Multi-Center Outcomes of Chlorhexidine Oral Decontamination in Intensive Care Units

Abstract

Background:

The efficacy of oral chlorhexidine (oCHG) for decontamination in intensive care unit (ICU) patients is controversial. The purpose of this study was to evaluate the effect of oCHG decontamination on the incidence of pneumonia, sepsis, and death in ICU patients.

Methods:

The Philips eICU database version 2.0 was queried for patients admitted to the ICU for ≥48 hours in 2014–2015. The primary outcome of interest was death in the ICU. Secondary outcomes were a diagnosis of pneumonia or sepsis. Patients with pneumonia or sepsis diagnosed within the first 48 hours of ICU admission were excluded from the outcome analyses. Univariable analysis was performed comparing age, gender, race, severity of illness scores, hospital characteristics, and oCHG order. Multivariable logistic regression was performed using univariable results with p < 0.05.

Results:

Of the 64,904 patients from 186 hospitals, 22.1% (n = 14,333) had oCHG ordered. The overall mortality rate was 6.9% (n = 4,449) and the mortality rate in patients receiving oCHG was 10.6% (n = 1,518; p < 0.001). After controlling for confounding factors, oCHG remained an independent risk factor for death (odds ratio [OR] 1.25; 95% confidence interval [CI] 1.16–1.34). After excluding patients with an early diagnosis of pneumonia, the overall pneumonia incidence was 2.6% (n = 1,431) and the incidence in patients having oCHG was 4.2% (n = 517; p < 0.001). However, multivariable logistic regression revealed no significant difference in the risk of pneumonia with oCHG (OR 0.97; 95% CI 0.85–1.09). After excluding patients with an early diagnosis of sepsis, the overall rate of sepsis was 1.8% (n = 949) and for patients with oCHG, the rate was 3.3% (n = 388; p < 0.001). After controlling for other confounders, oCHG remained an independent risk factor for sepsis (OR 1.37; 95% CI 1.19–1.59).

Conclusions:

A chlorhexidine mouthwash order is associated with increased odds of death and sepsis without decreased odds of pneumonia in a heterogeneous cohort of ICU patients. Additional studies are needed to understand better the effect of oCHG on outcomes.

Chlorhexidine is the most popular skin antiseptic primarily because of its extended duration of activity and broader coverage than iodophors [1,2]. Oral chlorhexidine gluconate (oCHG) has been formulated for the decontamination of oral flora specifically for the prevention of ventilator-associated pneumonia (VAP). Since the late 1990s, more than 20 studies have evaluated the efficacy of oCHG in reducing VAP rates. Most of the early trials were performed on cardiac surgery patients with brief intensive care unit (ICU) stays and therefore negligible opportunity for developing VAP. Only three of these studies reported significant reductions in the VAP rate [3 –5]. These limited early reports coincided with widespread adoption of guidelines and bundles for the prevention of VAP that eventually included oral care with an antiseptic agent [6]. Currently, the practice of oral decontamination has become so prevalent that it has led to the development of nursing protocols for using oCHG even for non-intubated patients despite the lack of evidence that this measure provides any benefit [7,8].

A Cochrane review of randomized controlled trials reported a decreased risk of VAP in patients treated with oCHG [9]. However, there is growing concern that this finding was the result of falsely negative respiratory cultures in patients treated with oCHG [10,11]. Apart from possibly missing a pneumonia diagnosis, the risks of using of oCHG are numerous. For example, erosive lesions, ulcerations, white/yellow plaque formation, and bleeding mucosa have been reported in as many as 10% of patients receiving oCHG [12]. The widespread use of antibacterial coating on medical devices such as central lines and urinary catheters has led to a surge in confirmed cases of chlorhexidine anaphylaxis [13 –16]. Bacterial resistance to CHG has been reported widely and has resulted in several outbreaks and healthcare-associated infections [17]. Additionally, a recent large single-center observational study of intubated and non-intubated patients both in and out of the ICU found that exposure to oCHG was associated with a higher risk of death, and the association was strongest in patients with the lowest risk of death [18]. Most concerning, however, are the two recent meta-analyses of randomized controlled trials that found that oCHG may actually be associated with an increase in the mortality rate [19,20].

