Patient choice and evidence based decisions: The case of complementary therapies

Abstract

Objective  Current government policies simultaneously pursue the development of ‘patient‐led’ and ‘evidence‐based’ approaches to healthcare. The objective of this study was to explore how primary care clinicians and Primary Care Trust (PCT) managers balance these potentially competing tensions when considering popular, controversial treatments, like complementary therapies, in consultations (clinicians) or funding decisions (PCT managers).

Setting and participants  We selected two case sites where complementary therapies were offered on NHS premises in England. We interviewed 18 PCT managers and clinicians, conducted an observation of a PCT meeting on complementary therapies and collected documentary data from referral databases and service funding bids. All interviews were taped, transcribed and analysed thematically. Interview, observation and documentary data were used to compare reported beliefs and behaviour to observed and documented behaviour.

Results  The majority of clinicians and PCT managers claimed that research evidence guided their decisions; those who did not felt increasingly marginalized. However, discrepancies between reported and observed behaviour suggest that perceptions of research evidence, rather than fact based knowledge, predominated when considering complementary therapies.

Conclusion  In the case of NHS complementary therapy service provision, patient preference may be largely insignificant in clinician and PCT managerial decisions, with decisions based mainly on ‘evidence rhetoric’ devised from collectively agreed, unchallenged, tacit perceptions of research literature. If a patient‐led NHS is to become a reality, NHS professionals need to cede the power that they wield with evidence rhetoric and acknowledge the legitimacy of patient preferences, views and alternative sources of evidence.

Introduction

Since the publication of the NHS Plan, government policy has increasingly emphasized the importance of prioritizing patients, with the goal of moving ‘from a service that does things to and for its patients to one which is patient led’. ¹ Although the term ‘patient led’ tends to dominate the policy discourse, within the clinical sphere, the term ‘patient‐centred’ is often used. Stewart defines ‘patient‐centred’ clinical care as fully informed patients, aware of their options, who agree a mutually acceptable management plan with their clinicians. ² Thus far, mechanisms to deliver patient‐led services focus principally on patients choosing a place of treatment, with initiatives such as ‘Choose and Book’ (see http://www.chooseandbook.nhs.uk), rather than opting for particular services or treatments themselves. However, critics of patient‐centredness have pointed out that this approach requires longer consultation times to discuss options ³ and may not improve health outcomes. ⁴ Nonetheless, it is a rising government mandate.

In tandem, evidence‐based medicine has also been enthusiastically promoted within the NHS. Originally, Sackett et al. proposed that research evidence, clinical judgment and patient preference all have a role in clinical decisions, but increasingly research evidence has gained prominence. ⁵ Evidence based policy making has also been introduced, defined as ‘set of rules and institutional arrangements designed to encourage transparent and balanced use of evidence in public policy making’. ⁶ Nonetheless, studies suggest that despite the quality of evidence or acceptability of the treatment, clinicians and Primary Care Trust (PCT) managers can be influenced by factors other than research evidence, such as respected colleagues and personal experiences. ⁷ , ⁸ , ⁹ Furthermore, some have suggested that NHS professionals use research evidence selectively, ¹⁰ often to legitimize clinical or funding decisions that have already been made. ¹¹

In characterizing the two models of evidence based and patient‐centred decision‐making, Bensing argues that

[Evidence based decision‐making] has basically a positivistic, biomedical perspective…which [considers] medicine merely as a cognitive‐rational enterprise…[while patient‐centred medicine] has basically a humanistic, biopsychosocial perspective… with medical care attuned to the patients’ needs and preferences. ¹²

Although Bensing polarizes the two models, she and others have attempted to find ways of integrating evidence based and patient‐centred approaches, mainly in the arena of patient consultations. One way advocated is evidence‐based patient choice consultations, in which patients select from a series of options with ‘robust’ evidence. ³ , ¹³ Other proposed models of clinical decision‐making, such as ‘shared decision‐making’ and ‘informed decision‐making’, give greater or lesser autonomy to patients in relation to research evidence. ¹⁴ Moving outside consultations, Rogers argues that patients’ views could be incorporated much earlier into the research process by engaging patients in the ‘entire process of evidence production, from commissioning appropriate research to developing guideline recommendations’. ¹⁵

