Background Clinical trials are required to strengthen the evidence base for prehospital care. This questionnaire study aimed to explore paramedics’ perceptions of prehospital research and barriers to conducting prehospital clinical trials.
Methods A self-completed questionnaire was developed to explore paramedic perceptions and barriers to undertaking prehospital trials based upon a review of existing research and semistructured qualitative interviews with five paramedics. The questionnaire was distributed by ‘research champions’ to 300 paramedics at randomly selected ambulance stations in Yorkshire.
Results Responses were received from 96/300 participants (32%). Interest in clinical trials was reported, but barriers were recognised, including perceptions of poor knowledge and limited use of evidence, that conducting research is not a paramedics’ responsibility, limited support for involvement in trials, concerns about the practicalities of randomisation and consent, and time pressures. No association was found between training route and perceived understanding of trials (p=0.263) or feeling that involvement in trials was a professional responsibility (p=0.838). Previous involvement in prehospital research was not associated with opinions on importance of an evidence base (p=0.934) or gaining consent (p=0.329). The number of years respondents had been practicing was not associated with opinions on personal experience versus scientific evidence (p=0.582) or willingness to receive training for clinical trials (p=0.111). However, the low response rate limited the power of the study to detect potential associations.
Conclusions Paramedics reported interest and understanding of research, but a number of practical and ethical barriers were recognised that need to be addressed if prehospital clinical trials are to increase.
- prehospital care
- research, clinical
Statistics from Altmetric.com
High quality evidence is required for effective prehospital care.1 Randomised trials (and meta-analysis of such trials) are widely accepted as providing the highest level of evidence for therapeutic interventions. However, a lack of randomised trials has been identified in prehospital care. A review undertaken in the USA in 1997 found that only 54 randomised controlled trials had been published in prehospital care, many less than in other areas of emergency care.2 Prehospital care thus often relies upon extrapolation of evidence from inhospital trials, even though provider training, diagnosis, conditions and time frames involved may differ between these settings.3 This lack of research has been considered responsible for holding back the development of prehospital care4 and may mean that the growing prehospital literature is based on lower forms of evidence, such as case reports.5
Various barriers to undertaking prehospital clinical trials have been identified, such as regulations and difficulties obtaining appropriate consent.6 ,7 Problems have been encountered in the few prehospital trials undertaken to date in the UK. For example, Woollard et al8 encountered technical difficulties during their trial of ambulance telemetry in delivering early thrombolysis and a trial by Turner et al9 investigating fluid resuscitation after traumatic haemorrhage was undermined by poor paramedic compliance with the alternative treatment protocols. Attempts to address such barriers have included cluster randomisation. For example, Perkins et al10 randomised ambulances to be equipped with an active compression/decompression device for cardiac arrest. Cluster randomisation provides a pragmatic approach to evaluation but carries risks of bias due to postrandomisation selection of participants.
Few studies have been undertaken into paramedic perceptions of clinical trials and barriers to their use in the prehospital setting. Schmidt et al11 used a questionnaire survey to explore paramedic and other emergency medical services provider's attitude towards research in Portland, USA. The results revealed an appreciation of the importance of research and identified a number of barriers. Burges Watson et al12 conducted semistructured focus groups to investigate paramedics’ perspectives on involvement in out-of-hospital research as part of a specific project in the UK: ‘Developing and Assessing Services for Hyper-acute Stroke’. The researchers found that three key themes emerged: patient benefit, professional identify and responsibility, and time. Issues surrounding patient benefit included balancing the paramedics’ perceived role of enabling access to treatment, and the research desire of helping to produce or prove a ‘cure’. Issues were identified with the randomisation process being seen as a breach of trust in the patient–paramedic relationship. If the paramedics do not know what treatment they are giving, especially if it is possibly a placebo, then this was felt to be not truthful. The rigid structure of randomisation triggered fears that paramedics’ professional responsibility and integrity of clinical decision making would be limited. There were concerns about additional time required to complete trial protocols and balancing the eventual benefits of the research completion with the need of individual patients to be treated and transported as quickly as possible.
Prehospital clinical trials are required for effective emergency healthcare but our understanding of paramedic perceptions of clinical trials is limited. We aimed to explore the perceptions of paramedics towards prehospital clinical trials and barriers to their use.
