No Decrease in Early Ventilator-Associated Pneumonia After Early Use of Chlorhexidine

Introduction

Ventilator-associated pneumonia (VAP) is a significant cause of morbidity and mortality in critically ill and injured patients.¹ VAP is one of the most common intensive care unit (ICU)-acquired nosocomial infection, affecting 4.8 – 7.5% of patients intubated > 24 hours with an estimated overall attributable mortality of 9%. VAP is associated with longer hospital stays with a resultant attributable cost of approximately $40,000 per case.^1–3 Several studies have demonstrated the effectiveness of oral chlorhexidine in preventing pneumonia among intubated patients, and in 2010 the Institute for Healthcare Improvement recommended routine chlorhexidine prophylaxis to prevent VAP.^4–10 Patients who are critically ill and intubated emergently have the highest risk of developing pneumonia,^5,10–13 and in a recent trial, a single dose of chlorhexidine early in the care of trauma patients significantly reduced subsequent VAP.¹⁴ Interestingly, pre-intubation chlorhexidine has not demonstrated a reduction in VAP.¹⁵

Oral care varies across intensive care units (ICUs).^12,16 Although some authors have advocated for routine application of chlorhexidine prior to ICU admission to reduce VAP,⁴ little is known about the importance of prophylaxis timing on clinical outcomes. This study was conducted to test the association between timing of chlorhexidine prophylaxis and VAP incidence.

Methods

Design, Setting and Sample

This retrospective cohort study was conducted in a 711-bed university hospital and involved all intubated adult patients (age ≥18 years) transferred to a 36-bed Surgical ICU by an air ambulance service between July 2011 and April 2013. Patients were excluded if they (1) died within 72 hours of hospital arrival, (2) presented with a diagnosis of pneumonia or infiltrates on chest imaging, or (3) were admitted to a transferring hospital prior to definitive transfer. This study is reported in accordance with the STROBE Statement,¹⁷ and was approved by the local Institutional Review Board under waiver of informed consent.

Procedures

The primary exposure was time to chlorhexidine (if given), and the primary outcome was early pneumonia (diagnosed within 5 days of admission).^18,19 Application of oral chlorhexidine is the standard procedure in this ICU, however no protocol exists for prehospital, ED nor pre-intubation administration of chlorhexidine. Time to chlorhexidine was defined as the time from helicopter departure from the scene or transferring hospital to chlorhexidine administration in the ICU. The definition of early prophylaxis was derived from the median time to chlorhexidine administration, but quartiles were used for sensitivity analysis.

Pneumonia was defined as a clinical suspicion for pneumonia by a board-certified intensivist with initiation of antibiotic treatment and with subsequent respiratory culture indicating a pathogenic organism. Statistical tests are reported using standard descriptive and comparative analyses, and significance was defined as p < 0.05 for two-tailed tests.

Results

There were 134 intubated patients included and 128 (95.5%) were treated with chlorhexidine, with 84% (n=113) treated prior to 12 hours (Table 1). Most were male (61.2%) who presented through inter-hospital transfer (89.6%), with 94.8% of these transfers through the receiving hospital emergency department (ED). The most common diagnostic category was trauma (26.1%). Median APACHE-II score was 22 and was not associated with VAP diagnosis (p = 0.117). Pneumonia was treated before 5 days in 11.2% of patients. The median time to chlorhexidine was not associated with early VAP diagnosis (6.1 hours vs. 5.2 hours, p = 0.232, Figure 1). Multivariate logistic regression failed to demonstrate APACHE II scores as a confounder on the relationship between time to chlorhexidine and VAP. Chlorhexidine prior to 6 hours was not associated with a decreased risk of VAP (15.2% vs. 8.1%, p = 0.213), and no significant difference was observed in sensitivity analysis for 3 hours or 9 hours (p = 0.617 and p = 0.188, respectively). There was no difference in baseline characteristics nor outcome when the six patients (4.5%) not treated with chlorhexidine were compared to patients receiving chlorhexidine; none of the patients who were not treated with chlorhexidine developed pneumonia.

Table 1.