However, the extent to which patients want to exercise choice is not clear. Patients may appreciate ‘having choices’, such as knowing about all the treatment options open to them, but be less enthusiastic about ‘making choices’, when deciding which particular option to pursue. ¹⁶ , ¹⁷ In making choices, where patients’ treatment preferences differ with those of their clinicians, one study found that, depending on the issue, GPs adopted a range of attitudes from controlling, where they advised the patient to comply with their clinical opinions, to respectful, where the patients’ wishes were not challenged. ¹⁸ Another study, which explored healthcare professionals’ behaviour in the field of heart failure, found that randomized controlled trials (RCT) fostered a dangerous certainty amongst clinicians, which led to doctors sidestepping, dismissing or not eliciting patient concerns. ¹⁹

Our intention was to explore what occurs when patient preferences and research evidence are almost inevitably in conflict. We decided to solicit the views of PCT managers as well as clinicians, as PCT managers have increasingly powerful roles in making decisions about the provision of services in the NHS. ²⁰ We chose to investigate the area of complementary therapies, as these treatments are popular with patients, but have disputed evidence bases.

Ninety percent of complementary therapy treatments are delivered privately, ²¹ although five NHS Homeopathic Hospitals have been NHS funded since 1948. About 75% of the adult population in England and Wales would like complementary therapies to be more widely available on the NHS. ²² Further provision of complementary therapies is hotly contested, usually because of a perceived lack of research evidence. ²³ Furthermore, the Department of Health, in responding to the House of Lord’s Select Committee on Complementary and Alternative Medicine, set ‘robust’ evidence of clinical effectiveness as a precursor to provision in the NHS. ²⁴ The actual quality and strength of research evidence varies tremendously; for example, the evidence for herbal medicine is generally considered the strongest, ²⁵ although it is one of the least available on the NHS. ²¹ But despite a perceived lack of convincing evidence, complementary therapies are used by nearly half of the adult population in England and Wales in their lifetime and about ten percent of the adult population visit a practitioner annually. ²⁶

Aim of this study

Given the popularity of complementary therapies and the debates about their clinical effectiveness and provision within the NHS, our objective was to explore how clinicians and PCT managers balance the tensions of being ‘patient‐led’ and ‘evidence‐based’ when considering complementary therapies. For clinicians, this scenario is most likely to occur during patient consultations, especially when referral to a NHS based complementary therapy service is possible. PCT managers may encounter this conflict when considering either the continued funding of current complementary therapy services or extending NHS funding to complementary therapy services that have been previously funded by other sources.

Methods

Design

We chose a case study methodology, whereby data were collected and compared across multiple sources, ²⁷ which was appropriate given our interest in comparing the reported beliefs and behaviour of PCT managers and clinicians with their observed and documented behaviour. The analysis was underpinned by a position in which ‘any given reality can be represented from a range of different perspectives (and) each of these representations may be treated as true’. ²⁸ We obtained ethical approval from the London Multi‐Centre Research Ethics Committee prior to fieldwork from July 2004 to June 2006. This study was part of a larger study exploring the changes necessary for complementary therapies to become mainstreamed in NHS primary care; some findings have been published elsewhere. ²⁹ , ³⁰ , ³¹

Settings

We chose two sites where complementary therapy services were available on NHS premises. Site selection was based on funding source (e.g. NHS or other), the degree to which the service appeared mainstreamed (or incorporated) within the NHS and willingness to take part (two sites declined). We wanted to include both NHS and non‐NHS funded sites, as we believed there might be variations in the extent to which the service was accepted, and consequently differences in the role of patient preferences and research evidence in decisions about referrals and funding.

The first site was supported by funding from the Office of the Deputy Prime Minister (non‐NHS), as part of a national community regeneration initiative for economically challenged neighbourhoods known as New Deal for Communities. Fifteen complementary therapists, without medical backgrounds, worked without NHS contracts to deliver therapies as diverse as osteopathy, massage, shiatsu and nutritional therapy. Line management responsibility lay with a community organization outside the NHS, but treatments were delivered in two GP surgeries in a deprived inner city area in England. There were no referral criteria; referral was not condition specific and any NHS professional could refer. Self‐referrals were also accepted. Because this service was based in an economically poorer neighbourhood, service users tended to be unemployed or low wage earners. After nearly 4 years of non‐NHS funding, in 2006 the local PCT (NHS) assumed funding for a re‐modelled service offering osteopathy, chiropracty and physiotherapy for musculoskeletal conditions only. Total session entitlement was reduced from eight to two and only NHS professionals could refer.

