Skip to main content

DEveloping A Complex Intervention for DEteriorating Patients using Theoretical Modelling (DECIDE study), Dr Duncan Smith, Senior Lecturer in Advanced Practice – City, University of London Hon. Nurse Consultant in Critical Care Outreach – University College London Hospitals NHS Foundation Trust – Wednesday 21 June 2023

By 23rd August 2023February 20th, 2024Previous Seminars

Abstract

Background

Patients who clinically deteriorate without recognition and/or response are at risk of unplanned admission to intensive care, cardiac arrest, and/or death (termed Serious Adverse Events (SAEs)). To mitigate SAEs, track-and-trigger tools are used internationally to prompt healthcare practitioners (typically nursing staff) to recognise physiological changes that signal deterioration, and to contact a practitioner with expertise in acute/critical illness. In the United Kingdom and parts of Europe, the National Early Warning Score (NEWS) (track-and-trigger tool) was developed and disseminated widely to standardise practice. Despite evidence track-and-trigger tools (like NEWS) improve patient outcomes, their translation into clinical practice is inconsistent. This is partly attributed to nursing staff failing to change their behaviour.

Aim

The aim of the project was to develop a theory-based behaviour change intervention to enhance enablers and overcome barriers, to Registered Nurses (RNs) and Healthcare Assistants (HCAs) enacting expected behaviours in recognising and responding to signs of patient deterioration.

Methods

A mixed methods design was used with three phases: 1. Focused ethnography on two clinical floors in an acute hospital to compare directly observed behaviours (of RNs and HCAs) with those specified in policy. From directed content analysis of field notes, target behaviours were identified, specified, and shortlisted; 2. Brief (not audio-recorded but recorded in field notes) interviews were conducted soon after direct observation of relevant behaviour. Some brief interview participants were recruited for an audio-recorded, semi-structured, interview informed by a Theoretical Domains Framework (TDF) topic guide. Interview data were analysed deductively (the 14 TDF domains were coding categories) and inductively to identify determinants (i.e. barriers and enablers) of target behaviours. TDF domains representing important determinants were identified using published criteria and linked to Behaviour Change Techniques (BCTs) from expert consensus literature; 3. BCTs were shortlisted by the research team and presented to clinical stakeholders alongside example applications (i.e. concrete strategies for operationalising BCTs). Using Nominal Group Technique, stakeholders ranked BCTs and their potential applications for acceptability and feasibility. Ranking data were used to inform the content of a preliminary intervention.

Results

During 300 hours of fieldwork, 499 items of data (i.e. an episode of observation or a set of vital signs from chart review) were recorded; 289 (58%) associated with expected (i.e. policy-specified) behaviour; 210 (42%) associated with unexpected behaviour (i.e. alternative behaviour or no behaviour). Ten behaviours were identified as potential behaviours for change; shortlisted to seven target behaviours. Brief interviews were conducted with 39 RNs and 50 HCAs, and semi-structured interviews with 16 RNs and 16 HCAs. Quotes from interviews were linked to nine (for brief interviews) and 14 (for semi-structured interviews) TDF domains. Nine TDF domains were identified as being of high importance: Knowledge, Social Professional Role and Identity, Beliefs about Consequences, Reinforcement, Intentions, Goals, Memory, Attention and Decision Processes, Environmental Context and Resources, Social Influences. These domains were linked to 50 BCTs; shortlisted to 14. Ranking data from two nominal groups held with 19 stakeholders were used to shortlist further, resulting in a preliminary intervention that includes an educational package and 12 BCTs that will be delivered through workshops and on acute wards, using 18 applications.

Conclusion

This research made a unique contribution to the international body of evidence, as it is the first study where a theoretical framework of behaviour change and a taxonomy of BCTs were used to model an intervention to improve responses to clinically deteriorating patients. The intervention is preliminary, as it is anticipated that it will be piloted in clinical practice and refined during a subsequent feasibility study.

Biography

Dr Smith  is a clinical-academic nurse with a background in acute and critical care spanning two decades. Duncan holds a Senior Lecturer post in the School of Health and Psychological Sciences at City, University of London where he is a module leader for the MSc in Advanced Clinical Practice and a contributor to the PGCert in Critical Care. Alongside his academic role, Dr Smith is an honorary Consultant Nurse with the Patient Emergency Response and Resuscitation Team (PERRT) at University College London Hospitals (UCLH) NHS Foundation Trust where he continues to work clinically. Duncan’s clinical experience as a critical care outreach nurse in several organisations across London shaped his research activities pre-doctorally and informed his PhD. In April 2018 Dr Smith was awarded a National Institute of Health Research (NIHR) Clinical Doctoral Research Fellowship to complete a programme of work titled: DEveloping a Complex Intervention for DEteriorating Patients using Theoretical Modelling (DECIDE study). The aim of his doctoral research was to develop a theory-based, complex, behaviour change intervention to optimise recognition of, and response to, patients with signs of clinical deterioration. Dr Smith has recently been awarded a NIHR Development and Skills (DSE) Enhancement grant to upskill ahead of applying for an advanced clinical academic fellowship. For his post-doctoral work, Dr Smith anticipates designing and delivering a feasibility study to pilot the intervention at multiple hospital sites and to test the feasibility of using a cluster RCT to evaluate its efficacy.