These findings indicate that further studies of decontamination with oCHG are warranted. The majority of studies with oCHG have been limited to culture-defined VAP, cardiac surgery patients, short ICU stays, and single centers. The purpose of this multi-center study was to evaluate the effect of oCHG decontamination on the incidence of pneumonia, sepsis, and death in critically ill patients.

Patients and Methods

In 2017, the first public release of the eICU Collaborative Research Database (eICU-CRD) ushered in an unparalleled era of high-quality, multi-center, ICU big data. The database was released as the result of a collaboration between Philips Healthcare's eICU Research Institute and the Massachusetts Institute of Technology Laboratory for Computational Physiology (LCP), which also produced the single-center Medical Information Mart for Intensive Care (MIMIC) database. The eICU-CRD builds on the success of MIMIC and greatly expands the capabilities for physiologic research by including multi-center data. The current release of the eICU-CRD contains data collected from 186 ICUs throughout the United States in 2014 and 2015 [21].

For this study, the eICU-CRD 2.0 was queried for patients admitted to the ICU for at least 48 hours. The primary outcome was death during the ICU stay. The secondary outcomes were a diagnosis of pneumonia or sepsis. Patients with pneumonia or sepsis diagnosed within the first 48 hours of ICU admission were excluded from their respective outcome analysis, as oCHG exposure is unlikely to affect this outcome after such a short time. Additionally, we sought to differentiate admissions attributable to pneumonia or sepsis from the nosocomial development of these diagnoses after ICU admission. Univariable analysis with a χ² test was performed for each outcome using age, gender, race, Acute Physiology Score (APS) III, Acute Physiology Age Chronic Health Evaluation (APACHE) IVa score, surgical specialty admission status, intubation status, teaching hospital status, hospital bed size, hospital region, ICU length of stay, and oCHG order. Multivariable logistic regression was performed using all univariable outcomes with p < 0.05 in order to limit variable selection bias. Continuous variables were compared using the mean with the standard deviation and Student's t-test. Statistical analysis was performed using SPSS Statistics V. 24 (IBM Corp., Armonk, NY). A p value of <0.05 was considered statistically significant. The Institutional Review Board of the University of Miami waived the requirement for approval of this study, as the eICU-CRD database contains deidentified, publicly available data, and its use is not considered human subjects research.

Results

Of the 64,904 patients from 186 hospitals, 22.1% (n = 14,333) had oCHG ordered, and 24.0% (n = 15,559) were intubated at some point during their ICU stay. From the intubated patients, 47.4% (n = 6,790) had oCHG ordered, while 17.3% (n = 8,769) of the non-intubated patients had oCHG ordered. The overall mortality rate was 6.9% (n = 4,449), but the mortality rate for patients with oCHG ordered was 10.6% (n = 1,518; p < 0.001). After controlling for confounding factors through multivariable logistic regression, oCHG remained an independent risk factor for death (OR 1.25; 95% CI 1.16–1.34) (Table 1). After excluding patients with a pneumonia diagnosis before 48 hours in the ICU (14.7%, n = 9,511), the overall pneumonia incidence was 2.6% (n = 1,431), and the rate of pneumonia in patients with oCHG ordered was 4.2% (n = 517; p < 0.001). However, multivariable logistic regression revealed no significant difference in the risk of pneumonia with oCHG (OR 0.97; 95% CI 0.85–1.09). After excluding patients with a sepsis diagnosis before 48 hours in the ICU (18.7%; n = 12,153), the overall rate of sepsis was 1.8% (n = 949), and for patients with oCHG ordered, the rate was 3.3% (n = 388; p < 0.001). After controlling for other confounders, an oCHG order was independently associated with higher odds of sepsis (OR 1.37; 95% CI 1.19–1.59).

Table 1.

