Abstract of thesis entitled. The Clinical Protocol in using High Flow Nasal Cannula in Critically ill patients. Submitted by YEUNG HIU LAM

Transcription

1 Abstract of thesis entitled The Clinical Protocol in using High Flow Nasal Cannula in Critically ill patients Submitted by YEUNG HIU LAM for the degree of Master of Nursing at The university of Hong Kong in August 2016 In the past two decades, high flow nasal cannula (HFNC) has been put under the spotlight in critically ill adult patients, as an alternative to standard oxygen delivery systems. HFNC is composed of an oxygen blender, an active heated humidifier, a single heated circuit and nasal prongs. It can deliver heated high-flow and humidified oxygen with inspiratory fraction of oxygen (FiO2) ranging from 0.21 to 1.0 at a flow of 60L/min (Dysart et al., 2009). It is extensively used in preterm neonates in times of apnea in prematurity, acute respiratory distress syndrome and post extubation (Holleman-Duray et al, 2007). Recently, more clinical trials have reported HFNC in adults in view of its potential clinical benefits in terms of oxygenation, ventilation and patient comfort. Literature search was performed in three databases while six studies met the inclusive criteria and were critically appraised. It was shown that HFNC effectively improves patient comfort and oxygenation. Afterwards, an evidence-based guideline was developed by transferring the current evidence to the local intensive care unit. A three month pilot study was planned then followed by a twelve-month implementation program. The outcomes will be measured and evaluated by PiO2/FiO2 ratio at 24hours after application of HFNC, patient comfort and nurse satisfaction scores. 1

2 The Clinical Protocol in using High flow nasal cannula in critically ill patients Submitted by YEUNG HIU LAM BNurs, RN A thesis submitted in partial fulfillment of the requirement for the degree of Master of Nursing at the University of Hong Kong August

3 Declaration I declare that this dissertation represents my own work, and that it has not been previously included in a thesis, dissertation or report submitted to this university or any other institution for degree, diploma or other qualification. (Sign) YEUNG HIU LAM August

4 ACKNOWLEDGEMENTS I would like to express my deep gratitude to Ms Kong Hoi Mei, Cecilia for her patient guidance, enthusiastic encouragement and useful critiques of this thesis. Her willingness to give her time so generously is greatly appreciated. My special thanks also extended to my family and friends for their support and encouragement throughout this two years of study. 4

5 Content Declaration... 3 Acknowledgements... 4 Table of contents Abbreviations... 8 Chapter 1: Introduction 1.1 Background Affirming the needs Objectives and significance Chapter 2: Critical Appraisal 2.1 Search and Appraisal strategies Research Question Selection Criteria Search strategy Appraisal strategy Result Search Result Study characteristics Methodological Quality Summary and synthesis Oxygenation and breathing effort Ventilation Comfort and compliance Mortality rate Potential role of HFNC in clinical setting

6 Chapter 3: Translation and application 3.1 Implementation potential Transferability of findings feasibility Cost benefit ratio of the innovation Evidence -based guideline Chapter 4: Implementation Plan 4.1 Communication plan Stakeholder analysis Communication process and strategies Pilot study plan Objectives Target population and enrollment Time frame Outcomes measurement Data collection Analysis of results and evaluation Evaluation Plan Outcomes identification Nature and number of target subjects Timing and frequency of evaluation Data analysis Basis of implementation Conclusion

7 Appendices Appendix 1. Table of evidence Appendix 2. PRISMA flow diagram Appendix 3. SIGN checklists for RCTs Appendix 4. Estimated material cost Appendix 5. Estimated non-material cost Appendix 6. Evidence-based guideline Appendix 7. SIGN grades of recommendation Appendix 8. Project timeline Appendix 9. Data collection sheet Appendix 10. Satisfaction survey References

8 Abbreviations ARF A-a gradient APN BMI HFNC FiO2 ICU PaO2 PaO2/FiO2 PaCO2 PEEP PICO RCT RN RR SaO2 SIGN Acute respiratory failure Alveolar- arterial oxygen tension gradient Advance Practice Nurse Body mass index High flow nasal cannula Fraction of inspired oxygen Intensive care unit Arterial oxygen tension Ratio of arterial oxygen tension to the inspiratory fraction of oxygen Partial pressure of carbon dioxide in the arterial blood Positive end expiratory pressure Patient Intervention Comparison Outcome Randomized Controlled Trial Registered Nurse Respiratory rate Arterial oxygen saturation Scottish Intercollegiate Guidelines Network 8

