Using Best-Worst Scaling to Understand the Preferences of Stroke Survivors, their Relatives/Carers and the Public on Providing Stroke Thrombectomy Services in England

Background Outcomes for acute ischaemic stroke patients with large artery occlusion are signi�cantly improved with endovascular thrombectomy (EVT). We aimed to identify preferences of stroke survivors, their relatives/carers and the public in England on: (i) EVT service organisation (localised versus centralised expert-delivered EVT services) and (ii) modelled outcomes for EVT effectiveness, equity of access and costs of EVT provision (based on increasing the number of comprehensive stroke centres across England that provide EVT from 24 to 30). Methods Stroke survivors, their relatives/carers and the public were engaged in an iterative co-design process to develop an accessible survey incorporating two Best-Worst Scaling (BWS) tasks: EVT service organisation (4 attributes each with 2 levels) and modelled outcomes (3 attributes each with 2 levels). On-line pilot testing was undertaken with 10 stroke survivors/carers. BWS data were transformed into standardised scores with corresponding 95% con�dence intervals (CIs). Results 105 respondents completed the survey (stroke survivors [18%]; relative/carer of stroke survivor [32%]; member of public [50%]). Centralised (local) EVT provision by experienced and specialised medical teams was associated with the strongest positive preference (0.76, 95% CIs = 0.69, 0.82), followed by secondary inter-hospital transfer for EVT; associated with positive preferences whether required (0.15, 95% CIs = 0.07, 0.22) or not (0.30, 95% CIs = 0.22, 0.38). Travel times by emergency ambulance > 45 min (-0.29, 95% CIs=-0.37, -0.22) were associated with stronger negative preferences than length of stay < 48 hours [-0.16, 95% CIs=- 0.22, − 0.10). Maximal effectiveness of EVT (52% recovering with none/mild disability) was associated with stronger positive preferences (0.40, 95% CIs = 0.27, 0.53) than equity of access (71% or 72% of eligible patients) to EVT within 7 hours since onset of stroke symptoms, which was associated with marginal positive preferences. Costs of EVT provision were associated with negative preferences; maximal cost had the strongest negative preference (-0.29, 95% CIs=-0.40, -0.18