Results of Multivariable Logistic Regression for Each Outcome

Characteristic		ICU Death		Pneumonia		Sepsis
Characteristic		OR (95% CI)	P Value	OR (95% CI)	P Value	OR (95% CI)	P Value
Chlorhexidine		1.25 (1.16–1.34)	<0.001	0.97 (0.85–1.09)	0.58	1.37 (1.19–1.59)	<0.001
Intubated		1.23 (1.14–1.32)	<0.001	1.56 (1.37–1.77)	<0.001	1.12 (0.96–1.30)	0.162
Surgical patient		0.79 (0.70–0.89)	<0.001	0.59 (0.48–0.74)	<0.001	0.57 (0.45–0.72)	<0.001
Cardiac surgery patient		0.49 (0.38–0.64)	<0.001	0.86 (0.58–1.29)	0.473	0.83 (0.51–1.35)	0.453
Trauma surgery patient		^a		1.58 (1.11–2.25)	0.011	^a
Age group (years)	<54	Reference		Reference		^a
	54–65	1.12 (1.01–1.24)	0.026	0.94 (0.81–1.09)	0.396
	66–75	1.15 (1.03–1.27)	0.011	0.89 (0.75–1.06)	0.18
	≥75	1.30 (1.17–1.44)	<0.001	0.78 (0.65–0.93)	0.005
Female		^a		0.85 (0.76–0.95)	0.004	^a
Race	White	Reference		Reference		Reference
	Black	0.90 (0.81–1.00)	0.046	0.69 (0.57–0.85)	<0.001	0.85 (0.67–1.08)	0.177
	Hispanic	0.70 (0.58–0.85)	<0.001	1.56 (1.23–1.98)	<0.001	0.94 (0.66–1.32)	0.71
	Other	1.07 (0.95–1.20)	0.298	0.85 (0.69–1.05)	0.131	1.17 (0.93–1.47)	0.189
APS group	<33	Reference		Reference		Reference
	33–45	0.93 (0.77–1.11)	0.412	0.87 (0.68–1.12)	0.276	1.38 (1.01–1.87)	0.043
	46–63	0.94 (0.75–1.16)	0.543	0.96 (0.70–1.33)	0.818	1.57 (1.08–2.29)	0.019
	≥64	1.31 (1.02–1.68)	0.034	0.94 (0.63–1.40)	0.763	2.20 (1.41–3.42)	<0.001
APACHE group	<44	Reference		Reference		Reference
	44–58	1.76 (1.46–2.14)	<0.001	1.22 (0.95–1.58)	0.123	1.16 (0.86–1.58)	0.335
	59–77	3.21 (2.54–4.05)	<0.001	1.30 (0.93–1.82)	0.124	1.24 (0.86–1.81)	0.251
	≥78	6.63 (5.08–8.66)	<0.001	1.47 (0.98–2.22)	0.065	1.26 (0.82–1.95)	0.294
Teaching hospital		^a		^a		^a
Hospital region	South	Reference		Reference		Reference
	Midwest	0.76 (0.70–0.83)	<0.001	0.77 (0.66–0.90)	0.001	1.02 (0.84–1.23)	0.849
	West	1.27 (1.16–1.40)	<0.001	1.96 (1.68–2.30)	<0.001	3.54 (2.94–4.26)	<0.001
	Northeast	1.26 (1.12–1.41)	<0.001	2.43 (2.02–2.91)	<0.001	2.12 (1.66–2.69)	<0.001
	Unknown	0.88 (0.73–1.06)	0.184	1.73 (1.26–2.37)	0.001	3.28 (2.49–4.34)	<0.001
Hospital beds	≥500	Reference		Reference		^a
	250–499	0.94 (0.86–1.02)	0.145	0.74 (0.63–0.86)	<0.001
	100–249	0.92 (0.84–1.01)	0.063	1.17 (1.00–1.36)	0.046
	<100	0.86 (0.72–1.03)	0.105	1.18 (0.89–1.57)	0.252
	Unknown	1.06 (0.93–1.21)	0.411	0.71 (0.55–0.92)	0.009
ICU LOS group (days)	2.0–2.6	Reference		Reference		Reference
	2.7–3.6	1.04 (0.93–1.15)	0.519	3.72 (2.32–5.96)	<0.001	7.49 (3.99–14.06)	<0.001
	3.7–5.9	1.17 (1.06–1.29)	0.002	10.85 (7.00–16.80)	<0.001	13.61 (7.38–25.10)	<0.001
	≥6.0	1.60 (1.45–1.76)	<0.001	50.19 (32.81–76.77)	<0.001	58.39 (32.11–106.18)	<0.001

Not significant on univariable analysis.

APACHE = Acute Physiology Age Chronic Health Evaluation; APS = Acute Physiology Score; CI = confidence interval; ICU = intensive care unit; LOS = length of stay; OR = odds ratio.

