- Email: [email protected]
Burns, inhalation injury and ventilator-associated pneumonia: Value of routine surveillance cultures
burns 38 (2012) 364–370
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/burns
Burns, inhalation injury and ventilator-associated pneumonia: Value of routine surveillance cultures Nele Brusselaers a,b,1,*, Dennis Logie b,1, Dirk Vogelaers a,b, Stan Monstrey b,c, Stijn Blot a,b,d a
General Internal Medicine & Infectious Diseases, Ghent University Hospital, De Pintelaan 185, Ghent 9000, Belgium Faculty of Medicine and Health Sciences, Ghent University, De Pintelaan 185, Ghent 9000, Belgium c Burn Unit, Ghent University Hospital, De Pintelaan 185, Ghent 9000, Belgium d Dept. of Healthcare, University College Ghent, Keramiekstraat 80, Ghent 9000, Belgium b
article info
abstract
Article history:
Purpose: Burn patients with inhalation injury are at particular risk for ventilator-associated
Accepted 1 September 2011
pneumonia (VAP). Routine endotracheal surveillance cultures may provide information
Keywords:
tive was to assess the incidence of VAP in burn patients with inhalation injury, and the
about the causative pathogen in subsequent VAP, improving antibiotic therapy. Our objecBurns
benefit of routine surveillance cultures to predict multidrug resistant (MDR) pathogens.
Ventilator-associated pneumonia
Procedures: Historical cohort (n = 53) including all burn patients with inhalation injury
Inhalation injury
requiring mechanical ventilation, admitted to the Ghent burn unit (2002–2010).
Endotracheal surveillance cultures
Main findings: Median (interquartile range) age and total burned surface area were 44y (39–
Predictive value
55y) and 35% (19–50%). Overall, 70 episodes of VAP occurred in 46 patients (86.8%). Median
Epidemiology
mechanical ventilation days (MVD) prior to VAP onset were 7d (4–9d). The incidence was 55 episodes/1000 MVD. In 23 episodes (32.9%) at least one MDR causative pathogen was involved, mostly Pseudomonas aeruginosa and Enterobacter spp. The sensitivity and specificity of surveillance cultures to predict MDR etiology in subsequent VAP was respectively 83.0% and 96.2%. The positive and negative predictive value was 87.0% and 95.0%, respectively. Conclusions: The incidence of VAP in burn patients with inhalation injury is high. In this cohort routine surveillance cultures had excellent operating characteristics to predict MDR pathogen involvement. # 2011 Elsevier Ltd and ISBI. All rights reserved.
1.
Introduction
Depending on the diagnostic criteria used, inhalation injury is reported in 0.3–43% of the patients with severe burn [1–3]. It is one of the major (independent) risk factors for mortality, as it is associated with up to 8–10 fold increased mortality (or 15–45%) [1–6]. Because of severe respiratory distress or because patent airway is compromised by laryngeal edema, endotracheal
intubation and mechanical ventilation may be necessary in severe inhalation injury, often for a prolonged period until resolution of edema. Due to their immunocompromised status, patients with severe burn are particularly prone to infections [7,8]. As all mechanically ventilated critically ill patients, burn patients are at risk for ventilator-associated pneumonia (VAP), a severe infection that contributes to an increase in morbidity with excess length of hospitalization, and increased mortality
* Corresponding author at: General Internal Medicine & Infectious Diseases, Ghent University Hospital, De Pintelaan 185 - 9000 Ghent, Belgium. E-mail addresses: [email protected] (N. Brusselaers), [email protected] (D. Logie), [email protected] (D. Vogelaers), [email protected] (S. Monstrey), [email protected] (S. Blot). 1 These authors are joint first authors. 0305-4179/$36.00 # 2011 Elsevier Ltd and ISBI. All rights reserved. doi:10.1016/j.burns.2011.09.005
365
11% 55
19% 57 31 (48) n.r.
37% 37% 40% n.r. n.r.
n.r. n.r.
19% 38% 9% 65% 77% 52% 48% 73% 27% (68%) 27% 37% 19.7% 87%
n.r., not reported; VAP, ventilator-associated pneumonia; MV, mechanical ventilation.
Retrospective cohort Brusselaers et al. (present study)
VAP
History + clinical diagnosis In mechanically ventilated pts also bronchoscopy History + clinical diagnosis (+bronchoscopy) + necessitating MV Pneumonia Retrospective cohort Edelman et al. [12]
Pneumonia Prospective cohort de la Cal et al. [27]
VAP Retrospective cohort Rue et al. [28]
Incidence of pneumonia (/1000 ventilation days) Prevalence of pneumonia Population
Burns (n = 1058): With inhalation injury (n = 373); Without inhalation injury (n = 685) Burns + MV (n = 370): With inhalation injury (n = 265); Without inhalation injury (n = 105) Burns (n = 56): With inhalation injury (n = 26); Without inhalation injury (n = 30) With MV: n = 41) Inhalation injury (n = 117): With burns (n = 51); Without burns (n = 66) Burns + inhalation injury + mechanical ventilation (n = 53) History + clinical diagnosis + bronchoscopy or ventilation/ perfusion scan History + clinical diagnosis + bronchoscopy or ventilation/ perfusion scan History + clinical diagnosis + bronchoscopy Pneumonia Retrospective cohort Shirani et al. [11]
We performed a retrospective cohort study (January 2002– March 2010) in the 6-bed burn unit at Ghent University Hospital. The unit serves a geographic area of about 2.6 million inhabitants. Approximately 60–80 patients are admitted to the unit each year [4]. The standard treatment protocol of the burn unit is described elsewhere [4,22]. This historical cohort study was approved by the Ethics Committee at Ghent University Hospital. Patients or their proxies were informed with an opting out possibility. We retrospectively analyzed the data of all burned patients admitted to our burn unit over an 8-year period (January 2002– March 2010). Patients admitted to the burn unit for other indications were excluded from the study (e.g. degloving injury, Lyell syndrome, Steven-Johnson syndrome). The appropriate patients were identified through the diagnosis related groups (DRGs) data (inhalation injury = admission criteria), and the prospectively collected information registered in the log book (describing characteristics on admission
Definition of inhalation injury
Study design and population
Pneumonia/ VAP
2.1.
Study design
Materials and methods
Author (year)
2.
Table 1 – Studies describing (ventilator-associated) pneumonia in populations with severe burn.
[9,10]. It can be presumed that the presence of an inhalation injury contributes to an even higher risk of VAP. Inhaled products of combustion cause de-epithelialisation in the tracheobronchial tree and lower respiratory tract lesions. This process results in an increase in extravascular lung water with decreased pulmonary compliance, inactivation of surfactant with micro-atelectasis, and pseudomembrane formation of mucus, cellular debris, fibrinous exudates, polymorphonuclear leucocytes, and clumps of bacteria. These factors contribute to an increased risk of pneumonia. However, data about the incidence of pneumonia, and VAP in particular, in burn patients with inhalation injury are scarce and difficult to compare due to different inclusion criteria and definitions (Table 1) [7,11,12]. Once pneumonia occurs, early adequate antibiotic therapy is essential to optimize the chance of survival [8,13]. Inappropriate empiric therapy, not covering the causative organism, has been associated with an increased mortality rate [14–16]. Routine endotracheal surveillance cultures (SC) have been advocated to steer empiric therapy [17,18] since they provide data about pathogens colonizing the lower respiratory tract and their susceptibility patterns [19–21]. Especially the prediction of multidrug resistant (MDR) pathogens is believed to be beneficial in improving early antibiotic therapy, since the results of the diagnostic cultures are only available 24–48 h after VAP onset, and the MDR pathogens may not be covered by the standard empiric therapy. In general ICU patients endotracheal SC have a high negative predictive value in VAP with regard to multidrug resistant pathogens [18]. The objectives of this study were to estimate the incidence of VAP in a cohort of severely burned patients with inhalation injury necessitating endotracheal intubation and mechanical ventilation, and to assess the value of routine endotracheal surveillance cultures in the prediction of MDR pathogens in VAP.
