Secondary Logo

A Comparison of Outcomes of Trauma Patients With Ventilator-Associated Events by Diagnostic Criteria Set

Younan, Duraid*; Griffin, Russell; Zaky, Ahmed; Pittet, Jean-Francois; Camins, Bernard§

doi: 10.1097/SHK.0000000000001214
Clinical Science Aspects
Editor's Choice

Background: The Centers for Disease Control and Prevention replaced the definition for ventilator-associated pneumonia with an algorithm comprised of three categories: ventilator-associated condition (VAC), infection-related ventilator associated complication (IVAC), and possible ventilator-associated pneumonia (PVAP). We sought to compare the outcome of trauma patients with VAEs to those with no VAEs.

Methods: Patients admitted from 2013 to 2017 were identified from trauma registry. Logistic regression was performed for the association between VAEs and mortality.

Results: Two thousand six hundred eighty patients were admitted to our trauma center, 2,290 had no VAE, 100 had VACs, 85 had IVACs, and 205 had PVAPs. Adjusted for race, sex, blunt injury mechanisms, and Injury Severity Score, all VAEs had a longer hospital length of stay, intensive care unit stay, and days of ventilator support when compared with those with no VAE (all P < 0.0001). Nosocomial complication rates were not different by VAE group. Compared with patients with no VAE, an over 2-fold increased mortality odds was observed for VAC (OR 2.39, 95% CI 1.50–3.80) and IVAC patients (OR 2.07, 95% CI 1.23–3.47), and a 50% mortality increased was observed for PVAP patients (OR 1.46, 95% CI 1.00–2.12). These associations became similar with an approximate 2.5-fold increased mortality odds among patients with at least 1 week on ventilator support.

Conclusion: VAEs increase the odds of mortality, particularly for patients with VACs and IVACs. Among patients on ventilator support for at least a week, the associations are similar among VAE types, suggesting no single VAE type is more severe than others.

*Division of Acute Care Surgery, Department of Surgery, University of Alabama at Birmingham, Birmingham, Alabama

Department of Epidemiology, University of Alabama at Birmingham, Birmingham, Alabama

Department of Anesthesiology and Perioperative Medicine, University of Alabama at Birmingham, Birmingham, Alabama

§Division of Infectious Diseases, Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama

Address reprint requests to Duraid Younan, MD, Division of Acute Care Surgery, University of Alabama at Birmingham, 701 19th Street S, LHRB #112, Birmingham, AL 35294. E-mail:;

Received 13 February, 2018

Revised 28 February, 2018

Accepted 22 June, 2018

DY: study design, data collection and interpretation, manuscript drafting, critical review.

RG: data analysis and interpretation, manuscript drafting, critical review.

AZ: interpretation of data, critical review.

J-FP: interpretation of data, critical review.

BC: study design, interpretation of data, manuscript drafting, and critical review.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

The authors report no proprietary or commercial interest in any product mentioned or concept discussed in this article.

The authors report no conflicts of interest.

Back to Top | Article Outline


Ventilator-associated pneumonia (VAP) is the most common nosocomial infection in patients with trauma requiring mechanical ventilation (1). As such, VAP is associated with prolonged hospital and intensive care unit (ICU) lengths of stay, days on mechanical ventilation, and an attributable morality of approximately 13% (2–4). As a result of its mortality and morbidity, VAP preventive bundles have been implemented and promoted and have been used to assess quality of care in mechanically ventilated patients (5).

In an attempt to increase the reliability, reproducibility, and inclusiveness of defining VAP, the Centers for Disease Control and Prevention (CDC) National Healthcare Safety Network (NHSN) has recently published a new surveillance algorithm for ventilator-associated events (6, 7). This algorithm mainly focuses on identifying mechanically ventilated patients who have worsening respiratory status resulting from various causes on the basis of changes in support, namely positive end-expiatory pressure (PEEP) and/or fraction of inspired oxygen (FiO2). The term infection-related ventilator-associated complications (IVACs) is applied when a ventilator-associated event is associated with leukocytosis/leukopenia or hypothermia/hyperthermia plus administration of new antimicrobial agents. Furthermore, the terms possible and probable VAP are applied to patients if there is evidence of pulmonary infection defined as a positive culture of lower respiratory tract specimens.

