Conceptualisation

There is no unified definition of SDM. A systematic review identified 40 SDM models currently available with 53 different elements clustered in 24 overarching components.1 Components present in more than half of models were: ‘describe treatment options’ (88% of models); ‘make a decision’ (75%); ‘patient preferences’ (68%); ‘tailor information’ (65%); ‘deliberate’ (58%); ‘create choice awareness’ (55%) and ‘learn about the patient’ (55%).

This study will conceptualise SDM using the ‘Three Talk model’ (2012),22 later refined to ‘Implement-SDM’ (2019).23 This single-component model provides a guide for enhancing health professional communication to deliver SDM, and is the most highly referenced model (>1800 citations, Web of Science). It involves three key steps consistent with other models of SDM: (1) introduction of choice, (2) describing options and (3) helping patients explore preferences and make decisions. This was the chosen model for an NHS MAGIC Programme7 and is recommended in NICE guidelines.24

Setting

Research will be conducted at three UK hospital trusts (North Bristol Trust/NBT, University Hospitals Bristol and Weston NHS Foundation Trust/UHBWFT and Bradford Teaching Hospitals NHS Foundation Trust/BTHFT) alongside quality improvement programmes to improve SDM. NBT is one of the largest acute NHS trusts in the UK.25 It provides a full range of acute clinical care for both local and regional clinical commissioning groups in South-West England. Specialised services are provided through NHS England, Welsh Health Boards and Welsh Specialist Commissioners. Services provided include elective and emergency gastrointestinal surgery, obstetrics and gynaecology, as well as specialist regional services in urology, neurosciences, trauma and orthopaedic and vascular surgery. One UHBWFT department is included as the South-West England regional cardiac surgical centre. BTHFT is an acute trust in the North of England with a full range of elective and emergency surgical services. Bristol and Bradford were purposively selected to maximise reach to ensure a diverse representative sample is included (eg, 26.8% classed as Asian or Asian British, compared with 5.5% in Bristol).

Feasibility of automated data collection

Feasibility assessment is designed to establish the feasibility of automated real-time evaluation of patient experiences of SDM at scale and will identify opportunities to optimise recruitment and data collection.

Participants

All patients over the age of 18 years who have been booked for planned vascular, gastrointestinal, urological, neurosurgical, gynaecological, breast, cardiac and orthopaedic surgical procedures at participating hospitals will be eligible to participate. Surgical departments have been selected to be broadly representative of a diverse range of surgical specialties. Excluded will be patients under the age of 18 years, those without decisional capacity to consent for medical procedures, or undergoing unplanned (emergency) surgery or endoscopic procedures. Data related to eligibility criteria are routinely collected through electronic patient record (EPR) systems.

Measurement of patient experience

Real-time measurement of patient experiences of the process of SDM will be facilitated by a secure, automated system procured through a third-party provider approved by NHS trusts. The system is a customisable off-the-shelf electronic patient-reported outcome measurement software and has previously been used for electronic data capture in other countries. Eligible patients will be identified through EPR using algorithms developed in collaboration with the software provider. Structured data queries will be designed to extract details of patients booked for eligible procedures. Queries will be designed to run automatically, securely transferring data from the hospital to the software provider daily to account for changes in scheduling. The automated system will send three validated SDM measurement instruments within 1 day after surgery booking (real-time baseline measurement). This time point in the decision-making process was chosen as a pragmatic point in time to represent patients’ cumulative experiences of SDM for surgery which may include discussions with surgeons, physicians, general practitioners, nurses, family and friends. The selected measurement instruments will be operationalised into an online survey and administered via short messaging service (SMS) or email. A reciprocal data feed will securely return patients’ survey responses immediately to the hospital data warehouse for secure storage (real-time analysis and feedback). Follow-up measures are sent within 1 day before surgery by either SMS or email (real-time follow-up measurement). A schematic of the process and intervention aims is illustrated in figure 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Schematic of measurement process. SDM, shared decision-making; SMS, short messaging service.