The most common age group in this study was 75 years or older (29.4%; n = 19,094). These older patients had a higher mortality rate (8.3%; n = 1,583; p < 0.001) and a lower rate of pneumonia (2.1%; n = 337; p < 0.001) than younger patients. The difference in rates of sepsis by age was not statistically significant (Table 2). The reference age group for the logistic regression was <54 years. Patients aged 75 years or older had a greater likelihood of death (OR 1.30; 95% CI 1.17–1.44) and a lower risk of pneumonia (OR 0.78; 95% CI 0.65–0.93) (Table 2) than the reference group. A surgical service admitted 14.3% (n = 9,293) of the patients, and their mortality rate was 4.4% (n = 411; p < 0.001). These patients also had lower rates of pneumonia (2.0%; n = 183; p < 0.001) and sepsis (1.2%; n = 102; p < 0.001; Table 2). After multivariable logistic regression, the odds were reduced for death (OR 0.79; 95% CI 0.70–0.89), pneumonia (OR 0.59; 95% CI 0.48–0.74), and sepsis (OR 0.57; 95% CI 0.45–0.72). Cardiac surgery patients also had reduced odds for death (OR 0.49; 95% CI 0.38–0.64), while the other outcomes were not statistically significant (pneumonia OR 0.86; 95% CI 0.58–1.29), and sepsis (OR 0.83; 95% CI 0.51–1.35) (Table 1). Patients admitted to the trauma surgery service represented only 2.1% (n = 1,363) of the patients in this study. The rates of mortality (5.9%; n = 81; p = 0.178) and sepsis (1.4%; n = 17; p = 0.228) were not significantly different for trauma patients, although the rate of pneumonia was elevated (4.2%; n = 55; p < 0.001) as were the adjusted odds (OR 1.58; 95% CI 1.11–2.25) (Table 2) compared with all other non-trauma patients.

Table 2.

Overall Patient and Hospital Characteristics by Outcome with Significance by χ² Test for Characteristic vs. Outcome

Characteristic		Total (%)	ICU Mortality Rate (%)		Pneumonia		Sepsis
Characteristic		n (%)	n (%)	P Value	n (%)	P Value	n (%)	P Value
Total		64,904 (100.0)	4,449 (6.9)		1,431 (2.6)		949 (1.8)
Chlorhexidine		14,333 (22.1)	1,518 (10.6)	<0.001	517 (4.2)	<0.001	388 (3.3)	<0.001
Intubated		15,559 (24.0)	1,908 (12.3)	<0.001	627 (4.8)	<0.001	388 (3.2)	<0.001
Surgical patient		9,293 (14.3)	411 (4.4)	<0.001	183 (2.0)	<0.001	102 (1.2)	<0.001
Cardiac surgery patient		2,830 (4.4)	71 (2.5)	<0.001	34 (1.2)	<0.001	21 (0.8)	<0.001
Trauma surgery patient		1,363 (2.1)	81 (5.9)	0.178	55 (4.2)	<0.001	17 (1.4)	0.228
Age group (years)	<54	15,118 (23.3)	785 (5.2)	<0.001	376 (2.9)	<0.001	213 (1.7)	0.098
	54–65	16,606 (25.6)	1,067 (6.4)		398 (2.8)		256 (1.9)
	66–75	14,086 (21.7)	1,014 (7.2)		320 (2.7)		229 (2.0)
	≥75	19,094 (29.4)	1,583 (8.3)		337 (2.1)		251 (1.7)
Female		29,555 (45.5)	1,976 (6.7)	0.119	576 (2.3)	<0.001	404 (1.7)	0.125
Race	White	49,882 (76.9)	3,469 (7.0)	0.002	1,120 (2.7)	<0.001	737 (1.8)	<0.001
	Black	7,891 (12.2)	496 (6.3)		121 (1.8)		84 (1.3)
	Hispanic	2,381 (3.7)	129 (5.4)		87 (4.3)		37 (2.0)
	Other	4,750 (7.3)	355 (7.5)		103 (2.5)		91 (2.4)
APS group	<33	16,036 (24.7)	321 (2.0)	<0.001	212 (1.5)	<0.001	108 (0.7)	<0.001
	33–45	15,921 (24.5)	519 (3.3)		231 (1.7)		171 (1.3)
	46–63	16,523 (25.5)	988 (6.0)		375 (2.7)		243 (1.9)
	≥64	16,424 (25.3)	2,621 (16.0)		613 (4.6)		427 (3.7)
APACHE group	<44	15,509 (23.9)	251 (1.6)	<0.001	208 (1.5)	<0.001	111 (0.8)	<0.001
	44–58	15,997 (24.7)	478 (3.0)		255 (1.8)		171 (1.2)
	59–77	17,074 (26.3)	1,031 (6.0)		383 (2.7)		267 (2.0)
	≥78	16,324 (25.2)	2,689 (16.5)		585 (4.5)		400 (3.6)
Teaching hospital		20,381 (31.4)	1,409 (6.9)	0.69	456 (2.6)	0.804	284 (1.7)	0.085
Hospital region	South	22,625 (34.9)	1,582 (7.0)	<0.001	416 (2.1)	<0.001	229 (1.2)	<0.001
	Midwest	23,792 (36.7)	1,254 (5.3)		337 (1.6)		236 (1.2)
	West	10,444 (16.1)	927 (8.9)		367 (4.1)		296 (3.9)
	Northeast	5,467 (8.4)	511 (9.4)		245 (5.8)		115 (2.8)
	Unknown	2,576 (4.0)	175 (6.8)		66 (3.0)		73 (3.6)
Hospital beds	≥500	27,933 (43.0)	2,005 (7.2)	<0.001	670 (2.7)	0.002	404 (1.7)	0.099
	250–499	15,118 (23.3)	1,119 (7.4)		300 (2.3)		226 (1.9)
	100–249	12,829 (19.8)	773 (6.0)		299 (2.8)		204 (2.0)
	<100	2,930 (4.5)	154 (5.3)		61 (2.7)		30 (1.4)
	Unknown	6,094 (9.4)	398 (6.5)		101 (1.9)		85 (1.7)
ICU LOS (d)	2.0–2.6	16,214 (25.0)	713 (4.4)	<0.001	22 (0.2)	<0.001	11 (0.1)	<0.001
	2.7–3.6	16,235 (25.0)	834 (5.1)		82 (0.6)		84 (0.6)
	3.7–5.9	16,228 (25.0)	1,095 (6.8)		245 (1.8)		160 (1.2)
	≥6.0	16,227 (25.0)	1,807 (11.1)		1,082 (8.3)		694 (5.6)