Despite initial encouraging reports (8, 9), recent data suggest poor correlation between IVAC and VAP diagnosed by the original definition in trauma population (10). This may be attributable to the fact that the recent CDC definition of IVAC does not take into account acutely deteriorating patients with no prior period of respiratory stability, patients with radiological signs of pneumonia and those on mechanical ventilatory modes not utilizing PEEP (9, 11).

In an attempt to further study the new ventilator-associated events (VAE) CDC definition in the trauma population, we sought to compare outcomes of patients with trauma with to those without VAE.

Back to Top | Article Outline


Study design and variable definitions

The study was approved by the University of Alabama at Birmingham Institutional Review Board. The study population included patients aged 18 years or older admitted to a tertiary care, academic Level-I trauma center between July 2013 to December 2015 and who had at least 3 days of mechanical ventilatory support.

For each patient, information regarding demographics (e.g., age and gender), injury, and clinical characteristics. Injury characteristics included injury mechanism (i.e., blunt or penetrating), Glasgow Coma Scale score at admission, and Injury Severity Score (ISS). Clinical data including respiratory rate and temperature at admission, incidence of nosocomial complications, hospital length of stay (LOS) in days, ICU length of stay in days, and ventilator support days were collected from the trauma registry. Other clinical data including red blood cell transfusions in the first 24 h following admission, ICD diagnosis codes (to identify comorbidities), PEEP, FiO2, white blood cell count (WBC), hematocrit, hemoglobin, platelet count, serum creatinine, blood urea nitrogen (BUN), glomerular flitration rate (GFR), temperature, antimicrobial agent administered and culture data (i.e., organism count, bacterial genus) were obtained from electronic medical records. The clinical data points used in the VAE algorithm were derived from directly querying the electronic medical record. For lab values of interest (i.e., BUN, creatinine, GFR, hematocrit, hemoglobin, platelet count, and WBC count), minimum and maximum values for the first 24-h from admission were calculated. Further, GFRs above 60 mL/min/1.73 m2 were documented in electronic medical records as “>60”; as a result, GFR was categorized as >60 (i.e., normal kidney function), 15 to 59 (i.e., kidney disease), and <15 (kidney failure).

VAEs as defined by the NHSN (12) was categorized as ventilator-associated condition (VAC), IVAC, and possible ventilator-associated pneumonia (PVAP). Patients categorized as VAC had a baseline period of 2 days of stable or improved FiO2 or PEEP while ventilated followed by an increased in FiO2 of greater than or equal to 0.2 over the daily minimum FiO2 in the baseline period that was sustained for 2 or more days or an increase in the daily minimum PEEP values of ≥3 cm H2O over the daily minimum in the baseline period sustained for 2 or more days as well. Those categorized as IVAC had to meet the criteria for VAC as well as have a core temperature >38°C or <36°C or had a WBC ≥ 12,000 or ≤4,000 cells/mm3, and were administered eligible antimicrobial agents that were continued for more than 3 days. For PVAP, in addition to meeting the criteria for VAC and IVAC, patients had to have >100,000 colony forming units (CFUs) for tracheal aspirates and >10,000 CFU of microorganism on cultures of broncho-alveolar lavage specimens.

The following were abstracted from the medical records: the infectious agents, the microbiological count on gram stain and/or culture, density of WBC on gram stain, and the antimicrobial agent administered if present. The NHSN criteria were also validated for the various VAE classifications (Fig. 1).

Fig. 1

Fig. 1

Back to Top | Article Outline

Statistical analysis

Demographic, injury, clinical characteristics, comorbidity, and nosocomial complication characteristics were compared among the No VAE, VAC, IVAC, and PVAP groups using a Kruskal–Wallis and Pearson chi-square test for continuous and categorical variables, respectively. A generalized linear model was used to estimate adjusted mean differences among the VAE groups for hospital LOS, ICU days, and ventilator support days adjusted for race, sex, blunt injury mechanisms, and ISS with pairwise comparisons among VAE groups using Tukey test. A logistic regression model adjusted for race, sex, blunt injury mechanism, and ISS was used to estimate odds ratios and associated 95% confidence intervals for the association between VAE group and in-hospital mortality. In a sensitivity analysis to examine the effects of the length of ventilator support on the observed mortality associations, separate logistic models were limited to patients with at least 3, 7, and 14 days on ventilator support. All tests of significance were 2-tailed, and a P value of <0.05 was considered statistically significant. Statistical analyses were performed using SAS version 9.4 (Cary, NC).

