Risk factors for mortality in ventilator-associated tracheobronchitis: a case-control study

ABSTRACT Objective To describe the microbiological characteristics and to assess the risk factors for mortality of ventilator-associated tracheobronchitis in a case-control study of intensive care patients. Methods This case-control study was conducted over a 6-year period in a 40-bed medical-surgical intensive care unit in a tertiary care, private hospital in São Paulo, Brazil. Case patients were identified using the Nosocomial Infection Control Committee database. For the analysis of risk factors, matched control subjects were selected from the same institution at a 1:8.8 ratio, between January 2006 and December 2011. Results A total of 40 episodes of ventilator-associated tracheobronchitis were evaluated in 40 patients in the intensive care unit, and 354 intensive care patients who did not experience tracheobronchitis were included as the Control Group. During the 6-year study period, a total of 42 organisms were identified (polymicrobial infections were 5%) and 88.2% of all the microorganisms identified were Gram-negative. Using a logistic regression model, we found the following independent risk factors for mortality in ventilator-associated tracheobronchitis patients: Acute Physiology and Chronic Health Evaluation I score (odds ratio 1.18 per unit of score; 95%CI: 1.05-1.38; p=0.01), and duration of mechanical ventilation (odds ratio 1.09 per day of mechanical ventilation; 95%CI: 1.03-1.17; p=0.004). Conclusion Our study provided insight into the risk factors for mortality and microbiological characteristics of ventilator-associated tracheobronchitis.

ventilação mecânica em um estudo caso-controle de pacientes de terapia intensiva. Métodos: Estudo realizado ao longo de 6 anos em uma unidade de terapia intensiva médico-cirúrgica de 40 leitos, em um hospital privado e de nível terciário em São Paulo, Brasil years, much less attention was given to ventilatorassociated tracheobronchitis (VAT), especially concerning mortality risk factors. This fact may be attributed to the reluctance of many authorities to consider VAT as an independent clinical entity. (1,2)

OBJECTIVE
To describe the microbiological characteristics and the risk factors for mortality of patients with ventilatorassociated tracheobronchitis in an intensive care unit.

METHODS
This case-control study was conducted over a 6-year period in a 40-bed medical-surgical intensive care unit (ICU) with the same physical layout as that in a tertiary care, private hospital in São Paulo, Brazil. The study was reviewed and approved by the hospital Research Ethics Committee, under protocol number CAAE: 04355112.1.0000.0071.
The patients were identified using the Infection Control Committee database. For the analysis of risk factors, matched control subjects were selected from the same organization at a 1:8.8 ratio, between January 2006 and December 2011. The control subjects were matched for six characteristics: sex, age (10-year maximum difference), time of mechanical ventilation (20-day maximum difference), use of tracheostomy, use of bilevel positive pressure airway (BiPAP), and number of days in the intensive care unit (20-day maximum difference). The medical records of all case patients and control subjects were reviewed.
Ventilator-associated tracheobronchitis was classified according to the Centers for Disease Control and Prevention (CDC) definitions. (3) Microorganisms were initially identified by Microbiology Department of the organization, using quantitative cultures of tracheal aspirates Colony Forming Units (CFU) (≥1x10 6 CFU/mL) or bronchoscopic specimens (≥1x10 4 CFU/mL).
All statistical analyses were performed using the R Project for Statistical Computing, version 3.1.1. Categorical variables were analyzed using the χ² test. Continuous variables were compared using the Wilcoxon-Mann-Whitney test. Univariate analysis was conducted for potential risk factors for mortality in tracheobronchitis associated with mechanical ventilation, and variables with p value of less than 0.15 were included in the multivariate model. A 2-tailed p value of 0.05 or less was considered statistically significant. During the 6-year study period, 42 microorganisms were identified (polymicrobial infections were 5%), and 88.2% of all the microorganisms identified were Gramnegative. The etiologic agents of the VAT episodes are shown in table 2.
We found very high levels of antibiotic resistance: 78.8% of the Staphylococcus aureus isolates were methicillin-resistant; and 52.5% of Pseudomonas aeruginosa isolates, and 48.7% of Klebsiella pneumoniae isolates were carbapenem-resistant. The rate of mortality due to VAT was 42.5%. Using a conditional logistic regression model (Table 3), we found the following independent risk factors for mortality in VAT patients: Acute Physiology and Chronic Health Evaluation I (APACHE I) score (odds ratio -OR: 1.18 per unit of score; 95% confidence interval -95%CI: 1.05-1.38; p=0.01), and duration of mechanical ventilation (OR: 1.09 per day of mechanical ventilation; 95%CI: 1.03-1.17; p=0.004).
However, there is controversy in the literature regarding whether increased mortality is associated with VAT. (6) The crude mortality in the Case Group was 42.5% similar from that found in the literature, which ranged from 21 to 55%. (6,7) The APACHE I score and duration of mechanical ventilation were the independent risk factors for mortality in our study, as already demonstrated in other studies on VAP. (8) In regard to APACHE I scores, for every increase by one unit in the score, there was an increase by 18% in the risk of death. As to duration of mechanical ventilation, for every increase by 1 day in mechanical ventilation, there was an increase by 9% in the risk of death. Information is lacking regarding the risk factors for mortality in VAT, perhaps because the subjective components within the VAT definition and diagnosis may impact the reliability and accuracy of case identification. In addition, the same risk factors for mortality in VAT and VAP could reinforce the hypothesis that VAT is an intermediate stage between colonization of the upper airways and VAP. (9) Regarding the etiologic agents of VAT, Gramnegative pathogens were the most common cause in many studies, accounting for more than 60% of isolates, similar to that found in our study (88.2%). Over the past 5 years, the incidence of VAT caused by multidrug resistant pathogens, such as carbapenem-resistant Gram-negative bacilli (e.g., P. aeruginosa, Acinetobacter baumannii, and K. pneumoniae) and methicillin-resistant S. aureus, has increased. (10) This study has notable limitations. First, we cannot exclude with certainty the possibility that some of our patients in the Case Group were misclassified, because we did not routinely perform lung computed tomography scans searching for occult infiltrates. Second, this is a retrospective observational study, and other unmeasured factors might have occurred coincidentally to the period. Third, data from a single unit and the small sample size limit the generalizability of findings.

CONCLUSION
Our study provided insight into the risk factors for mortality and microbiological characteristics of ventilator-associated tracheobronchitis over a 6-year study. Further studies on ventilator-associated tracheobronchitis, including risk factors for mortality, are necessary to define the best practices.

ACKNOWLEDGMENTS
We thank the Laboratory at Hospital Israelita Albert Einstein for assistance in obtaining microbiological data.

DISCUSSION
Ventilator-associated tracheobronchitis is recognized as a frequent complication of mechanical ventilation with rates ranging from 3.7 to 11.5%, according to the literature. (4,5) In addition, more recent data suggest that VAT may contribute to the need for longer stays at the intensive care unit and the need for a longer duration of mechanical ventilation, as demonstrated in our study.