Evaluation of prevention of ventilator-associated infections in four Australian intensive care units

Abstract

Background:

Effective approaches to practice improvement require development of tailored interventions in collaboration with knowledge users.

Objectives:

To explore critical care nurses’ knowledge and adherence to best practice guidelines for management of patients with an artificial airway to minimise development of ventilator-associated pneumonia.

Methods:

A cross-sectional study was undertaken across four intensive care units that involved three phases: (1) survey of critical care nurses regarding their current practice; (2) observation of respiratory care delivery; and (3) chart audit. Key care processes evaluated were: (1) technique and adherence to standard precautions when performing endotracheal suction, cuff pressure checks and extubation; and (2) frequency of endotracheal suctioning and mouth care.

Results:

Observational and chart audit data on the provision and documentation of respiratory care were collected for 36 nurse/patient dyads. Forty-six nurses were surveyed and the majority responded that endotracheal suctioning and mouth care should be performed ‘as required’ or every 2 hours (h). During observations of practice, no patient received mouth care every 2 h, nor had documentation of such. Inconsistent adherence to standard precautions and hand hygiene during respiratory care provision was observed. Chart audit indicated that nurses varied in the frequency of suctioning consistent with documented clinical assessment findings.

Conclusion:

Although nurses had good knowledge for the management of artificial airways, this was not consistently translated into practice. Gaps were identified in relation to respiratory related infection prevention, the prevention of micro-aspiration of oropharyngeal secretions and in the provision of mouth care.

Introduction

Ventilator-associated pneumonia (VAP) is one of the most common healthcare-associated infections (HAIs) affecting patients in intensive care units (ICUs) (Cason et al., 2007; Korhan et al., 2014; Mitchell et al., 2017; Ramirez et al., 2016; Torres et al., 2017). One Australian study reported VAP rates for patients intubated and ventilated for > 72 hours (h) of approximately 26 cases per 1000 ventilator bed-days or 17% of the study sample (Elliott et al., 2015). Optimal prevention of VAP requires a high degree of skills in managing the artificial airway as well as high standards of infection prevention and control (IPC). Although prolonged intubation is associated with an increased risk of VAP, it can occur in patients who are intubated and ventilated for < 24 h; optimal infection prevention strategies should therefore be instigated from the time of artificial airway insertion (Golia et al., 2013).

A ‘care bundle’ approach to VAP in which implementation of a range of evidence-based interventions are promoted to clinical staff has been associated with decreases in infection rates (Cason et al., 2007; Korhan et al., 2014; Resar et al., 2005; Sedwick et al. 2012). Care bundles vary across different centres; however, some evidence-based practices that are widely included are: positioning patients supine with the head of the bed raised > 30° (Blot et al., 2014; Wang et al., 2016); strategies to decrease micro-aspiration of oropharyngeal secretions such as regular oropharyngeal suctioning (Sole et al., 2011; Suys et al., 2013) and regular endotracheal tube [ETT]) cuff pressure checks (Blot et al., 2014; Lizy et al., 2014); and endotracheal suctioning as required using closed (in-line) suction devices (American Association for Respiratory Care, 2010). Despite current controversy over the effectiveness and safety of using products such as chlorhexidine mouth wash when performing mouth care for intubated patients, performing regular mouth care to decrease the bacterial load in the oropharynx is still considered to be an important component in the prevention of VAP (Branch-Elliman et al., 2015; Cutler and Sluman, 2014; El-Rabbany et al., 2015; Hua et al., 2017; Klompas et al., 2014). One source of exposure to potentially pathogenic bacteria that may colonise the lower airways of patients in the ICU is the use of contaminated equipment or contact with contaminated clothing worn by ICU staff and/or visitors (Halliwell and Nayda, 2016). For this reason, meticulous adherence to infection and prevention precautions when providing respiratory care to patients with an artificial airway is fundamental.

Effective approaches to practice improvement require engagement of the staff who will implement the practice change, identification of local gaps in the implementation of best practice and the development of tailored quality improvement interventions in collaboration with the knowledge users. Although VAP prevention guidelines and ‘care bundles’ are readily available, there is limited research available quantifying whether gaps exist in current best practice in Australian ICUs. The current study was undertaken to identify whether there were gaps in either: (1) critical care nurses’ knowledge of best practice when managing patients with an artificial airway; or (2) the implementation of evidence-based care for the prevention of VAP in intensive care patients in four Australian ICUs.

Methods

Design

A cross-sectional study design was used that included three modalities of data collection: observational data; chart audit; and a cross-sectional survey of critical care nurses.