The second site had been funded and line managed by the NHS since 1998 and was located in a former community trust in an inner city in England. The complementary therapy service was part of a women’s health service for menopause and premenstrual syndrome, and provided homeopathy, reflexology and aromatherapy. It was delivered by two medically trained professionals (a doctor and a nurse) and one professional complementary therapist. All therapists had NHS contracts. Self‐referrals or referrals from any NHS clinician across the city were accepted and consequently service users came from a range of socio‐economic backgrounds. To access the women’s health service, service users needed to meet a series of criteria based on their symptoms (e.g. three or more hot flushes a day). After referral, patients attended an educational session run by the nurses to learn about the various options available to them, including self‐care, the complementary therapy service and pharmacological treatments. If service users opted for conventional or complementary practitioner treatment, they were referred to a team of doctors, who then reviewed their case in a 30 minute consultation. Only these doctors could then refer on to the complementary therapy service. The referral criteria were that any woman who did not want or could not have pharmacological treatment could be referred to the complementary therapy service. Shortly after beginning fieldwork for this study, the reflexology/aromatherapy service was discontinued, as the therapist retired, and so fieldwork focused primarily on the homeopathy service. NHS funding for the homeopathy service ceased in the summer of 2006, just months after completing fieldwork.

Interviews

From the two sites, we purposively chose a range of participants for interviews using maximum variation techniques. Criteria included:

• • professional background (doctor, nurse or PCT manager)
• • current or past key role in developing, maintaining, referring to or funding the complementary therapy service (PCT managers and clinicians)
• • frequency of referral to the complementary therapy service (high or low for clinicians only)

All those approached at both sites agreed to be interviewed. In total, we interviewed 18 NHS professionals across the two sites, although no PCT managers were interviewed at the second site as clinicians feared this would raise the profile of the service and jeopardize its funding. We obtained informed written consent from all study participants and they were assured of anonymity and confidentiality.

We devised a semi‐structured topic guide, which included questions on attitudes towards complementary therapies, role of patient choice and research evidence in referral and funding deliberations and perceived referral behaviour (clinicians only). Interviews lasted between 15 and 75 minutes and were audio‐taped and transcribed verbatim. Eleven interviews were face to face and seven were by telephone. All of the telephone interviews took place with study participants from site 1, but they did not vary substantially in length from face to face interviews.

We used a constant comparative method to analyse interview data whereby data collection and analysis occur concurrently, with each informing the other. ³² Aided by Atlas‐ti software, we coded the interview transcripts using codes anticipated from previous literature and more emergent codes that arose from the data. We coded transcripts in batches following completion of a series of interviews (e.g. PCT managers at site 1, doctors at site 2 etc.). The coding framework developed as the interviews and analysis progressed, with codes being modified or merged, as appropriate, to account for new data. We re‐coded all transcripts using the final coding framework, once fieldwork ended, to ensure more recently developed codes were universally applied. In addition to coding within Atlas‐ti, we summarized each transcript into a single document of four to six pages noting significant points and quotations to better understand the overall ‘story’ of each interview. From these two processes (Atlas‐ti coding and transcript summaries), we developed key themes from the interviews.

Observation and documentary data

To elaborate our understanding of these themes, and compare reported behaviour with observed behaviour, we also observed a PCT meeting on complementary therapy service provision and collected and analysed documentary data from each site.

We observed a meeting of the Professional Executive Committee (PEC) of the PCT at the first site in April 2005. PEC members present included local nurse and GP representatives, all Directors of the PCT and the Chief Executive. The remit of this meeting was to review local NHS complementary therapy service provision. LW was a non‐participant observer at this meeting. Notes were taken, typed up in full, read and data relating to patient choice and evidence were highlighted.