A literature search was undertaken for studies identifying issues relevant to the feasibility of prehospital clinical trials. OVID, Scopus and Web of Knowledge were searched using the terms ‘prehospital’, ‘clinical trials’, ‘emergency medical services’, ‘paramedics’ and ‘health services research’ and relevant papers selected by a single reviewer (KH). The search was augmented by hand searches of the Emergency Medical Journal and the Journal of Paramedic Practice. The following themes were identified from the literature search: the paramedic role (changes to the role, professionalism, research roles and resistance to change), consent, randomisation, blinding, increased on-scene and transfer time, and the prehospital environment.
Face-to-face semistructured interviews were undertaken by a single researcher (KH) with five paramedics with an expressed interest in research who were identified by a senior manager at the Yorkshire Ambulance Service (PM). The interview schedule used nine open questions to elicit the interviewee's perceptions of the need for, importance of and uses of clinical trials; the anticipated difficulties of undertaking clinical trials (specifically considering recruitment, patient autonomy, treatment delays and additional workload); and potential solutions to difficulties. Written interview notes were taken, reviewed and informal analysis for themes undertaken.
The following themes were identified through the interviews: knowledge of clinical trials; knowledge of evidence based practice and its applications; use of and access to peer reviewed scientific journals; extent of independence and freedom to make clinical decisions; the evidence base of the ambulance service's clinical guidelines; paramedics as a professional body; the impact of the change in paramedic training; perceived or experienced problems with recruiting and randomising patients into clinical trials; ethical concerns with patient recruitment and consent; and paramedic willingness to complete additional training if required.
The themes identified in the interviews formed the basis of the questionnaire and were rationalised into grouped sections. Section A dealt with demographic details, experience, training and level of education achieved. Section B explored paramedic knowledge about clinical trials, use of scientific journals and integration of knowledge into clinical practice. Section C explored evidence based medicine within the ambulance service and what influences attitudes towards evidenced based practice. Section D explored the feasibility of different randomisation methods. Section E explored the practicalities of gaining patient consent, attitudes towards the ethical need for consent and opinions about the extra training or paperwork related to involvement in trials. The questionnaire used a combination of open and closed questions, and is included as an online supplementary appendix.
We selected 300 paramedics from the Yorkshire Ambulance Service to participate in the study. Assuming a 70% response rate, based on a similar previous survey of health professionals,13 this would give 200 responses, enough to estimate a typical proportion of around 20% with a reasonable degree of precision (ie, a 95% CI of 15% to 26%). Technicians and paramedic assistants were excluded as their training is unlikely to include enough relevant education about clinical trials, and they would be unlikely to be asked to independently recruit patients to trials. The sample of 300 was achieved by randomly selecting ambulance stations from within the Yorkshire Ambulance Service and including all paramedics in each sampled station until 300 was reached.
Research packs (including a covering letter, a participant information sheet, a supplementary information sheet defining a number of terms used in the questionnaire, a business return envelope and the questionnaire) were distributed to a ‘research champion’ at each of the sampled stations, who then distributed the packs to paramedics. As an incentive to participate paramedics were offered entry into a draw for a £50 gift voucher if they completed and returned the questionnaire.
The primary data analysis was descriptive: percentages were reported with a 95% CI for categorical data and mean with a SD for continuous data. Secondary analysis involved testing six prespecified hypotheses regarding the association between paramedic characteristics and perceptions. These hypotheses are described in the Results section.
Responses were received from 96/300 paramedics, giving a response rate of 32%. The respondents had a mean age of 39 years (range 24–63), 54 (56%) were male subjects and had been practicing for a mean of 8 years (range 1–30). The training route was through the Institute of Health Care Development (IHCD) training programme for 55 (57%), foundation degree/diploma for 35 (37%) and another route for 6 (6%). The highest educational level was GCSE/O level/equivalent for 16 (17%), A level/equivalent for 22 (23%), undergraduate degree for 43 (45%) and postgraduate degree for 15 (16%).