Mortality in pneumonia
burns 38 (2012) 364–370
366
burns 38 (2012) 364–370
and outcome data). The electronic and paper files were also consulted for all patients with flame burns of the face and upper limbs receiving mechanical ventilation. All consecutive patients with a severe inhalation injury, confirmed by laryngoscopy or bronchoscopy, and necessitating mechanical ventilation, were included. The following data were registered for each patient: age, total burned surface area (TBSA), mortality, microbiology, length of mechanical ventilation and length of stay, time of onset of VAP and outcome. Burn severity for this cohort was assessed by the Belgian Outcome in Burn Injury (BOBI) score, as well as the presence of acute kidney injury and need for vasopressive support [1,22,23]. This mathematical BOBI-model estimates the risk of mortality based on the three major risk factors for mortality in severe burns: age, total burned surface area (TBSA) and presence or absence of inhalation injury [1]. It divides age and TBSA in respectively 4 (0–3 points) and 5 (0–4 points) risk categories, and if present, inhalation injury scores three additional points; resulting in a score ranging 0–10 points, correlating with an increasing risk of mortality. The minimal score of patients in this cohort is therefore 3 points, since only patients with an inhalation injury are included.
2.2.
Definitions
VAP was suspected in patients receiving at least 48 h of mechanical ventilation, when a new or new or progressive infiltrate was identified on sequential daily chest X-ray, and when two of the following four criteria were met: fever (T8 > 38 8C) or hypothermia (T8 < 36 8C), leucocytosis (>10 109 cells/L) or neutropenia (<5 109 cells/L), purulent secretions, and gas exchange impairment affecting the partial pressure of oxygen and/or the fraction of inspired oxygen, together with a clinical pulmonary infection score (CPIS) 6 [24]. Since adult respiratory distress syndrome (ARDS) may be provoked by either inhalation injury or hypovolemic burn shock, the presence of ARDS was not considered in the calculation of the CPIS. For this study, the routine surveillance and diagnostic techniques of our intensive care units were used. Routine endotracheal aspirate (ETA) ‘surveillance’ cultures were performed thrice a week (Monday–Wednesday–Friday). Microbiological confirmation of VAP required ++ or +++ semiquantitative growth of a pathogen in a good quality endotracheal aspirate (‘diagnostic’ ETA) [25]. Coagulase-negative Staphylococcus and Candida species were considered non-pathogenic. Additionally, clinically likely VAP with limited (+) growth on endotracheal aspirate was considered as microbiologically confirmed if Gramstaining showed predominant presence of a pathogen and antibiotic therapy had been started or changed within 72 h. VAP was considered polymicrobial if more than one pathogen was shown, with growth above these thresholds. Late-onset VAP was defined as VAP diagnosed >5 days following intubation.
2.3.
Microbiology and antibiotic treatment
For each VAP episode, all available microbiologic specimen results were retrospectively reviewed and recorded. Routine surveillance included a microbiological analysis of thriceweekly endotracheal aspirates in all mechanically ventilated
patients. For convenience sake, only the most recent endotracheal aspirate cultures sampled before clinical diagnosis of VAP were taken into account. The following pathogens were considered as MDR: methicillin-resistant Staphylococcus aureus (MRSA), extended spectrum b-lactamase (ESBL) producing Enterobacteriaceae, MDR nonfermenting organisms such as Stenotrophomonas maltophilia, Acinetobacter baumannii and Pseudomonas aeruginosa resistant for at least one of the following antipseudomonal antibiotics: ceftazidime, ciprofloxacin, piperacillin-tazobactam, and imipenem [26]. Antimicrobial therapy was considered appropriate if it included at least one antimicrobial drug with in vitro activity against the causative pathogen.
2.4.
Statistical analysis
Statistical analyses are executed with PASW Statistics 18.0 (Chicago, Ill., USA). All used tests were two-tailed and statistical significance was defined as p < 0.05. The data are expressed as median (interquartile range [IQR]) or as n (%). Differences between groups were analyzed with Mann– Whitney U test and Chi square tests. The incidence of VAP is expressed as the number of episodes per total number of ventilation days, and per total number of ventilation days ‘at risk’, including only ventilation days before the onset of the VAP episode. The value of SC to predict MDR pathogens in VAP was expressed by the following operating characteristics (or ‘predictive parameters’): sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) (and respective 95% confidence intervals -CI). Results are reported as the proportion of MDR-VAP episodes on the total number of VAP episodes, and as the proportion of MDR microorganisms on total number of microorganisms recovered both from the SC cultures as the diagnostic VAP cultures. Microorganisms that were found both in the SC and in the diagnostic ETA sample (with development of VAP) were defined as ‘true positive’. When either found in the SC or in the diagnostic ETA, MDR microorganisms were respectively defined as ‘false positive’ and ‘false negative’. All other ‘non-MDR’ microorganisms were considered true negative. For subgroupanalyses, e.g. Pseudomonas spp., the other MDR microorganisms (e.g. MDR Enterobacteriaceae) were considered ‘true negative’.
3.
Results
During the study period 59 burn patients with inhalation injury were retrieved, of which 2 patients did not receive mechanical ventilation, and 4 patients were excluded because necessary patient information could not be retrieved. Consequently, 53 patients with burn and inhalation injury necessitating mechanical ventilation were included in the study (Table 2). Only one patient was not ventilated from the day of admission on (start mechanical ventilation on day 4). Vasopressive support was necessary in 6/53 (73.6%) patients, and acute kidney injury occurred in 6/53 patients (11.3%). The median (IQR) total BOBI score of this cohort of mechanically ventilated patients, was 5 (4–6), with a predicted mortality rate
367
burns 38 (2012) 364–370
Table 2 – Main characteristics of the total cohort of patients with severe burn and inhalation injury (n = 53). Variablea Age (years) Total burned surface area (%) Male gender Mechanical ventilation (days) Length of burn unit stay (days)
All patients (n = 53) 44.0 (38.5–54.5) 35.0 (18.0–50.0) 33/53 (62.3) 17.0 (7.0–30.0) 34.0 (18.0–49.0)
VAP, ventilator associated pneumonia. a Values are described as median (interquartile range) or n (%).
of 30% (20–50%), and an observed mortality of 17% (n = 9). The main characteristics of the overall population are reported in Table 2. Overall, 70 episodes of VAP occurred in 46 of the 53 patients (86.8%). Sixteen patients had two VAP episodes, and four patients three. Median (IQR) duration of mechanical ventilation prior to onset of the first VAP episode was 7 days (4–9 days). The incidence of VAP was 55 episodes/1000 ventilation days and 112 episodes/1000 ventilation days ‘‘at risk’’ (only including ventilation days before VAP onset). According to the VAP definition, all VAP episodes had a CPIS score 6 (max.10). Four
out of seven patients who did not develop VAP, died within 5 days limiting the period at risk for development of VAP. In view of the small number of patients without VAP, comparison of patients with and without VAP was considered not relevant. In patients with VAP, the rate appropriate therapy 24 h and 48 h after VAP onset was 67.1% and 85.7%, respectively.
3.1.
Microbiology in VAP
In 70 VAP episodes, 91 pathogens were isolated in the diagnostic VAP culture. In 21 (30.0%) of the VAP episodes, multiple microorganisms were isolated. In early VAP, P. aeruginosa, Haemophilus influenza and Staphylococcus pneumonia were most frequent. In late VAP, P. aeruginosa, Enterobacter spp. and S. aureus were most frequently isolated (Cf. Table 3). In 23 VAP episodes (32.9%), of which only 2 ‘early VAP’ episodes, at least one MDR pathogen was involved (24 isolated microorganisms) (Table 3).
3.2. VAP
Value of surveillance cultures to predict etiology in
In surveillance cultures, 23 MDR pathogens were isolated, whereof 20 were consequently found in the VAP as well (‘true
Table 3 – Microbiology in ventilator-associated pneumonia. Microorganisms
All VAP n MDR/total (%)
S. aureus S. pneumoniae Citrobacter spp. Enterobacter spp. E. coli H. influenza Klebsiella spp. Moraxella Proteus spp. P. aeruginosa Serratia spp. S. maltophilia
0/11 (0) 2/7 (28.6) 3/5 (60.0) 7/10 (70.0) 1/2 (50.0) 0/7 (0) 0/6 (0) 0/1 (0) 0/1 (0) 10/37 (27.0) 1/2 (50.0) 0/2 (0)
0/3 0/4 0/1 1/1 0/1 0/4 0/2 0/1 0/0 0/5 1/1 0/0
(0) (0) (0) (100) (0) (0) (0) (0) (0) (0) (100) (0)
0/8 (0) 2/3 (66.7) 3/4 (75.0) 6/9 (66.7) 1/1 (100) 0/3 (0) 0/4 (0) 0/0 (0) 0/1 (0) 10/32 (31.3) 0/1 (0) 0/2 (0)
24/91 (26.4)
2/23 (8.7)
22/68 (32.4)
Early-onset VAP n MDR/total (%)
Late-onset VAP n (% MDR/total (%))
VAP, ventilator-associated pneumonia; MDR, multidrug resistant.
