Ventilator-associated pneumonia caused by high risk microorganisms: A matched case-control study

(1)

by high risk microorganisms:

A matched case-control study

Melda AYBAR TÜRKOĞLU¹, Arzu TOPELİ İSKİT²

1 Gazi Üniversitesi Tıp Fakültesi, İç Hastalıkları Anabilim Dalı, Yoğun Bakım Bilim Dalı,

2Hacettepe Üniversitesi Tıp Fakültesi, İç Hastalıkları Anabilim Dalı, Yoğun Bakım Ünitesi, Ankara.

ÖZET

Yüksek riskli mikroorganizmalarla gelişen ventilatörle ilişkili pnömoni:

Eşleştirilmiş olgu-kontrol çalışması

Bu çalışmanın amacı, yüksek riskli mikroorganizmalarla gelişen ventilatörle ilişkili pnömoni (VİP)’nin sonuca etkisini ince- lemektir. Çalışmanın tipi: Eşleştirilmiş olgu-kontrol çalışması. Çalışma bir üniversite hastanesi dahiliye yoğun bakım üni- tesinde gerçekleştirilmiştir. Pseudomonas aeruginosa, Acinetobacter spp., Stenotrophomonas maltophilia ve/veya metisili- ne dirençli Staphylococcus aureus yüksek riskli mikroorganizmalar olarak tanımlanmış olup, yüksek riskli mikroorganiz- malarla VİP gelişen 35 hasta olgu grubu olarak kabul edilmiştir. VİP gelişmeyen 35 kontrol hastası APACHE II skoru, yaş, başvuru tarihi ve mekanik ventilasyon (MV) süresine göre eşleştirilmiştir. Yoğun bakım ve hastane mortalitesi olgu ve kont- rol grubunda benzer olarak tespit edilmiştir (sırasıyla p= 0.58 ve p= 1.00). Ancak, olgu grubunda kontrol grubuna göre MV süresi [sırasıyla 18 (10-25) gün (medyan -çeyrekler arası aralık-) ve 8 (6-11) gün, p< 0.01], yoğun bakım [sırasıyla 20 (11- 30) gün ve 13 (8-19) gün, p< 0.01] ve hastane yatış süresi [sırasıyla 29 (20-44) gün ve 22 (13-37) gün, p= 0.05] daha uzun olarak tespit edilmiştir. Yüksek riskli mikroorganizmalarla gelişen VİP’in her türlü etkenden bağımsız olarak yoğun bakım (OR: 6), hastane yatış süresi (OR: 4) ve MV süresini (OR: 11) uzatmakta olduğu tespit edilmiştir. Yüksek riskli mikroorga- nizmalarla gelişen VİP mortaliteyi önemli derecede etkilememektedir. Ancak, yoğun bakım ve hastane yatış süresini yedi gün, MV süresini 10 gün arttıran bağımsız bir risk faktörüdür.

Anahtar Kelimeler: Yüksek riskli mikroorganizmalar, mekanik ventilasyon, pnömoni, nozokomiyal, yatış süresi, morta- lite, sonuç.

Yazışma Adresi (Address for Correspondence):

Dr. Melda AYBAR TÜRKOĞLU, Gazi Üniversitesi Tıp Fakültesi, Hastanesi İç Hastalıkları Anabilim Dalı, Yoğun Bakım Ünitesi, 06510 Beşevler ANKARA - TURKEY

e-mail: meldaturkoglu@yahoo.com.tr

(2)

Ventilator-associated pneumonia (VAP) is the most frequently encountered nosocomial infection in the intensive care units (ICU) (1-3). It is also the leading cause of mortality due to nosocomial infections in ICUs, with a crude mortality ranging from 24% to 71% (4-8). The attributable mortality of VAP was reported to be 27% in general, but when high risk microorganisms (HRM) including Acinetobacter spp., Pseudomo- nas aeruginosa, Stenotrophomonas maltophilia and methicillin-resistant Staphylococcus aureus (MRSA) are considered, the rate might increase (4,6). Fagon et al. reported 43% attributable mortality rate with VAP due to Acinetobacter or Pseudomonas species (4). S. maltophilia, anot- her HRM causing increased mortality is beco- ming a frequently seen cause of VAP in ICUs (9,10). Although the most frequent bacteria causing VAP is gram-negative enteric bacteria, the frequency of MRSA as the cause of VAP is increasing and VAP due to MRSA carries a high mortality rate (1,2,11-13). Development of VAP, es-

pecially VAP due to HRM in the ICUs is also important, as it increases duration of mechanical ventilation (MV), length of ICU and hospital stay (4,5,13-17). However, one study reported that VAP caused by Acinetobacter baumannii was not significantly associated with an attributable mortality, nor increased ICU stay (18). In a re- cent study, it was also shown that, VAP due to A.

baumannii resulted in a mortality rate similar to that due to other microorganisms (19). So, we performed a matched case-control study to in- vestigate whether VAP caused by HRM causes increased mortality, length of ICU and hospital stay, and duration of MV.