Documentary data from case sites included: funding bids to secure funding for the service, e‐mails from PCT managers and referral databases of the services. We read funding bids and e‐mails and extracted relevant details regarding funding decisions.

In terms of referral data, the first site held a computerized database that generated data on referrals for over 5 years. Details included name of referrer, role (e.g. doctor, nurse etc.), therapy referred to and number of patients referred. At the second site, data were recorded manually and we collected details on name of referrer, therapy and month for nearly 2 years of referrals. To analyse both of these datasets, we calculated the frequency of referrals and type of therapy per referrer. For results of the analysis of these databases, see Box 1.

Box 1

At site 1, we found that from June 2001 to November 2006, 1250 referrals were made by 54 NHS professionals, 59% (n = 24) of whom were GPs and 7% (n = 30) were nurses. The remaining referrals were made by addiction counselors (8%) or were self‐referrals (26%). Twelve of the GPs were responsible for between 1–28% of the referrals, while 12 other GPs made six referrals or less and were responsible for <1% of the total number of referrals to the service. At site 2, findings from the referral database indicated that during the 22 month period under analysis, a total of 178 referrals were made from three referring doctors, who made 23%, 30% and 47% of the referrals respectively.

Comparison of data from all three sources

As the analysis progressed, we drew together and compared data from all three sources (interviews, observation and documentation). For example, we compared referral patterns reported in interviews to data from referral databases. Throughout the iterative data collection and analysis process, we recorded questions and analytical hypotheses in memos. LW also kept a journal to develop emerging ideas and reflect on the impact of the researchers on data collection and interpretation. We further interrogated emergent findings from interview, observation and documentary data by searching for disconfirming data and drawing mind maps. In the final stage of analysis, we re‐read the original data sources of interview transcripts, observation fieldnotes and documents to check for any data that were overlooked and to challenge interpretations.

Results

In interviews, all PCT managers and clinicians reported a belief that there was insufficient research evidence to support the clinical and cost effectiveness of complementary therapies. However, some of the clinicians said that they largely prioritized patient preferences over research evidence when considering referrals to complementary therapies.

Stated tendency to prioritize patient preferences

The most radical amongst this group were some doctors from the second site who described themselves as less ‘scientific’ and more ‘patient‐oriented’ than other doctors; they also commented on the limitations of evidence‐based medicine. Although aware of research evidence, they reported that they preferred to base decisions principally around patient preferences. Although certain that this was for the best, one said that evidence was being ‘forced’ onto her clinical practice (Site 2, doctor, RT) and another commented that prioritizing patient preferences ‘doesn’t go down particularly well these days’ (Site 2, doctor, WB). This same doctor recounted that she had to resign from a commissioning position at the local PCT because

I always put patients’ care first and if you’re in commissioning in the PCT level, you couldn’t [do that].’ (Site 2 doctor, WB)

The service model at the second site embedded these beliefs about patient‐led services, with its emphasis on self‐care, patient education and providing a broad range of complementary and conventional treatments. In addition, the patient‐centredness of this service was reflected in the discourse of the health professionals. All of the study participants interviewed from site 2 used the term ‘patient choice’ frequently during interviews. For example, as one doctor explained

So the first time I am counseling (sic) them about the homeopathy and the [pharmacological options] and the lifestyle changes, the dietary changes and then it’s their choice. It’s patient choice mainly. So after counseling, they can choose what they want. (Site 2, doctor, TX)

Although at the second site this patient‐centred style of consulting was reported to be universal in interviews, one doctor stated that the ‘vast majority of my prescribing would be done as evidence‐based’ (Site 2, doctor, KP). When asked how she could reconcile her belief in the lack of evidence for homeopathy and reflexology with her statement that her decisions were largely evidence based, she replied

I think you need to be very open to other options in terms of patient care and I think in many ways I probably do take a holistic approach to managing patient care. (Site 2, doctor, KP)

Thus, this particular doctor appeared to be able to integrate the demands of evidence based practice and patient‐centred care within her clinical practice, by acknowledging the limits of conventional care and pragmatically suggesting other options under the banner of ‘holistic’ care.

Although less overt, nurses from both sites also stated that they believed patient preferences should be given greater consideration with respect to referral to complementary therapy services.

