International Journal of Nursing Studies 48 (2011) 681–688

Contents lists available at ScienceDirect

International Journal of Nursing Studies

journal homepage: www.elsevier.com/ijns

Effects of three approaches to standardized oral hygiene to reduce bacterial

colonization and ventilator associated pneumonia in mechanically ventilated
patients: A randomised control trial
a,b, c d e f
A.M. Berry *, P.M. Davidson , J. Masters , K. Rolls , R. Ollerton
a
Intensive Care Unit, Westmead Hospital, Sydney, Australia
b
School of Nursing, University of Western Sydney, Australia
c d
School of Nursing and Midwifery, Curtin University of Technology, Centre for Cardiovascular and Chronic Care, Sydney, NSW, Australia Infection
Control Unit, Blacktown Mt Druitt Hospital, Sydney, Australia
e f
NSW Health Intensive Care Coordination and Monitoring Unit, Sydney, Australia
University of Western Sydney, Sydney, Australia
ARTICLE INFO ABSTRACT

Article history: Background: Ventilator associated pneumonia remains an important concern in the intensive care unit (ICU).
Received 11 February 2009 An increasing body of evidence shows that mortality and morbidity can be reduced by implementing a range
Received in revised form 14 November 2010 of preventive strategies, including optimizing oral hygiene.
Accepted 19 November 2010
Aim: The aim of this feasibility study was to test two oral hygiene strategies on the effects of microbial
Keywords: colonization of dental plaque with respiratory pathogens (primary outcome) and incidence of ventilator
Oral hygiene associated pneumonia (secondary outcome).
Ventilator associated pneumonia Methods: A single blind randomised comparative study was conducted in a 20-bed adult intensive care unit
Critical care
in a university hospital. Patients with an expected duration of mechanical ventilation more than 48 h were
Intensive care unit
eligible. Patients were randomised to one of three study regimens (Group A control, second hourly oral rinse
with sterile water, Group B sodium bicarbonate mouth wash second hourly, and Group C twice daily
irrigations with chlorhexidine 0.2% aqueous oral rinse and second hourly irrigations with sterile water). All
study options included cleaning with a toothbrush and non foaming toothpaste.

Results: Data from a total of 109 patients were analyzed. Group A 43, Group B 33 and Group C 33 (mean
age: 58 17 years, simplified acute physiology score II: 44 14 points). On admission no significant differences
were found between groups for all clinical data. While Group B showed a greater trend to reduction in
bacterial colonization no significant differences could be demonstrated at Day 4 of admission (p = 0.302).
The incidence of ventilator associated pneumonia was evenly spread between Groups B and C (5%) while
Group A was only 1%.
Conclusions: While a number of studies have advocated the use of various mouth rinses in reducing
colonization of dental plaque a standardized oral hygiene protocol which includes the use of mechanical
cleaning with a toothbrush may be a factor in the reduction of colonization of dental plaque with respiratory
pathogens. This feasibility study provides data to inform future adequately powered studies.

2010 Elsevier Ltd. All rights reserved.

* Corresponding author. Tel.: +61 2 9845 9450; fax: +61 2 9845 9395. E-mail address: angela_berry@wsahs.nsw.gov.au (A.M. Berry).

0020-7489/$ – see front matter 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijnurstu.2010.11.004
 What is already known about the topic?

Ventilator associated pneumonia is associated with significant morbidity and mortality.

682 A.M. Berry et al. / International Journal of Nursing Studies 48 (2011) 681–688
 Oral hygiene is an important nursing activity to prevent adverse outcomes in the critically ill, yet evidence to support oral hygiene in the
critically is sparse.

What this paper adds

Description of a randomised controlled trial to test approaches to oral hygiene.

Although this study failed to describe differences between groups, it suggests that standardized protocols for oral hygiene may improve
outcomes in the critically ill.