Setting

The present study was conducted at four metropolitan ICUs that included two not-for profit private hospitals and two public hospitals in metropolitan Melbourne, Australia. The two private ICUs comprised an eight-bed unit and a 26-bed unit; these units primarily provide care for elective surgical patients and admit a high proportion of patients who have undergone cardiothoracic surgery. The two public ICUs comprised an eight-bed unit and a 26-bed unit and provided care to a mixed population of both emergency and elective general medical and surgical patients. These units did not include patients who had undergone cardiothoracic surgery via sternotomy.

Participants

A convenience sample of critical care nurses across the four participating sites were recruited to the study. Inclusion criteria were critical care nurses who provided care for intubated and mechanically ventilated patients who were willing to give informed consent to study participation. Exclusion criteria were agency or casual staff members.

Before commencing the observational data collection, a member of the research team held information sessions at each of the participating ICUs to explain the broad aims and purpose of the study. On the day of observational data collection, each nurse caring for an intubated patient was approached individually and offered an opportunity to participate in the study. They were given a verbal summary of the study, what their participation would involve and a participant information and consent form.

Inclusion criteria for patients included in the observation and chart audit were patients who were intubated or had a tracheostomy who were receiving mechanical ventilation at the time of the observational data collection. A waiver of consent was granted by the institutional ethics committee at each site to collect observational and chart audit data from patients in relation to respiratory care provision. To maintain patient privacy and confidentiality, only limited demographic and treatment data were obtained from the chart audit.

Study procedures

The data collection tools used were developed based on current best practice guidelines. Drafts of the surveys and data collection tools were then further modified through an iterative process with co-researchers until an agreement was reached on each tool’s content. The data collection tools were trialled in clinical practice to confirm clarity and usability and were revised after an initial pilot (Appendix A).

Observation and chart audit: Between August and November 2016, two researchers collected observational and chart audit data over 11 separate days during morning and afternoon shifts.

On the days of observational data collection, nurses caring for intubated patients were approached to give consent and confirm that the collection of observational data collection was clinically appropriate. Wherever possible, observational and chart audit data were collected from all intubated patients and the primary nurse caring for them admitted at that time (known as a nurse/patient dyad). Observational and chart audit data on the provision and documentation of respiratory care were collected for 36 nurse/patient dyads across the four study sites. Observation of nurse/patient dyads and chart audit data collection per site were: site 1, n = 10; site 2, n = 10; site 3, n = 9; and site 4, n = 6. Each data collection period lasted for approximately 3 h (range = 3–11 h).

Critical nurses were observed conducting respiratory assessments, suctioning the endotracheal tube, checking ETT cuff pressures, performing extubation procedures and providing mouth care. Data on adherence to IPC standards and use of personal protective equipment (PPE) when performing respiratory care were also collected. Clinical practice was observed and recorded using purposely designed semi-structured observation tools. Although staff had consented to being observed while they practised, exactly what aspect of care was being observed was not divulged until after all study data collection had been completed at each site. This was done to decrease the possibility of bias, such as the Hawthorne Effect, being introduced by the observation process.

On completion of the observational data collection, patients’ charts for each observed patient were reviewed to evaluate documentation of the respiratory care provided over the previous 24 h.

Survey: On completion of the observational data collection, a cross-section of critical care nurses from four different ICUs was surveyed to obtain their views on current best practice when caring for patient with an artificial airway. The survey was made available to all nursing staff both as an electronic survey for online completion using Survey Monkey™ and in paper-based format. The surveys were obtained from August to November 2017 within a week of completing the observational data collection, so that the views expressed by staff were current and reflective of current practice within each unit.

The survey asked critical care nurses about the indications for and the frequency that key aspects of respiratory care (related to VAP prevention) should be provided. Likert scales were used and participants were asked to indicate agreement to a series of questions on a scale of 1 (strongly disagree) to 5 (strongly agree). Open-ended questions asked participants to identify key aspects of respiratory care provision when caring for intubated patients. The survey was developed based on best practice guidelines for VAP prevention (American Association for Respiratory Care, 2010; National Health and Medical Research Council, 2019) and was checked for face validity and useability by two expert clinicians. Survey questions were revised following this input. The final version of the survey was reviewed by the research team and agreed upon for wider distribution.

Data analysis

Descriptive statistics were used to summarise the observational data collection and chart audit, closed and open-ended survey questions were categorised, and the frequency of responses in each category were summarised. Concordance between the survey responses and observed and documented care provision were summarised using descriptive statistics. Unpaired t-tests were used to compare the documented frequency of suctioning according to admission type and the presence or absence of signs of lower respiratory tract infection (i.e. volume of lower respiratory tract secretions).