Responses showed that 30 (31%, 21%–40%) were absolutely sure, 62 (65%, 52%–72%) were reasonably sure, 3 (3%, 0%–6%) were neutral and 1 (1%, 0%–3%) was not at all sure what a clinical trial was. Some 34 (35%) had been involved in prehospital research in the past, of whom 16 (59%) had been involved in a clinical trial and 11 (41%) had not. Regarding scientific journals or articles, 4 (4%, 0%–7%) never read such material, 49 (51%, 41%–61%) read once or twice per year, 36 (38%, 28%–48%) read once or twice per month, and 7 (7%, 2%–12%) read every week. The main barriers were not knowing how to access the material (14 respondents), lack of time (61) and lack of computer access at work (20).
Table 1 shows responses to the statements regarding use of scientific journals or articles. Although the respondents were generally positive about their willingness and ability to use scientific evidence, they were less positive about their ability to put evidence into practice.
Table 2 shows responses to statements about evidence based clinical practice within the ambulance service. There was general acceptance of the importance of clinical trials but less agreement regarding the practicalities of undertaking trials.
Table 3 shows the responses to questions regarding the practicalities of randomising patients in clinical trials. The paramedics were asked to rate the practicality of various different randomisation methods. The full range of responses was reported for all statements, indicating a low level of agreement between the respondents. There was a tendency towards cluster randomised methods (ie, randomising groups of patients being treated on the same day or by the same paramedic) being considered more practical, but even these were considered to be impractical by a substantial proportion of the respondents.
Table 4 shows the responses to statements regarding potential problems with individual patient randomisation in clinical trials. Again, there was little consensus among the respondents but a substantial proportion responded that they would be concerned about the issues raised, particularly if randomisation might result in delays to treatment or switching between protocols.
Table 5 shows the responses to statements regarding potential problems with cluster randomisation. There was little consensus again, but a substantial proportion responded that they would be concerned about treatment delays, loss of skills if they were allocated to a protocol that did not require certain interventions and the ethics of randomisation before consent implicit in cluster randomised trials.
Table 6 shows the responses to statements regarding informed consent in clinical trials. There was general agreement on the importance of informed consent and willingness to follow a trial protocol that had been approved by an ethics committee. However, a substantial proportion disagreed with positive statements about randomising before consent, using proxy consent and not consenting individual patients when cluster methods are used.
Table 7 shows responses to statements regarding training and paperwork. The respondents were very positive about undertaking additional training to support clinical trials but most agreed that completing additional paperwork would cause operational problems.
The results of the six preplanned hypothesis tests are as follows:
Paramedics trained through the new foundation programme are more likely to be confident in their understanding of prehospital clinical trials: 9/35 (26%) were absolutely sure and 26/35 (74%) were reasonably sure of what a clinical trial is, compared with 19/55 (35%) and 32/55 (58%), with 3/55 (5%) neutral and 1/55 (2%) not at all sure among those trained via the IHCD route (p=0.263).
Paramedics trained through the new foundation programme are more likely to feel a professional responsibility to become involved in clinical trials: the proportions strongly agreeing, agreeing, neutral and disagreeing were 5/35 (14%), 14/35 (40%), 14/35 (40%) and 2/35 (6%), compared with 8/55 (15%), 25/55 (45%), 17/55 (31%) and 5/55 (9%) trained via the IHCD route (p=0.838).
Paramedics who have been involved in research will feel that it is important to have an evidence based ambulance service: the proportions strongly agreeing, agreeing and neutral were 15/34 (44%), 18/34 (53%) and 1/34 (3%) compared with 25/62 (40%), 34/62 (55%) and 3/62 (5%) who had not been involved in research (p=0.934).
Paramedics who have been involved in research will feel it is important to gain informed, written consent before recruitment into a clinical trial: the proportions strongly agreeing, agreeing, neutral and disagreeing were 6/34 (18%), 16/34 (47%), 7/34 (21%) and 5/34 (15%) compared with 18/62 (29%), 32/62 (52%), 7/62 (11%) and 5/62 (8%) who had not been involved in research (p=0.329).
Paramedics who have been practicing for a greater number of years will consider personal experience to be more important than evidence from clinical trials: linear regression showed no association between number of years as a paramedic and responses to the statement ‘personal opinion is more important than scientific evidence in clinical practice’ analysed as an interval scale (p=0.582).