MATERIALS and METHODS Study Population and Variables

A matched cohort study was conducted in the medical ICU of Hacettepe University Hospital, Ankara, Turkey from May 1, 1999 to April 30, 2001. Local ethics committee approved the SUMMARY

Ventilator-associated pneumonia caused by high risk microorganisms: A matched case-control study

Melda AYBAR TÜRKOĞLU¹, Arzu TOPELİ İSKİT²

1 Medical Intensive Care Unit, Department of Internal Medicine, Faculty of Medicine, Gazi University, Ankara, Turkey,

2Medical Intensive Care Unit, Department of Internal Medicine, Faculty of Medicine, Hacettepe University, Ankara, Turkey.

To determine the impact of ventilator-associated pneumonia (VAP) caused by high risk microorganisms (HRM) on patient outcome. Design: Matched case-control study. The study was conducted in a medical intensive care unit (ICU) of a univer- sity hospital. Thirty-five patients with VAP caused by HRM, including Pseudomonas aeruginosa, Acinetobacter spp., Ste- notrophomonas maltophilia and/or methicillin-resistant Staphylococcus aureus were accepted as the case the patients.

Thirty-five control patients, who did not develop VAP were matched to the case patients, according to APACHE II score, age, date of admission and duration of mechanical ventilation (MV). ICU and hospital mortality rates were similar between the case and the control patients (p= 0.58 and p= 1.00, respectively). However, length of ICU stay was longer in the case pati- ents than in the control patients [20 (11-30) days (median -interquartile range-) and 13 (8-19) days, respectively; p< 0.01].

Length of hospital stay was also longer in the case patients than in the control patients [29 (20-44) days and 22 (13-37) days, respectively; p= 0.05]. In addition, duration of MV was longer in the case patients than in the control patients [18 (10- 25) days and 8 (6-11) days, respectively; p< 0.01]. VAP caused by HRM independently prolonged ICU (OR: 6) and hospi- tal stay (OR: 4) and duration of MV (OR: 11). VAP caused by HRM was not significantly associated with mortality. Howe- ver, it was an independent risk factor, increasing length of ICU stay and hospital stay by seven days, and duration of MV by 10 days.

Key Words: High risk microorganisms, mechanical ventilation, pneumonia, nosocomial, length of stay, mortality, outcome.

(3)

study. Patients ventilated for more than 48 hours were enrolled and were followed daily, until death or discharge from the hospital. Patients who were admitted to the ICU, after receiving mechanical ventilation for more than 48 hours in another place were excluded.

Age, sex, admission diagnosis and underlying diseases were recorded at baseline. Acute Physi- ology Assessment and Chronic Health Evaluati- on II (APACHE II) score and Glasgow coma score (GCS) were calculated within the first 24 hours after admission (20,21). Risk factors for the development of VAP such as sedative and steroid use, albumin level, type of the nutritional sup- port and the stress ulcer prophylaxis, lenght of stay before the ICU admission, reentubation and antibiotic use before VAP were recorded. Appli- cation of tracheostomy; central venous catheterization; development of acute renal failure, acute respiratory distress syndrome (ARDS), disseminated intravascular coagulation (DIC) and shock requiring vasopressor therapy during the course of the stay were also noted. The antimic- robial treatment and antibiotic resistance patterns were examined retrospectively. The therapy was accepted as appropriate, when all ca- usative pathogens were in vitro susceptible to at least one of the antibiotics of the regimen and when the antibiotics were given in the appropriate dose.

Identification of Case Patients

The clinical diagnosis of VAP was established, when a new and persistent pulmonary infiltrate or a progress in the existing infiltration was observed on the chest X-ray, with the presence of any two of the following criteria: 1. Fever (>

38°C) or hypothermia (< 36°C), 2. Leukocytosis (≥ 10.000/mm³) or leukopenia (≤ 4000/mm³), 3. Presence of purulent sputum, 4. Increase in hypoxemia (22). For microbiologic diagnosis, endotracheal aspiration was performed. The endotracheal aspirate obtained was homogenized using repeated aspirations with a Pasteur’s pi- pette. Serial dilutions (0.1, 0.01 and 0.001) of each sample were prepared in sterile normal sa- line. One hundred milliliters of each dilution of endotracheal aspirate were inoculated into 5%

sheep blood and McConkey agar media, then processed as described elsewhere (23). Results were expressed as cfu/mL = number of colonies x dilution factor x inoculation factor. VAP was confirmed, when the culture of the endotracheal aspirate yielded ≥ 10⁵cfu/mL (23).