9 Chapter 1 Introduction 1.1 Background Acute respiratory failure (ARF) resulting in hypoxemia or hypercapnia is commonly encountered in hospitalized patients due to different triggers such as pneumonia, chronic obstructive pulmonary disorder, acute respiratory distress syndrome, and pulmonary edema. It is the most common reason for intensive care unit (ICU) admission. In the United States, there were around 2 million hospitalizations due to ARF in 2009 leading to nearly 380,000 deaths and inpatient costs over 54 billion (Stefan et al., 2013). Oxygen therapy is the first line treatment for ARF. There are different modalities of oxygen delivery such as a nasal cannula, simple face mask, Venturi mask and reservoir mask. Use of these devices often leads to drying of the nasal mucosa (O Driscoll, Howard & Davison, 2008). For more severe cases, these low flow oxygen devices may not be able to meet patients increased inspiratory flow rate and fraction of inspired oxygen (FiO2) (Pruitt & Jacobs, 2003; Wagstaff & Soni, 2007). Venturi masks deliver a constant flow of oxygen through ports with a precise size and entrainment of room air using the Bernoulli effect which states the inversely proportional relationship between the speed and pressure, and perform better than conventional nasal systems (Kallstrom, 2002). Nevertheless, masks are less well tolerated than nasal prongs because of inadequate humidification, the heat of the gas and the obtrusiveness of the mask. In addition, dilution of oxygen by entrained room air lowers the FiO2 when a patient s inspiratory flow rate is greater than the delivery rate (Chanques et al., 2009). In the past two decades, the high flow nasal cannula (HFNC) has been put under the spotlight for use 9

10 in critically ill adult patients as an alternative to standard oxygen delivery systems. HFNC is composed of an oxygen blender, an active heated humidifier, a single heated circuit and nasal prongs. It can deliver heated high-flow, humidified oxygen with an FiO2 ranging from 0.21 to 1.0 at a maximum flow of 60 L/min (Dysart, Miller, Wolfson, & Shaffer, 2009). It is used extensively in preterm neonates for apnea in prematurity, acute respiratory distress syndrome and post extubation (Holleman-Duray, Kaupie, & Weiss, 2007). Recently, clinical trials and case reports have reported on HFNC in adults in view of the potential clinical benefits in terms of oxygenation, ventilation and patient comfort. Oxygenation is the process of diffusion of oxygen from the alveoli to the pulmonary capillaries in which oxygen molecules bind to hemoglobin (Covelli, Nessan, & Tuttle, 1983). There are numerous ways to measure oxygenation to determine whether it is adequate to meet the metabolic requirements of body tissues, including arterial oxygen saturation (SaO2), arterial oxygen tension (PaO2), the alveolar- arterial oxygen tension gradient (A-a gradient), the ratio of arterial oxygen tension to the inspiratory fraction of oxygen (PaO2/FiO2) and the SaO2/FiO2 ratio. SaO2 indicates the proportion of hemoglobin in red cells saturated with oxygen and is usually measured noninvasively by pulse oximetry. PaO2 is the partial pressure of oxygen dissolved in arterial blood. The A-a gradient is the difference between the amount of oxygen in the alveoli and that dissolved in arterial blood, and can be measured by arterial blood gas analysis. The PaO2/FiO2 ratio is another common measure of oxygenation. A normal PaO2/FiO2 ratio is 300 to 500 mmhg. Impaired gas exchange is indicated by values less than 300 mmhg and severe hypoxemia is implied with values less than 200 mmhg. In one study, SpO2/FiO2 ratios of 235 and 315 were correlated with PaO2/FiO2 ratios of 10

11 200 and 300 respectively (Rice et al., 2007). 1.2 Affirming the needs Physiological benefits of HFNC HFNC is postulated to demonstrate advantages for hypoxemic patients. Clinical practice guidelines also suggest active humidification in noninvasive ventilation to improve adherence and comfort (Restrepo & Walsh, 2012). Firstly, heated and humidified gas delivered through an interface such as a nasal cannula instead of a mask can enhance patient comfort and tolerance with less leakage and dead space (Sztrymf et al., 2012). The increased water content of the mucus with humidification helps preserve mucosal functions, facilitates mobilization of secretions and avoids of desiccation and epithelial injury compared with other non-humidified high flow oxygen devices (Esquinas Rodriguez et al., 2012). Also, heated and humidified gas can avoid desiccation of the upper airways. It reduces the bronchoconstricting effect of cold, dry gas which helps reduce the work of breathing (Richards, Cistulli, Ungar, Berthon-Jones, & Sullivan, 1996). Secondly, high flow rates generated by HFNC exceed the patient s peak inspiratory flow resulting in less dilution of entrained room air with supplemental oxygen. Therefore, a more desirable FiO2 can be delivered. Moreover, HFNC is expected to improve the efficiency of ventilation and efficacy of oxygen delivery since its high flow rate can flush out carbon dioxide from the upper airway, eliminating nasopharyngeal dead space (Sztrymf et al., 2012). Spence, Buchmann, Jermy, and Moore (2010) showed that HFNC can significantly alter the flow pattern of gas in the nasopharynx, therefore raising the FiO2. In 11

12 addition, Groves and Tobin (2007) claimed that HFNC can generate positive end expiratory pressure (PEEP) with the mouth closed so that the work of breathing may be reduced. As aforesaid, HFNC has demonstrated different physiological benefits and it has been applied overseas. Two types of HFNC oxygen delivery devices have been available for a decade in Australian ICUs for management of acute respiratory failure in adults, but this new therapy was only introduced in Hong Kong a few years ago. However, there is still no evidence-based guideline from the Hospital Authority on the use of HFNC to optimize patient outcomes, and avoid complications and associated risks. Some important issues such as the indications, and the criteria for commencement or suspension of HFNC and for escalation of treatment remained unresolved for frontline staff. In the target setting, nearly three quarters of frontline nurses do not know the exact physiological benefits of HFNC, the differences between HFNC and conventional oxygen therapy, and the indications, complications and infection control issues. The HFNC device has been used intermittently since August, Therefore, it is of paramount importance to review the clinical effects and establish an evidence-based guideline on the application of HFNC in adult critical care patients. 1.3 Objectives and Significance The establishment of an evidence-based guideline on the use of HFNC can offer benefits to different stakeholders. For patients, HFNC provides respiratory assistance and improves patient comfort and compliance compared with standard oxygen therapy in the management of ARF. Critical care nurses who 12