I sometimes feel that we kind of go too far into sort of science and evidence based. I mean if it works and people benefit and they feel better, what more proof do we need? (Site 1, nurse, CM)

Nonetheless, unlike the doctors from the second site who referred to the complementary therapy service regularly, the referral database from the first site showed that none of the three nurses interviewed in this study had referred. One nurse at site 1, who was positive about the potential benefits of complementary therapies, recounted that when patients asked about complementary therapy treatments, she cautioned

Don’t use my name in terms of recommending it, because I don’t know. And we’re all very evidence‐based now, aren’t we? So we have to kind of back things up. (Site 1, nurse, PN)

Thus, nurses at the first site reported similar views to the doctors at the second site about the importance of patient preferences when considering complementary therapies, but articulating such a position appeared almost subversive. Moreover, they did not refer. Perhaps, this was because they perceived complementary therapies as lacking in evidence and the culture of their healthcare environment prioritized research evidence, at least rhetorically. This could also be because nurses reported some confusion around referral policies. Doctors were the primary focus of those marketing the service, so some nurses were not clear whether they could refer. Further more, historically nurses tend not to have responsibilities for making referrals.

Stated tendency to prioritize research evidence

Other study participants, mainly doctors from the first site and PCT managers, tended to define evidence as randomized controlled trials not ‘anecdotes or prescriptions (sic) of people’s experiences’ (Site 1, PCT manager and doctor, LP). Furthermore, in addition to data on therapeutic effectiveness, PCT managers also said they required service level evaluations that demonstrated the impact of complementary therapy services on NHS costs such as prescription and consultation rates. ³⁰ The perceived absence of either RCT or service level evidence was often cited as a barrier to referral to or funding of complementary therapy services. But PCT managers, in particular, conceded that the patient choice agenda sometimes made this position awkward.

The work I do is finding out what people in local communities want, what they think will improve their health, and whenever we do consultations around what people would like provided, complementary therapies almost always comes up as something people would like to have access to. And the health service and PCTs are supposed to be listening to what communities want and responding to that. (Site 1, PCT manager, YC)

However, a PCT manager predicted that this tension would be resolved in the future as the patient choice agenda, along with increased choice of service providers from amongst the private sector, would act as a potentially powerful lever for the inclusion of complementary therapy services.

My view is that the NHS is gearing up for full scale privatisation by the year 2015 or something like that. The whole idea about ‘Choose and Book’ (where patients choose from a range of providers) and ‘Practice Based Commissioning’ (where local primary care professionals such as GPs and nurses are devolved budgets to set up services) and bringing the whole budgeting process closer to the patient will ultimately lead to patient held benefit systems which will then be topped up by the private sector. And at that point people will say ‐ my friend went to a homeopath and she was much better and will say to the doctor,’ I want to do that on the scheme’. (Site 1, PCT manager, BE)

In the meantime, during fieldwork for this study, we found several discrepancies between reported and observed behaviour regarding the role of research evidence. First, although PCT managers uniformly said in interviews that NHS funding of complementary therapy provision was dependent on convincing RCT evidence, the funding bid for the re‐vamped service at Site 1 contained no mention of any research evidence, RCT or otherwise, yet was still funded by the PCT in 2006. An e‐mail from the PCT manager responsible for the bid confirmed that such evidence was not required. Second, PCT managers stated that they relied on the Directorate of Public Health to supply them with evidence on interventions, including complementary therapies. During the PCT meeting on complementary therapy provision that we observed, the Director of Public Health argued that there was no ‘good’ research evidence for complementary therapies. But the complementary therapy lead from this directorate told us during an informal conversation earlier in the same day that they had not yet searched for evidence on complementary therapies. Third, we were told by clinicians during interviews that research evidence guided the process of selecting the therapies offered at the first site. Yet herbal medicine was rejected, despite its stronger research evidence base, while massage and reflexology, with weaker research evidence, were offered. Fourth, a doctor at the first site, who stated during interview that her practice was guided by research evidence, and she believed there was little for complementary therapies, was shown to be the third highest referrer to the complementary therapy service of the 24 doctors on the referral database. Fifth, despite the rhetoric from all study participants about the importance of evidence, only three of the 18 participants in this study had undertaken a personal search to look up the research on complementary therapies. Thus, although during interviews many clinicians and PCT managers presented themselves as prioritizing scientific evidence, their behaviour suggested that perceptions of the evidence largely guided their decisions and other factors, such as a locally accessible complementary therapy service, may play a larger part.

