Is paediatric endotracheal suctioning by nurses' evidence based? An International Survey

Background: Endotracheal suction (ETS) is essential in intubated patients to prevent tube occlusion and is one of the most common nursing interventions performed in intensive care. Aims and objectives: To explore how paediatric ETS practices reflect evidence-based practice (EBP) recommendations in paediatric intensive care units (PICU) worldwide. Study design and methods: A cross-sectional electronic survey linked to a real patient suction episode. Nurses completed the survey following a recent ETS episode. Evidence-based practice (EBP) was defined based on four of the American Association for Conclusions: Considerable variation in paediatric endotracheal suctioning practices exists internationally. Although most nurses applied single components of evidence-based recommendations during ETT suctioning, just a quarter applied all four elements. Relevance to clinical practice: Nurses ’ need to consider and strive to apply EBP principles to common nursing interventions such as ETS.


| INTRODUCTION
Endotracheal suction (ETS) is essential for any child with an endotracheal tube (ETT), its primary aim being the removal of secretions and prevention of obstruction of the child's airway. 1 Failure to clear secretions can result in an obstructed or occluded ETT, which, if untreated, will impair oxygenation and ventilation and gas exchange, potentially resulting in cardiopulmonary arrest. 2 Although essential, ETS has established adverse effects including bradycardia, atelectasis, hypertension, hypoxaemia, and cardiac arrest, 3 and the risk of these complications may be increased in high-risk children. 4 The most recent published guidelines for ETS are those by the American Association of Respiratory Care (AARC) in 2010, 5 but many of these recommendations are based on limited evidence and extrapolated from adult data.
In 2015, Tume and Copnell 6 reviewed the paediatric ETS evidence, finding limited evidence for many aspects of ETS. Therefore, the aim of this study was to explore international nursing practices in ETS and to compare this to the best-available evidence on ETS.

| METHODS
We conducted an international, cross-sectional electronic survey of paediatric intensive care unit (PICU) nurses. Nurses were asked to complete the survey after performing ETS on their assigned child patient (age termed to 17 years) and to consider this patient when responding to the survey questions. The survey was initially open for a 5-day period in November 2016 but re-opened for another 5-day period at the end of 2016 to increase response rates. Nurses who had completed the survey in the first round did not complete round 2. We excluded suctioning on pre-term infants or adults and any suctioning not performed by nurses. The e-survey was translated into eight different languages by bilingual international PICU colleagues who generously gave their time to support this international survey (English into French, Spanish, Portuguese, Italian, Finnish, Slovakian, Dutch, and Latvian) using a recognized cultural translation and adaptation process. 7 These translators also predominantly acted as the country lead. All questions were multiple choice or close-ended to avoid the need for back-translation of responses. In

| Survey development and refinement
What is known about this topic • Endotracheal suctioning is one of the most common nurse-performed procedures in intensive care units.
• The application of best evidence-based principles to nursing other nursing interventions has been shown to be variable.

What this paper adds
• An international perspective on nurses' use of evidencebased principles during endotracheal suctioning in critically ill children.
• Evidence that despite published recommendations, there is considerable variability in nurses' use of evidencebased practice when performing endotracheal suctioning in children.
total, 22 versions of the survey instrument were built (one per country who agreed to take part). A translation record was kept for all languages so that all survey responses could be combined into one data set for analysis. Once countries agreed to participate, the country lead, who took responsibility for ensuring ethical requirements were fulfilled in their country and determined the best method for survey distribution in their country, was sent a link to the survey for distribution within their country. In some countries, the survey distribution was via a professional society, for others the survey link was sent from the individual to PICU nursing leads. Instructions for the survey completion were sent to country leads regarding the inclusion and exclusion criteria and instructions that the nurse who performed the suction should complete the survey as close to the suctioning episode as possible to reduce recall bias.
PICUs were only identifiable by country; no other unit-identifiable data were collected. Nurses completed the e-survey as soon as possible (within the same day) after their selected suction episode to ensure their account of the episode was as accurate as possible. For all participants, completion of the survey implied consent.

| Defining evidence-based practice
Pragmatically for the purpose of this study, the study team defined evidence-based practice (EBP) for paediatric suctioning, based on four of the AARC (2010) best evidence recommendations 5 and the latest review and recommendations for paediatric suction, 6 as pre-oxygenation prior to suction (evidence grade 2B), use of a suction catheter no more than half the diameter of the ETT (evidence grade 2C), the depth of suction to the length of the ETT or no more than 0.5 cm beyond (shallow suction) (evidence grade 2B), and the application of continuous suction pressure upon withdrawal of the catheter (no evidence grade). Our survey aimed to capture both planned and unplanned suction episodes; because of this our definition of EBP did not include other recommendations (eg, suction indications, use of saline) that might apply only to planned (non-urgent) suctioning episodes. We believe the four recommendations we chose to define EBP suction practice were applicable regardless of situation (planned or urgent) and all are recommended based on some evidence.
We categorized patients into low-and high-risk groups for analysis. High-risk patients in this survey were defined as follows: congenital heart disease requiring single ventricle repair, traumatic brain injury (TBI) with intracranial hypertension, and high-frequency oscillatory ventilation (HFOV).

