Introduction
Having investigated what is currently known about the decision-making process in our systematic reviews, and explored current practice and experience in our ethnographic study, we drew on our findings to develop an intervention that would support consistent, transparent and ethically justifiable patient-centred decision-making. In developing the DSI we were aware that it needed to be feasible to implement in daily practice.
Our specific objectives were to develop:
a DSF for decisions about referral and admission to ICU
support and training materials for clinicians using the framework
information for patients and families on decisions about referral and admission to ICU.
From our preceding work (systematic reviews, ethnography and choice experiment), we identified the key patient-related, situational and process-related factors that influenced (or were thought should influence) decisions about referral or admission to ICU. Using data from all three sources provided triangulation and strengthened our evaluation of these factors as being relevant to the decision-making process. These factors could then be emphasised, supported or mitigated within the intervention. From our ethnographic study, we also mapped the temporal and relational aspects of the decision-making process, the points of conflict, and the implicit and explicit values that informed it. This enabled us to map the DSI to current practice that would be recognisable to clinicians, and to address points of conflict and value positions in the associated educational support. Finally, the analysis in our ethnographic study specifically focused on what would constitute a good decision-making process and the normative framework to support this. This analysis fed directly into the development of the DSI and the implementation plan. For example, our analysis identified the importance of involving patients and their families in the decision-making process and of explicit balancing of harms and benefits, and the infrequency with which either of these occurred in practice. The choice experiment identified different preference patterns among ICU doctors, and this finding supported the need for explicit articulation of reasons when making a decision. Thus, in developing the DSI we needed to include prompts for both patient and family involvement and explicit balancing of reasons.
Development of the decision-support intervention
The development of the DSI (conceptual framework plus supporting resources) was in five phases:
initial development of a draft DSI (conceptual framework and supporting resources)
presentation of the DSI at a stakeholder conference including focus groups to identify areas for revision and refinement
post-conference refinement of the DSI based on this feedback
development of educational materials
refinement of the implementation strategy using normalisation process theory (NPT).
Initial development of the decision-support intervention
Developing a conceptual/cognitive framework
The first step was to develop a ‘best practice’ conceptual framework for decision-making, based on our previous work, that encapsulated what is important/essential to patients and clinicians in practice and what is ethically required.
Ethical reasoning is an integral part of clinical decision-making but is often implicit rather than explicit. Findings from our systematic review and ethnographic study show that clinicians’ values influence their decision-making, and that they do not explicitly balance different empirical and normative considerations when making a decision. There are some ethical frameworks and heuristics used in teaching medical ethics, most commonly a principlist approach. However, it seems unlikely from our observation work that providing a theoretical ethical framework for use in day-to-day practice would be successful. In seeking to develop a framework that could be embedded in daily practice, be congruent with clinical decision-making and prompt ethical reasoning, we identified the ‘accountability for reasonableness’ (AFR) framework157 as the most useful model on which to build our DSF. This is a commonly used ethical framework for considering the allocation of limited medical resources and focuses on the process of decision-making rather than relying on a specific moral theory. AFR has four requirements: decisions must be transparent, based on reasons that stakeholders can agree are relevant and revisable in the light of new evidence and arguments, and there should be an appeals process. Other authors have highlighted the requirement for priority setting decisions by clinicians (including ICU admissions) to be transparent (AFR requirement 1) and ethically justifiable (AFR requirements 2 and 3).173,174 Therefore, any framework to support decision-making should improve the transparency and ethical justification of decisions about referral and admission to ICU. A further ethical requirement is one of equity, which can be interpreted as consistency of process (the factors taken into account and the decision-making process should be consistent across all patients and all decision-makers).
Process of developing the framework
A generic model of the steps in the decision-making process was created based on the findings from our ethnographic study ().
The generic model decision-making process. Reproduced from Rees et al. This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and (more...)
Clinical and non-clinical members of the research team were then asked to map the factors identified in the qualitative study and systematic reviews (tabulated in Appendix 10) to relevant stages of the process. From this, a draft framework was developed, specifying content at each stage in the decision-making process and ensuring that overall it was consistent with the AFR ethical requirements. The resulting framework was discussed and refined in a series of meetings involving members of the research team and two members of our PPIAG. The draft framework was then used as a basis for developing the supporting materials ().
Development of supporting resources for clinicians
The supporting resources of the draft DSI were:
documentation to prompt and record best practice in the referral and decision-making process
information for patients and families to help them participate in the decision-making process
educational resources to support the implementation of the intervention.
