinflation) may be more effective in patients with less severe disease compared to advanced COPD stages. We acknowledge that definitive conclusion is hard to be drawn, particularly since there is luck of data on the relationship between ICU morbidity indices and COPD staging of severity.

Figure 1 ICU length of stay and mechanical ventilation duration in non-COPD patients and patients with COPD, according to the GOLD staging of COPD severity. Bars represent mean (SE) values.

1028 The present prospective study provides evidence suggesting that COPD is an independent factor for mortality in ICU patients who present VAP. Certainly, one might argue that this finding depends on the quality of adjustment to potential confounders, which are certainly present in COPD populations, such as advanced age or medical co-morbidities. Thus, in the present study we assessed several potential confounders by evaluating baseline characteristics of ICU patients, including co-morbidities (i.e. cardiac disease, diabetes), severity of critical illness (SAPSII, LOD, MacCabe score) and parameters related to VAP severity. All factors which were associated with mortality at the 0.1 level in univariate analysis were forced into the multivariable analysis model. Our results (Table 4) suggested that COPD remained an independent factor of mortality when confounders were taken into account. Furthermore, it was revealed that mortality in VAP patients was also independently associated to the severity of critical care illness (SAPSII) and VAP severity (presence of shock at diagnosis). This might be not surprising since these findings confirm results reported previously.7,34 However, our study points out that the significant impact of COPD in the mortality of patients who present VAP in the ICU. Inappropriate initial antibiotic treatment was common (37%) in our population. This could be explained by the high rate of MDR bacteria (57%) which are frequently resistant to initial antibiotic regimen. Our study did not find a significant relationship between inappropriate initial antibiotic treatment for VAP and ICU mortality. Two potential explanations could be suggested. First, we did not take into account timing of antibiotic administration or dose of antimicrobial therapy. Second, our study was probably underpowered to detect such an effect. The present study suggests that outcome in VAP patients is strongly related to the underlying respiratory disease, such as COPD. In this respect, our findings might have several clinical implications in the prevention and management of VAP. First, existing preventive strategies for VAP should be further enhanced in COPD population and the effectiveness of more rigorous protocols should be evaluated, especially in advanced disease. Furthermore, our findings support the rigorous use of non-invasive-ventilation in critically ill COPD.35e37 On the other hand, physicians should be alert not only for the presence of COPD in the medical history of ICU patients but also for prompt detection of signs suggestive of the presence of airflow limitation during mechanical ventilation (i.e. autoPEEP). Moreover, mechanical ventilation strategies that reduce hyperinflation and early adequate antibiotic treatment should have high priority in these patients. In addition, parameters reflecting flow limitation might be helpful in assessing critical illness severity at baseline (i.e. by including such parameters in ICU severity scores) and finally, the presence of COPD should be especially considered in resource planning and family communication issues. However, at this point we wish to underline that our data should not be used to withdraw life support in these patients. Our study has some limitations. First, this study was performed in a single center. Therefore, our results may not be generalizable to other ICU patients. Second, several significant differences were found between COPD and nonCOPD patients. However, multivariable analysis allowed, at least in part, adjustment for these confounders. Third,

D. Makris et al. one might argue that the incidence of COPD may have been underestimated in this study. We certainly agree that the presence of spirometry or a prior diagnosis of COPD was compulsory for a patient to be treated as a COPD patient in our study and thus, a selection bias may be possible. Not all patients who actually have COPD are diagnosed before critical illness. It is not uncommon for COPD to be diagnosed for the first time on the basis of clinical and radiological findings during treatment in the ICU. For example, patients who have a history of smoking, have distinct signs of hyperinflation and/or emphysema in the chest X-ray or those who have expiratory flow limitation during controlled mechanical ventilation have a very high clinical probability of actually having COPD. However, such radiographic or physiologic criteria have not been used in our study. Thus, the true prevalence of COPD may even be higher and in this respect the relationship of COPD with outcome reported in our study is only an estimate; these issues should be taken into consideration in the interpretation of our results. In addition, it should be noted that in the present ICU study spirometry could not be performed at baseline. Thus, recent spirometry findings were not available for 14 (21%) COPD patients. In those patients diagnosis was based on official medical records and medical history obtained by interviews in ICU. Previous studies in critical care patients addressing the same issue with the present study have not provided information with regard the severity of COPD, therefore any comparison is difficult to be made. This is a pragmatic limitation of observational ICU studies that include COPD patients. However, we have repeated univariate and multivariable analyses after exclusion of these patients. COPD was still independently associated with ICU mortality (data not shown). In conclusion, COPD, SAPS II, and presence of shock at VAP diagnosis are independently associated with ICU mortality in patients who present VAP. These data outline the need to improve the treatment and prevention of VAP in COPD patients.