Table 4 – Value of surveillance cultures to predict multidrug resistant pathogens in a subsequent episode of ventilatorassociated pneumonia.
All episodes Gram-positive bacteriaa Gram-negative bacteria P. aeruginosa Enterobacteriaceae
Sensitivity
Specificity
PPV
NPV
20/24 (83.3) (64.1–93.3) 2/2 (83.3) (31.0–98.2) 17/21 (81.0) (60.0–92.3) 9/10 (90.0) (59.6–98.2) 9/12 (75.0) (46.8–91.1)
76/79 (96.2) (89.4–98.7) 17/18 (92.1) (71.9–98.2) 58/60 (96.7) (88.6–99.1) 90/91 (98.9) (94.0–99.8) 88/89 (98.9) (93.9–99.8)
20/23 (87.0) (67.9–95.5) 2/3 (62.5) (21.9–90.8) 17/19 (89.5) (68.6–97.1) 9/10 (90.0) (59.6–98.2) 9/10 (90.0) (59.6–98.2)
76/80 (95.0) (87.8–98.0) 17/17 (97.2) (78.1–99.7) 58/62 (93.5) (84.6–97.5) 92/93 (98.9) (94.0–99.8) 88/91 (96.7) (90.8–98.9)
Data are presented as a proportion of n (%) microorganisms presenting the operating characteristics (and corresponding 95% confidence intervals) of routine endotracheal surveillance cultures in the prediction of multidrug resistance. PPV, positive predictive value; NPV, negative predictive value. a Since no false negatives occurred, 0.5 was added to each cell for these calculations.
368
burns 38 (2012) 364–370
positive’). Three ‘false positive’ MDR microorganisms were found in the surveillance cultures and four (‘false negative’) MDR microorganisms that caused the VAP were not found in the surveillance cultures. These numbers correspond with a sensitivity and specificity of the surveillance cultures of respectively 83.3% (95% CI: 64.1–93.3) and 96.2% (89.4–98.7). The PPV and NPV were 87.0% (67.9–95.5%) and 95.0% (87.8– 98.0), respectively. Subgroup analyses are shown in Table 4. When calculated on the total number of VAP-episodes (n = 70) instead of the total number of microorganisms (n = 101), sensitivity and specificity were slightly lower (respectively 82.6% and 95.7%), and positive and negative predictive values were higher (respectively 90.5% and 91.8%).
4.
Discussion
In an 8-year study period, only a limited number of patients admitted to the burn suffered both severe burns and inhalation injury requiring mechanical ventilation. Nevertheless, these 53 patients represented an extremely vulnerable group, shown by the high risk to develop pneumonia. The incidence of VAP in these burn patients with inhalation injury is high: 55 episodes/1000 ventilation days or 112 episodes/1000 ventilation days ‘‘at risk’’. P. aeruginosa spp. was the most prevalent microorganism, and was present in more than half of the episodes (37/70 episodes, 52.9%), and 43.5% of episodes with MDR pathogens (10/23 episodes). Overall, 86.8% of patients developed VAP, which is significantly higher compared with other ‘critically ill’ patients, as described by Safdar et al. in a systematic review comprising 38 studies on VAP in general ICUs (pooled cumulative incidence: 9.7% (95% CI: 7.0–12.5) [9]. Only a few studies focusing on burn patients have been published, although most of these studies focused on pneumonia instead of VAP (cf. Table 1). Despite the large heterogeneity in definitions and inclusion criteria, our study results seem comparable with the incidence and prevalence rates of de la Cal et al. and Rue et al. [27,28]. Two other studies reported pneumonia rates in patients with inhalation injury either with or without the necessity of mechanical ventilation [11,12]. However, in contrast to the current study, these studies did not consider the stringent criterion of the necessity of mechanical ventilation for the diagnosis of inhalation injury. As the need for endotracheal intubation and mechanical ventilation implies more severe injury, and as such a higher risk of respiratory tract infection, we assume this may have contributed to their lower occurrence rate of pneumonia. Despite the 8-year study period, only 53 patients met the inclusion criteria. The weakness of the study is the lack of sound criteria to diagnose VAP. Diagnosis of VAP in burn patients is certainly problematic. In patients with ARDS the diagnosis of a new or worsening infiltrate is difficult. Also the systemic inflammatory response provoked by the burn might hamper a clinical diagnosis of pneumonia. Therefore it is not impossible that the occurrence rate of VAP may be overestimated. In order to strengthen the diagnosis of VAP we additionally evaluated the CPIS [24]. In all patients that were judged to have VAP the CPIS was 6 or higher. Particular attention was given to microbiology in the diagnosis of VAP
and the clinician’s decision to start empiric antimicrobial therapy. Yet, microbiological diagnosis of VAP was made on semi-quantitative cultures of endotracheal aspirates, as broncho-alveolar lavage is not routinely performed in the burn unit. As a consequence it cannot be ruled out that the incidence of VAP was overestimated in our cohort. Overestimation of the VAP occurrence rate might be suggested as well on basis of the low mortality observed, which is substantially lower than the mortality as expected on basis of the BOBI score (approximately 30%). As this model predicts mortality on burn unit admission, survival bias might have influenced this observation as many extremely severely burned patients die within the first 48 h and as such before nosocomial infection may have developed [10,22]. Furthermore, the adverse impact of VAP in traumatized patients is milder compared to other ICU patients [29–31] and this may also – at least in part – explain the favorable outcome figure. Routine endotracheal surveillance cultures demonstrated excellent operating characteristics to predict MDR pathogens in VAP, illustrated by an overall 87% PPV and a 95% NPV, in particular for P. aeruginosa and MDR Enterobacteriaceae. Several methodological and ecological elements influence the predictive value of surveillance cultures. First, to judge the value of surveillance cultures one should focus rather on the NPV than on PPV. On the basis of a high NPV certain potential pathogens can be ruled out. This allows a restrictive empiric antimicrobial scheme with reduced consumption of antipseudomonal antibiotic agents compared to strict empiric schemes as proposed in guidelines [26,32]. The second goal is to assess the odds of involvement of MDR pathogens that pose a challenge to the clinicians in the appropriate empiric choice. Evaluating the predictive value of surveillance cultures by taking into account MDR and non-MDR pathogens will result in poor operating characteristics as demonstrated by the studies of Hayon et al. and Papadomichelakis et al. [20,33]. Third, the frequency of sampling should be at least twice weekly. A onceweekly sampling frequency provided unfavorable predictive values in two studies, albeit that these studies also focused on all microorganisms, instead of exclusively on MDR pathogens reinforcing a negative influence on operating characteristics [17,34]. Fourth, a higher prevalence of MDR pathogens in the unit results in a larger benefit of surveillance culture guided strategy as compared to strictly risk factor driven empiric regimens (as in classic guidelines) in terms of rates of appropriate empiric therapy and/or saving of antibiotic cost and classes used [25]. Finally, it can be assumed that some pathogens are more frequently detected prior to infection than others, maybe reflecting different rates and risks of progression from colonization to infection, leading to differences in the likelihood of potential pathogen detection in sequential surveillance cultures prior to the development of VAP [35]. In a 2-year prospective study to evaluate the endemicity of P. aeruginosa in ICUs, Bertrand et al. found surveillance cultures to have a high specificity and negative predictive value, as colonization preceded infection in almost all patients who experienced P. aeruginosa infection during their ICU course [36]. As Gram-negative bacteria, and P. aeruginosa in particular, are predominant pathogens in burn units [37,38], the value of regular surveillance cultures to steer empiric antimicrobial therapy may be higher in this particular setting.
burns 38 (2012) 364–370
5.
Conclusion
The incidence of VAP in burn patients with inhalation injury is high but its impact in terms of mortality appears to be rather limited. Yet, incidence estimates as well as mortality may be biased due to the pitfalls in the diagnosis of pneumonia in this particular patient population. In this cohort routine surveillance cultures appear to have excellent operating characteristics to predict MDR pathogens in VAP.
Conflicts of interest None.