Development of the decision support documentation
We developed a referral form (for the referring team to use) and a decision-support form (for ICU doctors to use). The aim of the forms was to provide the clinician with a framework within which to structure their decision, and to prompt them to include the key steps identified as necessary to ensure an ethically justified decision-making process. A series of drafts were produced and revised at a number of meetings of the research team and members of the PPIAG. Feedback was also sought on their format and usability from clinicians at sites that had participated in the ethnographic study. The final versions, following revisions informed by the stakeholder conference, are presented in Appendices 11 and 12.
Electronic format
As part of the development work, we worked with the Health and Social Care Information Centre, now NHS Digital, to develop an electronic version of the referral and decision-support forms. Draft electronic copies of the referral and decision-support forms were created in the Lorenzo electronic health record system (DXC Technology, Tysons, VA, USA) by collaborators at the NHS Digital. This team also developed system agnostic criteria for electronic versions of the forms that IT developers in individual NHS trusts could use to incorporate the forms into local operating systems.
Development of patient and family information leaflets
The patient and family information leaflets were initially developed by a member of the study PPIAG. The research team and other members of the PPIAG then refined and expanded these drafts, drawing on the findings of the ethnographic study, particularly interviews with patients and families. These were presented at the stakeholder conference. The final versions, including revisions following the stakeholder conference, can be found in Appendices 13 and 14.
Consensus conference
A 1-day conference was held in July 2016 at the University of Warwick to gain the views of a wider stakeholder population regarding the content of the DSI and the facilitators of and barriers to its implementation.
Delegates were invited from the following stakeholder groups (see Appendix 15 for the list of delegate representatives):
Patients –
people who had survived a critical illness
patient groups who may be particularly likely to need intensive care treatment (e.g. those with heart disease or lung disease)
patient groups who may be disadvantaged in terms of access to health care (e.g. those with a learning disability)
ICU doctors (including professional body representation)
ICU nurses and CCOR nurses (including professional body representation)
referring specialty clinicians (e.g. cardiology, renal, elderly medicine, respiratory, surgery)
regulatory authority, legal and health-care management representation.
Method
Meeting sessions were chaired by a PPI co-investigator (CW) and by a senior member of the Faculty of Intensive Care Medicine. Before the event, delegates were sent copies of the draft DSI and a summary of the findings of the research to date. The event included presentations of the research on which the DSI was based and a demonstration of the electronic version of the referral and decision-support form. Delegates then participated in focus groups in which they were asked to consider specific elements of the intervention and answer a series of questions relating to these elements (see Report Supplementary Material 13 for topic guides). Each focus group included a cross-section of delegates to ensure a balance of disciplines and backgrounds within each group. The focus groups were facilitated by a member of the study team and the discussion was captured by a medical student acting as a scribe. The discussion was recorded with participants’ consent and recordings used to check accuracy when interpreting the written notes of the discussion. All focus groups were asked to consider the referral form and the decision-support form, and either patient or family information leaflet or learning outcomes for educational resources. Two key points from each group on each element considered were transcribed onto flip chart paper and placed around the main meeting room. In a subsequent plenary session all delegates were asked to read the flipcharts and write any further thoughts or suggestions on sticky notes and attach them to the flip charts.
Analysis
Following the conference, the scribes’ notes and points from the sticky notes were circulated to two members of the research team (AS and HH) for analysis. The data was categorised under the headings: referral form and process; DSF; patient and family information; education; and barriers and facilitators to implementation. A summary of this analysis was used by the research team to refine the documentation and develop a strategy for implementation of the DSI.
Post-consensus conference development of the decision-support intervention
A conference report, including the summary of key points and the changes made to the intervention in response, was sent to all delegates with an invitation to provide further comment (see Appendix 16). The DSI was then finalised for use in the implementation feasibility study. The changes made to the components of the DSI are outlined below.
Refining the components of the intervention
Development of the referral and decision-support documentation
The focus of all of the components of the DSI was broadened to include decision-making regarding all forms of critical care support. The term ‘family member’ was changed to ‘person closest to the patient’. The referral form was modified to be closer to the SBAR (Situation, Background, Assessment and Recommendation) format. Specific questions related to ICU bed availability and ward capacity were removed and reframed to ask first, what was the optimum treatment for the particular patient, and then to ask where this care could be safely delivered.