13 monitor patients airways and respiratory status should be able to make clinical judgements concerning the most effective oxygen therapy for patients in the acute setting. This guideline would enable them to ensure the best evidence-based care with the most pertinent skills and optimize outcomes with safety concerns. Macroscopically, for the healthcare system, reductions in morbidity, mortality and the length of ICU stay could relieve service demands and financial burden in the healthcare sector. To translate the evidence and make recommendations for application of HFNC in critically adult patients in this guideline, the following objectives should be achieved. 1. To review and critique current evidence on the application of HFNC in adult critical care patients 2. To develop an evidence-based guideline for the application of HFNC in adult critical care patients. 3. To assess the transferability and feasibility of implementing the guideline in an adult ICU in a local public hospital. 4. To devise an implementation and evaluation plan to assess the utilization of the proposed guideline in the local setting. 13

14 Chapter 2 Critical Appraisal 2.1 Search and Appraisal Strategies Research Question The Patient Intervention Comparison Outcome (PICO) framework was used to guide the development of a search strategy and to compose the research question (Courtney & McCutcheon, 2010). In this study, the patient (P) component was critically ill patients aged over 18 years old admitted to the ICU; intervention (I) was application of HFNC which delivers humidified oxygen at a flow greater than 15L/min. Comparison (C) was standard oxygen therapy including low flow and high flow oxygen delivery devices such as a nasal cannula, Venturi mask and non-rebreathing mask; outcome (O) was improved oxygenation and patient comfort Selection criteria Studies were included which met all of the following criteria: 1. The target population was adult patients over 18 years old admitted to the ICU requiring oxygen therapy. 2. The intervention was an HFNC system at a flow greater than 15L/min, compared with standard oxygen therapy including a nasal cannula, Venturi mask and nonrebreathing masks. 3. The study design included randomization. 14

15 4. The outcome measures included oxygenation as indicated by the PaO2, PaO2/FiO2 ratio obtained from arterial blood gas analysis, pulse oximetry, the intubation rate, mortality and length of ICU stay and patient comfort as presented by a subjective comfort score and physiological parameters such as the heart rate and respiratory rate (RR). Studies were excluded which met any of the following criteria: 1. Target population was non-icu patients or pediatric patients. 2. The articles were systematic reviews, editorials, or authors comments Search strategy To identify the studies, a literature search was performed in three databases, PubMed, Ovid Medline and Cochrane Library. The search keywords were high flow nasal oxygen, humidified, heated, oxygen therapy, nasal oxygen and nasal cannula. Firstly, titles and abstracts were screened to identify articles which met the selection criteria and then the full texts of the selected articles were screened to ensure eligibility. Only articles in the English and Chinese literature were included in the study. In view of developing a guideline reflecting current clinical needs, only articles published from 2005 to 2015 were selected. Finally six studies were included in the dissertation. A PRISMA flow diagram (Moher, Liberati, Tetzlaff, Altman & The PRISMA Group, 2009) has been used to show the details of the literature search (Appendix 2). 15

16 2.1.4 Appraisal strategy The selected studies were reviewed in a systematic way with the guidance of the checklist from the Scottish Intercollegiate Guidelines Network (SIGN) which is a tool to critique control trials in terms of methodological quality and performance of the overall assessment of the study (Scottish Intercollegiate Guidelines Network, 2012). Studies are graded 1++, 1+, 1-, 2++, 2+, 2-, 3 or 4 according to the coding system suggested in the SIGN handbook (Scottish Intercollegiate Guidelines Network 2011). 2.2 Results Search results A keyword search limited to randomized clinical trials from 2005 and 2015 was conducted from September to November 2015 in the three databases. A total of 48 papers were retrieved of which 9 papers were duplicates. All studies were in English. After screening the titles and abstracts, 20 papers were excluded as potentially irrelevant to the topic. Then of the remaining 10 full text papers, 4 were excluded due to inappropriate populations, i.e. patients in the emergency department or having dental care and patients who had chronic obstructive pulmonary disease but were not critically ill. In the end, 6 randomized controlled trials were selected as eligible studies Study characteristics These 6 studies were carried out in Western and Asian countries, including France and Belgium (Frat 16