Acknowledgments The authors have no potential conflicts of interest to declare and no involvement in any organization with a direct financial interest in the subject of the manuscript.

Authors’ contribution DM, AD, EZ, and SN designed and motivated this study. DM, BD, and SN collected data. DM, and SN wrote the manuscript, and all authors participated in its critical revision. DM had full access to all data in the study and had final responsibility for the decision to submit for publication. All authors read and approved the final manuscript.

Conflict of interest All authors declare no conflict of interest.

Mortality in COPD patients with VAP

References 1. Chastre J, Fagon JY. Ventilator-associated pneumonia. Am J Respir Crit Care Med 2002;165:867e903. 2. Safdar N, Dezfulian C, Collard HR, Saint S. Clinical and economic consequences of ventilator-associated pneumonia: a systematic review. Crit Care Med 2005;33:2184e93. 3. Melsen WG, Rovers MM, Bonten MJ. Ventilator-associated pneumonia and mortality: a systematic review of observational studies. Crit Care Med 2009;37:2709e18. 4. Nguile-Makao M, Zahar JR, Francais A, Tabah A, Garrouste-Org B, Allaouchiche B, et al. Attributable mortality of ventilator-associated pneumonia: respective impact of main characteristics at ICU admission and VAP onset using conditional logistic regression and multi-state models. Intensive Care Med 2010;36:781e9. 5. Bercault N, Boulain T. Mortality rate attributable to ventilatorassociated nosocomial pneumonia in an adult intensive care unit: a prospective case-control study. Crit Care Med 2001;29:2303e9. 6. Leone M, Bourgoin A, Giuly E, Antonini F, Dubuc M, Viviand X, et al. Influence on outcome of ventilator-associated pneumonia in multiple trauma patients with head trauma treated with selected digestive decontamination. Crit Care Med 2002; 30:1741e6. 7. Combes A, Luyt CE, Fagon JY, Wolff M, Trouillet JL, Chastre J. Early predictors for infection recurrence and death in patients with ventilator-associated pneumonia. Crit Care Med 2007;35:146e54. 8. Depuydt PO, Vandijck DM, Bekaert MA, Decruyenaere JM, Blot SI, Vogelaers DP, et al. Determinants and impact of multidrug antibiotic resistance in pathogens causing ventilatorassociated-pneumonia. Crit Care 2008;12:R142. 9. Teixeira PJ, Seligman R, Hertz FT, Cruz DB, Fachel JM. Inadequate treatment of ventilator-associated pneumonia: risk factors and impact on outcomes. J Hosp Infect 2007;65:361e7. 10. Lisboa T, Diaz E, Sa-Borges M, Socias A, Sole-Violan J, Rodriguez A, et al. The ventilator-associated pneumonia PIRO score: a tool for predicting ICU mortality and health-care resources use in ventilator-associated pneumonia. Chest 2008;134:1208e16. 11. Nseir S, Di Pompeo C, Soubrier S, Cavestri B, Jozefowicz E, Saulnier F, et al. Impact of ventilator-associated pneumonia on outcome in patients with COPD. Chest 2005;128:1650e6. 12. Torres A, Aznar R, Gatell JM, Jimenez P, Gonzalez J, Ferrer A, et al. Incidence, risk, and prognosis factors of nosocomial pneumonia in mechanically ventilated patients. Am Rev Respir Dis 1990;142:523e8. 13. Rello J, Ausina V, Ricart M, Castella J, Prats G. Impact of previous antimicrobial therapy on the etiology and outcome of ventilator-associated pneumonia. Chest 1993;104:1230e5. 14. Celli BR, MacNee W. Standards for the diagnosis and treatment of patients with COPD: a summary of the ATS/ERS position paper. Eur Respir J 2004;23:932e46. 15. Niederman MS, Craven DE. Guidelines for the management of adults with hospital-acquired, ventilator-associated, and healthcare-associated pneumonia. Am J Respir Crit Care Med 2005;171:388e416. 16. Dellinger RP, Levy MM, Carlet JM, Bion J, Parker MM, Jaeschke R, et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008. Crit Care Med 2008;36:296e327. 17. Nseir S, Di Pompeo C, Diarra M, Brisson H, Tissier S, Boulo M, et al. Relationship between immunosuppression and intensive care unit-acquired multidrug-resistant bacteria: a case-control study. Crit Care Med 2007;35:1318e23. 18. Luna CM, Blanzaco D, Niederman MS, Matarucco W, Baredes NC, Desmery P, et al. Resolution of ventilator-associated pneumonia: prospective evaluation of the clinical pulmonary infection score as an early clinical predictor of outcome. Crit Care Med 2003;31:676e82.