Demographics and baseline characteristics, stratified by VAP

	VAP (n = 15)	No VAP (n = 119)	p
Baseline characteristics
Male	11 (73.3%)	71 (59.6%)	0.404
Age (mean ± SD)	47.9 ± 21.4	51.8 ± 19.6	0.474
APACHE II score (mean ± SD)	25.6 ± 8.9	22.3 ± 10.2	0.242
BMI (mean ± SD)	27.5 ± 6.4	28.1 ± 5.6	0.729
Scene flight	0 (0%)	13 (10.9%)	0.373
Inter-facility transfer	15 (100%)	105 (88.2%)
Diagnoses (not mutually exclusive)
Trauma, n (%)	6 (40%)	45 (37.8%)	0.999
Cerebral Hemorrhage	5 (33%)	49 (41.2%)	0.781
Brain Injury	4 (26.6%)	33 (27.7%)	0.999
Respiratory failure	3 (20%)	18 (15.2%)	0.705
Pulmonary contusion, n (%)	3 (20%)	7 (5.9%)	0.085
Spinal cord injury, n (%)	2 (13.3%)	12 (15.9%)	0.657
Rib fracture, n (%)	2 (13.3%)	13 (10.9%)	0.676
Sepsis, n (%)	1 (6.6%)	3 (2.5%)	0.382
Timing of chlorhexidine amongst recipients
Median time to chlorhexidine (IQR)*	5.2 (3.0, 6.3) hours	6.1 (3.8, 9) hours	0.232
< 3 hours	2 (13.3%)	8 (6.7%)	0.310
<6 hours	10 (66.7%)	56 (47.1%)	0.179
<12 hours	14 (93.3%)	98 (82.4%)	0.464
<24 hours	14 (93.3%)	104 (87.4%)	0.999
≥24 hours	1 (6.7%)	9 (7.6%)
No chlorhexidine**	0 (0%)	6 (5.0%)	0.999
Outcomes
Hospital Days (LOS) (mean ± SD)	19.5 ± 10.4 days	11.1 ± 9.3 days	0.002
Death, n (%)	3 (20%)	13 (10.9%)	0.389

P values derived from t-tests or Mann-Whitney U-tests for continuous variables and Chi square tests for categorical variables.

^*IQR: intra quartile range, 25^th, 75^th percentile

^**No difference in baseline characteristics, diagnoses, nor outcome in recipients vs non-recipients of chlorhexidine

Figure 1.

Proportion of patients treated with chlorhexidine by time, stratified by early pneumonia status.

Discussion

Hospital-acquired infection has gained importance as a marker of health care quality, yet many recognize that patient-oriented factors predispose some patients to infection. The effects of prehospital intubation on subsequent pneumonia are controversial,^33–35 but most agree that severity of illness is associated with developing pneumonia.^19,36,37

Administration of oral chlorhexidine to intubated patients has been shown to prevent VAP^{4,6,8,9,12,14}, but few studies have defined a clear goal for treatment initiation time. Munro et al have recently shown that treatment with chlorhexidine immediately prior to endotracheal intubation does not reduce CPIS scores.¹⁵ Ours is now the second study demonstrating that early prophylaxis does not decrease the incidence of VAP, in our study using a different, clinical definition of VAP.

The CPIS score has been criticized for its test characteristics,^20–26 perhaps due to the lack of a true gold-standard diagnosis of VAP. The Centers for Disease Control and Prevention definition of VAP has also been criticized as too insensitive for clinical practice.^27–29 Stringency of diagnosis for VAP is highly variable among published diagnostic criteria for VAP; in one group, the diagnosis of VAP varied between 4–42%, depending on which of six diagnostic criteria (including CDC and CPIS) was utilized, with resultant 4–8 days of delay in diagnosis of VAP.³⁰ Not surprisingly, the more stringent the diagnostic algorithm, the higher the mortality, which ranged from 50–80%. A clinical outcome as used in our analysis may capture a more meaningful patient-oriented event and the importance of early antibiotic initiation has been shown to reduce mortality.^31,32

In addition to early chlorhexidine prophylaxis, our institution has standardized oral care including elevation of the head of bed ≥30 degrees, tooth brushing and subglottic suctioning twice daily, and chlorhexidine and oral suction every 4 hours. Similar oral care has demonstrated a 46% decrease in VAP at other institutions.³⁸ It is challenging to separate the effect of chlorhexidine from standard oral care, so for institutions striving for early VAP prophylaxis, nursing education should stress chlorhexidine as a component of overall oral care.

Our study has several limitations. First, as a retrospective study only data recorded in the medical record is available. We selected variables that are likely recorded accurately and utilized standardized data abstraction. Second, it is challenging to divorce oral chlorhexidine from other prophylactic oral care delivered in the ICU. Third, our center has relatively short times to ICU admission, and most (84%) patients were treated with chlorhexidine within 12 hours of presentation. As such, the absence of effect in our cohort may not be replicated in centers where chlorhexidine treatment is less routine or rapid.

Finally, our clinical definition of VAP, which combined the treating clinician’s suspicion for pneumonia and bacteriologic evidence for diagnosis, differs from prior studies which utilized CPIS, making direct comparison difficult. Use of a retrospective dataset prevented our use of CPIS scores, however the test characteristics of the CPIS to predict VAP are not ideal. Use of our clinical diagnosis of VAP, however, has shown similar lack of protective effects of early chlorhexidine against VAP to very recent studies using CPIS as an outcome.¹⁵

Conclusions

Oral chlorhexidine prophylaxis administered within the initial 12 hours of treatment is not associated with a decreased incidence of early VAP in mechanically ventilated surgical ICU patients. The role of VAP prevention, with respect to timing of interventions and the interventions themselves, should be explored further.