1. Introduction

Ventilator associated pneumonia (VAP) is a major source of morbidity and mortality in the intensive care unit (ICU) (Hingston et al., 2010;
Ganz et al., 2009). Reported mortality rates range from 24% to 50% with potentially higher figures in the immunocompromised patient group or
when multi-resistant organisms are involved (Chaste and Fagon, 2002).Ventilator associated pneumonia is characterized by a constellation of
clinical and diagnostic criteria, resulting from pulmonary inflammation following intubation (Chas-tre and Fagon, 2002). The mechanical
process of intubation not only compromises the natural barrier between the oropharynx and trachea but may also facilitate the entry of bacteria
into the lungs by pooling and leakage of con-taminated secretions around the endotracheal tube cuff (Fitch et al., 1999). A major cause of
contamination of oral secretions is due to colonization of dental plaque with respiratory pathogens (Fourrier et al., 1998; Scannapieco, 1996).
Dental plaque is the product of adherence on the tooth surface of a diverse accumulation of microorganisms and their end-products. These
organisms may pose a serious health risk to the critically ill patient (Bagg et al., 1999). Dental plaque adheres tenaciously to the teeth and is not
easily removed (Spraycar, 1995). Critically ill patients are at increased risk of dental plaque accumulation. This is due to the development of
increased levels of proteases in their oral secretions which deplete a glycoprotein called fibronectin. Decreased levels of this substance enhance
the adherence of organisms such as Pseudomonas aeruginosa to buccal epithelial cells (Gibbons, 1989).
Oral hygiene, together with subglottal suctioning and head of bed elevation has been proposed as an important strategy in reducing the incidence
of VAP (Tablan et al., 2003). Our systematic review of oral hygiene practice identified the effectiveness of various oral hygiene routines but none
offered a definitive solution to this important nosocomial infection (Berry etal., 2007). However an oral hygiene regimen which does not include
mechanical cleaning of the oral cavity results in accumulation of dental plague and resultant colonization by pathogens (Fourrieret al., 2005).
Despite this a number of studies have been limited to evaluating the effects of only antiseptic oral rinses. Fourrier et al. (2005) compared 0.2%
chlorhexidine gel with a placebo gel. Toothbrushing was not permitted in the study. They reported no difference in the incidence of
VAP and Pseudomonas, Acinetobacter and Enterobacter species identified on initial dental plaque culture were not eliminated by the antiseptic
agent. In contrast Koeman et al. (2006) determined that topical oral decontamination with chlorhexidine or chlorhexidine/colistin reduced the
incidence of VAP. Again, this study did not include the use of toothbrushing. The study gels were applied with a gloved fingertip after cleaning
with a saline soaked gauze. Both these studies used a benign gel as a control so in effect did not provide oral hygiene to the control group. Other
studies supporting the relationship between oral hygiene using chlorhexidine mouth rinse have shown a decreased trend in the development of
VAP (Grap et al., 2004; De Riso et al., 1996; Genuit et al., 2001). Chlorhexidine gluconate mouthwash is an antiplaque agent with potent
antimicrobial activity that is effective at low concentrations while allowing little, if any, absorption from the gastrointestinal tract (Dodd et al.,
2000; Jones, 2000). Due to the incongruity in outcomes related to the use of chlorhexidine we aimed to compare chlorhexidine with an active
oral cleaning solution instead of a benign solution. Sodium bicarbonate mouth rinse is a cleaning agent which has the ability to dissolve mucus
and loosen oral debris (Carl et al., 1999; Dodd et al., 2000). It has reported similar efficacy as hydrogen peroxide in the treatment of mucositis
due to radiation (Dudjak, 1987). In an earlier study by Fourrier et al. (2000), sodium bicarbonate was used as a control substance compared to a
0.2% chlorhexidine gel. Although the frequency of colonization of plaque on Day 5 was higher in the sodium bicarbonate group, by Day 10 no
significant difference could be detected between groups. Again, mechanical cleaning with a toothbrush was not included in the study protocol. To
date, there are no reports of randomised controlled studies that support the use of sodium bicarbonate over any other mouth rinse in patients who
are critically ill. While untested in the intensive care environment, the mechanical action of a toothbrush to remove dental plaque has been
recommended and in fact is considered a standard in oral hygiene by several authors (Franklin et al., 2000; Griffiths et al., 2000; Pearson, 2006).
The elimination of dental plaque has been demonstrated to be more effective by the use of a toothbrush compared with foam swabs (Pearson,
2006). The oral cavity of the mechanically ventilated patient is often difficult to access due to the presence of an endotrachael tube and
occasionally an orogastric tube. Therefore a small, pedia-tric toothbrush has been recommended by Griffiths et al. (2000) because it can reach all
aspects of the mouth and can be used to clean the tongue and gums in edentulous patients. While the incidence of bacteraemia associated with
tooth brushing in healthy persons is rare (Hartzell et al., 2005), the incidence in critically ill patients is less certain. Tooth brushing was included
in the study protocol because it is standard oral care and because it assists in removing plaque and other debris from tooth surfaces which can
readily become a host for pathogens (American Dental Association, 2005; Berry et al., 2007). The protocol for the study was developed
following extensive consulta-tion with experienced periodontologists and microbiolo-
gists including those who specialise in dental and critical care practice as well as intensive care specialists, both nursing and medical and a
biomedical statistician.