Back to Top | Article Outline


From 2013 to 2017, 2,680 patients were admitted to our Level 1 trauma center had at least 2 days of ventilator support, of whom 2,290 (85.4%) had no VAE, 100 (3.7%) met the definition for VAC, 85 (3.2%) met the criteria for IVAC, and 205 (7.6%) met the criteria for PVAP. Those with PVAP were more likely to be white (P = 0.0024) and male (P = 0.0103) (Table 1). In addition, those with PVAP were most likely to have a blunt injury mechanism (P = 0.0060), and had the highest mean hospital length of stay (P < 0.0001), ICU days (P < 0.0001), and ventilator support days (P < 0.0001). Those with no VAE were less likely to have comorbid conditions including myocardial infarction (P = 0.0009), peptic ulcer disease (P = 0.0022), and diabetes without chronic complications (P = 0.0007) (Table 2). Within the first 24 h from admission, patients with VAC and IVAC had higher minimum and maximum BUN values (P < 0.0001) and maximum serum creatinine values (P = 0.0022) (Table 3). In addition, those with IVAC had lower minimum platelet count (P = 0.0467), and those with IVAC or PVAP were less likely to have normal kidney function as measured by the GFR (P = 0.0100). Those with IVAC were more likely to have HIV (P = 0.0451). The incidence of nosocomial complications was lowest among those with no VAE for all selected complications (all P < 0.0001) (Table 4). Notably, VAC patients had the highest rate of cardiac arrest (14.0%); IVAC had the highest rates of ARDS (21.2%), pulmonary embolism (8.2%), and sepsis (10.6%); and PVAP patients had the highest rates of acute kidney injury (13.7%) and deep vein thrombosis (15.6%).

Table 1

Table 1

Table 2

Table 2

Table 3

Table 3

Table 4

Table 4

Adjusted for race, sex, blunt injury mechanism, and ISS, patients with no VAE had lower hospital lengths of stay compared with VAC, IVAC, and PVAP patients (all P < 0.0001) (Table 5). Patients with PVAP had higher adjusted lengths of stay compared with VAC patients (mean difference 9.3 days, P = 0.0003) but not IVAC patients (mean difference 5.1 days, P = 0.1576). A similar pattern was observed for ICU length of stay, with all VAEs associated with a higher adjusted ICU stay difference (all P < 0.0001) and longer stay for PVAP compared with VAC patients (patients (mean difference 8.5 days, P < 0.0001); however, those with PVAP also had longer lengths of ICU stay compared with IVAC patients (mean difference 6.3 days, P = 0.0022). For ventilator support days, all VAEs again had a higher number of days compared with patients with no VAE (all P < 0.0001), and those with PVAP had longer days on ventilator support compared with VAC (mean difference 8.6 days, P < 0.0001) and IVAC patients (mean difference 4.1 days, P = 0.0419).

Table 5

Table 5

For the cohort overall—compared with patients with no VAE and adjusting for race, sex, blunt injury mechanism, and ISS—the odds of mortality was 2-fold higher for those with VAC (OR 2.39, 95% CI 1.50–3.80), IVAC (OR 2.07, 95% CI 1.23–3.47), and PVAP (OR 1.46, 95% CI 1.00–2.12) (Table 6). There was no significant difference in the odds of mortality among VAC, IVAC, and PVAP patients. In the sensitivity analysis by time on ventilator support, among patients with at least 3 days of mechanical ventilation, the associations became stronger for all VAEs with a 3-fold observed increased in the likelihood of mortality for those with VAC (OR 3.02, 95% CI 1.87–4.89), a near 3-fold increase for those with IVAC (OR 2.78, 95% CI 1.65–4.67), and a 2-fold increase for those with PVAP (OR 1.93, 95% CI 1.32–2.83). These associations were similar for those on the ventilator at least 7 days and at least 2 weeks, though, of note, the association for mortality for PVAP compared with VAC and IVAC, though not statistically significant across all sensitivity groups, was estimated to be between a 30% to 40% reduction through day three of ventilation, but was close to a null association by day 7.

