Identifying and Understanding Risk Factors for Instability and Adverse Events Associated With Chest Physiotherapy in Ventilated Children

Research Questions

1. What is current chest physiotherapy practice within UK paediatric intensive care units?
2. How do physiotherapists make decisions regarding provision of chest physiotherapy in UK paediatric intensive care units and what other factors influence this decision making?
3. What do physiotherapists perceive to be risk factors for physiological instability and adverse events and how do they manage these?
4. What is the prevalence of physiological instability and adverse events associated with chest physiotherapy in ventilated children?
5. What are the risk factors/characteristics of children who display instability and/or adverse events associated with chest physiotherapy?
6. What is the long term impact on the child of instability and adverse events associated with chest physiotherapy?

WORK PACKAGE 1

Phase 1 A descriptive, cross sectional study investigating current chest physiotherapy practice in paediatric intensive care units.

Sample The subject group is physiotherapists working within the 28 NHS PICUs in the UK. Using information from a local benchmarking study which investigated physiotherapy staffing levels on seven UK PICUs, a total of 33 physiotherapy staff was reported, providing an average of four per site. Extrapolating this for the 28 national centres gives an approximate population size of 112 physiotherapists.

Method An electronic survey, created in Smart-survey will be used. The first page of the questionnaire will include a covering statement and a link to the more comprehensive participant information sheet. The survey includes dichotomous and multiple choice closed questions, and open questions where more detail is required. It will be self-administered and anonymous.

Key topics covered include:

• Chest physiotherapy techniques used
• Chest physiotherapy referral criteria
• Decision to treat/not treat processes
• Risk factors monitored
• Demographic information The survey will be open for eight weeks and a reminder email will be sent after four weeks. Consent will be implied if the survey is completed and returned.

Data Analysis The anonymous data will be imported into Excel and descriptive statistics will be used including frequency distributions, statistics of central tendency, and dispersion. Qualitative data from the open questions will be thematically analysed, as described by Braun and Clarke (2006).

Phase 2 A qualitative study to explore and understand the decision making processes that guide delivery of chest physiotherapy, including the perceived risk factors for physiological instability and adverse events, and risk management strategies.

Sample Nine paediatric critical care units have been pre-selected, these are representative of all 28 UK NHS PICUs. Up to three physiotherapists will be purposively sampled from each centre to ensure a heterogeneous sample in terms of agenda for change banding and years of experience. This strategy will generate a sample size of 18-27 and ensure saturation of themes.

Method Semi-structured interviews will be conducted. Eligible participants will be provided with appropriate study information and all participants asked to give written consent prior to taking part. The interviews will take place at a time and location convenient to the participant, outside of their contracted working hours. The interviews will last between 30-60 minutes. Where possible the interviews will be face to face, however Skype/telephone interviews will also be offered. A topic guide for the interviews will be developed using the data from phase 1, existing literature, and advice from the steering group. The topic guide will be piloted in the first three interviews. The pilot data will be included in the data set unless significant revisions to the topic guide are required. The interviews will be audio recorded, with permission, and transcribed verbatim. Field notes will be taken to provide context and information on non-verbal cues.

Data Analysis Thematic analysis will be used to analyse the interview transcripts, to identify, organise and report themes within the data. The themes will be presented logically and meaningfully, evidenced and illustrated with quotes as appropriate. A visual display of the findings will also be developed.

WORK PACKAGE 2 An observational (case-note review) study to assess the variability and prevalence of physiological instability and adverse events associated with chest physiotherapy, and identify risk factors that predict instability.

Sample The sampling frame will consist of patients on the three paediatric critical care units at a tertiary children's hospital (neonatal, general paediatric, cardiac).

Subject Recruitment Patients who meet the inclusion criteria will be identified from physiotherapy records on a weekly basis by the chief investigator.

Data Collection This study will involve the collection of four categories of variables: primary and secondary outcomes, risk factors/demographics, and general PICU outcomes. These are routinely monitored and recorded for all patients on the PICUs. They are retrieved from high resolution monitor data (T3) and electronic patient records (EPR). A workspace within the digital platform Aridhia will be designed collaboratively with the GOSH Digital Research, Informatics and Virtual Environments (DRIVE) team. This will provide a comprehensive and secure database to collate, store and analyse the data. The DRIVE team will downloaded the de-identified data into the workspace.

Sample Size and Planned Recruitment Rate The use of digital methods allows a larger sample than in previous experimental and retrospective studies. Retrospective data will be collected over a 12 month period. An audit of chest physiotherapy across the three PICUs was completed over one week. Twenty two patients, accounting for 77 CPT events, were identified who would meet the study inclusion criteria. Scaling this up, data collection over a 12 month period would provide an approximate sample of 1144 patients and 4000 CPT treatments. Completing the data collection over a full year will avoid the impact of seasonal bias on the sample.

Data Analysis The analysis will be completed using the R Project for Statistical Computing. Multilevel analysis will be used to investigate change in the physiological variables and allow risk modelling.

Data Synthesis The findings of both work packages will be synthesised to determine similarities and disparities between the perceived risks of physiological instability and adverse events associated with CPT and those demonstrated by the objective data. Side by side comparison for merged data analysis will be used in a summary table and through visual data representation. This will clarify the new knowledge created and areas for service improvement.