2. Study aim

This feasibility study sought to assess a standardized oral hygiene protocol on the colonization of dental plaque and gums with respiratory
pathogens and the incidence of ventilator associated pneumonia. The results of this study would be used to test the study protocol and to
determine power calculations for a subsequent large, multi-centered randomised control trial.
3. Methods

3.1. Study setting and patients

The site was a 1000 bed, adult metropolitan university hospital. The 20 bed surgical-medical ICU provides mechanical ventilation to over
800 adult patients annually. A range of protocols were present in the ICU ensuring that the baseline conditions of standard care were the same
across all three treatment arms. High dependency patients were managed in separate high dependency units within the hospital. Due to a
concurrent antibiotic cycling study implemented in the study setting, there was no significant difference between the groups in relation to
antibiotic administration. Selective decontamination of the digestive tract or continuous subglottic suctioning was not standard practice in the
study setting during the randomisation period. Subglottic suctioning was attended second hourly. All intubated patients admitted to the unit were
con-sidered for inclusion in the study provided they met the following criteria: able to be randomised within 12 h of intubation, aged over 15
years and next of kin able to give informed consent. Patients were ineligible for study participation if they: required specific oral hygiene
procedures in relation to facio-maxillary or dental trauma/surgery; had been in the ICU previously during the current period of hospitalization;
received irradiation or chemotherapy on admission to the ICU or in the preceding 6 weeks; or suffered an autoimmune disease. Informed consent
was obtained for all subjects and agreement to participate could be withdrawn at any time.

3.2. Study design and blinding

Due to the difficulty in achieving identical appearance in study agents, the study could not be performed in a double-blinded manner. The
hospital pharmacy depart-ment was responsible for the storage of all study solutions. On enrollment the patients were randomised into one of
three groups according to a balanced randomisation table prepared by a biostatistician. The study packs were assembled by the principal
investigator and all study packs were identical in outward appearance. Treatment allocations remained blinded until the study pack was opened
by the attending nurse. The microbiologist super-vising bacteriological sampling was blinded to the treat-ment allocation code. Designation of
VAP radiological criteria was verified by radiologists blinded to treatment allocation. Other VAP criteria were ascertained from a review of the
patients’ records and validation of diagnosis of VAP was obtained from the ICU data base.
The study commenced following implementation of a training program provided to the ICU registered nurses. The training program involved
detailed description of the study procedure, techniques for oral hygiene and use of an oral assessment score. The study agents selected were: (A)
sterile water rinsed second hourly; (B) sodium bicarbonate mouth wash rinsed second hourly; or (C) twice daily irrigation with chlorhexidine
0.2% aqueous oral rinse with second hourly irrigation with sterile water. Mouth rinses were applied using a curved tip dental syringe. All
treatment options included a comprehensive cleaning of the mouth using a soft, pediatric toothbrush three times a day. Using a cuff manometer,
as specified by the ICU protocol, the occlusive pressure of the endotracheal cuff was checked immediately prior to initiating mouth care and the
patient’s subglottic area was suctioned using a pliable suction catheter on completion of the procedure. This protocol was used until the patient
was extubated, decannulated, if they had a tracheostomy, or upon discharge from the ICU or death.
On enrolment and immediately prior to initiating oral cleaning, oral swabs were taken to determine baseline microbial status. The oral swabs
were taken from the surfaces of both top and bottom teeth or the surface of gums in edentulous patients. The samples were plated on
chromogenic agar for normal respiratory flora and combination agar for pathogenic flora. These included multiresistant Staphylococcus,
multiresistant Acinetobacter, vancomycin resistant Enterococcus, Pseudomonas and others. Repeated oral swabs were taken at day four to assess
5
effectiveness of the oral hygiene regimen. Semi-quantitative analysis classified positive oral colonization as 10 cfu/ml.

3.3. Study outcome measures

The primary outcome of the feasibility study was the change from Day 1 to Day 4 in microbial colonization of dental plaque or gums in
edentulous patients, of mechani-cally ventilated patients in the ICU. The secondary outcome was the incidence of VAP during the period of
mechanical ventilation.