Acknowledgement None.
references
[1] The Belgian Outcome in Burn Injury Study Group. Development and validation of a model for prediction of mortality in patients with acute burn injury. Br J Surg 2009;96:111–7. [2] Brusselaers N, Monstrey S, Blot S. The FLAMES score accurately predicts mortality risk in burn patients (Gomez M, et al. 2008). J Trauma 2009;67:415. [3] Brusselaers N, Monstrey S, Vogelaers D, Hoste E, Blot S. Severe burn injury in Europe: a systematic review of the incidence, aetiology, morbidity and mortality. Crit Care 2010;14:R188. [4] Brusselaers N, Hoste EA, Monstrey S, Colpaert KE, De Waele JJ, Vandewoude KH, et al. Outcome and changes over time in survival following severe burns from 1985 to 2004. Intensive Care Med 2005;31:1648–53. [5] Thaler U, Kraincuk P, Kamolz LP, Frey M, Metnitz PG. Inhalation injury – epidemiology, diagnosis and therapy. Wien Klin Wochenschr 2010;122:11–21. [6] Edelman DA, White MT, Tyburski JG, Wilson RF. Factors affecting prognosis of inhalation injury. J Burn Care Res 2006;27:848–53. [7] Wurtz R, Karajovic M, Dacumos E, Jovanovic B, Hanumadass M. Nosocomial infections in a burn intensive care unit. Burns 1995;21:181–4. [8] Wahl WL, Taddonio MA, Arbabi S, Hemmila MR. Duration of antibiotic therapy for ventilator-associated pneumonia in burn patients. J Burn Care Res 2009;30:801–6. [9] Safdar N, Dezfulian C, Collard HR, Saint S. Clinical and economic consequences of ventilator-associated pneumonia: a systematic review. Crit Care Med 2005;33:2184–93. [10] Myny D, Depuydt P, Colardyn F, Blot S. Ventilatorassociated pneumonia in a tertiary care ICU: analysis of risk factors for acquisition and mortality. Acta Clin Belg 2005;60:114–21. [11] Shirani KZ, Pruitt Jr BA, Mason Jr AD. The influence of inhalation injury and pneumonia on burn mortality. Ann Surg 1987;205:82–7. [12] Edelman DA, Khan N, Kempf K, White MT. Pneumonia after inhalation injury. J Burn Care Res 2007;28:241–6.
369
[13] Blot S. Limiting the attributable mortality of nosocomial infection and multidrug resistance in intensive care units. Clin Microbiol Infect 2008;14:5–13. [14] Iregui M, Ward S, Sherman G, Fraser VJ, Kollef MH. Clinical importance of delays in the initiation of appropriate antibiotic treatment for ventilator-associated pneumonia. Chest 2002;122:262–8. [15] Rello J, Gallego M, Mariscal D, Sonora R, Valles J. The value of routine microbial investigation in ventilator-associated pneumonia. Am J Respir Crit Care Med 1997;156:196–200. [16] Luna CM, Vujacich P, Niederman MS, Vay C, Gherardi C, Matera J, et al. Impact of BAL data on the therapy and outcome of ventilator-associated pneumonia. Chest 1997;111:676–85. [17] Jung B, Sebbane M, Chanques G, Courouble P, Verzilli D, Perrigault PF, et al. Previous endotracheal aspirate allows guiding the initial treatment of ventilator-associated pneumonia. Intensive Care Med 2009;35:101–7. [18] Blot S, Depuydt P, Vogelaers D. Maximizing rates of empiric appropriate antibiotic therapy with minimized use of broad-spectrum agents: are surveillance cultures the key? Intensive Care Med 2008;34:2130–3. [19] A‘Court CH, Garrard CS, Crook D, Bowler I, Conlon C, Peto T, et al. Microbiological lung surveillance in mechanically ventilated patients, using non-directed bronchial lavage and quantitative culture. Q J Med 1993;86:635–48. [20] Hayon J, Figliolini C, Combes A, Trouillet JL, Kassis N, Dombret MC, et al. Role of serial routine microbiologic culture results in the initial management of ventilatorassociated pneumonia. Am J Respir Crit Care Med 2002;165:41–6. [21] Felton T, Mount T, Chadwick P, Ghrew M, Dark P. Surveillance non-directed bronchial lavage allows confident use of focused antibiotics in the management of ventilator-associated pneumonia. J Infect 2010;60:397–9. [22] Brusselaers N, Monstrey S, Snoeij T, Vandijck D, Lizy C, Hoste E, et al. Morbidity and mortality of bloodstream infections in patients with severe burn injury. Am J Crit Care 2010;19:e81–7. [23] Brusselaers N, Monstrey S, Colpaert K, Decruyenaere J, Blot SI, Hoste EA. Outcome of acute kidney injury in severe burns: a systematic review and meta-analysis. Intensive Care Med 2010;36:915–25. [24] Pugin J, Auckenthaler R, Mili N, Janssens JP, Lew PD, Suter PM. Diagnosis of ventilator-associated pneumonia by bacteriologic analysis of bronchoscopic and nonbronchoscopic blind bronchoalveolar lavage fluid. Am Rev Respir Dis 1991;143:1121–9. [25] Depuydt P, Benoit D, Vogelaers D, Decruyenaere J, Vandijck D, Claeys G, et al. Systematic surveillance cultures as a tool to predict involvement of multidrug antibiotic resistant bacteria in ventilator-associated pneumonia. Intensive Care Med 2008;34:675–82. [26] Depuydt PO, Blot SI, Benoit DD, Claeys GW, Verschraegen GL, Vandewoude KH, et al. Antimicrobial resistance in nosocomial bloodstream infection associated with pneumonia and the value of systematic surveillance cultures in an adult intensive care unit. Crit Care Med 2006;34:653–9. [27] de La Cal MA, Cerda E, Garcia-Hierro P, Lorente L, SanchezConcheiro M, Diaz C, et al. Pneumonia in patients with severe burns: a classification according to the concept of the carrier state. Chest 2001;119:1160–5. [28] Rue 3rd LW, Cioffi WG, Mason Jr AD, McManus WF, Pruitt Jr BA. The risk of pneumonia in thermally injured patients requiring ventilatory support. J Burn Care Rehabil 1995;16:262–8. [29] Cook A, Norwood S, Berne J. Ventilator-associated pneumonia is more common and of less consequence in
370
[30]
[31]
[32]
[33]
burns 38 (2012) 364–370
trauma patients compared with other critically ill patients. J Trauma 2010;69:1083–91. Magret M, Amaya-Villar R, Garnacho J, Lisboa T, Diaz E, Dewaele J, et al. Ventilator-associated pneumonia in trauma patients is associated with lower mortality: results from EU-VAP study. J Trauma 2010;69:849–54. Eckert MJ, Wade TE, Davis KA, Luchette FA, Esposito TJ, Poulakidas SJ, et al. Ventilator-associated pneumonia after combined burn and trauma is caused by associated injuries and not the burn wound. J Burn Care Res 2006;27:457–62. Michel F, Franceschini B, Berger P, Arnal JM, Gainnier M, Sainty JM, et al. Early antibiotic treatment for BAL-confirmed ventilator-associated pneumonia: a role for routine endotracheal aspirate cultures. Chest 2005;127:589–97. Papadomichelakis E, Kontopidou F, Antoniadou A, Poulakou G, Koratzanis E, Kopterides P, et al. Screening for resistant Gram-negative microorganisms to guide empiric therapy of subsequent infection. Intensive Care Med 2008;34:2169–75.
[34] Lampati L, Maggioni E, Langer M, Malacarne P, Mozzo R, Pesenti A, et al. Can routine surveillance samples from tracheal aspirate predict bacterial flora in cases of ventilator-associated pneumonia? Minerva Anestesiol 2009;75:555–62. [35] Langer M, Carretto E, Haeusler EA. Infection control in ICU: back (forward) to surveillance samples? Intensive Care Med 2001;27:1561–3. [36] Bertrand X, Thouverez M, Talon D, Boillot A, Capellier G, Floriot C, et al. Endemicity, molecular diversity and colonisation routes of Pseudomonas aeruginosa in intensive care units. Intensive Care Med 2001;27:1263–8. [37] Oncul O, Ulkur E, Acar A, Turhan V, Yeniz E, Karacaer Z, et al. Prospective analysis of nosocomial infections in a burn care unit, Turkey. Indian J Med Res 2009;130: 758–64. [38] Branski LK, Al-Mousawi A, Rivero H, Jeschke MG, Sanford AP, Herndon DN. Emerging infections in burns. Surg Infect (Larchmt) 2009;10:389–97.