Table 6

Table 6

Back to Top | Article Outline


We report that trauma patients meeting the criteria for VAE (VAC, IVAC, and PVAP) had almost a 2-fold increase in mortality compared with those with no VAE by day 3 of mechanical ventilation; they also had longer hospital length of stay, ICU length of stay and ventilator support compared with those with no ventilator associated events. To the best of our knowledge, this is first study to address outcomes of VAE in the trauma population and highlight the fact that early interventions before the progression to VAE may improve outcomes in this population.

This study is in trauma patients where traumatic brain injury and chest injury have been associated with an increased risk of ventilator-associated events. The incidence of VAEs in this study (14.5%) is comparable to published reports (13–15), even though the incidence of PVAPs (7.6%) is substantially higher. While Klompas et al. (16) found the incidence rate of VAP of 1.3% among 20,356 episodes of mechanical ventilation in a group of medical, surgical, and neuroscience patients, the study group was heterogeneous and not limited to trauma patients. Our patients had longer days on the ventilator across all patients regardless of VAE status, when compared with Klompas’ study, were severely injured (median ISS is 20 among No VAEs and 24 in PVAP) and have a different pathology, which can in part explain the difference in incidence of PVAP. This is a single-center study with a largely blunt trauma population which could have affected these results.

There was a significant difference in hospital stay, ICU days, and ventilator days in our study between patients with VAEs (VAC, IVAC, and PVAP) and those with No VAE. This is in agreement with published reports confirming the worse outcomes associated with VAEs. (16, 17) There was no difference in these outcomes between those with VAC and IVAC. Although Zhu et al. (18) in a prospective multicenter study conducted in medical-surgical ICUs, demonstrated longer duration of mechanical ventilation and ICU stay in patients with IVAC compared with those VACs alone, the study included a different patient population and was not limited to trauma patients. Also of note in this study, patients with PVAP had longer hospital, ventilator, and ICU days compared with those with VAC, which is in agreement with published literature (16).

In this study, patients with VAEs also had twice the mortality of those with no VAEs. While VAPs have a reported associated mortality in the range of 15% to 50% in critically ill patients (19), the group of patients with PVAP had almost twice the odds of death compared of those with no VAE. Certain authors showed that VACs are associated with increased mortality (9); Kobayashi et al. (20) in a single-center retrospective study found patients with VAC to have a higher mortality when compared with those with no VAEs and Muscedere et al. (17), in another study conducted in 11 ICUs in North America observed that patients with IVAC had a trend toward higher mortality compared with those with no IVAC. We did not find a difference in mortality between patients with VAC, IVAC, and those with PVAP. Our data are in agreement with these published studies (16).

Our study should be appraised in then light of its limitations. First, the clinical practice during the duration of the study was based on the older definition of VAP and hence it is possible that some VAEs might have been misclassified if cultures or antibiotics were not ordered despite the apparent need for either. Additionally, the reported PEEP and FIO2 may as well be taken during a spontaneous breathing trial and rather may not reflect actual ventilator settings. Second, this study was performed only at a single institution, decreasing the generalizability of our findings. As the only Level-I trauma center in the region, our institution receives a greater proportion of sicker (i.e., more severely injured) patients; thus, it is possible that these results overestimate the true association between VAE and adverse clinical outcome. Third, the definition of PVAP was designed for surveillance rather than retrospective research so this may lead to misclassification bias. This is a single-center study with a largely blunt trauma population.

Back to Top | Article Outline


VAEs increase the odds of mortality, particularly for patients with VACs and IVACs. Among patients on ventilator support for at least a week, the associations are similar among VAE types, suggesting no single VAE type is more severe than others. Future research should investigate whether prevention protocols for VAEs are needed in this population of critically injured patients.