3.4. Data collection

Data collected relating to baseline characteristics included age, gender, acute physiology and chronic health evaluation (APACHE) score,
simplified acute physiology score (SAPS), past medical history, primary diagnosis, presence of dentures and location of endotracheal tube
(oral/nasal), on enrolment. APACHE and SAPS scores include a number of physiological variables which provide a score used to stratify patients
by severity and gives an estimation of probability of survival. APACHE scores greater than 24 indicate a poor prognosis and SAPS scores greater
than 34 indicate a 20% probability of death. Results of daily blood tests and chest X-rays, done as part of the
usual management of the patient, were recorded to determine, with other parameters, the diagnosis of VAP. Data relating to ventilation hours,
length of ICU stay, medications including antibiotics administered in ICU and incidence of VAP were also collected. There is an absence of
consensus in diagnosing VAP (Heyland et al., 2006). Pugin’s criteria, a widely accepted diagnostic approach, were adopted for the study (Pugin
et al., 1991). The criteria used were new or worsening radiological infiltrates, together with at least two of the following: (a) temperature greater
3 3
than 38.5 8C or less than 35.0 8C, (b) white cell count greater than 11,000/mm or less than 4000/mm , (c) change in characteristics of bronchial
secretions from mucoid to muco-purulent or purulent, (d) increase in fraction of inspired oxygen or positive end expiratory pressure requirement
by more than 20% to maintain oxygen saturation above 92%. While firm diagnoses of VAP were not established using quantitative cultures,
participants were assessed by the ICU medical clinicians and radiologists blinded to the treatment allocation.

3.5. Statistical analysis

Microbial growth data were categorized into no change, more growth or less growth, and comparisons made using ANOVA, multinomial
logistic regression and chi-squared tests. Data were assessed for normality and underlying test assumptions examined graphically and by use of
appro-priate statistics. Where needed parametric test results were confirmed by their non-parametric equivalents. Statistical significance was
accepted at p < 0.05. Data were analyzed using SPSS Version 13.

3.6. Sample size calculation

In order to inform power calculations and address the pragmatic considerations of a larger multi-center research project, the study was
designed as a feasibility study. Based on similar studies such as De Riso et al.’s (1996) study
comparing chlorhexidine gluconate 0.12% oral rinse with a control group, a sample of 350 subjects was the planned sample size. This study
recruiting 353 subjects, reported a 69% reduction in the incidence of total respiratory tract infections in the chlorhexidine group and significantly
less gram-negative organisms associated with these infections. We used a two sample power calculation with an a = 0.05 and b = 0.8 and with a
sample size of seventy-five in each group, gives a minimum detectable difference in inhibition of microbial growth from Day 1 to Day 4 of 0.32
(units), assuming a standard deviation of 0.45. The resulting sample size is thus sufficient to compare both active treatment groups with the
control, but possibly not to distinguish between active treatments.

4. Results

4.1. Participants

During the 15 month study period a total of 225 mechanically ventilated patients were enrolled in the study. Unfortunately the study was
prematurely termi-nated due to market withdrawal of one of the study’s products, by the Australian
Therapeutic Goods Adminis-tration. At this point, a total of 109 patients had sufficient data for analysis of the primary endpoint, microbial
growth/inhibition, at Day 4 of enrollment. Of the 225 patients enrolled, a total of 87 patients were mechanically ventilated less than 96 h with
numbers spread evenly throughout the three study groups. Twenty-two patients divided between Group A N = 4, Group B N = 13 and Group C N
= 5 died within 96 h of enrollment. A further 11 patients were withdrawn for breach of protocol such as the use of other oral care products other
than the assigned treatment. See Fig. 1 for participant flow showing the study’s enrolment, randomisation, follow-up and cases available for
analysis.

4.2. Baseline characteristics

No statistically significant differences between baseline characteristics of demographics, severity of illness and pre-existing medical
conditions were observed between the three groups as shown in Table 1.
The mean ICU length of stay was 9.8 10.56 days and the average length of ventilation was 7.2 8.29 days. One patient in Group C was nasally
intubated, the remaining orally intubated. There was no significant difference between the groups in relation to drug, including antibiotic, admin-
istration.

4.3. Primary endpoints

Microorganisms isolated, with the exception of normal flora on enrolment (Day 1) are compared with colonization at Day 4 (Table 2).
5
Colonization was considered positive at 10 cfu/ml.