APACHE = Acute Physiology Age Chronic Health Evaluation; APS = Acute Physiology Score; ICU = intensive care unit; LOS = length of stay.

For patients with pneumonia and an order for oCHG, the mean time from ICU admission to pneumonia diagnosis was 7.0 ± 5.9 days. For patients with pneumonia without an order for oCHG, the mean time to pneumonia diagnosis was 5.6 ± 4.6 days (p < 0.001). For patients with sepsis and an order for oCHG, the mean time from ICU admission to sepsis diagnosis was 6.5 ± 5.7 days. For patients with sepsis without an order for oCHG, the mean time to sepsis diagnosis was 5.5 ± 5.7 days (p = 0.003).

Discussion

This study represents the largest single evaluation of patients receiving oCHG decontamination in ICUs across hospitals in the United States. The finding of an increased odds of death associated with oCHG is consistent with two other recent meta-analyses of randomized controlled trials [19,20]. The earliest studies of oCHG for the prevention of VAP were performed on cardiac surgery patients, the majority of whom were in the ICU for <48 hours [3,22,23]. With such a short length of stay, the risk of developing nosocomial pneumonia was much lower than that for other ICU patients [10]. This selection bias was addressed in this study first by excluding patients who were in the ICU for <48 hours. This resulted in a sample that was more reflective of a truly critically ill patient population. Additionally, all surgical and cardiac surgery patients were identified, and their length of stay was stratified. Using these variables in the logistic regression model revealed a decreased risk for death of surgical patients as well as patients with shorter lengths of stay.

Another limitation of the early studies of oCHG for the prevention of VAP was their lack of blinding. This is particularly important in the diagnostic criteria for VAP, which includes the subjective determination of the “purulence” of the respiratory secretions. When this determination is made by individuals with an awareness of the oCHG administration status of each patient, there is the potential to introduce ascertainment bias [24]. This study has the advantage that a diagnosis of pneumonia was used from the eICU-CRD as documented by clinical staff without any awareness of future comparative use of the data. Thus, although the clinical staff were aware of secretions and the oCHG administration status, they were unaware of any future comparative use of these data. This could reduce the likelihood of subconscious bias by clinicians in this study.