Back to Top | Article Outline


1. Cohen MJ, Bir N, Rahn P, Dotson R, Brohi K, Chesebro BB, Mackersie R, Carles M, Wiener-Kronish J, Pittet JF. Protein C depletion early after trauma increases the risk of ventilator-associated pneumonia. J Trauma 67 6:1176–1181, 2009.
2. Melsen WG, Rovers MM, Groenwold RH, Bergmans DC, Camus C, Bauer TT, Hanisch EW, Klarin B, Koeman M, Krueger WA, et al. Attributable mortality of ventilator-associated pneumonia: a meta-analysis of individual patient data from randomised prevention studies. Lancet Infect Dis 13 8:665–671, 2013.
3. Siempos Ii, Vardakas KZ, Kyriakopoulos CE, Ntaidou TK, Falagas ME. Predictors of mortality in adult patients with ventilator-associated pneumonia: a meta-analysis. Shock 33 6:590–601, 2010.
4. Siempos Ii, Athanassa Z, Falagas ME. Frequency and predictors of ventilator-associated pneumonia recurrence: a meta-analysis. Shock 30 5:487–495, 2008.
5. Klompas M, Berra L. Should ventilator-associated events become a quality indicator for ICUs? Respir Care 61 6:723–736, 2016.
6. Klompas M. Ventilator-associated conditions versus ventilator-associated pneumonia: different by design. Curr Infect Dis Rep 16 10:430, 2014.
7. Cdc: Available at: Vol. 2016. Accessed December 5, 2017.
8. Stevens JP, Silva G, Gillis J, Novack V, Talmor D, Klompas M, Howell MD. Automated surveillance for ventilator-associated events. Chest 146 6:1612–1618, 2014.
9. Boyer AF, Schoenberg N, Babcock H, Mcmullen KM, Micek ST, Kollef MH. A prospective evaluation of ventilator-associated conditions and infection-related ventilator-associated conditions. Chest 147 1:68–81, 2015.
10. Piriyapatsom A, Lin H, Pirrone M, De Pascale G, Corona De Lapuerta J, Bittner EA, Schmidt UH, De Moya M, Berra L. Evaluation of the infection-related ventilator-associated events algorithm for ventilator-associated pneumonia surveillance in a trauma population. Respir Care 61 3:269–276, 2016.
11. Kerlin MP, Trick WE, Anderson DJ, Babcock HM, Lautenbach E, Gueret R, Klompas M. Interrater reliability of surveillance for ventilator-associated events and pneumonia. Infect Control Hosp Epidemiol 38 2:172–178, 2017.
12. Magill SS, Klompas M, Balk R, Burns SM, Deutschman CS, Diekema D, Fridkin S, Greene L, Guh A, Gutterman D, et al. Executive summary: developing a new, national approach to surveillance for ventilator-associated events. Ann Am Thorac Soc 10 6:S220–S223, 2013.
13. Selvan K, Edriss H, Sigler M, Nugent KM. Complications and resource utilization associated with mechanical ventilation in a medical intensive care unit in 2013. J Intensive Care Med 32:146–150, 2015.
14. Chang HC, Chen CM, Kung SC, Wang CM, Liu WL, Lai CC. Differences between novel and conventional surveillance paradigms of ventilator-associated pneumonia. Am J Infect Control 43 2:133–136, 2015.
15. Klein Klouwenberg PM, Van Mourik MS, Ong DS, Horn J, Schultz MJ, Cremer OL, Bonten MJ, Consortium M. Electronic implementation of a novel surveillance paradigm for ventilator-associated events. Feasibility and validation. Am J Respir Crit Care Med 189 8:947–955, 2014.
16. Klompas M, Kleinman K, Murphy MV. Descriptive epidemiology and attributable morbidity of ventilator-associated events. Infect Control Hosp Epidemiol 35 5:502–510, 2014.
17. Muscedere J, Sinuff T, Heyland DK, Dodek PM, Keenan SP, Wood G, Jiang X, Day AG, Laporta D, Klompas M, et al. The clinical impact and preventability of ventilator-associated conditions in critically ill patients who are mechanically ventilated. Chest 144 5:1453–1460, 2013.
18. Zhu S, Cai L, Ma C, Zeng H, Guo H, Mao X, Zeng C, Li X, Zhao H, Liu Y, et al. The clinical impact of ventilator-associated events: a prospective multi-center surveillance study. Infect Control Hosp Epidemiol 36 12:1388–1395, 2015.
19. Blot S, Koulenti D, Dimopoulos G, Martin C, Komnos A, Krueger WA, Spina G, Armaganidis A, Rello J, Investigators E-VS. Prevalence, risk factors, and mortality for ventilator-associated pneumonia in middle-aged, old, and very old critically ill patients. Crit Care Med 42 3:601–609, 2014.
20. Kobayashi H, Uchino S, Takinami M, Uezono S. The impact of ventilator-associated events in critically ill subjects with prolonged mechanical ventilation. Respir Care 62:1376–1386, 2017.

Infection-related ventilator-associated complication; outcome; trauma; ventilator-associated condition

© 2019 by the Shock Society