Background
Acute stroke is a medical emergency [1], the fourth leading cause of premature death in the United Kingdom [2][3][4], and second globally [5], with two thirds of survivors left with long-term disability [6][7].The standard emergency care pathway for acute stroke in the UK is shown in Fig. 1.
Pre-hospital assessment of patients with suspected stroke involves administration of a standard clinical checklist upon arrival at the scene by paramedics -the Face Arm Speech Test (FAST) which records facial weakness, arm weakness and/or speech disturbance [8].FAST positive patients are conveyed by emergency medical services (EMS) ambulance to the nearest hyper-acute stroke unit (HASU), which ideally is pre-alerted.Brain (CT) imaging and stroke specialist review is undertaken to con rm stroke and sub-type (ischaemic or haemorrhagic).In the case of con rmed ischaemic stroke (accounting for approximately 85% of stroke [9]), intravenous thrombolysis (IVT) is usually indicated when time from symptom onset is less than 4.5 hours [10].In addition, patients with large artery occlusion (LAO) within 6 hours of symptom onset are eligible for secondary transfer via EMS ambulance (referred to as drip and ship) to a specialist comprehensive stroke centre (CSC) for further treatment with endovascular thrombectomy (EVT) [11].
EVT in the treatment of acute ischaemic stroke caused by LAO signi cantly decreases functional disability, with no increase in mortality or symptomatic intracranial haemorrhage [12].Between 10,140 and 11,530 stroke patients in the UK are eligible for EVT annually, equating to 11-12% of UK stroke hospital admissions [13,14].In 2016, there were 24 CSCs in England that provided EVT, most did not provide it 24/7, with variable use of imaging scoring systems and criteria in uencing patient selection for EVT [15].With reference to an individual patient data meta-analysis of EVT [16], it is estimated that an additional 2,420 stroke patients per year could achieve functional independence with comprehensive implementation of EVT in the United Kingdom (UK), with annual savings of ~£73 million over the lifetime of patients [13].
There are multiple options for increasing provision of EVT across England.For example, a mothership model involving bypassing the nearest HASU with direct transfer of stroke patients with large artery occlusion to CSC for EVT.However, due to a range of issues such as the lack of robust strati cation tools for use in the pre-hospital phase by paramedics [17], a drip and ship model (treatment with IVT at the nearest HASU before transfer to a CSC for EVT) was found to be the most preferred service con guration amongst professionals [18].In a previous modelling study [19] using genetic algorithms, one optimal solution to expedite EVT within the drip and ship paradigm would be to increase the number of CSCs providing EVT in England.A modest geographically targeted increase, for instance from 24 to 30 centres, has been indicated as maximising the population gains [19,20].Modelled outcomes for 24 versus 30 CSCs across England using a discrete event simulation indicated this service recon guration would deliver health gains within current thresholds of cost-effectiveness with a very high probability of being cost-effective [20].
It is important to capture and integrate patient and public preferences with other research evidence in health economic modelling to inform commissioning decision-making.Understanding patient and public preferences on planned or potential service re-organisation and the impact on outcomes can help to optimise the acceptability of decision making by commissioners and service providers, including the subsequent uptake and e ciency of healthcare provision [21].
Attributes associated with a drip and ship model of EVT provision such as: safety of treating a patient with IVT at an acute stroke unit, and then immediately conveying the patient via emergency ambulance to a CSC for EVT (referred to as a secondary hospital transfer); the impact on equity of access to EVT for eligible patients within 7 hours across England; treatment e cacy and costs are all likely to be sensitive to patient and public preferences.
Patients and public should be involved in the co-design of methods used to capture their preferences and values.This is imperative for optimising accessibility by people with different health literacy levels [22], and in the case of stroke survivors, post-stroke communication di culties (aphasia) can inhibit their capability to respond meaningfully.
We aimed to identify preferences related to localised versus more centralised EVT services from the perspective of stroke survivors, their relatives/carers and the public with speci c reference to (i) service organisation for EVT in England; and (ii) outcomes derived from health economic modelling.

Methods
Best-Worst Scaling (BWS) is a technique that belongs to the conjoint analysis family of methods that aim to identify peoples' preferences and tradeoffs that contribute to their choices about products or services.BWS surveys are arguably less cognitively demanding than other conjoint analysis methods such as discrete choice experiments (DCEs).BWS offer a better way of identifying the relative trade-offs between levels of attributes (compared with making judgements on groups of attributes within scenarios as in DCEs or simple comparisons using Likert scales) [23,24].DCEs typically present a series of alternative scenarios that convey information on multiple attributes with a range of variable levels, whereby respondents choose their preferred scenario [25].Conjoint analyses make explicit the trade-off between the bene ts of increasing one attribute at a cost of decreasing another.After careful consideration and expert advice, BWS was selected in this study as the preferred technique to elicit patient, relative/carer and public preferences.
In a BWS task, respondents are asked to select both the "best" (i.e.most preferred) and "worst" (i.e.least preferred) items in different subsets.The participant chooses "the pair that exhibits the largest perceptual difference on an underlying continuum of interest" [24,26].BWS has been used in a variety of health and health care settings.For instance, recently published studies have used BWS to understand patient, carer and public preferences on multiple aspects of healthcare, particularly on medical treatments [27][28][29][30].In this study, we used the BWS pro le/attribute case method (also referred to in the literature as 'Case 2'), in which the importance of each contributing attribute of a decision is evaluated relative to all other attributes.Two separate BWS tasks were designed.One focused on critical attributes of service organisation for EVT, and the second explored preferences on modelled outcomes for increasing the number of CSCs across England.