Patients having positive culture results for HRM including P. aeruginosa, Acinetobacter spp., S.

maltophilia and/or MRSA, in addition to clinical findings were accepted as the case patients.

Identification of Control Patients

Control patients were selected from the ventilated patients who had no clinical and microbiolo- gical evidence of VAP. Each case patient was matched to one control patient according to the following criteria: 1. Control patient had to be ventilated for at least as long as the case patient prior to the onset of VAP, 2. APACHE II score (±

5 points), 3. Age (± 5 years), 4. Date of admission of the case and the control patient had to be within 12 months.

Statistical Analysis

Statistical calculations were performed using the Statistical Package for Social Sciences (SPSS version 7.5). The Wilcoxon’s rank sum test and Mc Nemar’s test were used to compare continuous and categorical variables, between the case and the control patients, respectively.

Bivariate and multivariate analyses were conducted to determine the factors influencing mortality, prolonged ICU and hospital stay, and duration of MV. Prolonged length of stay and duration of MV were considered to be present, when length of stay or duration of MV exceeded the median value for each parameter. In the bivariate analyis, t-test and chi-square test were used to compare continuous and categorical variables, respectively. Variables with a p value < 0.10 in the bivariate analysis were used in the logistic regression model to determine the independent factors influencing outcome. The results were expressed as median [interquartile range], n (%), odds ratio (OR) and 95% confidence inter- vals (CI). A p value less than 0.05 was considered as statistically significant.

(4)

RESULTS

Study population is seen Figure 1. Among 216 patients who received MV, 156 patients were included in the study. VAP developed in 61 (39%) patients and VAP caused by HRM developed in 45 (29%) patients. Ninety-five patients did not have VAP. Thirty-five patients with VAP caused by HRM (case patients) were matched with 35 patients who did not develop VAP. The matching process failed in 10 cases because of longer duration of MV of the case patients prior to the development of VAP compared to the control patients. Among the 35 pairs, 28 were matched according to the predefined criteria. In the remaining seven pairs, the differences in APACHE II scores were six and eight points in two pairs; difference in age was six years in one pair and 13 years in two pairs; and differences in the admission dates were 15 and 16 months in two pairs. These seven pairs were not excluded, because the differences were thought not to be clinically significant.

As seen in Table 1, two groups were similar with respect to age, APACHE II score, GCS, sex, ad-

mission diagnosis, the type of underlying disease, sedative and steroid use, and use of enteral nutrition. Duration of MV prior to the development of VAP in the case patients was 6 (3.5-9.5) days and total duration of MV in control patients was 8 (6-11.5) days (p< 0.01).

In the 35 case patients, 43 microorganisms we- re isolated. P. aeruginosa was isolated in 15 pa- tients, A. baumannii in 11 patients, Acinetobac- ter lwoffii in four patients, MRSA in 10 patients and S. maltophilia in three patients. Of these 35 patients, six patients had infection with two of these microorganisms and one patient had infection with three of these microorganisms.

Among 32 case patients whose data about antibiotic resistance patterns was available, 17 (53%) patients received appropriate and 11 (34%) patients received inappropriate antibiotic therapy. Isolated Acinetobacter or Pseudomonas species in four patients were resistant to all ami- noglycosides, carbapenems, third and fourth ge- neration cephalosporins, piperacilin and quino- lones and these patients died. Of these four pa- tients, two patients had VAP due to P. aerugino-

3 patients were diagnosed as brain dead

35 matched patients used as control patients 35 patients used as case

patients 45 patients had

VAP with HRM

156 patients were included 536 admissions

60 patients were excluded

21 patients received MV > 48 hour prior to this admission

36 patients received MV < 48 hour 95 patients did

not have VAP 61 patients had VAP

320 patients did not receive MV 216 patients received

MV with intubation

MV: Mechanical ventilation, VAP: Ventilator-associated pneumonia, HRM: High risk microorganisms.

Figure 1. Study population.

(5)

sa. One of them died before the culture resulted.