Paramedics who have been practicing for a greater number of years will have more concerns about completing additional training: linear regression showed no association between number of years as a paramedic and responses to the statement ‘if I had to attend training sessions about the trial protocol, I would be happy to do so’ analysed as an interval scale (p=0.111).
The respondents to our questionnaire reported good understanding and appreciation of the need for clinical trials to develop a sound evidence base, but a relatively low use of published evidence and poor confidence in applying it in practice. Responses suggested a perception of lack of support and opportunities in research, but significant interest in becoming involved and undergoing further training was expressed, if support were increased. Significant ethical and practical barriers were recognised, especially fears that randomisation and consent processes could delay patient treatment. We found no evidence that paramedic opinions depended on their training route, previous involvement in research or number of years practicing as a paramedic.
A similar questionnaire survey was undertaken in the USA in the context of a specific trial. Schmidt et al11 found that 93% of respondents agreed that prehospital research was important, almost identical to the 92% agreement in our study. There were other similarities, for example, both studies found concerns about research processes causing delays in patient treatment. In contrast to our study, they found a significant difference in opinion based on the number of years practicing, especially when asking about the acceptability of enrolling patients before consent. Schmidt et al11 concluded that one of the key steps to addressing the identified barriers is to improve paramedic education and involve paramedics at every stage in trial planning.
In common with our project, a qualitative study undertaken in the UK alongside a stroke trial by Burges Watson et al12 identified that time limitations were a key barrier and that ethical concerns were raised, particularly about offering a treatment that might be a placebo. They concluded that education and paramedic involvement at every stage of the trial design and implementation were required to overcome barriers. The study also reported enthusiasm among respondents that did not necessarily lead to willingness to participate.
The study had a number of strengths and weaknesses. The survey was developed using face-to-face interviews to ensure that it was based upon the perceptions of paramedics rather than those of the researchers. It was piloted to ensure understanding and designed using the principles of good survey design. The main limitation was the low response rate. This means that the sample may not be representative of paramedics in general and may not have been powered adequately to test the preplanned hypotheses. We might speculate that paramedics with an interest in research may be more likely to respond, in which case the reporting good understanding and appreciation of the need for trials may be overestimated. Conversely, the reservations and potential barriers to research may be underestimated. Another limitation that a questionnaire can only measure reported behaviour rather than actual behaviour may also lead to an overestimate of support for trials and underestimate the problems. Only 35% of the respondents had experience of involvement in prehospital research, so the perceptions reported are unlikely to be based on direct experience.
It is not clear why the response rate was lower than anticipated. The distribution of questionnaires through research champions was anticipated to be a strength due to their rapport with paramedics. However, distribution was slow and response rates varied markedly between ambulance stations, indicating varied commitment to the project from research champions. Furthermore, removing control of distribution from the lead researcher (KH) meant that non-responders could not be systematically sent reminders or repeat surveys, which have been shown to improve response rates.14 It is also possible that the low response rate reflects limited engagement in research among paramedics and thus low saliency of the questionnaire subject. Despite efforts to make the questionnaire approachable, understandable and of a manageable length, it may have been found irrelevant, too long or too complex, reducing the return rate.15
Despite these limitations a number of conclusions can be drawn from the responses. Randomisation appears to be a significant challenge and some commonly used methods were considered impractical. Time delays incurred by trial protocols were perceived to be a major operational barrier. Many paramedics perceived a lack of support for involvement in clinical trials, in terms or time, funding and career progression opportunities. Despite these problems, it was generally felt that clinical trials and developing the evidence base for clinical practice are important. There was general agreement that being involved in clinical trials is part of a paramedic's role and there was interest in completing additional required training, although this may be over-represented due to responder bias.
We thank Jane Shewan, the research champions and paramedics of Yorkshire Ambulance Service.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online appendix
Contributors KH developed the questionnaire, analysed the data and wrote the content for the paper. SG conceived and designed the study, supervised KH, and wrote the first draft of the paper. PM supervised administration of the questionnaire. All authors contributed to redrafting the paper and approved the final draft.
Funding The project was undertaken as a BMedSci project and supported by the University of Sheffield. All authors had full access to all of the data (including statistical reports and tables) in the study and can take responsibility for the integrity of the data and the accuracy of the data analysis.
Competing interests None.
Ethics approval Granted by the Sheffield Research Ethics Committee.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.