Selection of attributes and levels for the service organisation BWS task
Attributes and their levels for service organisation were derived from a survey (N = 147 responses) of stroke survivor (n = 27/18%); relatives/carer of stroke survivors (n = 51/35%) and other members of the public (n = 69/47%) views on service organisation for EVT in England (Appendix 1a & b).This resulted in four attributes with 2 × 2 × 2 × 2 = 16 possible combinations of attribute levels for the service organisation BWS task (Table 1).

Selection of attributes and levels for the modelled outcomes BWS task
Previous modelling studies indicated that increasing the number of CSCs from 24 (current number in England) to 30 would deliver both greatly enhanced population access to EVT plus health gains with a very high probability of being cost-effective [19,20].Therefore, three outcomes with levels informed by 24 and 30 CSCs were included as attributes in the modelled BWS task: (1) clinical effectiveness of EVT; (2) cost of setting up, maintaining and running CSCs; and (3) equity of access to EVT for eligible patients.Each modelled outcome attribute had two levels, although to reduce the cognitive burden on respondents the attribute level descriptors did not refer to numbers of CSCs.
Levels for the modelled clinical effectiveness of EVT attribute were described as modi ed Rankin Scale (mRS) outcomes derived from those published by the academic Highly Effective Reperfusion evaluated in Multiple Endovascular Stroke trials (HERMES) collaboration [11].The mRS outcomes were de ned as the proportion of patients that would likely be independent (mRS 0 to 2 -mild or no disability), dependent (mRS 3 to 5moderate or severe disability) or dead (mRS 6) after treatment with EVT within 4 hours (30 CSCs) or 4 to 6 hours (24 CSCs) at 90 days follow-up.
The levels for effectiveness (41% and 52% recover with no/mild disability) re ected proportions of patients that would reach a CSC within the treatment window, and receive treatment with EVT, due to increasing the number of CSCs from 24 to 30 across England.These outcomes were taken from Saver et al [11] Fig. 1B, where the reduced effectiveness of EVT over time is estimated.
As opposed to use of currency values, the cost of EVT provision was represented using a metaphor to ensure that costs were evaluated against another hi-tech and potentially highly valued health service provision.After considerable consideration, the metaphor chosen was the equivalent cost of setting up, maintaining and running MRI scanners for 2 years.Like angiography suites for EVT, additional MRI machines involves the provision of hi-tech expensive imaging infrastructure with multiple different staff groups required to run them, yet is an item which would be at least somewhat familiar to many, even most, respondents.Costs were estimated via in-house costing work (HL) based on published data [11,19,[31][32][33][34][35][36][37][38] and expert opinion.Equity was operationalised as proportions of patients with and without access to thrombectomy within 7 hours since onset of stroke symptoms.This resulted in three attributes with 2 × 2 × 2 = 8 attribute level combinations for inclusion in the modelled outcomes BWS task (Table 2).

BWS Survey Design
The research programme PPI representative (DB) facilitated recruitment of 7 stroke survivors and their relatives/carers from the NE CRN (Stroke) PPI Panel in an iterative co-design process to develop the form and content of the BWS survey informed by best evidence on presentation of probabilities (numerical and graphical [39]) and guidelines on the design of information for people with aphasia [40].Online pilot testing was conducted with 10 stroke survivors/carers and two researchers with expertise in choice experiments (TR, DC).
By their very design and construction, BWS tasks are repetitive; therefore, attempts were made to present the fewest possible number of questions to maximise both participation in and completion rates of the survey.However, to have enough degrees of freedom to provide individual parameter estimates across key subgroups, a minimum of 8 responses per participant for each BWS task is required [41].
A randomised block design was used to generate two blocks of 8 questions for the service organisation BWS task (with each participant only completing 1 of 2 blocks) and one block of 8 questions (completed by all participants) for the modelled outcomes BWS task.Attribute levels within each block was controlled by ensuring even numbers of levels for each attribute were included between blocks.Maximum difference scaling was used, whereby respondents stated their most and least (best or worst) preferred options for four attributes (with 2 levels each) in the service organisation BWS task (Fig. 2) and three attributes (with 2 levels each) in modelled outcomes BWS tasks (Fig. 3).In total, individual respondents were required to answer 2 × 8 (16) BWS questions The nal survey (Appendix 2) was delivered via on-line QualtricsXM platform using the 'maximum difference scaling' question type.A URL to the anonymous survey was generously hosted on the Stroke Association website and a link to the survey, along with information about the study was distributed to all Healthwatch services in England and for national Clinical Research Networks to distribute to patient groups.The survey also appeared in June 2019 edition of the NHS In Touch Newsletter.The introduction of the survey provided an overview of thrombectomy and aims of the study, including a link to a short lm about thrombectomy produced by the BBC [42].In order to help completion of the BWS questions, a short information lm was embedded at the start explaining the repetitive nature of BWS surveys and the rationale for using this methodology in appropriate language and style for the intended PPI respondents [43].
Anonymised demographic data were collected at the end of the survey on age, gender, region of England, and respondent type (stroke survivor, relative/carer of a stroke survivor or member of the public).