For the other one, antibiotic therapy was modi- fied according to the other microorganisms seen in the same endotracheal aspirate culture (MRSA and A. lwoffii). Of these four patients, two patients had VAP due to Acinetobacter spe- cies. For these patients carbapenem and ami- noglycoside combination therapy was used.

Patterns of antibiotic resistance for P. aeruginosa and Acinetobacter species are shown in Table 2.

Very high rates of resistance to several antibi- otics were observed for P. aeruginosa and Acine- tobacter species.

Table 1. Distribution of the general characteristics of the case and the control patients.

Case patients (n= 35) Control patients (n= 35) p

Age, years* 69 (50-74) 67 (52-75) 0.58

APACHE II score* 20 (16-27) 19 (17-25) 0.07

GCS* 12 (8-15) 14 (9-15) 0.50

Male 19 12 0.17

Admission diagnosis

Respiratory causes 18 13 0.38

Sepsis/multi-organ failure 5 5 1.00

Neurologic causes 3 2 1.00

Cardiovascular causes 2 6 0.22

Other 7 9 0.79

Underlying diseases**

Respiratory diseases*** 11 7 0.42

Cardiovascular diseases 16 17 1.00

Diabetes mellitus 5 3 0.73

Chronic renal failure 4 5 1.00

Neurologic diseases 4 1 0.38

Malignancy 4 5 1.00

Rheumatologic diseases 3 1 0.63

Chronic liver failure 1 0 1.00

Other 2 1 1.00

Sedative use 15 10 0.23

Steroid use 10 7 0.51

Enteral feeding 15 10 0.23

Stress ulcer prophylaxis with H₂receptor 30 31 1.00

antagonists or proton pump inhibitors

Lenght of stay before the ICU admission, days 2 (1-5) 3 (1-9) 0.65

Reentubation before VAP 10 17 0.17

Antibiotic use before VAP 19 21 0.79

* Values represent median (interquartile range).

** Total numbers exceed 35, because each patient might have multiple underlying diseases.

*** Respiratory diseases in the case patients were chronic obstructive pulmonary disease or asthma, interstitial lung disease and tubercu- losis in nine patients, one patient and one patient, respectively. In the control patients respiratory diseases were chronic obstructive pulmonary disease or asthma, interstitial lung disease and bronchiectasia in five patients, one patient and one patient, respectively.

APACHE: Acute Physiology Assessment and Chronic Health Evaluation, GCS: Glasgow coma score, ICU: Intensive care unit, VAP: Ven- tilatory-associated pneumonia.

Table 2. Patterns of antibiotic resistance for PP..

aaeerruuggiinnoossaaand AAcciinneettoobbaacctteerrspp.

A

Acciinneettoobbaacctteerr P

P.. aaeerruuggiinnoossaa spp.

(n= 15) (n= 15)

Carbapenem resistance 8 11

Tobramycin resistance 14 2

Amikacin resistance 10 13

Ceftazidime resistance 13 13 Piperacillin resistance 12 14 Ciprofloxacin resistance 15 14

(6)

There was no difference between the two groups in terms of the development of organ failures (Table 3). However, tracheostomy was more frequently performed in case patients than in control patients (p= 0.02), and there was a tendency for more frequent use of central venuos catheterization in case patients than in control patients (p= 0.08).

ICU mortality rate was similar between case and control patients (80% and 71%, respectively, p= 0.58). Hospital mortality rate was 80% for both groups (p= 1.00). However, length of ICU stay was longer in case patients than in control patients [20 (11-30) days and 13 (8-19) days,

p< 0.01]. Length of hospital stay was also longer in case patients than in control patients [29 (20- 44) days and 22 (13-37) days, p= 0.05]. In addition, duration of MV was longer in case patients than in control patients [18 (10-25) days and 8 (6-11) days, p< 0.01]. Therefore, VAP caused by HRM resulted in an increase in length of ICU stay and hospital length of stay by seven days, and in duration of MV by 10 days.

In bivariate analysis, factors with a p value

< 0.10, resulting in prolonged length of ICU stay, i.e. > 16 days (median value), were GCS; presence of underlying chronic renal failure, underlying malignancy; development of VAP caused

Table 3. Development of organ failures and application of invasive procedures in the case and the control patients.

Case patients (n= 35) Control patients (n= 35) p

Acute renal failure 18 19 1.00

ARDS 6 2 0.29

DIC 12 12 1.00

Shock requiring vasopressor therapy 20 21 1.00

Tracheostomy 11 2 0.02

Central venuos catheterization 30 22 0.08

ARDS: Acute respiratory distress syndrome, DIC: Disseminated intravascular coagulation.