17 et al., 2015; Vourc h et al., 2015), New Zealand (Parke, McGuinness, Dixon, & Jull, 2012; Parke, McGuinness & Eccleston, 2011), Italy (Maggiore et al., 2014) and Thailand (Rittayamai, Tscheikuna & Rujiwit, 2014). Two of the six RCTs were graded 1++, three were 1+ and one was 1-. The target subjects of all studies was were critically ill adults in the ICU suffering from different extents of hypoxemia under different circumstances. Some of them were post cardiac surgery patients (Parke et al., 2013; Parke et al., 2011) while some were recently extubated patients (Rittayamai et al., 2014; Maggiore et al., 2014). Participants from one study were recruited 4 hours before endotracheal intubation (Vourc h et al., 2015) and in one study were patients with hypoxemia defined by a PaO2/FiO2 ratio <=300mmHg (Frat et al., 2015). In all studies, the effectiveness of HFNC was evaluated compared with standard oxygen therapy, which included all kinds of low and high-flow oxygen devices such as nasal cannulas, Venturi masks and reservoir masks. The brands and manufacturers of the HFNC devices were clearly reported in all studies. In addition, oxygenation was the primary outcome in all studies. Assessment of oxygenation was done using the SpO2, PaO2/FiO2 ratio or SpO2/FiO2 ratio in the Parke et al. (2013), Maggiore et al. (2014) and Parke et al. (2011) studies, the intubation rate in the Frat et al. (2015) study and analysis of the work of breathing using the RR in all studies. Patient comfort was evaluated in all studies except Vourc h et al. (2015). Three studies were partly sponsored by the medical company producing the HFNC device (Parke et al., 2013; Maggiore et al., 2014; Vourc h et al., 2015). This may have exposed them to funding bias although the authors claimed that the sponsor had no access to either the study design or the data. Others 17

18 were funded by university research funds or government funding Methodological Quality The SIGN checklist for controlled trials was used as an appraisal tool to ensure the validity and reliability of these six studies, with details in Appendix 4. Based on the SIGN checklist in assessing internal validity and overall assessment, all studies clearly addressed the focus question and target population with reasonable inclusion and exclusion criteria which ensured the samples validly reflected the population of interest. To reduce extraneous influences, randomized assignment of subjects to treatment groups was adopted. All studies used computerized methods for randomization such as Clinsight, Ennov software, sequence numbers or spreadsheet software except Rittayamai et al (2014) who did not mention any details of randomization. Blocked randomization was performed by Frat et al. (2015), Parke et al. (2013), Vourc h et al. (2015) and Parke, et al. (2011). Stratified randomization was noted in some studies to ensure balance between groups for possible prognostic factors. Parke et al. (2013) stratified for body mass index (BMI) at randomization as a positive association between expiratory lung impedance and BMI was noted in another observational study of cardiac surgery patients (Corley, Carunna, Barenett, Tronstad, & Fraser, 2011). Vourc h et al. (2015) stratified for center in a multicenter trial and Frat et al. (2015) stratified according to center and history of cardiac insufficiency. Parke et al. (2013), Maggiore et al. (2014) and Parke et al. (2011) used sequential, opaque and sealed 18

19 envelopes for concealment. The design meant that subjects and investigators were not blind to the treatment but third parties with no direct involvement in the study procedures were mentioned in Parke et al. (2013), Vourc h et al. (2015) and Maggiore et al. (2014) to minimize bias. In five eligible studies, consent was obtained from the patient or next of kin before or after the procedure and the corresponding regional review board or ethics committee approved the studies. Parke et al. (2011) did not mention informed consent or approval from a concerned committee. The total number of recruited patients and the number of patients assigned to each group were well documented in all studies. Moreover, basic characteristics such as patient data, comorbidities, respiratory failure aetiology and physiologic variables were collected as baseline measurements. All studies reported no significant differences between the intervention and control groups except for the mean heart rate in Parke et al. (2011). The outcomes were measured in a standard, valid and reliable way in all studies. Oxygenation was assessed by measurement of oxygen saturation in all studies while the PaO2/FiO2 ratio as a surrogate outcome was used in Frat et al. (2015), Parke et al. (2013), Maggiore et al. (2014) and Parke et al. (2011). Other outcomes such as mortality, comfort, and length of stay were also mentioned in 6 studies according to the experimental design. Parke et al. (2013) and Parke et al. (2011) suggested the need for escalation of respiratory support to define the success of therapy. According to the SIGN guidelines (2015), drop-out rates should not exceed 20%. All studies met this criteria except Rittayamai et al. (2014), in which 8 out of 25 subjects were excluded since they could not 19

20 tolerate the spontaneous breathing trial for 120 minutes. Frat et al. (2015), Parke et al. (2013), Maggiore et al. (2014) and Vourc h et al. (2015) applied intention-to-treat analysis in which subjects are analyzed according to their randomized treatment assignments to avoid overoptimistic estimates of the effectiveness of the intervention due to removal of noncompliance and protocol deviations (Gupta, 2011). In terms of study design, only Rittayamai et al. (2014) was a randomized crossover study while the others were randomized controlled trials; half of the eligible studies (Frat et al., 2015, Maggiore et al., 2014, Vourc h et al., 2015) were carried out at more than one site, which demonstrated higher transferability. All studies used power analysis to calculate the minimum sample size except Parke et al. (2011) since it was a preliminary trial. Only Parke et al. (2013) did a pretrial audit of patients at the study site to determine the minimum effect size. 2.3 Summary and synthesis All the eligible clinical trials had level 1+ or above evidence except Rittayamai et al. (2014) which was graded 1-. The mean ages of the 965 subjects were from 59 to 66.8 in the 6 eligible studies. The studies were performed in critically ill adults under various scenarios such as pre-oxygenation before intubation, postextubation, postoperative cardiac surgery patients and hypoxemic respiratory failure. All studies examined the effectiveness of HFNC compared with standard oxygen therapy with the use of various 20