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/burns
Burns, inhalation injury and ventilator-associated pneumonia: Value of routine surveillance cultures Nele Brusselaers a,b,1,*, Dennis Logie b,1, Dirk Vogelaers a,b, Stan Monstrey b,c, Stijn Blot a,b,d a
General Internal Medicine & Infectious Diseases, Ghent University Hospital, De Pintelaan 185, Ghent 9000, Belgium Faculty of Medicine and Health Sciences, Ghent University, De Pintelaan 185, Ghent 9000, Belgium c Burn Unit, Ghent University Hospital, De Pintelaan 185, Ghent 9000, Belgium d Dept. of Healthcare, University College Ghent, Keramiekstraat 80, Ghent 9000, Belgium b
article info
abstract
Article history:
Purpose: Burn patients with inhalation injury are at particular risk for ventilator-associated
Accepted 1 September 2011
pneumonia (VAP). Routine endotracheal surveillance cultures may provide information
Keywords:
tive was to assess the incidence of VAP in burn patients with inhalation injury, and the
about the causative pathogen in subsequent VAP, improving antibiotic therapy. Our objecBurns
benefit of routine surveillance cultures to predict multidrug resistant (MDR) pathogens.
Ventilator-associated pneumonia
Procedures: Historical cohort (n = 53) including all burn patients with inhalation injury
Inhalation injury
requiring mechanical ventilation, admitted to the Ghent burn unit (2002–2010).
Endotracheal surveillance cultures
Main findings: Median (interquartile range) age and total burned surface area were 44y (39–
Predictive value
55y) and 35% (19–50%). Overall, 70 episodes of VAP occurred in 46 patients (86.8%). Median
Epidemiology
mechanical ventilation days (MVD) prior to VAP onset were 7d (4–9d). The incidence was 55 episodes/1000 MVD. In 23 episodes (32.9%) at least one MDR causative pathogen was involved, mostly Pseudomonas aeruginosa and Enterobacter spp. The sensitivity and specificity of surveillance cultures to predict MDR etiology in subsequent VAP was respectively 83.0% and 96.2%. The positive and negative predictive value was 87.0% and 95.0%, respectively. Conclusions: The incidence of VAP in burn patients with inhalation injury is high. In this cohort routine surveillance cultures had excellent operating characteristics to predict MDR pathogen involvement. # 2011 Elsevier Ltd and ISBI. All rights reserved.
1.
Introduction
Depending on the diagnostic criteria used, inhalation injury is reported in 0.3–43% of the patients with severe burn [1–3]. It is one of the major (independent) risk factors for mortality, as it is associated with up to 8–10 fold increased mortality (or 15–45%) [1–6]. Because of severe respiratory distress or because patent airway is compromised by laryngeal edema, endotracheal
intubation and mechanical ventilation may be necessary in severe inhalation injury, often for a prolonged period until resolution of edema. Due to their immunocompromised status, patients with severe burn are particularly prone to infections [7,8]. As all mechanically ventilated critically ill patients, burn patients are at risk for ventilator-associated pneumonia (VAP), a severe infection that contributes to an increase in morbidity with excess length of hospitalization, and increased mortality
* Corresponding author at: General Internal Medicine & Infectious Diseases, Ghent University Hospital, De Pintelaan 185 - 9000 Ghent, Belgium. E-mail addresses: [email protected] (N. Brusselaers), [email protected] (D. Logie), [email protected] (D. Vogelaers), [email protected] (S. Monstrey), [email protected] (S. Blot). 1 These authors are joint first authors. 0305-4179/$36.00 # 2011 Elsevier Ltd and ISBI. All rights reserved. doi:10.1016/j.burns.2011.09.005
365
11% 55
19% 57 31 (48) n.r.
37% 37% 40% n.r. n.r.
n.r. n.r.
19% 38% 9% 65% 77% 52% 48% 73% 27% (68%) 27% 37% 19.7% 87%
n.r., not reported; VAP, ventilator-associated pneumonia; MV, mechanical ventilation.
Retrospective cohort Brusselaers et al. (present study)
VAP
History + clinical diagnosis In mechanically ventilated pts also bronchoscopy History + clinical diagnosis (+bronchoscopy) + necessitating MV Pneumonia Retrospective cohort Edelman et al. [12]
Pneumonia Prospective cohort de la Cal et al. [27]
VAP Retrospective cohort Rue et al. [28]
Incidence of pneumonia (/1000 ventilation days) Prevalence of pneumonia Population
Burns (n = 1058): With inhalation injury (n = 373); Without inhalation injury (n = 685) Burns + MV (n = 370): With inhalation injury (n = 265); Without inhalation injury (n = 105) Burns (n = 56): With inhalation injury (n = 26); Without inhalation injury (n = 30) With MV: n = 41) Inhalation injury (n = 117): With burns (n = 51); Without burns (n = 66) Burns + inhalation injury + mechanical ventilation (n = 53) History + clinical diagnosis + bronchoscopy or ventilation/ perfusion scan History + clinical diagnosis + bronchoscopy or ventilation/ perfusion scan History + clinical diagnosis + bronchoscopy Pneumonia Retrospective cohort Shirani et al. [11]
We performed a retrospective cohort study (January 2002– March 2010) in the 6-bed burn unit at Ghent University Hospital. The unit serves a geographic area of about 2.6 million inhabitants. Approximately 60–80 patients are admitted to the unit each year [4]. The standard treatment protocol of the burn unit is described elsewhere [4,22]. This historical cohort study was approved by the Ethics Committee at Ghent University Hospital. Patients or their proxies were informed with an opting out possibility. We retrospectively analyzed the data of all burned patients admitted to our burn unit over an 8-year period (January 2002– March 2010). Patients admitted to the burn unit for other indications were excluded from the study (e.g. degloving injury, Lyell syndrome, Steven-Johnson syndrome). The appropriate patients were identified through the diagnosis related groups (DRGs) data (inhalation injury = admission criteria), and the prospectively collected information registered in the log book (describing characteristics on admission
Definition of inhalation injury
Study design and population
Pneumonia/ VAP
2.1.
Study design
Materials and methods
Author (year)
2.
Table 1 – Studies describing (ventilator-associated) pneumonia in populations with severe burn.
[9,10]. It can be presumed that the presence of an inhalation injury contributes to an even higher risk of VAP. Inhaled products of combustion cause de-epithelialisation in the tracheobronchial tree and lower respiratory tract lesions. This process results in an increase in extravascular lung water with decreased pulmonary compliance, inactivation of surfactant with micro-atelectasis, and pseudomembrane formation of mucus, cellular debris, fibrinous exudates, polymorphonuclear leucocytes, and clumps of bacteria. These factors contribute to an increased risk of pneumonia. However, data about the incidence of pneumonia, and VAP in particular, in burn patients with inhalation injury are scarce and difficult to compare due to different inclusion criteria and definitions (Table 1) [7,11,12]. Once pneumonia occurs, early adequate antibiotic therapy is essential to optimize the chance of survival [8,13]. Inappropriate empiric therapy, not covering the causative organism, has been associated with an increased mortality rate [14–16]. Routine endotracheal surveillance cultures (SC) have been advocated to steer empiric therapy [17,18] since they provide data about pathogens colonizing the lower respiratory tract and their susceptibility patterns [19–21]. Especially the prediction of multidrug resistant (MDR) pathogens is believed to be beneficial in improving early antibiotic therapy, since the results of the diagnostic cultures are only available 24–48 h after VAP onset, and the MDR pathogens may not be covered by the standard empiric therapy. In general ICU patients endotracheal SC have a high negative predictive value in VAP with regard to multidrug resistant pathogens [18]. The objectives of this study were to estimate the incidence of VAP in a cohort of severely burned patients with inhalation injury necessitating endotracheal intubation and mechanical ventilation, and to assess the value of routine endotracheal surveillance cultures in the prediction of MDR pathogens in VAP.