Table 4. Factors causing increased length of ICU stay, length of hospital stay and duration of MV in multivari- ate analysis.

OR (95% CI) p

Length of ICU stay

Underlying chronic renal failure 12 (1.6-87.5) 0.02

Underlying malignancy 7 (1.1-40.7) 0.04

VAP caused by HRM 6 (1.8-19.7) < 0.01

GCS 1 (1.1-1.3) 0.04

Length of hospital stay

Underlying chronic renal failure 9 (0.9-93.5) 0.06

VAP caused by HRM 4 (1.2-16.4) 0.03

Hospital stay prior to ICU admision 1 (1.1-1.5) < 0.01

Duration of MV

Tracheostomy 55 (1.7-1806.7) 0.02

Underlying cardiovascular disease 21 (3.5-130.7) < 0.01

Santral venous catheterization 18 (1.9-173.0) 0.01

VAP caused by HRM 11 (2.1-54.5) < 0.01

Hemodialysis 5 (1.0-26.9) 0.06

OR: Odds ratio, CI: Confidence interval, ICU: Intensive care unit, VAP: Ventilatory-associated pneumonia, HRM: High risk microorganisms, GCS: Glasgow coma score, MV: Mechanical ventilation.

(7)

by HRM; development of ARDS; presence of shock requiring vasopressor treatment; application of central venous catheterization and tracheostomy (data not shown). When these factors were put into the logistic regression analysis, VAP caused by HRM was found to be an independent risk factor for increased length of ICU stay with an OR of 6 (1.8-19.7) (Table 4). The other independent risk factors for increased length of ICU stay were also shown in Table 4.

Factors with a p value < 0.10, resulting in prolonged length of hospital stay, i.e. > 26 days (median value), were albumin level; presence of underlying chronic renal failure, underlying malignancy; development of VAP caused by HRM;

length of hospital stay prior to ICU admission and number of antibiotics used prior to VAP (data not shown). In the logistic regression analysis, VAP caused by HRM was also found to be an independent risk factor for increased length of

hospital stay with an OR of 4 (1.2-16.4) (Table 4). The other independent risk factors for increased length of hospital stay were also shown in Table 4. For prolonged duration of MV, i.e. > 11 days (median value), factors with a p value <

0.10, were age; presence of underlying cardiovascular disease; development of VAP caused by HRM; application of central venous catheterization, hemodialysis, tracheostomy and enteral feeding (data not shown). In the logistic regression analysis, it was shown that VAP caused by HRM was an independent risk factor for increased duration of MV with an OR of 11 (2.1-54.5) (Table 4). The other independent risk factors for prolonged duration of MV were also shown in Table 4.

Almost similar results were observed when outcome variables were evaluated separately for each type of isolated microorganism, except S.

maltophilia (Table 5). VAP caused by P. aerugi-

Table 5. Outcome variables according to the type of the isolated microorganism.

Case patients (n= 35) Control patients (n= 35) p P

P.. aaeerruuggiinnoossaa (n= 15)

ICU mortality 12 12 1.00

Hospital mortality 12 13 1.00

Length of ICU stay* 22 (11-30) 13 (8-16) 0.05

Length of hospital stay* 29 (16-46) 19 (13-37) 0.21

Duration of MV* 20 (6-24) 9 (6-15) 0.09

A

Acciinneettoobbaacctteerr species (n= 15)

Length of ICU stay* 17 (11-31) 12 (7-17) < 0.01

Duration of MV* 15 (8-31) 7 (5-11) < 0.01

MRSA (n= 10)

Duration of MV* 21 (9-32) 7 (5-8) < 0.01

SS.. mmaallttoopphhiilliiaa (n= 3)

Duration of MV* 24 (13-25) 11 (10-16) 0.11

* Results are shown as median [interquartile range] and in days.

ICU: Intensive care unit, MV: Mechanical ventilation.

(8)

nosa resulted in an increased length of ICU stay and in a tendency for increased duration of MV (p= 0.05 and p= 0.09, respectively). In VAP ca- used by Acinetobacter species, length of ICU and hospital stay and duration of MV increased significantly (p< 0.01, p= 0.03 and p< 0.01, respectively). VAP caused by MRSA resulted in an increased length of ICU stay and increased duration of MV (p= 0.01 and p< 0.01, respectively).