21 outcome measures Oxygenation and breathing effort PaO2/FiO2 ratio When quantifying oxygenation, the PaO2/FiO2 ratio was a common oxygenation indicator in mechanically ventilated patients. The higher the ratio, the better the oxygenation. The PaO2/FiO2 ratio was measured at different time intervals in three studies, 1 hour, 24 hours, 36 hours and 48 hours after treatment. All results favored HFNC. Frat et al. (2015), Maggiore et al. (2014) and Parke et al. (2011) showed significantly higher PaO2/FiO2 ratios with the use of HFNC than with standard oxygen therapy. Intubation rates and ventilator- free days Some studies also used the intubation rate and ventilator-free days as outcome measures. Only Frat et al. (2015) stated that the intubation rate on day 28 for the overall population was not significant but fewer patients with a PaO2/FiO2 ratio < 200 mmhg needed intubation on day 28 (p=0.009) in post-hoc analysis. Frat et al. (2015) also demonstrated a significant result for ventilator-free days with HFNC in 310 hypoxemic patients with PaO2/FiO2 ratios less than 300 mmhg. However, in Vourc h et al. s (2015) study of 124 subjects with acute hypoxemic respiratory failure (PaO2/FiO2 ratio less than 300 mmhg), the difference in median number of days ventilator free on day 28 was not significant. Respiratory rate HFNC also reduced the RR compared with standard oxygen therapy which may be deemed as 21

22 reduced work in the breathing effort (Frat et al., 2015; Maggiore et al., 2014; Rittayamai et al, 2014). These three studies showed consistent statistically significant results while the other three studies didn t use this as a secondary outcome Ventilation In terms of ventilation, it s hypothesized that anatomical dead space is reduced with high flow rates and thus alveolar ventilation would be enhanced with a reduction of the carbon dioxide level in the arterial blood. This is reflected by the partial pressure of carbon dioxide in the arterial blood (PaCO2). The higher the PaCO2, the less effective the alveolar ventilation. The PaCO2 was measured in two studies, Frat et al. (2015) and Maggiore et al. (2014). Maggiore et al. (2014) stated that the PaCO2 was lower with HFNC at all time intervals from 0 to 48 hours after treatment, but the result was statistically significant only at 3 hours. However, in the Frat et al study, (2015) no significant difference was detected at 1 hour or 6 hours after the intervention.. Both intervention groups were given HFNC at a flow of 50L/min to keep the SpO2 >92% but control groups were a bit different with non-rebreathing masks (Frat et al., 2015) and Venturi masks (Maggiore et al., 2014). The subjects with PaO2/FiO2 ratios around 240 mmhg in Maggiore et al. (2014) were given treatment after extubation while hypoxemic patients in Frat et al. s study were more ill with PaO2/FiO2 ratios around 160 mmhg. This have may contributed to the differences in the results, therefore the conclusions could not be drawn in terms of efficiency of ventilation with HFNC. 22

23 2.3.3 Comfort and compliance Assessment of patient comfort and compliance was also done in four studies (Frat et al., 2015; Parke et al., 2013; Maggiore et al., 2014; Rittayamai et al., 2014). Improving comfort and thus compliance with corresponding oxygen therapy can enhance the therapeutic effect with fewer interruptions in therapy. Only Maggiore et al. s (2014) study mentioned the details of the comfort score which was related to the interface or airway discomfort, such as mouth dryness, throat dryness, difficult swallowing and throat pain. This could be explained by better humidification with HFNC compared with standard oxygen therapy. The comfort score was measured on a numerical scale which provided high validity and reliability (Frat et al., 2015; Parke et al., 2013; Maggiore et al., 2014). Only a visual analog scale was used in Rittayamai et al. s (2014) study. Significantly higher comfort scores and fewer episodes of interface displacement which prevented desaturation were demonstrated with the use of HFNC in all the studies except Parke et al. (2013). The participants in Parke et al. (2013) were a group of reasonably well patients compared with the other studies since they were a group of patients prone to respiratory complications who had undergone cardiac surgery in a critical care unit Mortality rate In three studies, Frat et al. (2015), Parke et al. (2013) and Vourc h et al. (2015), HFNC was compared with standard oxygen in terms of mortality and length of stay but only Frat et al. (2015) showed a significant result for mortality in the ICU and mortality on day 90 in favor of HFNC with odds ratios of 23

24 (95% CI) 1.85 ( and 2.01 ( ) respectively Potential role of HFNC in the clinical setting Moderate to severe hypoxemia All in all, by virtue of the multiple physiologic and subjective benefits of HFNC compared with standard oxygen therapy, current evidence supports the role of HFNC in the management of hypoxemia and respiratory failure. According to Frat et al. s (2015) high quality RCT with 310 participants in multiple centers, the promising results were the lower intubation rate in patients with a PaO2/FiO2 ratio < 200 mmhg, improved ventilator-free days, lower mortality rate in the ICU, lower mortality rate on day 90, reduced respiratory discomfort and reduced RR. The research suggested that patients with moderate to severe hypoxemia would derive great benefits from HFNC. Nevertheless, HFNC may not offer advantages in patients with mild hypoxemia who are already doing well on low flow oxygen devices, as routine use of HFNC was not associated with an increase in oxygenation in postoperative patients in Parke et al. s (2013) study. Routine use of HFNC in post cardiac surgery patients Parke et al. (2013) aimed to determine if HFNC can reduce complications after cardiac surgery so the number of patients with a SpO2/FiO2 ratio >=445 on day 3 after surgery was defined as a primary outcome. However, the results were not significant. Among these six studies, only Parke et al. (2013) used escalation of therapy as one of the secondary outcomes. When patients had impaired oxygenation 24