Cross-tabulation was used to summarize the variable of changes in microbial growth from Day 1 to Day 4 in respect to no result, no change
in growth, more growth, less growth and missing data (Table 3). A comparison of change in microbial growth from Day 1 to Day 4 in relation to
2
each treatment group demonstrated no significant differences between the groups ( x (6, N = 225) = 7.209, p = 0.302). This indicates, while
there is a trend to greater effect by the use of a sodium bicarbonate oral rinse, there was not a significant relationship between the treatment
group and microbial growth.
 A one-way analysis of variance was undertaken to assess whether the means of the dependent variable, microbial difference between Day 1
and Day 4 were significantly different between treatment groups. While significant difference was not observed, Group B’s mean of 0.03 (SD =
0.46) indicates that this treatment option has a bias towards improvement over Group A (mean = 0.23, SD = 0.52) and Group C (mean = 0.27, SD
= 0.51).

The nonparametric, Kruskal–Wallis test was under-taken to confirm the above analysis and again indicated no significant difference in the
2
medians, x (2, N = 109) = 4.34, p = 0.114. Because the test failed to reach significance, pairwise comparisons using the Mann–Whitney U test,
among the three groups were not done.
]GIF$DT)([ A.M. Berry et al. / International Journal of Nursing Studies 48 (2011) 681–688 685

Fig. 1. Participant flow.

 Finally multinomial logistic regression was used in an attempt to classify subjects based on values of a set of predictor variables. This was to
explore for any interaction between the predictor variables. If interaction is present then the predictor variables cannot be used separately as
predictors, as their effect must be assessed as the simultaneous effect of both factors (Kleinbaum et al., 1988). This type of regression was
selected because while similar to logistic regression, it is more general. That is, the dependent variable is not restricted to two categories.
The predictor variables included age, APACHE scores, oral assessment scores at Day 1 and dentures. The dependent variable was treatment
groups and the out-come measure was change in microbial growth from Day 1 to Day 4. Nominal regression failed to demonstrate significant
differences between the treatment groups.

the study. There were four each in Groups B and C and one in Group A. The bacteria cultured from the oral cavities of these patients were
5
multiresistant Acinetobacter bauman-nii, P. aeruginosa, Serratia and Enterobacter species. All organisms identified were >10 cfu/ml and are
consistent with VAP associated organisms identified in other studies (Aurora et al., 2002; Collard et al., 2003).
One participant’s sputum culture sampled by non-bronchoscopic bronchial lavage (NBBAL) produced the same result, i.e. normal respiratory
flora as the oral sample. The other eight participants who met the criteria for VAP grew different microorganisms in their NBBAL or sputum
samples from those identified in their oral isolates. At Day 4 three participants had an increased growth of micro-organisms on dental plaque.

4.5. Adverse events

There were no adverse events relating to any of the treatment options identified during this study.

5. Discussion

Over the last two decades, there has been emerging evidence to demonstrate a relationship between coloniza-tion of dental plaque with
respiratory pathogens and the development of pneumonia (De Riso et al., 1996; Houston, 2002; Scannapieco and Stewart, 1992; Robriquet and
Fourrier, 2010). Despite this evidence, there has been little
advancement in the development of standardized oral hygiene protocols in the ICU. The results of this study indicate that a considerable number
of patients admitted to ICU have poor oral health on admission. The implications of this for subsequent morbidity are an area of increasing
concern. The extent of colonization of dental plaque by respiratory pathogens and the clinical differences as a result of the three oral hygiene
regimens in this study could not be determined due to the inadequate sample size.
An important finding of this study is that the protocol was acceptable and feasible within a single ICU. The lack of differences between
solution groups with respect to colonization of the oral cavity may indicate that the type of mouth rinse is less important than mechanical
cleaning of the teeth with a toothbrush. In a number of studies reported in the literature which claim effectiveness of one oral rinse over another,
many did not report the use of a toothbrush to remove plaque (Fourrier et al., 2000; Houston et al., 2003; Seguin et al., 2006) and in fact some
control groups received diminished oral hygiene. For example, in the trial undertaken by Koeman et al. (2006)
comparing chlorhexidine 2% in vaseline and chlorhexidine/ colistin 2% in vaseline with a control group who received only vaseline. Seguin et al.
(2006) also reported better outcomes with respect to the development of VAP when they compared the use of povidine–iodine oropharyngeal
decontamination with saline or an unspecified normal regimen. In contrast to these studies, a historical study where no oral care was compared to
the use of a protocol which included cleaning with a tooth brush, reported a significant reduction in the incidence of VAP in the oral care group
(Mori et al., 2006). More recently Garcia et al. (2009) demonstrated a reduction in the incidence of VAP by 4 per ventilator days and a significant
difference in duration of mechanical ventilation, ICU length of stay and mortality following the introduction of a comprehensive oral care
protocol which included tooth brushing twice a day. This suggests that the agent used is not as important as the actual undertaking of oral
hygiene (Berry et al., 2007).
Given that VAP is the most frequently occurring nosocomial infection in the ICU and is associated with increased morbidity and mortality
(Koeman et al., 2001; Robriquet and Fourrier, 2010), any strategy which has the potential to reduce its incidence should be explored. According
to the Centers for Disease Control (CDC) the prevention of VAP requires a multifaceted approach ( Tablan et al., 2003). They further recommend
that this approach should include the development and implemen-tation of a comprehensive oral hygiene program. Standar-dized protocols
which include the use of an antiseptic agent, can potentially reduce the risk for patients in acute care settings of developing VAP. Thus the CDC
supports interventions to improve oral health, and subsequent reduction in the colonization of dental plaque with respiratory pathogens which
may result in the develop-ment of VAP (Tablan et al., 2003). Published series report VAP rates of 28% or higher (Chaste and Fagon, 2002). The
low rate of VAP in this study suggests that the use of a standardized protocol incorpor-ating mechanical cleaning with a toothbrush cannot be
overstated and needs to be tested in adequately powered studies. Dental plaque, a reservoir for respiratory patho-gens (Scannapieco et al., 2003)
is a very thick biofilm, and as such requires the mechanical action of cleaning with a toothbrush. This strategy is reported to be superior to
cleaning with an antiseptic mouth rinse or foam swabs (Scannapieco et al., 2003; Mori et al., 2006).
The study findings parallel that of Fitch and colleagues who also reported that an oral hygiene protocol which included the use of a
and effect of a standardized oral
toothbrush resulted in improved oral health of ICU patients (Fitch et al., 1999). The development
hygiene protocol directed towards improving oral health and its subsequent relation-
ship with systemic disease is yet to be fully explored. This study has provided valuable insight into the possible gains to be made by the inclusion
of tooth brushing, in detailed protocols directed towards improving the oral health of mechanically ventilated patients in ICU.