1029 19. Weinberg PF, Matthay MA, Webster RO, Roskos KV, Goldstein IM, Murray JF. Biologically active products of complement and acute lung injury in patients with the sepsis syndrome. Am Rev Respir Dis 1984;130:791e6. 20. McCabe WR, Jackson GG. Gram-negative bacteremia. Etiology and ecology. Arch Intern Med 1962;110:847e55. 21. Le Gall JR, Klar J, Lemeshow S, Saulnier F, Alberti C, Artigas A, et alICU Scoring Group. The Logistic Organ Dysfunction system. A new way to assess organ dysfunction in the intensive care unit. J Am Med Assoc 1996;276:802e10. 22. Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P. Acute renal failure e definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 2004;8:R204e12. 23. Molinos L, Clemente MG, Miranda B, Alvarez C, del Busto B, Cocina BR, et al. Community-acquired pneumonia in patients with and without chronic obstructive pulmonary disease. J Infect 2009; 58:417e24. 24. Restrepo MI, Mortensen EM, Pugh JA, Anzueto A. COPD is associated with increased mortality in patients with community-acquired pneumonia. Eur Respir J 2006;28:346e51. 25. Rello J, Rodriguez A, Torres A, Roig J, Sole-Violan J, GarnachoMontero J, et al. Implications of COPD in patients admitted to the intensive care unit by community-acquired pneumonia. Eur Respir J 2006;27:1210e6. 26. Makris D, Lazarou S, Alexandrakis M, Kourelis TV, Tzanakis N, Kyriakou D, et al. Tc2 response at the onset of COPD exacerbations. Chest 2008;134:483e8. 27. Makris D, Vrekoussis T, Izoldi M, Alexandra K, Katerina D, Dimitris T, et al. Increased apoptosis of neutrophils in induced sputum of COPD patients. Respir Med 2009;103:1130e5. 28. Rabe KF, Hurd S, Anzueto A, Barnes PJ, Buist SA, Calverley P, et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary. Am J Respir Crit Care Med 2007;176:532e55. 29. Scott A, Wang X, Road JD, Reid WD. Increased injury and intramuscular collagen of the diaphragm in COPD: autopsy observations. Eur Respir J 2006;27:51e9. 30. Nseir S, Di Pompeo C, Cavestri B, Jozefowicz E, Nyunga M, Soubrier S, et al. Multiple-drug-resistant bacteria in patients with severe acute exacerbation of chronic obstructive pulmonary disease: prevalence, risk factors, and outcome. Crit Care Med 2006;34:2959e66. 31. Ferrer M, Ioanas M, Arancibia F, Marco MA, de la Bellacasa JP, Torres A. Microbial airway colonization is associated with noninvasive ventilation failure in exacerbation of chronic obstructive pulmonary disease. Crit Care Med 2005;33:2003e9. 32. Nseir S, Deplanque X, Di Pompeo C, Diarra M, RousselDelvallez M, Durocher A. Risk factors for relapse of ventilatorassociated pneumonia related to nonfermenting gram negative bacilli: a case-control study. J Infect 2008;56:319e25. 33. Kollef MH. Broad-spectrum antimicrobials and the treatment of serious bacterial infections: getting it right up front. Clin Infect Dis 2008;47(Suppl. 1):S3e13. 34. Cunningham-Rundles S. Nutrition and the mucosal immune system. Curr Opin Gastroenterol 2001;17:171e6. 35. Tejerina E, Frutos-Vivar F, Restrepo MI, Anzueto A, Abroug F, Palizas F, et al. Incidence, risk factors, and outcome of ventilator-associated pneumonia. J Crit Care 2006;21:56e65. 36. Lightowler JV, Wedzicha JA, Elliott MW, Ram FS. Non-invasive positive pressure ventilation to treat respiratory failure resulting from exacerbations of chronic obstructive pulmonary disease: Cochrane systematic review and meta-analysis. BMJ 2003;326:185. 37. Burns KE, Adhikari NK, Keenan SP, Meade M. Use of non-invasive ventilation to wean critically ill adults off invasive ventilation: meta-analysis and systematic review. BMJ 2009;338:b1574e9.