Selection of three SDM measures was made through discussions within the study team and was informed by a systematic review of SDM measurement instruments using COSMIN (consensus-based standards for the selection of health measurement instruments) methods,26 national guidelines19 and recommendations and use within NHS clinical practice.27–29 The CollaboRATE instrument is a validated three-item patient-reported measure assessing the extent of SDM experienced by patients.30 Assessment of the instrument using COSMIN methods demonstrated acceptable discriminative validity, concurrent validity, intrarater reliability and sensitivity to change.26 It has been used in excess of 40 studies,31 including evaluations of quality improvement projects in surgery.32 The SDM-Q-9 instrument is a validated nine-item patient-reported measure that evaluates their perceptions of involvement in the decision-making process.33 It has been widely used in interventional studies and demonstrates good reliability and structural validity.26 33 Systematic review evidence recommended use of SDM-Q-9 for surgery.34 The SHARED-Q10 instrument is a 10-item patient-reported measure to assess patient perceptions of information provided, involvement in consultations and agreement with the decision made.35 This measure is included because it was developed, validated and used in an NHS quality improvement programme36 37 and evaluates domains beyond patient perception of professional communication. Complete measurement instruments can be found in online supplemental figures 1–3.

Analysis

Feasibility of real-time monitoring will be evaluated by analysis of overall recruitment rate, response rates and time to response for the SDM measures at baseline and follow-up. Response rates will be presented as a number and percentage based on patients who completed the measures (eg, completed all three items of the CollaboRATE instrument). Issues of equality, diversity and inclusion will be explored by examining the correlation between responders/non-responders and sociodemographic patient variables extracted from EPR.

Relationships between responders and non-responders and clinical and sociodemographic details will be explored using univariable and multivariable logistic regression. Included will be age, sex, ethnicity, Index of Multiple Deprivation, and clinical and treatment parameters (eg, operation (three-digit Classification of Interventions and Procedures (OPCS) code), diagnosis (International Statistical Classification of Diseases and Related Health Problems (ICD-10) code), date of booking, specialty, number of outpatient appointments in relevant specialty, number and length of hospital inpatient episode in relevant specialty). Index of Multiple Deprivation will be derived by lower layer super output area for individuals’ postcode. All variables will be extracted from routinely collected data in EPR.

In addition, the study team will document any relevant technical, financial, administrative and logistical observations throughout the study and pertinent challenges using shared electronic records (eg, Microsoft Office suite). Any learning points will be descriptively summarised.

Usability of the electronic measurement system

Post-deployment usability testing will be conducted according to International Standards Organisation (ISO) standards for human–systems interaction (9241-11:2018) to evaluate the system’s use in this context.38 39 System users, defined as anyone who is a current or prospective surgical patient, will be invited to participate in a mixed-methods usability evaluation to assess system effectiveness, system efficiency and user satisfaction (see box 1). To complete usability evaluation, a process map will be created to define the number and type of task required to complete the measurement system.

System effectiveness

One-to-one user testing sessions will be used to assess system effectiveness by evaluating task completion and error rates. Sessions will involve completing the automated system in a simulated environment, applying concurrent think aloud techniques.40–42 A topic guide will be developed and will structure the testing session discussions (online supplemental table 1).

Patient and public representatives will be invited to participate in online user testing sessions. They will be eligible if they are over the age of 18 years. Individuals from two patient experience panels (NBT, BTHFT) will be recruited through respective panel coordinators. Sampling will be purposive to maximise variation in geographical location, ethnicity and sex, and will aim to include individuals whose first language is not English.

User testing will be completed using video-conferencing software (eg, Zoom, Microsoft Teams) and audio-recorded. Two researchers familiar with the automated system and trained in qualitative research will conduct the user testing sessions. Observational notes will be taken to collect further information about challenges or errors encountered during task completion.43 44

Task completion rates will be calculated as percentage of tasks completed. Error rates will be calculated based on number of user errors encountered. User errors will be deviations or problems experienced that will interfere with successfully completing the task. Number and type of non-critical errors (successfully addressed by testers themselves following instructions from the observer) and critical errors (require the observer to intervene or take remedial actions) will be noted. Results will be presented using descriptive statistics.