Ethical considerations

Ethics approval was obtained from the Human Research Ethics Committee (HREC) of each of the participating health services and from the appropriate university ethics committee (EH2016-32; LR26/2016; DUREC 2016-074).

Results

Observation and chart audit

Direct respiratory care was observed for 32 patients; chart audit data were obtained for four patients only as there were no respiratory care-related activities performed during the observational period. Of the 36 nurse/patient dyads included in the study, 22 were general medical patients and 14 were surgical admissions; of these, nine patients had undergone cardiac surgery (Table 1). Lower respiratory tract infections were identified in 16 (73%) of the 22 general medical patients.

Table 1.

Observational data collection by site.

	Site 1	Site 2	Site 3	Site 4
	n (%)	n (%)	n (%)	n (%)
Clinical observation
General ICU patients	2 (6)	5 (16)	7 (22)	4 (13)
Postoperative cardiac patients	6 (19)	3 (3)
Endotracheal tube extubation	2 (6)	2 (6)		1 (3)
Total (n = 32 (100%))	10 (31)	10 (31)	7 (22)	5 (16)

Survey respondents

Forty-six critical care nurses from the following sites completed the survey: site 1, n = 13; site 2, n = 14; site 3, n = 10; and site 4, n = 9. In total, 70 surveys were distributed across the four sites and 46 surveys were completed, giving a response rate of 68%. Respondents’ years of experience working in ICU ranged from < 12 months to > 20 years, 11 (24%) had worked in the ICU < 5 years, 16 (35%) 5–10 years and 19 (41%) > 10 years. Critical care nurses’ opinions on the frequency and indications for endotracheal suction and mouth care are summarised in Table 2.

Table 2.

Critical care nurses’ opinions on frequency and indications for endotracheal suction.

Survey question	Response	General ICUs	Surgical ICUs	Total
		n (%)	n (%)	n (%)
Frequency that e ndotracheal t ube s uctioning should be performed
A postoperative surgical patient	As required	17 (89.5)	17 (63)	35 (76.1)
	Every 1–2 h	0	1 (4)	1 (2.2)
	⩾ 2 to < 4 h	2 (10.5)	9 (33)	11 (23.9)
	⩾ 4 h	0	0	0
	Unsure	0	0	0
A postoperative cardiac surgical patient	As required	12 (63.2)	15 (55.6)	27 (58.7)
	Every 1–2 h	0	2 (7.4)	2 (4.4)
	⩾ 2 to < 4 h	1 (5.3)	9 (33.3)	10 (21.7)
	⩾ 4 h	6 (31.6)	0	6 (13)
	Unsure	0	0	0
A patient with a chest infection or pneumonia	As required	12 (63.2)	11 (40.7)	23 (50.0)
	Every 1–2 h	1 (5.3)	9 (33.3)	10 (21.7)
	⩾ 2 to < 4 h	5 (26.3)	6 (22.2)	9 (19.6)
	⩾ 4 h	1 (5.3)	0	1 (2.2)
	Unsure	0	1 (3.7)	1 (2.2)
Indications for endotracheal tube suctioning
If secretions are visible in the endotracheal or tracheostomy tube?	Agree / Strongly agree	18 (94.7)	27 (100)	45 (97.8)
	Unsure	1 (5.3)	0	1 (2.2)
	Disagree / Strongly disagree	0	0	0
If you plan to extubate the patient?	Agree / Strongly agree	18 (94.8)	27 (100)	45 (97.8)
	Unsure	1 (5.3)	0	1 (2.2)
	Disagree / Strongly disagree	0	0	0
If you are planning to reposition the endotracheal tube?	Agree / Strongly agree	15 (78.9)	26 (96.3)	43 (93.5)
	Unsure	2 (10.5)	1 (3.7)	3 (6.5)
	Disagree / Strongly disagree	2 (10.5)	0	2 (4.4)
Frequency that mouth care is indicated for ventilated patients
Intubated and ventilated < 24 h	> 2 h	3 (15.8)	6 (22.2)	9 (19.6)
	Every 2 h	13 (68.4)	14 (51.9)	27 (58.7)
	Every hour	2 (10.5)	2 (7.4)	4 (8.7)
	Unsure	1 (5.3)	1 (3.7)	2 (4.4)
	As required		4 (14.8)	4 (8.7)
Total		n = 19	n = 27	n = 46

Implementation of key elements of best practice for artificial airway management

Patient positioning

Survey: Positioning the patient with the head of the bed raised ⩾ 30° was identified as important for all intubated patients by 7 (15%) participants and for patients diagnosed with pneumonia it was considered important by 19 (41%).

