This survey provides evidence of HCPs and P&Cs’ views on the use of ATs in upper limb stroke rehabilitation and a benchmark from which changing attitudes can be measured. The results are relevant to all healthcare systems and particularly those that are free at the point-of-care such as the UK NHS. ATs will become increasingly important in the drive to deliver cost-effective improvements in stroke rehabilitation outcomes and to satisfy, for example, the UK National Clinical Guidelines for Stroke [
8] that state: “Patients who have some arm movement should be given every opportunity to practise activities within their capacity.” Despite the increasing reference to ATs in healthcare policy, research into effectiveness and investment in commercial development, no previous survey has sampled or compared HCPs’ and P&Cs’ views of ATs. Our survey has generated new information about factors that influence clinical use of ATs and the opportunities for and barriers to translation of ATs into clinical practice. Based on our findings, we discuss how opportunities can be exploited and barriers overcome. We also discuss the influence our sample may have had on our findings, the strengths and limitations of the survey and how it will impact on future work.
Factors that determine clinical use
Strength of evidence for clinical effectiveness and usability has been cited as an important factor influencing translation of rehabilitation research into clinical practice [
33,
34]. The results of our survey generated evidence to support this: HCPs cited evidence-base as the most important of ten factors for an ideal technology compared to P&Cs who ranked it 5
th; and who considered both ease of set-up and comfort as more important. However, despite ranking evidence-base as the most important factor, many HCPs prescribed and used ATs that were not evidence-based. FES for example was cited as the most commonly prescribed AT by both HCPs (96/152 [63%]) and P&Cs, (16/45 [36%]), yet research evidence is equivocal. The Evidenced-Based Research in Stroke Rehabilitation, EBRSR) [
35], considers the evidence for FES in improving upper extremity function in acute stroke to be strong (level 1a), but the UK National Institute for Clinical Excellence (NICE) and the National Clinical Guidelines for Stroke [
8], do not consider it strong enough to recommend that FES is used other than in the context of a clinical trial. VR, the second most prescribed AT by HCPs (87/152 [57%]), currently has little evidence for its benefit, and is not recommended by any guidelines, but HCPs scored it highly on being ‘fun’ and ‘easy’ to use. Robot therapy, the least cited technology used by either HCPs or P&Cs, by comparison receives a more positive recommendation from both the EBRSR and the UK National Guidelines in terms of evidence for reduction of impairment and improvement of shoulder and elbow function which state that: Robot-assisted movement therapy should only be used as an adjunct to conventional therapy when the goal is to reduce arm impairment or in the context of a clinical trial. CIMT, for which there is strong (level 1a) research evidence and which is recommended by both EBRSR and the UK Stroke Guidelines was cited as prescribed by HCPs 77/152 [51%] but the AT most frequently prescribed by only 28/154 (18%) of HCPs. For P&C, CIMT was cited as the most frequently used AT by 4/45 [<10%] of P&Cs who responded.
Conflict between clinical use and research evidence may reflect the absence of definitive research evidence for any ATs. Other reasons for why some ATs are not more widely used could be limited applicability – CIMT is only suitable for people with >20 degrees of wrist extension and 10 degrees of finger extension, or the unwillingness of HCPs or P&Cs to use it for practical reasons, or, in the case of robots and biofeedback, factors such as cost, unsuitability for home-use, or use without supervision. In summary therefore, research evidence and evidence-based recommendations currently have little influence on choice of AT by either P&Cs or HCPs and other factors, such as usability, may be regarded by them as more important. Awareness of the evidence may also be a factor. All HCPs who used CIMT and robots and most (94%) who used FES thought that those ATs were evidence-based; yet only CIMT is recommended by both the UK and EBRSR guidelines.
These findings highlight the need to improve the evidence-base for ATs, particularly those that are currently being used and that satisfy the requirements of the respondents such as: ease of set up; fun to use; safe; comfortable and durable.
Mismatch between strength of research evidence and clinical use has been reported previously [
34] and the survey identified examples of technologies that are becoming widely used in clinical practice for which there has been strong commercial drive and investment. For example, the Saeboflex (
http://www.saebo.com) is used in many UK centres, Bioness Inc. (
http://www.Bioness.com) who manufacture upper and lower limb FES devices report increasing sales and reimbursement, and O
2 have launched a project to make the ‘Wii fit’ available to stroke patients as part of their Global e-health strategy (
http://www.oz.com). None of these devices has undergone large-scale clinical trials.