Statistical Analysis
Standardised scores were calculated for each attribute level that ranged from − 1 to + 1, which represented the strength of preferences (best [positive scores] or worst [negative scores]) for the attribute level across the sample.These were calculated by establishing the difference between the frequency of an attribute level being chosen as best versus worst divided by the 'availability' of each attribute (the number of times it appeared across the design).Each attribute level was shown four times to each of the 105 respondents who fully completed the survey, giving a total attribute availability of 420.For example, if one level of an attribute was chosen as best (n = 320) and worst (n = 2), the difference (B-W) is 318, and the standardised score = 318/420 = 0.76 (positive preference).Corresponding 95% con dence intervals for standardised scores were generated by iteratively resampling with replacement each of 105 respondent's data 1,000 times (referred to as bootstrapping).All analyses were conducted in Stata version 15 [44] and Microsoft Excel [45].In other BWS studies, standardised scores have been shown to provide similar information to regression coe cients from a conditional logistic regression model, but with a greater ease of both interpretation and calculation [46,47].

Service Organisation BWS Task
Standardised scores for the service organisation attribute levels (Fig. 4) showed that an experienced/specialist medical team had the strongest positive (best) preferences (0.76, 95% CIs = 0.69, 0.82), with a local service staffed by less experienced medical teams the strongest negative (worst) preferences (-0.62, 95% CIs= -0.72, -0.52
There were no statistically signi cant differences between standardised scores strati ed in terms of stroke survivor/relative status versus the public for service organisation or modelled outcomes (Results not shown).