Furthermore, because toothbrushes must be treated as potential sources of contamination, thorough cleaning and protected storage of the
brushes after each use should be mandatory.

5.1. Study limitations

The major limitation of this study was the small sample size and the premature cessation of the trial due to withdrawal of a study agent, an
uncommon occurrence and one which is difficult to anticipate These factors reduce the power of the study to determine significant differences
between the treatment options. As noted in previous studies the heterogeneity of the study population is a consideration, yet reflects the
characteristics of the ICU environment. In spite of this, adherence to the study protocol was closely monitored and protocol violations clearly
documented. This study, like that of Cutler and Davis (2005) also failed to examine the duration and frequency of oral cleaning which may be an
important aspect in reducing colonization of dental plaque with respiratory pathogens.

5.2. Study strengths

In spite of the above limitations, this study has several strengths. Firstly, it has tested an empirically derived study protocol in a nurse directed
study and these data provide an important foundation for larger scale clinical trials. The heterogeneity of the ICU population is both a strength
and a weakness in this study. The use of standardized protocols, including the use of antibiotic cycling provided good control of the ICU
environment to observe differences between groups. Secondly it has summarized the oral characteristics of the ICU population and lastly
identified pragmatic considerations in the implementation of an oral hygiene protocol in the ICU. This feasibility study has derived baseline data
using rigorous methods to inform progression to an adequately powered, randomised con-trolled clinical trial. It also contributes to the emerging
evidence that standardized protocols can assist in improv-ing patient outcomes.

6. Conclusion

This project was designed as a feasibility study to provide data for informed power calculations and as such did not have sufficient power to
discriminate a difference in the incidence of VAP. This study has demonstrated that the study procedures were sufficiently rigourous and
demonstrate proof of principle in proceeding to a larger randomised controlled clinical trial. With the caveat of a limited sample size, the
emerging trends support the findings of other studies which advocate the use of a toothbrush in the oral care of ICU patients to improve oral
health and minimize the risk of developing VAP and also the importance of standardized protocols ( Ganz et al., 2009). It further identifies that
given the complexity of undertaking research in the ICU, the sample size to demonstrate a difference in VAP is likely to be beyond the resources
of a single setting and the importance of close monitoring to ensure protocol adherence. Issues to be addressed in larger studies include the
frequency and duration of oral care together with further testing of specific tools and mouth rinses. Also, larger trials are required for the
development and testing of reliable and valid oral assessment tools promoting intra-rater relia-bility to facilitate standardized monitoring of
measures focused on optimizing oral health.
Conflicts of interest