Understanding of system effectiveness will be supplemented by analysis of response rates generated through feasibility work in 1.1.

System efficiency

System efficiency will be assessed by calculating task completion time and task efficiency. Task completion time is defined as the time participants took from the first activity (starting the survey by following the hyperlink) to the last activity (submission of the survey). Task efficiency is defined as the time spent to complete each task. Analyses will be based on those who completed the automated system and for whom first and last activity timestamps were available.

User satisfaction

One-to-one user interviews will be conducted to assess user satisfaction in depth. Interviews will explore issues including ease of use/navigation, satisfaction with instructions, satisfaction with the visual display, ease of access, burden and likelihood of using the system again. Barriers and facilitators to completing the measurement system will also be explored. A topic guide will be tested and refined and used to direct discussions.

A subset of eligible patients and participants of the user testing sessions will be invited to take part. A purposive sampling strategy will be adopted to ensure that insights are drawn from a range of perspectives. Sampling characteristics will be (1) experience with surgery (vascular, gastrointestinal, urological, neurosurgical, gynaecological, breast or orthopaedic surgery) and good/bad SDM experience, (2) sex, (3) age, (4) ethnicity and (5) individuals whose first language is not English. Participant characteristics will be assessed as the study progresses and recruitment efforts will focus to target under-represented patients as necessary. Recruitment of the subset of patient participants will be undertaken by the principal investigator, research nurse or clinical collaborators via email or telephone. User testing participants will be recruited by researchers during the user testing sessions and interviews will be conducted immediately following the user testing session.

Interviews will be conducted primarily remotely (eg, telephone or video conference) by experienced and trained qualitative researchers. All audio-recorded interviews will be transcribed and anonymised. Transcripts will be thematically analysed (see below).

Views of diverse stakeholders to inform the use of the system to improve SDM

Qualitative research with a wide range of stakeholders (including patients, healthcare professionals and members of the community) will be conducted to understand views of multiple stakeholders to inform the use of the system to improve SDM. Opinions about the acceptability and potential impact of real-time monitoring of SDM will be sought. Views on potential intervention components (activities), mechanisms of change, intermediate outcomes, assumptions and indicators will be explored. Results will be used to co-develop initial programme theory to inform phase 2.

Patients and members of the public and community over the age of 18 years will be eligible to take part. The sample will include people who are disproportionately affected by a poor SDM process and outcomes of surgery: those who are economically disadvantaged, from minority ethnic groups and in older age.45–48 Professionals working in participating trusts will be eligible for inclusion and may include surgeons, anaesthetists, nurses, perioperative care physicians, allied health professionals and hospital managers.

Recruitment

Eligible participants will be identified through existing networks, collaborations with local hospital patient panels, community leaders and patients who have participated in feasibility (1.1) and usability (1.2) data collection. We will seek to recruit individuals who experience multiple intersecting inequalities to ensure the views of those with barriers to accessing healthcare are incorporated.45 Recruitment of members of the community will be conducted using techniques developed and successfully applied by the Born in Bradford team49 50 and the patient and public involvement and engagement (PPIE) group of the National Institute for Health and Care Research Bristol Biomedical Research Centre. Recruitment materials will be translated into most spoken languages within the local areas.

Purposive sampling will seek to achieve diversity in relation to sociodemographic characteristics (eg, age, gender), experience with surgery or SDM (eg, surgical specialty, good/bad SDM experience) or underserved groups (economically disadvantaged, older age, ethnic minority). Where appropriate, snowball sampling will also be used, whereby individuals who participate in the study are asked about other potentially interested participants. The sample size will ultimately depend on theoretical saturation (eg, when no new insights are identified from the data and sufficient data are collected to address the research question).51 52 It is anticipated that approximately 130 participants (around 105 patients and members of the community, and 25 professionals) will be required.