| Data analysis
Data were exported from SurveyMonkey (San Mateo, California) in a CSV file into Microsoft Excel (Microsoft Corp, Washington DC) and then into IBM SPSS v22 (IBM Corp., Armonk, New York) for analysis. Most of the survey data were categorical and are presented descriptively as percentages, or for continuous variables with median (interquartile range [IQR]) or mean (SD) depending on the normality of the data. Non-parametric tests were used to test both the relationship and correlation between the categorical variables. The dependent variable was the use of EBP (as previously defined as meeting the four criteria) and the independent variables tested were nurse experience (under or over 5 years), specialist paediatric intensive care qualification, the presence of a local unit suction guideline, and an English-speaking country. Chi-square test and Spearman's rank test were used to test these relationships. Spearman ρ was used for categorical data to examine whether there was a correlation between the use of evidence-based guidelines and key variables. A P value of <.05 was considered significant and two-tailed tests were used.

| RESULTS
In total, 446 questionnaires were completed in nine different languages by participants from 20 countries.  (Figure 1). In total, 59% (259/443) reported pre-suction oxygen saturations (SaO 2 ) were above 95% and 76% (335/443) stated they were 90% or above. Only 4.7% (6/443) reported that SaO 2 saturations were between 70% and 80% pre-procedure and the remaining 1.6% (7/443) stated that SaO 2 were less than 70%. The lowest reported SaO 2 was 25%. We did not ask for specific patient diagnoses and therefore cannot relate which proportion of these are children with cyanotic congenital heart disease.  No relationship was found between compliance with EBP and nurse experience (>5 years) (P = .253) or specialist PICU nurse education (P = .171). Nor was there any relationship between application of EBP and the presence of local suction guidelines (P = .487) or between English-speaking countries and non-English speaking (P = .587). In addition, there was no relationship between suctioning the "higher risk" patients (single ventricle repair, TBI or HFOV) and

| Deterioration associated with suctioning
nurses' use of EBP (P = .839). In the higher risk groups, there was also no significant relationship between nurse experience (>5 year) (P = .200), specialist nurse education (P = .307), presence of local guidelines (P = .87), or an English-speaking country (P = .407) to the application of EBP.

| DISCUSSION
This study is the largest and only international survey of nurses' suction practice in paediatric intensive care, which specifically links a real patient suction episode to the nurses' practice.
Nurse experience or PICU education did not enhance the use of evidence-based suction practice. This contrasts with the findings of a survey of ICU nurses in Tanzania, 8  index of mortality score. 19 Statistical modelling revealed that saline instillation was significantly associated with an increased risk of arterial desaturation.

| Strengths and limitations
There are some limitations to our study that warrant mentioning.
Although we asked nurses to relate their suction episode to a specific child, the data were self-reported, cannot be verified, may be affected by recall bias, and the nurses' perception of deterioration is subjective.
The possibility of selection bias may also reflect better motivated nurses and we cannot know whether this is representative of all PICU nurses. Additionally, we combined both mild and severe events and all patient deterioration events, which may have impacted our findings.
Nurses' selection of patients to include may also have induced bias towards the less sick children and in those there was no deterioration.
Our pragmatic definition of EBP criteria is based on the 2010 AARC recommendations, much of which lacks robust paediatric evidence, and this may have impacted on our results. However, we used the most common criteria for EBP and "best practice" across all patient types, based on these published recommendations. The overrepresentation from North America and Europe may introduce some bias and limit generalizability. Due to our survey distribution methods, and an unknown denominator, we are unable to calculate a response rate. Despite these limitations, we believe by conducting the survey in this way and relating nursing practice to a specific patient we have captured a realistic international picture of paediatric suctioning practice by nurses.

| CONCLUSIONS
This is the largest international survey of nurses' suctioning practice of critically ill children; as such it provides important new information about nurses' use of EBP recommendations during suction. We found that despite these widely available published guidelines (in 2010), the evidence was not being utilized in practice by all nurses at the bedside. Further work needs to explore how nurses use evidence-based recommendations, why guidelines are not followed for endotracheal suctioning, and how guidelines have been implemented into units.