25 and met the criteria in this study s protocol, therapy was escalated according to an algorithm. For example, the standard care group was treated with HFNC, noninvasive ventilation or even intubation if these hypoxemic patients met certain criteria. They claimed that a well-developed protocol for escalation of therapy was one of the strengths of this study. They had previously published a detailed description of methods of data collection and criteria for escalation of therapy in the HOT-AS study (Parke, McGuinness, Dixon, & Jull, 2012). Although significantly fewer patients using HFNC than standard oxygen therapy required escalation of respiratory support on postoperative days 1 to 3, the primary outcome, the SpO2/FiO2 was not significantly different so routine use of HFNC in all postoperative cardiac surgery patients after extubation was not recommended in this study. Hypoxemic patients after extubation Maggiore et al. (2014) and Rittayamaiet al. (2014) suggested a potential role of HFNC in patients with PaO2/FiO2 ratios less than 300 mmhg after extubation. Their studies showed statistically significant results in improving the PaO2/FiO2 ratio by 31mmHg at 24 hours, 36 hours, and 48 hours, with a higher oxygen saturation, lower RR, less discomfort and fewer episodes of desaturation. It is believed that the alveoli could be kept open as high flow gas could theoretically prevent lung decruitment after extubation following anesthesia and paralysis, but these studies didn t measure the end expiratory pressure in spontaneously breathing patients. Although HFNC is a safe alternative, it was not recommended as a preoxygenation device before endotracheal intubation to prevent desaturation in another multicenter study (Vourc h et al., 2015) as outcomes comparing it with standard oxygen therapy with facial masks were not 25

26 statistically significant. In conclusion, in this review, HFNC with humidification at a high flow rate results in better oxygenation and ventilation and enhances comfort for patients with moderate to severe hypoxemia but routine use in postoperative cardiac surgery patients and as a pre-oxygenation device before intubation is not suggested. Patients with PaO2/FiO2 ratios less than 300 mmhg could be considered for HFNC after extubation. Since PEEP was not demonstrated in these six studies and non-invasive ventilation was not compared in this review, it is suggested that HFNC could be used in the middle spectrum of therapy to manage acute respiratory failure. 26

27 Chapter 3 Translation and application 3.1 Implementation potential The current systemic literature review supports the use of HFNC in critically ill adults to improve clinical outcomes including oxygenation and patient comfort. After reviewing the appraised evidence, it is recommended an evidence based guideline could be developed through assessment of the implementation potential of the clinical innovation in the target setting in terms of transferability, feasibility and the cost benefit ratio (Polit & Beck, 2012) Transferability of the findings Although the innovation was shown to be clinically effective in the studies, the results are not directly transferred to any setting. So ensuring that different aspects of the setting such as the types of target clients, philosophy of care, and benefits to target clients are basically congruent with the innovation is a main issue with regard to transferability. Type of target clients The target setting is a mixed surgical and medical ICU consisting of 26 beds in a local public hospital. All patients over 18 years old in an unstable or deteriorating condition or are admitted to the ICU mainly from the accident & emergency department, operating theatre or general wards for close 27

28 monitoring and further management. Around 80% of patients in the target setting experience mild to severe respiratory failure. The subjects in the six reviewed studies were also under care in the ICU with mild to severe respiratory failure at a median age ranging from 59 to 67 years. The causes of respiratory failure were mainly pneumonia, sepsis, pulmonary edema and atelectasis, similar to the situation in the target setting. Philosophy of care The hospital authority is dedicated to providing the best possible service and patient-centered care. The proposed protocol will not only include study of the clinical significance of the innovation but also emphasize the comfort level of the patient, which is totally entrenched in the core values and beliefs held by the hospital authority in a view to help people regain health. In addition, the hospital authority also acknowledges the importance of professionalism by promoting staff education, so nurses can keep abreast of the latest developments in their profession to support evidence-based clinical practice. Since the underlying philosophy of care of this innovation is in line with the hospital authority s views, an evidence-based protocol could be established with support. Benefits to target clients Approximately 1700 patients are admitted to the target intensive care unit annually. Respiratory failure is the most common reason for ICU admission. Nearly 95% of patients require oxygen therapy and 28

29 even ventilatory support. So there is a sufficient number of clients in the target setting who could benefit from the innovation. High flow nasal cannulas could improve oxygenation and patient comfort, reduce escalation of therapy, and hopefully reduce the length of ICU stay. Duration of implementation and evaluation The implementation and evaluation of the proposed guideline would take eight months in total. The preparation stage of this pilot program includes collaboration with the management team and medical professionals, needs assessment, promotion of the program, protocol introduction, manpower training and material preparation. Then data would be collected for three months in the target setting. Feedback will be collected from the frontline staff and patients at the same time. The findings will be gathered and synthesized the following month. Lastly, an evaluation will be carried out for one month and a modified evidence- based guideline will be established with reports to the management team Feasibility Although the findings are transferable to the target clinical setting after considering the implementation potential, the feasibility of the innovation should be taken into account as well. Four aspects should be assessed (1) the organizational climate, (2) method, (3) availability of staff, and (4) availability of resources (Polit & Beck, 2012). 29