Observational data collection: For general ICU patients, a kinetic bed was used for 34 (92%) patients admitted to the ICU. The frequency of turns was documented more frequently for medical patients (22, 100%) than surgical patients (11, 79%), but the average number of turns per 12 h did not differ between groups (χ² [1, n = 36] = 5.14, P = 0.021). The majority of medical patients (18, 82%) had the head of the bed elevated, and in 19 (87%) patients, routine turns were performed at intervals > 2 h. For postoperative cardiac surgery patients, 6 (67%) did not have the head of the bed elevated to ⩾ 30° within 30 min of return to the ICU from the operating room.

Frequency and indications for endotracheal suctioning

Survey: Data showed that the majority of critical care nurses (n = 35, 76.1%) agreed that endotracheal suctioning should be performed ‘as required’ for postoperative surgical patients. In contrast, for patients with signs of a chest infection, 23 (50.0%) agreed that suctioning should be performed as required, 10 (21.7%) every hour, 11 (24.0%) every 2–4 h and 2 (4.4%) at > 4 h (Table 2).

Chart audit

The chart audit showed that the mean frequency of artificial airway suctioning was 1.44 (SD = 1.83) times per 12 h. The results of the chart audit provided evidence that critical care nurses were adjusting the frequency at which endotracheal suctioning was performed based on their clinical assessment of patient need. An unpaired t-test was used to compare the documented frequency of suctioning between general medical patients (average 2.5 [SD = 2.10]) per 12 h versus surgical patients (average 0.5 [SD = 1.83]) per 12 hours (t = 2.68 (34), P = 0.006). Within the general medical group, 16 (72.7%) patients had signs of a lower respiratory tract infection; in this group, the mean frequency of airway suctioning was 3.6 (SD 2.3) times per 12 h versus 6 (27.3%) patients with no signs of lower respiratory tract infection, 0.6 (SD 0.5) times per 12 h (t = −6.45 (34), P < 0.001).

Endotracheal tube cuff pressure checks

On admission to ICU postoperatively, ETT cuff pressures were checked in 6 (67%) postoperative cardiac surgery patients. For other ICU patients, ETT cuff pressure was observed being performed on one occasion and recorded on the chart for one patient. Performing cuff pressure checks was not identified as an important aspect of respiratory care in the survey responses.

Provision of mouth care

Survey: Survey responses showed that 31 (67%) critical care nurses agreed that mouth care should be performed every 1–2 h for short-term intubated patients; 9 (19.6%) indicated > 2 h, 4 (8.7%) as required and 2 (4.4%) were unsure (Table 2).

Observational data collection: Only six patients were observed to have mouth care provided once each as part of routine respiratory care; of these six episodes, chlorhexidine mouthwash was used in three episodes and oropharyngeal suctioning was performed for three episodes. Observation of five endotracheal extubation events showed that 4 (80%) patients had their mouths suctioned before this procedure.

Chart audit data showed the frequency of mouth care was not documented in 19/36 (53%) cases. Comparing the medical and surgical patient groups, a higher proportion of general medical 14/22 (64%) versus 4/14 (29%) surgical patients had provision of mouth care documented at least once on their chart over a 12-h period (χ² [1, n = 36] = 4.21, P = 0.040). No patient was documented as receiving mouth care 1-2 hourly. These findings showed discordance between the survey responses for the frequency that mouth care should be performed for intubated patients and observed and documented clinical practice.

Adherence to infection prevention and control standards

Survey: In the open-ended survey questions asking participants to identify important aspects of respiratory care, no participants commented on the importance of IPC measures or the use of PPE when caring for intubated patients.

Observational data collection: Endotracheal suctioning of 17 patients was observed as part of providing general respiratory care. When performing this procedure, critical care nurses’ adherence to the principles of the five moments of hand hygiene was low, only 2 (12%) washed their hands once before performing respiratory care procedures and nurses often donned plastic gloves instead (15, 90%). The use of PPE was also inconsistent (Figure 1a). Five extubation procedures (each involved two critical nurses) were observed. In preparation for the extubation procedure, all observed nurses donned non-sterile gloves, a gown was worn by 8 (80%) nurses, goggles were worn by 6 (60%) members of staff and 4 (40%) staff washed their hands (Figure 1b).

Figure 1.