Development of the educational package
Key suggestions were incorporated into the development of training materials including the importance of communication, relevant knowledge of legal and regulatory frameworks, a clear ethical framework, and guidance for using the forms (see Report Supplementary Material 14 for copies of educational materials).
Development of the patient and family support material
New drafts of the patient and family information leaflets were written by members of the PPIAG and our PPI co-investigators. Notably the use of separate leaflets for patients and family/friends was developed rather than a single combined leaflet. The language was simplified and the layout improved. The finalised leaflets received a Crystal Mark from the Plain English Campaign.176
Refining implementation of the intervention
To identify potential difficulties in the implementation of the DSI and to improve the chances of its successful implementation, eight intensive care consultants and CCOR nurses from sites that had been involved in the ethnographic study attended an implementation planning meeting. The meeting used an adapted NPT177 (see Report Supplementary Material 15) toolkit: a way of exploring changes to practice to understand how an innovation is implemented and integrated into the work of an organisation.
The DSI was presented, and how it was perceived was explored using adapted questions from the NPT toolkit. Responses were categorised into the four constructs of NPT: coherence (could they make sense of the intervention), cognitive participation (could they envisage their colleagues and other key stakeholders actively engaging in this process), collective action (could they envisage this working in practice and what might be the barriers and facilitators) and reflexive monitoring (how could they envisage the intervention changing practice and how would they evaluate this) (see Report Supplementary Material 15). This analysis revealed particular concerns regarding the active engagement of all stakeholders and uncertainty around organisational support for the intervention creating difficulties for implementation.
In response to this feedback, we adapted our implementation strategy to include approaching the medical directors of each NHS trust participating in the implementation study to secure organisational support for the intervention and research. We also adapted our educational resources and the train-the-trainer sessions to encourage and support stakeholder engagement.
The decision-support intervention
The final DSI is described briefly below.
Guidance for referral for intensive care support
We developed guidance to improve communication between clinical teams around referral. The guidance is summarised as follows:
Referral to intensive care should not be delegated to junior members of the clinical team.
Referring teams should clearly articulate what support they are seeking from the intensive care team.
The benefits and burdens of intensive care should be considered for each patient.
Ensure that the patient and/or their family has been consulted in the referral process.
The referral should follow recognised best practice using the SBAR format.
The referral form (see Appendix 11) provides a structured format for information gathering and communication with the intensive care clinicians to facilitate a collaborative decision-making process.
A decision-support framework to guide intensive care unit doctors when making decisions
The framework describes best practice for decision-making in three domains: (1) evidence – the collection of information relevant to the decision; (2) reasoning – using this evidence to identify relevant outcomes, balance the burdens and benefits of treatments and make recommendations for treatment; and (3) implementation – ensuring that decisions are acted on and communicated effectively. The framework aims to integrate ethical reasoning within the decision-making process in a practical and accessible manner. The decision-support form (see Appendix 12) enables the decision-making process to be recorded to facilitate transparency and consistency of decision-making.
A pocket-sized aide-mémoire (), which has the referral guidance on one side and the framework for decision-making on the other, was provided to act as a reminder of the referral guidance and decision framework to referring teams and ICU doctors.
A pocket-sized summary of the cognitive framework to act as an aide-mémoire.
Patient and family support information
Patient and family information leaflets (see Appendices 13 and 14) to be given to patients and/or their families/someone close to them when they are recognised as being critically unwell and a decision regarding ICU referral is being considered. The leaflets are intended to support discussion between clinicians and the patient/family and not to be given in isolation.
Educational resources
Educational resources to support the implementation of the DSI included (see Report Supplementary Material 14):
‘train-the-champion’ session for implementation champions (including an introductory presentation to the legal and ethical background)
a presentation of the DSI and its rationale at hospital Grand Rounds or other formal meetings
a shorter presentation for use at more informal or smaller departmental meetings, for example quality improvement meetings
a brief ‘key points’ presentation for opportunistic teaching of individuals or small groups
a table-top case-based simulation exercise to familiarise clinicians with the intervention.
The final version of all components of the DSI formed the basis of our implementation feasibility study (see Chapter 7).
Summary
This chapter has described the development of an intervention to support best practice in decision-making for escalation of treatment and admission to intensive care. The development was informed at each stage by the findings of our preceding WPs and input from our PPI co-investigators and advisory group. Accountability for Reasonableness provided the normative theoretical framework for the intervention. It was developed to fit with established clinical practice while offering support for improvement in that practice, and supported by resources to increase the likelihood of successful implementation.