30 Organizational climate Seeking approval and support from the management level and medical professionals is essential in the preparation phase of the innovation. A high level of evidence should be shown and positive outcomes and a cost-effective strategy should be promoted to persuade the management level and medical team to support the proposed guideline. Our department emphasizes evidence-based practice as shown by strict adherence to the latest international guidelines of the ventilator care bundle and sepsis management bundle. The frontline nurses and medical team in our department are empowered to set up guidelines. Previously they have initiated protocols for sedation and inotrope weaning. Although this is not a teaching hospital, staff are given opportunities for on-the-job training, motivated to foster innovation, prepared to embrace change and involved in clinical research. Although each staff member has a different job within the organization, unity and a sense of purpose are always needed to link to the organization as a whole, especially in the health care industry, ultimately contributing to society and benefitting public health. Method A project team will be set up to facilitate implementation of this new oxygen device. Since we have a respiratory committee (2 doctors, 2 APNs and 8 RNs) which safeguard the use of ventilator and respiratory equipment and promote strong adherence to the ventilator care bundle in our department, the proposed guideline will first be introduced to them to gain their consensus and support. The nurse 30

31 consultant will take charge of the program and promote communication among physicians, the administration team and clinical staff. Other core members will be responsible for liaison with the medical device company, carrying out staff training sessions, resource arrangement, data collection and analysis, project evaluation and reports to the administration team. It may not be feasible for a respiratory team member to cover problems every work shift, but there is always at least one advanced practice nurse (APN) on duty. So all APNs will be trained first. If the pilot program is successful, a clinical audit will be done to ensure the guideline is carried out in a standardized way. Availability of staff The availability of manpower is a main concern in executing the proposed guideline. This not only involves the number of staff but also their motivational readiness, knowledge and technical skills when performing the change. Firstly, a platform should be provided for any enquiry about the innovation as all nurses have the freedom and right to voice their opinions about the new practice. Nurses opinions should be collected and their knowledge about current conventional oxygen therapy assessed in the preliminary phase of the pilot program. This shows respect to frontline users and allows them to know the importance of adopting the change. Secondly, all nurses should be equipped with knowledge and technical skills through provision of training sessions. The training session will include didactic and practical components. It will take about 30 minutes and be held in the ward. In the didactic component, teaching materials which include the physiological effects of HFNC, examples of evidence-based clinical trials, 31

32 structure of the high flow nasal device and accessories, explanation of the guideline, after-use issues and infection control concerns will be presented in power point. In the practical component, the operation of the device and trouble shooting will be demonstrated, followed by return demonstrations. At the end of the training session, a short quiz will be given to make sure participants know the critical points of the device. Cue cards will be attached to the device and the guideline and reference materials will be provided in the nursing station in case any staff member encounters problems in setup. During the evaluation stage, nurses are welcomed to comment on the device and the proposed guideline through anonymous evaluation forms and focus group discussions. It is anticipated that some staff may be unwilling or not confident to in handling the new technology. After comments are collected, the training sessions emphasizing evidence-based care and the guideline will be modified. Thus, we hope this innovation can maximize the benefits for staff by building a win-win, supportive working environment. Availability of resources Manpower and external materials must be managed in this project. As all members are from our department, no extra manpower is required. For materials, the manufacturer should be contacted for the product trial and quotations for the HFNC device and related equipment, including the heated tube kit, nasal cannula interface, tracheostomy connection, specialized filters, disinfection filters, sterile water and pole stands. For the training sessions, computer hardware, a printer, and A4 paper for hardcopies of the didactic materials and quiz will be needed. Photocopies of reference materials placed in the nursing 32

33 station and evaluation forms are also needed in the implementation period Cost benefit ratio of the innovation A critical part of initiating an evidence-based project is assessment of the costs and benefits of the innovation. Potential risks, benefits to different parties and costs will be explained below. Potential benefits to clients HFNC flushes anatomical dead space so oxygenation is improved. Moreover, it delivers a constant FiO2 despite an increase in the work of breathing and increased inspiratory demand due to less entrainment of air. Active humidification of oxygen through the nasal cannula enhances patient comfort and compliance with therapy (Frat et al., 2015; Parke et al., 2013; Maggiore et al., 2014; Sztrymf et al., 2012). As the days of ventilator use, length of stay in the ICU and mortality in the ICU are also reduced (Frat et al., 2015), the incidence rate of ventilator-associated pneumonia and relevant healthcare expenditures may be also reduced. Potential benefits to nurses and the organization Positive clinical outcomes could enhance nurses job satisfaction while improved patient comfort and compliance with therapy could reduce their workload. The success of implementing the evidence- based guideline could also enhance healthcare professionalism, cultivate an atmosphere of evidence- 33