Discussion
This is the rst study to use BWS to investigate stroke survivor, relative/carer and public preferences for service organisation and outcomes in the context of stroke services.Our data indicate that expertise of centralised teams is of paramount importance for patients and the public.Combined with a positive preference for secondary transfer (to EVT capable hospital), our study suggests that patient and public preferences are broadly aligned with those of professionals for the drip and ship paradigm of EVT provision [18].Travel times of up to 45 minutes are likewise identi ed as acceptable to patients and the public, which concur reasonably well with transport time to nearest EVT centres derived from modelling studies (30 to 54 minutes in urban/suburban settings; 49 to 141 minutes in rural setting) [48].Mitigation of longer transport times for those residing in sparsely populated remoter rural settings could be achieved to a considerable extent with secondary transfer via helicopters instead of ground-based ambulance [49].
As expected, protracted length of stay in hospital was associated with negative preferences; although this could be offset to some extent by communicating to patients and public that EVT as well as improving clinical outcomes, reduces length of stay compared with medical therapy alone [50].Furthermore, 76% of respondents in the thrombectomy service organisation survey, which informed the corresponding BWS task would be willing to stay in EVT hospital for > 48 hours.
With regards to modelled outcomes, based on an increase in number of CSCs from 24 to 30 in England, improved clinical e cacy and higher costs were the most and least preferred outcomes respectively.The strong preference for improved e cacy is consistent with studies reporting that even relatively mild de cits from stroke result in negative patient preferences for health states due to stroke [51].
Our ndings suggest that, whilst increasing the number of centres may be acceptable to patients and public in terms of increased effectiveness, this may be offset by negative perceptions of the higher costs involved.Health care in England is publicly funded, and costs in the BWS task may have represented concerns that other highly valued services (such as MRI scanners used as a metaphor for costs in our study) would not be funded.
Respondents in our sample preferred effectiveness over equity, which were both valued more than costs.The levels for the equity attribute had an absolute difference of only 1%, which was informed by robust statistically modelling, but may have obfuscated the impact of this attribute in the BWS task.However, it may equally be the case that increasing equity of access for a minority of stroke patients was not considered to be su ciently important by patients and the public.Findings for both BWS tasks were consistent across respondent type, which indicated that experience with acute stroke may not be a factor involved in forming preferences about service organisation and outcomes in the context of EVT and thus allows these ndings to be considered more generalisable.
The advantages of our approach were engaging stroke survivors and carers in a meaningful, iterative co-design process, which enabled the development of an inclusive and accessible survey, which enhanced comprehensibility of the information and thus meaningfulness of responses.A potential disadvantage that could be inferred from our approach is the modest sample size achieved despite wide dissemination, and a preponderance of respondents from the north east of England.The inherent repetitive nature of BWS may have reduced the completion rate.
Furthermore, the attribute levels in the service organisation BWS tasks were explicit, whereas those in the modelled outcomes BWS tasks were implicit (i.e. it was not stated that the levels of attributes corresponded with 24 and 30 CSCs).This was done to decrease cognitive burden, but this restricts the extent that ndings can be used to justify public and patient support for increasing the number of CSCs across England.
Future work might explore patient and public preferences on helicopter emergency medical service transfers for EVT, including the method of preference elicitation using BWS and other commonly used conjoint analysis methods such as DCEs.We chose BWS on the basis that it is less demanding; however, people use a range of decision-making strategies [52].Consequently, preference-elicitation tasks may be susceptible to patient and public preferences and be variably acceptable to respondents when they are presented as BWS or a DCE format.Studies support the concept that responses to both survey methods may arise from identical underlying preferences [e.g.53].Indeed, studies have demonstrated the validity of assessing preferences of caregivers of people with Alzheimer's Disease for 17 attributes using a DCE and a BWS task, with two overlapping attributes in both preference-elicitation formats [54].A further study found similar patterns of preference weights from DCE and BWS for social care outcomes [55].

Conclusions
Our data indicate that preferences of stroke survivors, their relatives and the public for critical aspects of service organisation of EVT services are broadly consistent with preferences of stroke professionals and modelling studies.Secondary transfer to a CSC with travel time via emergency ambulance for EVT of up to 45 mins was acceptable to our sample of respondents.Clinical effectiveness was prioritised over equity of access with costs of EVT evaluated negatively.Commissioning decision-making should consider patient and public preferences on trade-offs between maximising health gains from EVT and preferences for costs.

Ethics approval and consent to participate
All respondents furnished with detailed information about the rationale for the study, including that participation is completely anonymous, data would be securely stored, and the voluntary nature of participation (see supplemental le).Participants indicated their consent by completing the anonymous survey.This survey was completed voluntarily and anonymously by respondents, did not include any sensitive topics, and did not involve recruitment from the NHS.

Supplementary Files
This is a list of supplementary les associated with this preprint.Click to download. BWSversion1Appendices.docx

Figure 2 .
Figure 2. Attributes and levels for the service organisation BWS task

Table 1
Attributes and levels in the service organisation BWS task

Table 2
Attributes and levels in the modelled outcomes BWS task Based on the NHS Health Research Authority decision tool (http://www.hra-decisiontools.org.uk/ethics/), approval by a Local Research Ethics Committee was not required [https://www.hra.nhs.uk/approvals-amendments/what-approvals-do-i-need/research-ethicscommittee-review/non-nhs-research-projects/].