Mortality in pneumonia
burns 38 (2012) 364–370
366
burns 38 (2012) 364–370
and outcome data). The electronic and paper files were also consulted for all patients with flame burns of the face and upper limbs receiving mechanical ventilation. All consecutive patients with a severe inhalation injury, confirmed by laryngoscopy or bronchoscopy, and necessitating mechanical ventilation, were included. The following data were registered for each patient: age, total burned surface area (TBSA), mortality, microbiology, length of mechanical ventilation and length of stay, time of onset of VAP and outcome. Burn severity for this cohort was assessed by the Belgian Outcome in Burn Injury (BOBI) score, as well as the presence of acute kidney injury and need for vasopressive support [1,22,23]. This mathematical BOBI-model estimates the risk of mortality based on the three major risk factors for mortality in severe burns: age, total burned surface area (TBSA) and presence or absence of inhalation injury [1]. It divides age and TBSA in respectively 4 (0–3 points) and 5 (0–4 points) risk categories, and if present, inhalation injury scores three additional points; resulting in a score ranging 0–10 points, correlating with an increasing risk of mortality. The minimal score of patients in this cohort is therefore 3 points, since only patients with an inhalation injury are included.
2.2.
Definitions
VAP was suspected in patients receiving at least 48 h of mechanical ventilation, when a new or new or progressive infiltrate was identified on sequential daily chest X-ray, and when two of the following four criteria were met: fever (T8 > 38 8C) or hypothermia (T8 < 36 8C), leucocytosis (>10 109 cells/L) or neutropenia (<5 109 cells/L), purulent secretions, and gas exchange impairment affecting the partial pressure of oxygen and/or the fraction of inspired oxygen, together with a clinical pulmonary infection score (CPIS) 6 [24]. Since adult respiratory distress syndrome (ARDS) may be provoked by either inhalation injury or hypovolemic burn shock, the presence of ARDS was not considered in the calculation of the CPIS. For this study, the routine surveillance and diagnostic techniques of our intensive care units were used. Routine endotracheal aspirate (ETA) ‘surveillance’ cultures were performed thrice a week (Monday–Wednesday–Friday). Microbiological confirmation of VAP required ++ or +++ semiquantitative growth of a pathogen in a good quality endotracheal aspirate (‘diagnostic’ ETA) [25]. Coagulase-negative Staphylococcus and Candida species were considered non-pathogenic. Additionally, clinically likely VAP with limited (+) growth on endotracheal aspirate was considered as microbiologically confirmed if Gramstaining showed predominant presence of a pathogen and antibiotic therapy had been started or changed within 72 h. VAP was considered polymicrobial if more than one pathogen was shown, with growth above these thresholds. Late-onset VAP was defined as VAP diagnosed >5 days following intubation.
2.3.
Microbiology and antibiotic treatment
For each VAP episode, all available microbiologic specimen results were retrospectively reviewed and recorded. Routine surveillance included a microbiological analysis of thriceweekly endotracheal aspirates in all mechanically ventilated
patients. For convenience sake, only the most recent endotracheal aspirate cultures sampled before clinical diagnosis of VAP were taken into account. The following pathogens were considered as MDR: methicillin-resistant Staphylococcus aureus (MRSA), extended spectrum b-lactamase (ESBL) producing Enterobacteriaceae, MDR nonfermenting organisms such as Stenotrophomonas maltophilia, Acinetobacter baumannii and Pseudomonas aeruginosa resistant for at least one of the following antipseudomonal antibiotics: ceftazidime, ciprofloxacin, piperacillin-tazobactam, and imipenem [26]. Antimicrobial therapy was considered appropriate if it included at least one antimicrobial drug with in vitro activity against the causative pathogen.
2.4.
Statistical analysis
Statistical analyses are executed with PASW Statistics 18.0 (Chicago, Ill., USA). All used tests were two-tailed and statistical significance was defined as p < 0.05. The data are expressed as median (interquartile range [IQR]) or as n (%). Differences between groups were analyzed with Mann– Whitney U test and Chi square tests. The incidence of VAP is expressed as the number of episodes per total number of ventilation days, and per total number of ventilation days ‘at risk’, including only ventilation days before the onset of the VAP episode. The value of SC to predict MDR pathogens in VAP was expressed by the following operating characteristics (or ‘predictive parameters’): sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) (and respective 95% confidence intervals -CI). Results are reported as the proportion of MDR-VAP episodes on the total number of VAP episodes, and as the proportion of MDR microorganisms on total number of microorganisms recovered both from the SC cultures as the diagnostic VAP cultures. Microorganisms that were found both in the SC and in the diagnostic ETA sample (with development of VAP) were defined as ‘true positive’. When either found in the SC or in the diagnostic ETA, MDR microorganisms were respectively defined as ‘false positive’ and ‘false negative’. All other ‘non-MDR’ microorganisms were considered true negative. For subgroupanalyses, e.g. Pseudomonas spp., the other MDR microorganisms (e.g. MDR Enterobacteriaceae) were considered ‘true negative’.
3.
Results
During the study period 59 burn patients with inhalation injury were retrieved, of which 2 patients did not receive mechanical ventilation, and 4 patients were excluded because necessary patient information could not be retrieved. Consequently, 53 patients with burn and inhalation injury necessitating mechanical ventilation were included in the study (Table 2). Only one patient was not ventilated from the day of admission on (start mechanical ventilation on day 4). Vasopressive support was necessary in 6/53 (73.6%) patients, and acute kidney injury occurred in 6/53 patients (11.3%). The median (IQR) total BOBI score of this cohort of mechanically ventilated patients, was 5 (4–6), with a predicted mortality rate
367
burns 38 (2012) 364–370
Table 2 – Main characteristics of the total cohort of patients with severe burn and inhalation injury (n = 53). Variablea Age (years) Total burned surface area (%) Male gender Mechanical ventilation (days) Length of burn unit stay (days)
All patients (n = 53) 44.0 (38.5–54.5) 35.0 (18.0–50.0) 33/53 (62.3) 17.0 (7.0–30.0) 34.0 (18.0–49.0)
VAP, ventilator associated pneumonia. a Values are described as median (interquartile range) or n (%).
of 30% (20–50%), and an observed mortality of 17% (n = 9). The main characteristics of the overall population are reported in Table 2. Overall, 70 episodes of VAP occurred in 46 of the 53 patients (86.8%). Sixteen patients had two VAP episodes, and four patients three. Median (IQR) duration of mechanical ventilation prior to onset of the first VAP episode was 7 days (4–9 days). The incidence of VAP was 55 episodes/1000 ventilation days and 112 episodes/1000 ventilation days ‘‘at risk’’ (only including ventilation days before VAP onset). According to the VAP definition, all VAP episodes had a CPIS score 6 (max.10). Four
out of seven patients who did not develop VAP, died within 5 days limiting the period at risk for development of VAP. In view of the small number of patients without VAP, comparison of patients with and without VAP was considered not relevant. In patients with VAP, the rate appropriate therapy 24 h and 48 h after VAP onset was 67.1% and 85.7%, respectively.
3.1.
Microbiology in VAP
In 70 VAP episodes, 91 pathogens were isolated in the diagnostic VAP culture. In 21 (30.0%) of the VAP episodes, multiple microorganisms were isolated. In early VAP, P. aeruginosa, Haemophilus influenza and Staphylococcus pneumonia were most frequent. In late VAP, P. aeruginosa, Enterobacter spp. and S. aureus were most frequently isolated (Cf. Table 3). In 23 VAP episodes (32.9%), of which only 2 ‘early VAP’ episodes, at least one MDR pathogen was involved (24 isolated microorganisms) (Table 3).
3.2. VAP
Value of surveillance cultures to predict etiology in
In surveillance cultures, 23 MDR pathogens were isolated, whereof 20 were consequently found in the VAP as well (‘true
Table 3 – Microbiology in ventilator-associated pneumonia. Microorganisms
All VAP n MDR/total (%)
S. aureus S. pneumoniae Citrobacter spp. Enterobacter spp. E. coli H. influenza Klebsiella spp. Moraxella Proteus spp. P. aeruginosa Serratia spp. S. maltophilia
0/11 (0) 2/7 (28.6) 3/5 (60.0) 7/10 (70.0) 1/2 (50.0) 0/7 (0) 0/6 (0) 0/1 (0) 0/1 (0) 10/37 (27.0) 1/2 (50.0) 0/2 (0)
0/3 0/4 0/1 1/1 0/1 0/4 0/2 0/1 0/0 0/5 1/1 0/0
(0) (0) (0) (100) (0) (0) (0) (0) (0) (0) (100) (0)
0/8 (0) 2/3 (66.7) 3/4 (75.0) 6/9 (66.7) 1/1 (100) 0/3 (0) 0/4 (0) 0/0 (0) 0/1 (0) 10/32 (31.3) 0/1 (0) 0/2 (0)
24/91 (26.4)
2/23 (8.7)
22/68 (32.4)
Early-onset VAP n MDR/total (%)
Late-onset VAP n (% MDR/total (%))
VAP, ventilator-associated pneumonia; MDR, multidrug resistant.