The lack of a valid and reliable definition of VAP has resulted in wide differences in VAP rates being reported from across the world. The rates of VAP in the U.S. are more than 20 times higher than in the rest of the world in some cases [25,26]. By using data from 186 hospitals across the U.S. in this study, there is less likelihood that a single center's or region's practice patterns would be accounting for the differences in outcomes observed.

A few of the early studies [4,22] of oCHG added respiratory cultures to the diagnostic criteria of VAP in order to reduce the subjective nature of the diagnosis. However, this further complicates reporting the performance of oCHG, as the primary source of falsely positive respiratory cultures is bacteria from the upper airway. Patients receiving an oral antiseptic should have a reduced bacterial load resulting in fewer false-positive VAP cultures. This could explain why the reduced VAP rate with oCHG has not been shown to lead to overall improvement in outcomes [11,27]. This study found that there was no difference in the rate of pneumonia with oCHG use. The clinical diagnosis of pneumonia used in this study was as documented in the ICU stay by clinical staff and is separate from diagnostic coding for professional billing or hospital reimbursement purposes [21]. Thus, the diagnosis of pneumonia in this study should be less influenced by administrative pressures or research biases.

The time to diagnosis also is an important consideration in this study. This was about 24 hours later for both pneumonia and sepsis in patients receiving oCHG and was statistically significant. This could be reflective of the fact that patients in the ICU longer have a greater likelihood of receiving oCHG. This is demonstrated in the multivariable logistic regression that shows ICU length of stay ≥6 days had the highest odds ratio for death, pneumonia, and sepsis. Future studies should control for oCHG being ordered simply because the patients were sicker and in the ICU longer. By using the severity of illness scores and length of stay in the multivariable logistic regression, this study addresses these considerations to the greatest extent possible given the available data.

This study is limited by the data in the eICU-CRD. These data were intended for use during routine clinical care and may contain inconsistencies that are inconsequential for care [21]. However, these inconsistencies may introduce information bias as a result of measurement error and misclassification. The drugs recorded in the eICU-CRD reflect the active medication orders but do not necessarily reflect administration to the patient. Additionally, the criteria used by each center for the diagnosis of sepsis are unknown and could be influenced by confounders that are unaccounted for. The current version of the database does not contain enough information to determine the cause of death or to confirm the diagnoses using microbiological data; however, future versions of the database will likely include these data. Despite these limitations, this study represents the largest multi-center study to date of oCHG throughout the U.S. The findings uncovered here raise important questions about the effectiveness of oCHG.

In conclusion, this study demonstrates that chlorhexidine mouthwash is associated with an increased risk of death and sepsis in ICU patients while not significantly affecting the risk of pneumonia. Additional randomized, multi-center trials are needed to better understand the effect of oral chlorhexidine on outcomes.

Footnotes

Authors' Contributions

RR, SB, HS, and JP contributed to the study design. JP and RR carried out the statistical analysis and drafted the manuscript. All authors contributed to the critical review and manuscript revisions and approve the final version.

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

Author Disclosure Statement

No competing financial interests exist.

References

Marino

MD.

Marino's The Little ICU Book. Second edition. Philadelphia. LWW; 2016.

Fiser

The ABSITE Review. Fifth edition. Philadelphia. LWW; 2016.

Segers

, Speekenbrink

RGH

, Ubbink

, et al. Prevention of nosocomial infection in cardiac surgery by decontamination of the nasopharynx and oropharynx with chlorhexidine gluconate: A randomized controlled trial. JAMA, 2006; 296:2460–2466.

Fourrier

, Cau-Pottier

, Boutigny

, et al. Effects of dental plaque antiseptic decontamination on bacterial colonization and nosocomial infections in critically ill patients. Intensive Care Med, 2000; 26:1239–1247.

Özçaka Ö, Başoğlu

, Buduneli

, et al. Chlorhexidine decreases the risk of ventilator-associated pneumonia in intensive care unit patients: A randomized clinical trial. J Periodontal Res, 2012; 47:584–592.

Institute for Healthcare Improvement. Getting started kit: Prevent ventilator-associated pneumonia: How-to guide. Crit Care Nurs Q, 2006; 29:157–173.