Data collection

Data collection will apply a flexible strategy to minimise perception that the research is ‘hard to engage with’.49 A range of qualitative research methods are planned remotely and/or face-to-face including semi-structured interviews, focus groups and participatory approaches (eg, community events, discussion groups). It is anticipated that a minimum of 30 one-to-one interviews and 6 focus groups are required, complemented by recruitment through community events and discussion groups. However, these methods may be adapted based on evolving best-practice evidence from citizen science53 and feedback from PPIE stakeholders. For example, evidence suggests that some British Asian people may be more willing to participate in a focus group in a familiar setting (eg, community centre) than other settings.54

Interviews and focus groups will be facilitated by experienced qualitative researchers based in Bristol and Bradford. Topic guides for interviews and focus groups will be developed to direct discussions. This will be iteratively refined during data collection to explore emergent views. Interviews and focus groups will be held face-to-face, over the telephone or using a secure video-conference service (eg, Zoom, Microsoft Teams) but will ultimately depend on participant preference. Data collection will primarily be conducted in English. However, where data are collected from non-English speaking members of the community, additional support will be provided by interpreters and specialist researchers who conduct relevant foreign language interviews and focus groups. All interviews and focus groups will be audio-recorded and transcribed verbatim. Field notes will be taken during the interviews.

Qualitative analyses

Transcripts and field notes will be analysed using a thematic approach with the help of qualitative data management software (NVivo). Principles of thematic analysis will be applied to the data whereby (1) transcripts and notes will be read and reread, (2) codes are generated and assigned to relevant excerpts within the transcripts, (3) themes will be identified by collating similar codes, (4) accuracy of themes will be checked and (5) detailed analysis of themes will take place.55 Analysis will involve linking transcripts and observational notes by integrating relevant data from both sources to gain a more comprehensive understanding of key findings. This process will primarily be inductive, with codes developed and iteratively refined through interpretation of the data. There will, however, be an a priori interest in examining data in relation to the study aims. For example, information to support evidence for the acceptability of monitoring the SDM process and impact of monitoring on clinical care will actively be sought.

Analyses will be conducted separately for different stakeholder groups (patients, professionals, community) to help ascertain different viewpoints or experiences reported by each participant group. Depending on findings, an additional layer of analysis may be conducted to contrast results for several subgroups (eg, different underserved groups; different specialties) to ensure differing perspectives and experiences by population and context are accounted for in later intervention development. At least two experienced qualitative researchers will perform analysis independently and meet regularly to discuss impressions of the data. A subset of transcripts will be double-coded by another experienced qualitative researcher. Any discrepancies in coding or interpretation of data will be referred to the wider study team for further discussions.

Summaries of findings from the analyses (descriptive reports) will be written, combining preliminary findings from the various data sources in relation to the study objectives. Drafts of these summaries will be prepared following rounds of recruitment and analyses and discussed within the study team. The summaries will be iteratively developed as analysis proceeds and will inform discussions about saturation.

Dedicated multidisciplinary meetings involving public contributors will be held to articulate an initial programme theory to inform the future development of the intervention to be more inclusive of recognised underserved groups. A summary of key findings from qualitative data collection will be prepared. We will draw on behavioural (Capability-Opportunity-Motivation Behaviour (COM-B) model)56 and organisational (Normalisation Process Theory)57 change theory to identify theory of how the intervention will work for underserved groups. Summaries will be combined to form a comprehensive report, providing a basis for phase 2.

Data management

All data will be generated and handled in accordance with relevant directives and regulations (eg, Data Protection Act 2018). Any data collected as part of qualitative data collection will be recorded using encrypted devices. Audio files will be securely transferred and transcribed by transcription services approved by the University of Bristol. Transcripts will ensure anonymity of participants (eg, in future study outputs) by assigning pseudonyms or participant IDs to replace any names or identifiable information. All electronic data files will be saved in restricted folders only accessible to the research team, on secure University of Bristol network space that adheres to the University of Bristol’s data security policies. Files containing any personal information (eg, contact details) will exclusively use the linked participant ID and will be encrypted and stored securely on the university servers.