DISCUSSION

We found that VAP caused by HRM was not associated with an increased mortality. However, it resulted in an increase in length of ICU stay and hospital stay by seven days, and in duration of MV by 10 days. In addition, these patients were exposed more frequently to invasive procedures such as tracheostomy and central venous catheterization.

In our study, 43 HRM were isolated in the case patients. The most common HRM was P. aerugi- nosa, isolated in 15 patients. Other HRM were A.

baumannii, MRSA, A. lwoffii, S. maltophilia in 11, 10, four and three patients, respectively. Ex- cept A. lwoffii, all of these microorganisms we- re well known microorganisms causing VAP. A.

lwoffii (formerly A. calcoaceticus var. lwoffii) has been recognized as normal flora of the skin, oropharynx and perineum of healthy individuals.

In humans there heve been few reports of A.

lwoffii infections, nearly all of which were cent- ral intravasculer catheter-related blood straem infections or bacteremia particularly in immuno- compromised hosts (24,25). Pneumonia due to A. lwoffii was reported only in one case report which describes community-acquired pneumo- nia due to A. lwoffii in a patient infected with the human immunodeficiency virus (26). In our study, we reported four patients with VAP due to A. lwoffii. Similar with the other reports, all of the patients infected with A. lwoffii were immun- compromised (24-26). Two patients were receiving steroid and immunosupressive therapies;

one for systemic lupus erythematosus and the other one for kidney transplantation. The other two patients were receiving only steroid therapy, one for interstital lung disease and the other one

for acute exarbation of chronic obstructive pulmonary disease.

Attributable Mortality in VAP Caused By HRM Attributable mortality rate of several infectious problems, especially VAP, has been the focus of interest in various studies with conflicting results, since it is important to differentiate the mortality due to the underlying process or the infection itself (4-7,15,18,27). When VAP caused by HRM are concerned, there are few studies looking at the attributable mortality rate for these microorganisms in matched case-control studies (4-7). Similar to our study Papazian and coworkers, could not demonstrate an increased mortality caused by VAP (5). In their analysis of mortality according to the responsible microorganism, they did not show any significant difference between the mortality rate observed in pa- tients with VAP caused by P. aeruginosa and tho- se with VAP caused by other microorganisms (5). Heyland and coworkers, reported a slightly (5.8%) increased mortality in VAP but they could not demonstrate an increased mortality attribu- table to HRM, - i.e., Pseudomonas, Acinetobac- ter or Stenotrophomonas species, and MRSA (6). Contrary to our study and the two studies mentioned above, Fagon and coworkers, found a 27% attributable mortality rate of VAP, and this rate increased to 43% when Pseudomonas or Acinetobacter species were concerned (4-6). Si- milarly, Bercault and Boulain reported an attributable mortality rate of 27% in VAP and found that VAP caused by HRM was an independent cause of death (7). Contrast to our study, in most studies mentioned above, attributable mortality rate for HRM was not investigated pri- marily, but rather in subgroup analysis.

There are very few studies looking at the atribu- table mortality of HRM, separately (18,28). Rel- lo and coworkers, reported a high attributable mortality rate in VAP caused by P. aeruginosa, whereas in a recent study, Garnacho and coworkers, investigated the attributable mortality rate in VAP caused by A. baumannii and found that VAP caused by A. baumannii was not associ- ated with an increased mortality rate (18,28).

Attributable mortality rate of VAP caused by

(9)

MRSA or S. maltophilia has not been investiga- ted, yet.

Attributable Length of Stay and Duration of MV in VAP Caused By HRM

The major finding of this matched case-control study was a considerably increased length of ICU and hospital stay and duration of MV in VAP caused by HRM. Similar to the mortality issue, increase in length of stay and duration of MV attributable to VAP caused by HRM have not been studied extensively (4-7,18). Attributable increase in length of ICU stay in VAP caused by HRM varies from six to 12 days (4,6,7). Papazian and coworkers, reported an increased duration of MV by 13 days in VAP caused by P. aeruginosa, compared to VAP caused by gram-positive microorganisms (5). These results are similar to our results where we demonstrated a seven day increase in length of ICU and hospital stay and 10 day increase in duration of MV in VAP caused by HRM. In addition to these findings, we have also shown that development of VAP due to HRM is an independent risk factor for prolonged length of ICU and hospital stay, and duration of mechanical ventilation. Although increase in length of stay and duration of mechanical ventilation ha- ve been demonstrated, when VAP due to P. aeru- ginosa, Acinetobacter species and MRSA were considered separately, sample size for each microorganism was not sufficient to draw definite conclusions.