34 based practice and establish an image of dedication to deliver the best possible care to the public. Potential risks A retrospective study mentioned that failure of HFNC may result in delayed intubation and worse clinical outcomes in patients with respiratory failure but no conclusion could be drawn (Kang et al., 2015). As with other types of oxygen therapy, HFNC can alleviate hypoxemia but may not improve ventilation. It does not ensure definite airway protection, so respiratory parameters still need to be monitored. Potential complications of oxygen use including carbon dioxide retention and oxygen toxicity must also be considered. Oxygen can support combustion so smoking and use of aerosol sprays should be avoided in the vicinity of the device. Cost The tangible costs include both material and non-material costs. The material costs consist of equipment, training materials and evaluation materials. We will purchase one AIRVO 2 humidifier from Fisher & Paykel Healthcare Ltd., which costs $20,000. We expect there will be around 20 patients in the 3 month pilot program. The costs including consumables for 20 patients would be $31,210. There are 120 staff in our department. Printed materials including quizzes for the training sessions and evaluation forms cost $12.5. For details, please refer to Appendix 4. Manpower expenditure will be the major component of the non-material costs. 34

35 According to the Civil Pay Scales (Civil Service Bureau, 2015), the estimated median monthly salary for APNs is HK$51,805 and for registered nurses (RN) is HK$32,560. The hourly pay rates for APNs and RNs are approximately HK$324 and HK$204, respectively. The training session takes 30 minutes. That means the total manpower cost for the whole department is around $10,080. The total cost of the 3 month pilot program will be $41, The cost will be $74,612.5 if the program runs for a year. For details, please refer to Appendix Evidence-based guideline To develop the evidence-based practice guideline, recommendations were made based on the summary and analysis of the data and results in the six eligible studies (Frat et al., 2015; Parke et al., 2013; Parke et al., 2011; Maggiore et al., 2014; Rittayamai et al. 2014;Vourc h et al., 2015). Furthermore, the recommendations were graded based on the guideline published by the Scottish Intercollegiate Guidelines Network (2012) (Appendix 7). The grade of most recommendations was high (i.e A/B) as all the studies were randomized control trials and their target populations were very similar to the population for this innovation. The details of the guideline are shown in Appendix 6. 35

36 Chapter 4 Implementation plan 4.1 Communication Plan After assessing the transferability and feasibility of the innovation, the implementation plan will be carried out with prior identification of stakeholders, thereby developing a comprehensive communication plan Stakeholder analysis In his classic book Strategic Management: A Stakeholder Approach, Freeman (1984) defined a stakeholder as any group or individual who can affect or is affected by the achievement of the organization s objectives. Stakeholder analysis is undeniably imperative in adopting change since a problem encompasses or affects numerous groups in a shared-power world. Taking stakeholders into account helps address needs, solve problems, fulfill purposes and facilitate change. The stakeholders in this innovation can be categorized into three levels: (1) administrators, (2) managerial staff and (3) operational staff. The administrators of the ICU include the chief of service, department operations manager, medical consultants, a nursing consultant and ward managers. They play an important role in approving and supervising the progress of the project, determining the success of the project and ensuring reasonable project funding and resource distribution. The managerial staff includes the project team members and all APNs who will liaise with the 36

37 medical device company, carry out staff training sessions, allocate resources, collect and analyze data, perform clinical supervision, evaluate the project and report to the administration team. The operational staff consists of all medical officers and frontline nurses working in the ICU. The medical officers can prescribe this therapy to patients if indicated. The nurses execute orders according to the guideline, monitor the patients clinical conditions and evaluate the intervention Communication process and strategies Communication is a process of conveying messages and information. A communication plan can help organize the actions that address the concerns of stakeholders and bring long-term benefits of the innovation to the target audience. A timeframe of the study plan is shown in table form in Appendix 8. As we initiate this innovation, we have to prepare a proposal and seek approval from the administrators in the first phase (Week 1). The proposal should state the existing problems in current practice, significance of the intervention, literature review, transferability and feasibility of the intervention and potential benefits of different stakeholders. Seeking resource assistance such as booking of venues and clerical support from the administrators is also important. We may have to make appointments with nursing consultants, the department operations manager, and ward managers and present the proposal in a meeting. Once the proposal is approved, the pragmatic part of the project begins in Phase 2 (Week 2). The nursing consultant will be invited to be the project director and a project team consisting of one APN as team leader and four RNs (from the respiratory committee) will be set up. 37

38 Scheduled meetings will be arranged among team members and the respiratory committee will be contacted in the upcoming meeting to gain consensus and support. In Phase 3 (Weeks 3-5), the first meeting of project team members and the director will be held. The significance and vision of this project will be shared in the meeting, followed by job allocation including staff training, liaison with the medical device company, guideline promotion, the pilot study and equipment preparation. In Phase 4 (Weeks 6-8), staff training sessions will be held for all operational staff with collection of feedback to refine the guideline before the pilot study. A pilot study will be done over the next three months (Phase 5). Data and feedback will be collected and analyzed during Phase 5 and Phase 6. The data will be evaluated and an evaluation report will be released to the administrators by presentation and in Phase 6. The guideline will be refined according to the pilot study and applied in the implementation phase which will last for six months (Phase 7). After this, a final project and evaluation report will be presented to the administrators (Phase 8). 4.2 Pilot Study Plan Objectives A pilot study is a preliminary trial carried out on a small scale to test the feasibility, discover adverse events and help predict the performance of the full study. The objectives of this pilot study are as follow: 38