Oral Care for Acutely and Critically Ill Patients. Crit Care Nurse, 2017; 37:e19–e21.

Munro

, Grap

, Sessler

, et al. Preintubation application of oral chlorhexidine does not provide additional benefit in prevention of early-onset ventilator-associated pneumonia. Chest, 2015; 147:328–334.

Hua

, Xie

, Worthington

, et al. Oral hygiene care for critically ill patients to prevent ventilator-associated pneumonia. Cochrane Database Syst Rev Available at: https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD008367.pub3 Cited May 15, 2019.

10.

Klompas

Oropharyngeal decontamination with antiseptics to prevent ventilator-associated pneumonia: Rethinking the benefits of chlorhexidine. Semin Respir Crit Care Med, 2017; 38:381–390.

11.

Bouadma

, Klompas

Oral care with chlorhexidine: Beware!. Intensive Care Med, 2018; 44:1153–1155.

12.

Plantinga

, Wittekamp

BHJ

, Leleu

, et al. Oral mucosal adverse events with chlorhexidine 2% mouthwash in ICU. Intensive Care Med, 2016; 42:620–621.

13.

Garvey

, Krøigaard

, Poulsen

, et al. IgE-mediated allergy to chlorhexidine. J Allergy Clin Immunol, 2007; 120:409–415.

14.

Kow

, Low

, Ruben

, et al. Life-threatening chlorhexidine anaphylaxis: A case report. Malays Orthop J, 2017; 11:72–74.

15.

Moka

, Argyra

, Siafaka

, Vadalouca

. Chlorhexidine: Hypersensitivity and anaphylactic reactions in the perioperative setting. J Anaesthesiol Clin Pharmacol, 2015; 31:145–148.

16.

Teixeira de Abreu

, Ribeiro de Oliveira

, Teixeira de Abreu

, et al. Perioperative anaphylaxis to chlorhexidine during surgery and septoplasty. Case Rep Otolaryngol, 2017; 2017:9605804.

17.

Kampf

Acquired resistance to chlorhexidine: Is it time to establish an “antiseptic stewardship” initiative?. J Hosp Infect, 2016; 94:213–227.

18.

Deschepper

, Waegeman

, Eeckloo

, et al. Effects of chlorhexidine gluconate oral care on hospital mortality: A hospital-wide, observational cohort study. Intensive Care Med, 2018; 44:1017–1026.

19.

Klompas

, Speck

, Howell

, et al. Reappraisal of routine oral care with chlorhexidine gluconate for patients receiving mechanical ventilation: Systematic review and meta-analysis. JAMA Intern Med, 2014; 174:751–761.

20.

Price

, MacLennan

, Glen

, SuDDICU

Collaboration

. Selective digestive or oropharyngeal decontamination and topical oropharyngeal chlorhexidine for prevention of death in general intensive care: Systematic review and network meta-analysis. BMJ, 2014; 348:g2197.

21.

Pollard

, Johnson

AEW

, Raffa

, et al. The eICU Collaborative Research Database, a freely available multi-center database for critical care research. Sci Data, 2018; 5:180178.

22.

DeRiso

2nd , Ladowski

, Dillon

, et al. Chlorhexidine gluconate 0.12% oral rinse reduces the incidence of total nosocomial respiratory infection and nonprophylactic systemic antibiotic use in patients undergoing heart surgery. Chest, 1996; 109:1556–1561.

23.

Houston

, Hougland

, Anderson

, et al. Effectiveness of 0.12% chlorhexidine gluconate oral rinse in reducing prevalence of nosocomial pneumonia in patients undergoing heart surgery. Am J Crit Care, 2002; 11:567–570.

24.

Klompas

Interobserver variability in ventilator-associated pneumonia surveillance. Am J Infect Control, 2010; 38:237–239.

25.

Nora

, Póvoa

. Antibiotic consumption and ventilator-associated pneumonia rates, some parallelism but some discrepancies. Ann Transl Med, 2017; 5:450.

26.

Rosenthal

, Al-Abdely

, El-Kholy

, et al. International Nosocomial Infection Control Consortium Report, data summary of 50 countries for 2010–2015: Device-associated module. Am J Infect Control, 2016; 44:1495–1504.

27.

Klompas

The paradox of ventilator-associated pneumonia prevention measures. Crit Care, 2009; 13:315.