Organ Failures and Invasive Procedures in VAP Caused By HRM

There was no difference between the case and the control patients in terms of the development of organ failures in this study. Studies are lac- king showing development of organ failures in VAP caused by HRM. In a matched case-control study, Blot and coworkers, found that P. aerugi- nosa bacteremia was associated with a higher frequency of acute respiratory failure and he- modynamic instability (29). In their cohort study comparing pneumonia caused by MRSA and methicillin-sensitive S. aureus (MSSA), Rello and coworkers, also found that septic shock was more common in VAP due to MRSA than in VAP

due to MSSA (13). On the other hand, in another cohort study, comparing VAP due to MRSA with VAP due to MSSA, Gonzales and coworkers, showed that development of septic shock and organ failures during MV were similar in both groups (30). Tracheostomy was performed more frequently in the case patients than in the control patients in this study, due to the increased duration of MV in the case patients, indica- ting that VAP caused by HRM increases morbidity of the patients considerably.

Limitations of the Study

Small sample size is one limitation of this study.

This might be one of the reasons why we could not observe a difference in mortality rate between case and the control patients. Another reason for similar mortality rates in the case and the control patients might be similar frequencies of organ failures and disease severities observed in our case and the control patients. The mortality rate in our control patients is higher than that reported in various studies (4-7,15,18). APACHE II score has been used as a matching criteria in three of those studies (5,6,18). In two of those studies, mean APACHE II scores of the control group varied from 16.9 to 17.6 and the observed ICU mortality rate ranged from 28.3% to 38.8%

(5,18). The mean APACHE II score of our control patients was 20 and the ICU mortality rate was 71%. In these studies the patients were me- dico-surgical ICU patients, whereas in our study the patients were medical patients only. The mortality rate in our control patients could be considered to be slightly higher than the rate of 60% reported by Knaus, in mechanically ventilated patients with an APACHE II score ranging from 16 to 20 (31). Differences in various factors, such as patient to nurse ratio, differences in the case-mix, could lead to differences in mortality rates.

Investigating different kinds of microorganisms in the same study might considered to be another limitation of our study. But to our knowled- ge, our study is important for being the first matched case-control study investigating the attributable mortality and morbidity of VAP ca-

(10)

used by HRM which revealed that VAP caused by HRM did not increase mortality, however it increased ICU and hospital length of stay and duration of MV.

REFERENCES

1. Richards MJ, Edwards JR, Culver DH, Gaynes RP. Noso- comial infections in combined medical-surgical intensive care units in the United States. Infect Control Hosp Epi- demiol 2000; 21: 510-5.

2. Urli T, Perone A, Acquarolo S, et al. Surveillance of infec- tions acquired in intensive care: Usefulness in clinical practice. J Hosp Infect 2002; 52: 130-5.

3. Legras A, Malvy D, Quinioux AI, et al. Nosocomial infec- tions: Prospective survey of incidence in five French in- tensive care units. Intensive Care Med 1998; 24: 1040-6.

4. Fagon JY, Chastre J, Hance AJ, et al. Nosocomial pne- umonia in ventilated patients: A cohort study evaluating attributable mortality and hospital stay. Am J Med 1993;

94: 281-8.

5. Papazian L, Bregeon F, Thirion X, et al. Effect of ventila- tor-associated pneumonia on mortality and morbidity.

Am J Respir Crit Care Med 1996; 154: 91-7.

6. Heyland DK, Cook DJ, Griffith L, et al. The attributable morbidity and mortalitiy of ventilator associated pne- umonia in the critically ill patient. Am J Respir Crit Care Med 1999; 159: 1249-56.

7. Bercault N, Boulain T. Mortality rate attributable to ven- tilator-associated nosocomial pneumonia in an adult in- tensive care unit: A prospective case-control study. Crit Care Med 2001; 29: 2303-9.

8. Rosenthal VD, Guzman S, Orellano PW. Nosocomial in- fections in medical-surgical intensive care units in Ar- gantina: Attributable mortality and length of stay. Am J Infect Control 2003; 31: 291-5.

9. Hanes SD, Demirkan K, Tolley E, et al. Risk factors for la- te-onset nosocomial pneumonia caused by Stenotropho- monas maltophilia in critically ill trauma patients. Clin Infect Dis 2002; 35: 228-35.

10. Nicholson AM, Castle D, Akpaka P, et al. The emergence of Stenotrophomonas maltophilia as a significant noso- comial pathogen at the University Hospital of the West Indies. West Indian Med J 2004; 53: 17-22.