Conflict between key sources of information regarding effectiveness suggests that other factors are important in determining clinical use of ATs. Inevitably, in the absence of clear research evidence and guidance, HCPs’ decision-making is influenced by anecdotal evidence and their own experience. These may incorporate a complex integration of factors including observed clinical effectiveness, what they rate as important and the views of the patient and sometimes the carer. Our findings suggest that therapists take a pragmatic view when it comes to using ATs. They prioritise evidence, but also acknowledge that usability is crucial to avoid AT abandonment. One reason for this, that has been suggested, is lack of consideration of user opinion in design and selection of ATs [
36].
Cost effectiveness is a key factor in the adoption of any new treatment into clinical practice. The cost-effectiveness of using robots for upper limb rehabilitation, for example, has been reported in a single randomised controlled trial that also demonstrated modest clinical benefits compared with usual care at 36 weeks [
37], but further empirical research using economic analysis is required to demonstrate consistent results. In the UK a 3-arm RCT of robot therapy with an expected sample size of N = 720 is currently being undertaken (
http://www.nets.nihr.ac.uk/projects/hta/112605). The trial will generate the required evidence, however, cost-effectiveness in the use of expensive technologies is heavily dependent on how much devices are used and the proportion of patients who benefit. Our survey has shown that cost-effectiveness does not guarantee that new treatments, especially those that rely on a change in practice (such as ATs), will automatically be adopted [
8].
Barriers to clinical use
The survey has identified barriers to clinical use of ATs, but understanding them is fundamental to overcoming them. Much can be learnt from the wider field. For example, Reynaud (2008), proposed different models of acceptance (an attitude) and adoption (a process) of technologies from the fields of information systems and sociology [
38]. These models can be applied to any technology type, but can also be specialised to individual technologies. Acceptance models, such as the Technology Acceptance Model (TAM) [
39,
40], suggests that when users are presented with a new technology a number of factors influence their decision about how and when they will use it, including “perceived usefulness” (the extent to which a person believes that using a particular system would enhance their performance) and “perceived ease-of-use” (the extent to which a person believes that using a particular system would be free from effort). This model, which has been updated several times [
41‐
43], corresponds with the findings from the survey presented in Table
4 for the high priority given by both HCPs and P&Cs to the ease of set up and use for an ideal technology. Technology adoption processes reported included those developed by Rogers [
44] and Silverstone and Haddon [
45]. In the former, the model aims to provide a framework for understanding how technology innovations change, and are changed, by their social contexts. In the latter, the model focuses mostly on the individual’s decision to buy or not to buy. At a systems level the Normalization Process model was developed to assist both service providers and research constituencies in understanding how health care interventions, technologies, and practices are implemented, embedded, and integrated in everyday life [
46]. These models could be applied more widely within the AT field to assess key barriers to the translation of individual technologies and should be integrated into clinical effectiveness trials. Based on the survey’s findings, barriers reported by HPC and P&C were a lack of knowledge and access to ATs the following are recommendations to address these.
Knowledge-transfer and changes in service provision - key opportunities to translate ATs into clinical practice, are currently not exploited and therefore remain barriers. For example 41% of HCPs and 64% of P&Cs had not used ATs. When asked why not, 98/114 (86%) HCPs cited lack of access and 27/114 (24%), lack of knowledge. Among P&C respondents 48/78 (62%) knew nothing about them and 21/78 (27%) said they could not get one on the NHS. Similar responses were expressed in the ‘views’ section of the questionnaire, for example only 10% of P&Cs and 21% HCPs disagreed or strongly disagreed with the statement that it was ‘difficult to access training or advice’. Training was also identified as an important area for further development and could be used to raise HCPs awareness of research evidence.