(a) Shows nurses’ adherence to infection prevention and control standards while completing episodes of general respiratory care. (b) Use of personal protective equipment when performing endotracheal tube extubation.

Discussion

Overall, the results of this study suggest there were consistent gaps in the implementation and documentation of VAP prevention strategies across all four ICUs. The main gaps in care were in implementation of procedures to decrease the bacterial load in oropharyngeal secretions (performing regular mouth care) and in adherence to standard infection prevention precautions when providing respiratory care to intubated patients.

Critical care nurses were observed to implement some elements of VAP prevention consistently, although there were observed differences across the four ICUs. The practice of head of bed elevation was the most consistently observed VAP prevention strategy used. The majority (82%) of general ICU patients were nursed with the head of the bed raised and this practice was on par with previous reports (Wang et al., 2016). In contrast, only 33% postoperative cardiac patients had the head of the bed raised within 30 min of admission to the ICU. Results for all methods of data collection showed relative concordance in the practice of ETT suctioning. Of note, patients with discoloured sputum were documented to have more frequent ETT suctioning than those with clear and scanty sputum; indicating that nurses may have performed suctioning in response to changes in the patients’ respiratory status (Stockley et al., 2001).

The frequency at which mouth care was provided was observed to be very low; for some patients there was no documentation of mouth care being provided while for others mouth care was only documented as having taken place once or twice in 12 h. The discrepancy between nurses’ reports of how frequently mouth care should be performed and what was observed in practice indicates that although nurses knew that mouth care was important, they did not prioritise this task in their clinical practice. Previous reports have also found that the perception by some nurses that these procedures were ‘unnecessary’ were also thought to contribute to the low rates of adherence in practice (Beuret et al., 2013; Flodgren et al., 2013; Hamishehkar et al., 2014; Jansson et al., 2013). These findings reflect the outcomes of a recent randomised controlled trial evaluating the impact of increasing mouth care and tooth brushing rates to decrease the occurrence of hospital-acquired pneumonia, which found that even under clinical trial conditions implementation of mouth care protocols did not reach the recommended targets (McNally et al., 2019). These findings indicate that acute care staff underestimate the impact of oral hygiene on clinical outcomes.

Observations of nursing practice demonstrated consistent gaps in the implementation of IPC standards across all four sites. Importantly, the majority of critical care nurses were found not to adhere to appropriate hand hygiene measures (World Health Organization, 2009) when providing episodes of respiratory care and there was inconsistent use of PPE. Previous studies suggest that nurses have a good knowledge of the importance of hand hygiene and know when to complete the procedure; however, despite this knowledge, they continue to exhibit low compliance rates (De Wandel et al., 2010).

Although the survey responses indicated that critical care nurses had reasonable knowledge of best practice for the management of the artificial airway, this was not consistently translated into practice. This finding is in line with previous studies that have found that education and knowledge is insufficient to change practice (Bagheri-Nesami et al., 2015; Hamishehkar et al., 2014; Labeau et al., 2008; Tompson and Stapley, 2011). Studies conducted on barriers to adherence to VAP prevention strategies found that nurses reported a lack of time, skills or knowledge as barriers to consistent implementation of VAP prevention strategies (Ladbrook et al., 2019; Westwell et al., 2008), while some nurses acknowledged that infection prevention was not a priority in their day-to-day practice (Beuret et al., 2013; Flodgren et al., 2013; Hamishehkar et al., 2014; Jansson et al., 2013).

Strengths and limitations

One strength of the present study was the use of multiple methods of data collection, which allowed the researchers to compare and contrast nurses’ opinions about some aspects of evidence-based practice in respiratory care provision with their actual practice. This allowed the researchers to identify practice gaps and provided some indication of key areas in which quality improvement initiatives should be targeted.

One limitation of the present study is the short period of time over which each period of clinical observation was conducted. It is possible that some respiratory procedures may have been provided but were not captured during the observational data collection and that there may have been delays in documentation of some care practices. Convenience sampling was used in this study and the results therefore provide a snapshot of nurses’ opinions and current practice when providing respiratory care to mechanically ventilated patients; however, the reported findings may not be representative of all nurse’s practice or be generalisable to all ICUs in Australia. The study findings highlight a need for further translational research to be conducted to improve implementation of evidence-based VAP prevention strategies and to explore barriers to adherence to IPC.

Conclusion

In the present study, variability in critical care nurses’ implementation of evidence-base practice when providing respiratory care was evident. Practice gaps were identified in relation to respiratory-related infection prevention, the prevention of micro aspiration of oropharyngeal secretions and in the provision of mouth care.