11. Torres A, Carlet J (European Task Force on ventilator-as- sociated pneumonia). Ventilator-associated pneumonia.

Eur Respir J 2001; 17: 1034-45.

12. Kollef MH, Morrow LE, Niederman MS, et al. Clinical characteristics and treatment patterns among patients with ventilator-associated pneumonia. Chest 2006; 129:

1210-8.

13. Rello J, Torres A, Ricart M, et al. Ventilator-associated pneumonia by Staphylococcus aureus. Comparison of methicillin-resistant and methicillin-sensitive episodes.

Am J Respir Crit Care Med 1994; 150(6 Pt 1): 1545-9.

14. Jimenez P, Torres A, Rodriguez-Roisin R, et al. Incidence and aetiology of pneumonia acquired during mechani- cal ventilation. Crit Care Med 1989; 17: 882-5.

15. Rello J, Ollendorf DA, Oster G, et al. VAP Outcomes Sci- entific Advisory Group. Epidemiology and outcomes of ventilator-associated pneumonia in a large US database.

Chest 2002; 122: 2115-21.

16. Warren DK, Shukla SJ, Olsen MA, et al. Outcome and attributable cost of ventilator associated pneumonia among intensive care unit patients in a suburban medi- cal center. Crit Care Med 2003; 31: 1312-7.

17. Shorr AF, Combes A, Kollef MH, Chastre J. Methicillin-re- sistant Staphylococcus aureus prolongs intensive care unit stay in ventilator-associated pneumonia, despite ini- tially appropriate antibiotic therapy. Crit Care Med 2006;

34: 700-6.

18. Garnacho J, Sole-Violan J, Sa-Borges M, et al. Clinical impact of pneumonia caused by Acinetobacter bauman- nii in intubated patients: A matched cohort study. Crit Care Med 2003; 31: 2478-82.

19. Garnacho-Montero J, Ortiz-Leyha C, Fernández-Hinojosa E, et al. Acinetobacter baumannii ventilator-associated pneumonia: Epidemiological and clinical findings. Inten- sive Care Med 2005; 31: 649-55.

20. Knaus WA, Draper EA, Wagner DP, Zimmerman JE.

APACHE II: A severity of disease classification system.

Crit Care Med 1985; 13: 818-29.

21. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale. Lancet 1974; 13: 81-4.

22. Grossman RF, Fein A. Evidence-based assessment of di- agnostic tests for ventilator-associated pneumonia: Exe- cutive summary. Chest 2000; 117 (Suppl 2): 117-81.

23. El-Ebiary M, Torres A, González J, et al. Quantitative cul- tures of endotracheal aspirates for the diagnosis of venti- lator-associated pneumonia. Am Rev Respir Dis 1993;

148(6 Pt 1): 1552-7.

24. Tega L, Raieta K, Ottaviani D, et al. Catheter-related bac- teremia and multidrug-resistant Acinetobacter lwoffii.

Emerg Infect Dis 2007; 13: 355-6.

25. Ku SC, Hsueh PR, Yang PC, et al. Clinical and microbi- ological characteristics of bacteremia caused by Acineto- bacter lwoffii. Eur J Clin Microbiol Infect Dis 2000; 19:

501-5.

26. Domingo P, Muñoz R, Frontera G, et al. Community-ac- quired pneumonia due to Acinetobacter lwoffii in a pati- ent infected with the human immunodeficiency virus.

Clin Infect Dis 1995; 20: 205-6.

(11)

27. Bregeon F, Ciais V, Carret V, et al. Is ventilator-associated pneumonia an independent risk factor for death? Anest- hesiology 2001; 94: 554-60.

28. Rello J, Juber P, Vallés J, et al. Evaluation of outcome for intubated patients with pneumonia due to Pseudomo- nas aeruginosa. Clin Infect Dis 1996; 23: 973-978.

29. Blot S, Vandewoude K, Hoste E, Colardyn F. Reappraisal of attributable mortality in critically ill patients with no- socomial bacteraemia involving Pseudomonas aerugino- sa. J Hosp Infect 2003; 53: 18-24.

30. Gonzáles C, Rubio M, Romero-Vivas J, et al. Bacteremic pneumonia due to Staphylococcus aureus: A compari- son of disease caused by methicillin-resistant and met- hicillin-susceptible organisms. Clin Infect Dis 1999; 29:

1171-7.

31. Knaus WA. Prognosis with mechanical ventilation: The influence of disease, severity of disease, age and chronic health status on survival from an acute illness. Am Rev Respir Dis 1989; 140: 8-13.