Table 4 – Value of surveillance cultures to predict multidrug resistant pathogens in a subsequent episode of ventilatorassociated pneumonia.
All episodes Gram-positive bacteriaa Gram-negative bacteria P. aeruginosa Enterobacteriaceae
Sensitivity
Specificity
PPV
NPV
20/24 (83.3) (64.1–93.3) 2/2 (83.3) (31.0–98.2) 17/21 (81.0) (60.0–92.3) 9/10 (90.0) (59.6–98.2) 9/12 (75.0) (46.8–91.1)
76/79 (96.2) (89.4–98.7) 17/18 (92.1) (71.9–98.2) 58/60 (96.7) (88.6–99.1) 90/91 (98.9) (94.0–99.8) 88/89 (98.9) (93.9–99.8)
20/23 (87.0) (67.9–95.5) 2/3 (62.5) (21.9–90.8) 17/19 (89.5) (68.6–97.1) 9/10 (90.0) (59.6–98.2) 9/10 (90.0) (59.6–98.2)
76/80 (95.0) (87.8–98.0) 17/17 (97.2) (78.1–99.7) 58/62 (93.5) (84.6–97.5) 92/93 (98.9) (94.0–99.8) 88/91 (96.7) (90.8–98.9)
Data are presented as a proportion of n (%) microorganisms presenting the operating characteristics (and corresponding 95% confidence intervals) of routine endotracheal surveillance cultures in the prediction of multidrug resistance. PPV, positive predictive value; NPV, negative predictive value. a Since no false negatives occurred, 0.5 was added to each cell for these calculations.
368
burns 38 (2012) 364–370
positive’). Three ‘false positive’ MDR microorganisms were found in the surveillance cultures and four (‘false negative’) MDR microorganisms that caused the VAP were not found in the surveillance cultures. These numbers correspond with a sensitivity and specificity of the surveillance cultures of respectively 83.3% (95% CI: 64.1–93.3) and 96.2% (89.4–98.7). The PPV and NPV were 87.0% (67.9–95.5%) and 95.0% (87.8– 98.0), respectively. Subgroup analyses are shown in Table 4. When calculated on the total number of VAP-episodes (n = 70) instead of the total number of microorganisms (n = 101), sensitivity and specificity were slightly lower (respectively 82.6% and 95.7%), and positive and negative predictive values were higher (respectively 90.5% and 91.8%).
4.
Discussion
In an 8-year study period, only a limited number of patients admitted to the burn suffered both severe burns and inhalation injury requiring mechanical ventilation. Nevertheless, these 53 patients represented an extremely vulnerable group, shown by the high risk to develop pneumonia. The incidence of VAP in these burn patients with inhalation injury is high: 55 episodes/1000 ventilation days or 112 episodes/1000 ventilation days ‘‘at risk’’. P. aeruginosa spp. was the most prevalent microorganism, and was present in more than half of the episodes (37/70 episodes, 52.9%), and 43.5% of episodes with MDR pathogens (10/23 episodes). Overall, 86.8% of patients developed VAP, which is significantly higher compared with other ‘critically ill’ patients, as described by Safdar et al. in a systematic review comprising 38 studies on VAP in general ICUs (pooled cumulative incidence: 9.7% (95% CI: 7.0–12.5) [9]. Only a few studies focusing on burn patients have been published, although most of these studies focused on pneumonia instead of VAP (cf. Table 1). Despite the large heterogeneity in definitions and inclusion criteria, our study results seem comparable with the incidence and prevalence rates of de la Cal et al. and Rue et al. [27,28]. Two other studies reported pneumonia rates in patients with inhalation injury either with or without the necessity of mechanical ventilation [11,12]. However, in contrast to the current study, these studies did not consider the stringent criterion of the necessity of mechanical ventilation for the diagnosis of inhalation injury. As the need for endotracheal intubation and mechanical ventilation implies more severe injury, and as such a higher risk of respiratory tract infection, we assume this may have contributed to their lower occurrence rate of pneumonia. Despite the 8-year study period, only 53 patients met the inclusion criteria. The weakness of the study is the lack of sound criteria to diagnose VAP. Diagnosis of VAP in burn patients is certainly problematic. In patients with ARDS the diagnosis of a new or worsening infiltrate is difficult. Also the systemic inflammatory response provoked by the burn might hamper a clinical diagnosis of pneumonia. Therefore it is not impossible that the occurrence rate of VAP may be overestimated. In order to strengthen the diagnosis of VAP we additionally evaluated the CPIS [24]. In all patients that were judged to have VAP the CPIS was 6 or higher. Particular attention was given to microbiology in the diagnosis of VAP
and the clinician’s decision to start empiric antimicrobial therapy. Yet, microbiological diagnosis of VAP was made on semi-quantitative cultures of endotracheal aspirates, as broncho-alveolar lavage is not routinely performed in the burn unit. As a consequence it cannot be ruled out that the incidence of VAP was overestimated in our cohort. Overestimation of the VAP occurrence rate might be suggested as well on basis of the low mortality observed, which is substantially lower than the mortality as expected on basis of the BOBI score (approximately 30%). As this model predicts mortality on burn unit admission, survival bias might have influenced this observation as many extremely severely burned patients die within the first 48 h and as such before nosocomial infection may have developed [10,22]. Furthermore, the adverse impact of VAP in traumatized patients is milder compared to other ICU patients [29–31] and this may also – at least in part – explain the favorable outcome figure. Routine endotracheal surveillance cultures demonstrated excellent operating characteristics to predict MDR pathogens in VAP, illustrated by an overall 87% PPV and a 95% NPV, in particular for P. aeruginosa and MDR Enterobacteriaceae. Several methodological and ecological elements influence the predictive value of surveillance cultures. First, to judge the value of surveillance cultures one should focus rather on the NPV than on PPV. On the basis of a high NPV certain potential pathogens can be ruled out. This allows a restrictive empiric antimicrobial scheme with reduced consumption of antipseudomonal antibiotic agents compared to strict empiric schemes as proposed in guidelines [26,32]. The second goal is to assess the odds of involvement of MDR pathogens that pose a challenge to the clinicians in the appropriate empiric choice. Evaluating the predictive value of surveillance cultures by taking into account MDR and non-MDR pathogens will result in poor operating characteristics as demonstrated by the studies of Hayon et al. and Papadomichelakis et al. [20,33]. Third, the frequency of sampling should be at least twice weekly. A onceweekly sampling frequency provided unfavorable predictive values in two studies, albeit that these studies also focused on all microorganisms, instead of exclusively on MDR pathogens reinforcing a negative influence on operating characteristics [17,34]. Fourth, a higher prevalence of MDR pathogens in the unit results in a larger benefit of surveillance culture guided strategy as compared to strictly risk factor driven empiric regimens (as in classic guidelines) in terms of rates of appropriate empiric therapy and/or saving of antibiotic cost and classes used [25]. Finally, it can be assumed that some pathogens are more frequently detected prior to infection than others, maybe reflecting different rates and risks of progression from colonization to infection, leading to differences in the likelihood of potential pathogen detection in sequential surveillance cultures prior to the development of VAP [35]. In a 2-year prospective study to evaluate the endemicity of P. aeruginosa in ICUs, Bertrand et al. found surveillance cultures to have a high specificity and negative predictive value, as colonization preceded infection in almost all patients who experienced P. aeruginosa infection during their ICU course [36]. As Gram-negative bacteria, and P. aeruginosa in particular, are predominant pathogens in burn units [37,38], the value of regular surveillance cultures to steer empiric antimicrobial therapy may be higher in this particular setting.
burns 38 (2012) 364–370
5.
Conclusion
The incidence of VAP in burn patients with inhalation injury is high but its impact in terms of mortality appears to be rather limited. Yet, incidence estimates as well as mortality may be biased due to the pitfalls in the diagnosis of pneumonia in this particular patient population. In this cohort routine surveillance cultures appear to have excellent operating characteristics to predict MDR pathogens in VAP.
Conflicts of interest None.