Until ATs become a core element of therapist’s training they are unlikely to be used in routine clinical practice, but without evidence for their effectiveness, inclusion is hard to justify, despite the fact that many approaches to stroke rehabilitation currently taught are not evidence-based. A more pragmatic approach would be to include opportunities for learning about ATs (as well as training in using them) at post-qualifying level. Collaboration between universities, healthcare providers and the commercial sector may be an effective way of providing this. The wider study (of which this survey was part) began with an interactive exhibition of ATs that brought together commercial companies demonstrating a wide range of technologies, and patients, carers, researchers and clinicians. In doing so we increased clinicians’ awareness, knowledge and understanding of ATs and established a communication network between clinicians, researchers and the commercial sector that has been influential in the formation of the International Industrial Society in Advanced Rehabilitation Technologies (IISART). Awareness may be increased by similar exhibitions, run by trade organisations, in collaboration with universities providing healthcare courses nationally, or with regional or national specialist interest groups for HCPs. Keeping up-to-date with new information, given the time constraints in clinical practice, has been recognised as a barrier to achieving evidence-based stroke rehabilitation [
47]. Professional bodies can play a role in removing barriers by providing independent advice, based on new evidence, and employing a variety of accessible Internet-based methods such as webinars and podcasts.
Access to ATs is posited as a key factor that influences whether therapists have used or prescribed them. Currently, access to ATs in the UK is through commissioners of stroke services, who require a business case to justify their provision that identifies the cost benefits of using ATs, both in the short and long term. This would ideally include service level benefits: proven implementation in other regions, a description of need, number of people suitable for treatment, level of resource required (e.g. reduced costs through fewer repeat visits to the service; reduced level of social care and reusability of equipment). It would also include plans for auditing use such as number of suitable patients and duration of use. In addition, commissioners need estimates of benefits for patients, such as quality of life outcomes and increased social activity. Increased collaboration between HCPs and manufacturers is an additional way of providing better access to ATs, through for example, training and ATs ‘on-loan’, enabling HCPs to be more knowledgeable and experienced in using them.
The survey identified a conflict between research evidence and clinical use of ATs, and thus a failure in translation of research into clinical practice. Translation may be more efficient if research studies are pragmatic – testing ATs in the environment in which they will be used clinically and incorporating robust examination of user’s views and the burden that the AT makes on users. Furthermore, involving end users, both HCPs and P&Cs, into the design of clinical trials, may facilitate translation.
Strengths and limitations of the study
The study has several strengths. It is based on a nationwide survey of a large number of HCPs (n = 292) who work with stroke in a range of settings and 123 P&C who have experienced a stroke. The questions were developed using data generated by focus groups and subsequently pre and pilot tested to ensure relevance, comprehensiveness and to minimise bias. However there were limitations. There is a selection bias in the way the questionnaires were designed and potentially responded to. The methodology was designed to recruit a sample of the population of P&Cs and HCPs working in stroke rehabilitation. However, it is likely that, being a self-selected sample, the results are biased towards the views of people interested in ATs. Factors that will have contributed to this are that some respondents would have attended the exhibition. Whilst the questionnaire data were from a national sample, the majority of whom had not attended the exhibition, and as such were likely to be more representative of national views and experiences, people may have been more likely to look for and complete the questionnaire if they had an interest in ATs. In that respect knowledge and use of ATs reported here might be greater than in the population as a whole and the views expressed may be more positive. Additionally, the response rate cannot be specified for the online version of the questionnaire, and the fact the data were combined with the paper version raises issues about whether the data are directly comparable. The number of individuals involved is uncertain, for instance we cannot guarantee that individuals have not completed more than one questionnaire, and the representativeness of the responses is uncertain. Motivation to respond to the questionnaire may have differed across the different professional groups, leading to biased estimates of the popularity of some interventions. Another limitation of the study is that it was based on self-report: HCPs reports of their practice may not reflect actual practices. Some patients had suffered their first stroke many years before completing the questionnaire; their experiences may therefore not be up-to-date. Moreover, a social desirability bias (adapting responses to meet what people believe they should be thinking) cannot be ruled out especially when considering attitudes and opinions. Test re-test reliability and internal consistency were not tested, which along with cognitive interviewing processes would have strengthened the data.
Future work
This work has been conducted in parallel with surveys of stroke services to determine current AT rehabilitation methods and systematic literature reviews to provide both narrative descriptions and quantitative comparisons of each AT for upper limb function. This information will be used to determine which combination of ATs has the greatest probability of significantly improving upper limb rehabilitation following stroke, is cost effective and acceptable for use by patients and in health services. Results from this work have directly informed the design of a clinical trial which will be used to propose a new service delivery model, as well as provide an operational framework for future studies. Engagement with all stakeholders has been embedded through the whole programme of work.
Further work could explore the results of the current survey with, for example, qualitative interviews, with a purposive sample of respondents. Longitudinal studies would help to determine how quickly attitudes to arm and hand rehabilitation, and use of ATs in practice are changing.