Acknowledgement None.
references
[1] The Belgian Outcome in Burn Injury Study Group. Development and validation of a model for prediction of mortality in patients with acute burn injury. Br J Surg 2009;96:111–7. [2] Brusselaers N, Monstrey S, Blot S. The FLAMES score accurately predicts mortality risk in burn patients (Gomez M, et al. 2008). J Trauma 2009;67:415. [3] Brusselaers N, Monstrey S, Vogelaers D, Hoste E, Blot S. Severe burn injury in Europe: a systematic review of the incidence, aetiology, morbidity and mortality. Crit Care 2010;14:R188. [4] Brusselaers N, Hoste EA, Monstrey S, Colpaert KE, De Waele JJ, Vandewoude KH, et al. Outcome and changes over time in survival following severe burns from 1985 to 2004. Intensive Care Med 2005;31:1648–53. [5] Thaler U, Kraincuk P, Kamolz LP, Frey M, Metnitz PG. Inhalation injury – epidemiology, diagnosis and therapy. Wien Klin Wochenschr 2010;122:11–21. [6] Edelman DA, White MT, Tyburski JG, Wilson RF. Factors affecting prognosis of inhalation injury. J Burn Care Res 2006;27:848–53. [7] Wurtz R, Karajovic M, Dacumos E, Jovanovic B, Hanumadass M. Nosocomial infections in a burn intensive care unit. Burns 1995;21:181–4. [8] Wahl WL, Taddonio MA, Arbabi S, Hemmila MR. Duration of antibiotic therapy for ventilator-associated pneumonia in burn patients. J Burn Care Res 2009;30:801–6. [9] Safdar N, Dezfulian C, Collard HR, Saint S. Clinical and economic consequences of ventilator-associated pneumonia: a systematic review. Crit Care Med 2005;33:2184–93. [10] Myny D, Depuydt P, Colardyn F, Blot S. Ventilatorassociated pneumonia in a tertiary care ICU: analysis of risk factors for acquisition and mortality. Acta Clin Belg 2005;60:114–21. [11] Shirani KZ, Pruitt Jr BA, Mason Jr AD. The influence of inhalation injury and pneumonia on burn mortality. Ann Surg 1987;205:82–7. [12] Edelman DA, Khan N, Kempf K, White MT. Pneumonia after inhalation injury. J Burn Care Res 2007;28:241–6.
369
[13] Blot S. Limiting the attributable mortality of nosocomial infection and multidrug resistance in intensive care units. Clin Microbiol Infect 2008;14:5–13. [14] Iregui M, Ward S, Sherman G, Fraser VJ, Kollef MH. Clinical importance of delays in the initiation of appropriate antibiotic treatment for ventilator-associated pneumonia. Chest 2002;122:262–8. [15] Rello J, Gallego M, Mariscal D, Sonora R, Valles J. The value of routine microbial investigation in ventilator-associated pneumonia. Am J Respir Crit Care Med 1997;156:196–200. [16] Luna CM, Vujacich P, Niederman MS, Vay C, Gherardi C, Matera J, et al. Impact of BAL data on the therapy and outcome of ventilator-associated pneumonia. Chest 1997;111:676–85. [17] Jung B, Sebbane M, Chanques G, Courouble P, Verzilli D, Perrigault PF, et al. Previous endotracheal aspirate allows guiding the initial treatment of ventilator-associated pneumonia. Intensive Care Med 2009;35:101–7. [18] Blot S, Depuydt P, Vogelaers D. Maximizing rates of empiric appropriate antibiotic therapy with minimized use of broad-spectrum agents: are surveillance cultures the key? Intensive Care Med 2008;34:2130–3. [19] A‘Court CH, Garrard CS, Crook D, Bowler I, Conlon C, Peto T, et al. Microbiological lung surveillance in mechanically ventilated patients, using non-directed bronchial lavage and quantitative culture. Q J Med 1993;86:635–48. [20] Hayon J, Figliolini C, Combes A, Trouillet JL, Kassis N, Dombret MC, et al. Role of serial routine microbiologic culture results in the initial management of ventilatorassociated pneumonia. Am J Respir Crit Care Med 2002;165:41–6. [21] Felton T, Mount T, Chadwick P, Ghrew M, Dark P. Surveillance non-directed bronchial lavage allows confident use of focused antibiotics in the management of ventilator-associated pneumonia. J Infect 2010;60:397–9. [22] Brusselaers N, Monstrey S, Snoeij T, Vandijck D, Lizy C, Hoste E, et al. Morbidity and mortality of bloodstream infections in patients with severe burn injury. Am J Crit Care 2010;19:e81–7. [23] Brusselaers N, Monstrey S, Colpaert K, Decruyenaere J, Blot SI, Hoste EA. Outcome of acute kidney injury in severe burns: a systematic review and meta-analysis. Intensive Care Med 2010;36:915–25. [24] Pugin J, Auckenthaler R, Mili N, Janssens JP, Lew PD, Suter PM. Diagnosis of ventilator-associated pneumonia by bacteriologic analysis of bronchoscopic and nonbronchoscopic blind bronchoalveolar lavage fluid. Am Rev Respir Dis 1991;143:1121–9. [25] Depuydt P, Benoit D, Vogelaers D, Decruyenaere J, Vandijck D, Claeys G, et al. Systematic surveillance cultures as a tool to predict involvement of multidrug antibiotic resistant bacteria in ventilator-associated pneumonia. Intensive Care Med 2008;34:675–82. [26] Depuydt PO, Blot SI, Benoit DD, Claeys GW, Verschraegen GL, Vandewoude KH, et al. Antimicrobial resistance in nosocomial bloodstream infection associated with pneumonia and the value of systematic surveillance cultures in an adult intensive care unit. Crit Care Med 2006;34:653–9. [27] de La Cal MA, Cerda E, Garcia-Hierro P, Lorente L, SanchezConcheiro M, Diaz C, et al. Pneumonia in patients with severe burns: a classification according to the concept of the carrier state. Chest 2001;119:1160–5. [28] Rue 3rd LW, Cioffi WG, Mason Jr AD, McManus WF, Pruitt Jr BA. The risk of pneumonia in thermally injured patients requiring ventilatory support. J Burn Care Rehabil 1995;16:262–8. [29] Cook A, Norwood S, Berne J. Ventilator-associated pneumonia is more common and of less consequence in
370
[30]
[31]
[32]
[33]
burns 38 (2012) 364–370
trauma patients compared with other critically ill patients. J Trauma 2010;69:1083–91. Magret M, Amaya-Villar R, Garnacho J, Lisboa T, Diaz E, Dewaele J, et al. Ventilator-associated pneumonia in trauma patients is associated with lower mortality: results from EU-VAP study. J Trauma 2010;69:849–54. Eckert MJ, Wade TE, Davis KA, Luchette FA, Esposito TJ, Poulakidas SJ, et al. Ventilator-associated pneumonia after combined burn and trauma is caused by associated injuries and not the burn wound. J Burn Care Res 2006;27:457–62. Michel F, Franceschini B, Berger P, Arnal JM, Gainnier M, Sainty JM, et al. Early antibiotic treatment for BAL-confirmed ventilator-associated pneumonia: a role for routine endotracheal aspirate cultures. Chest 2005;127:589–97. Papadomichelakis E, Kontopidou F, Antoniadou A, Poulakou G, Koratzanis E, Kopterides P, et al. Screening for resistant Gram-negative microorganisms to guide empiric therapy of subsequent infection. Intensive Care Med 2008;34:2169–75.
[34] Lampati L, Maggioni E, Langer M, Malacarne P, Mozzo R, Pesenti A, et al. Can routine surveillance samples from tracheal aspirate predict bacterial flora in cases of ventilator-associated pneumonia? Minerva Anestesiol 2009;75:555–62. [35] Langer M, Carretto E, Haeusler EA. Infection control in ICU: back (forward) to surveillance samples? Intensive Care Med 2001;27:1561–3. [36] Bertrand X, Thouverez M, Talon D, Boillot A, Capellier G, Floriot C, et al. Endemicity, molecular diversity and colonisation routes of Pseudomonas aeruginosa in intensive care units. Intensive Care Med 2001;27:1263–8. [37] Oncul O, Ulkur E, Acar A, Turhan V, Yeniz E, Karacaer Z, et al. Prospective analysis of nosocomial infections in a burn care unit, Turkey. Indian J Med Res 2009;130: 758–64. [38] Branski LK, Al-Mousawi A, Rivero H, Jeschke MG, Sanford AP, Herndon DN. Emerging infections in burns. Surg Infect (Larchmt) 2009;10:389–97.