Discussion

Summary of main findings

Our objective was to explore how clinicians and PCT managers balance the demands of patient choice and research evidence when making decisions about referrals to and funding of complementary therapy services. Although we only drew on two case sites and were unable to interview any PCT managers at the second site, by focusing on two services, we were able to conduct detailed examination of expressed views and recorded behaviour. The findings from this study may be conceptually transferable to other similar settings, but this needs future research.

We found that the doctors at site 2 said that they prioritized patient preferences over scientific evidence and regular referrals to the complementary therapy service suggested that their behaviour was consistent with their discourse. Nurses at both sites also said they prioritized patient preferences, but at site 1, where nurses could refer, the referral database indicated that none of the nurses interviewed had referred. In contrast, doctors at site 1 and PCT managers stated that research evidence was prioritized, however their behaviour contradicted this as uninformed (and sometimes erroneous) perceptions of the evidence appeared to have greater influence. These findings have several ramifications.

Implications

First, those who stated their decisions were guided by patient preferences felt increasingly marginalized, as their practice was not congruent with the reported approach of the wider NHS community. To manage this, two of the doctors, who were situated in powerful roles, fostered a predominantly patient‐centred culture to which recently recruited staff members, with more evidence‐based inclinations, adapted. Unsurprisingly, complementary therapy provision, with its similar patient‐centred values, thrived in such an environment. This suggests that complementary therapists seeking to establish NHS complementary therapy services should identify and cultivate relationships with powerful clinicians within these NHS enclaves. However, in the face of the growing momentum of the evidence based movement, such clinicians may be a dying breed.

Second, the general perception amongst all those interviewed in this study was that positive research evidence for complementary therapies was scarce ‐ a view that appears to be widespread. ³³ Thus, since it appears that many NHS professionals base their decisions on their perceptions, rather than the facts, those seeking to introduce complementary therapy services into the NHS have two options. One, they can either only offer those therapies where the perception is that research evidence is more robust (e.g. osteopathy, chiropractic, acupuncture) ³⁰ or two they need to develop effective strategies for creating an alternative perception that the evidence base for some complementary therapies is reasonably robust. Relying on the production of further clinical trials of effectiveness to change beliefs may not be enough, as it appears positive studies on complementary therapies may rarely penetrate the mainstream press. ³⁴

Third, the PCT managers in this study utilized a strategy of ‘rationing by exclusion’, ³⁵ justifying their decisions with vague references to clinical effectiveness (some of which would be difficult to uphold under scrutiny) to counter local demands for access to complementary therapies. This occurs nationally as well as locally. For example, one of the recommendations of the 2005 Department of Health public consultation on well‐being was provision of complementary therapies (p. 31). ³⁶ But this was not even mentioned in the final White Paper, Our Health, Our Care, Our Say. ³⁷ Thus, patients are not being offered unfettered choice either nationally or, in the case of this study, locally. Instead, a policy of ‘managed’ choice exists in which a menu of options, reportedly based on research evidence, are offered. ³⁸ Given that resources are limited, ‘managed’ choice may be inevitable. Nonetheless, national policymakers and local PCT managers risk creating cynical service users, who believe that the patient choice agenda is a mere tick box exercise, ¹⁶ if they are not explicit that the patient choice agenda has restraints. Moreover, they also need to manage the expectations of potential providers of complementary therapies, and perhaps other possible third sector providers, by being honest about the terms for inclusion on the managed choice menu.

Fourth, by claiming that research evidence steers decisions, clinicians and PCT managers may be at variance with patients, who report drawing on different sources to guide their decisions. For example, in a study of men’s use of complementary therapies for cancer, influential sources included: the personal stories of friends, family and colleagues, the long history and popularity of some of the therapies, the plausibility of the mechanism of action and a belief or trust in individual therapies or their providers. ³⁹ Scientific evidence ranked low for these men, yet it was reported as the guiding criterion for many of the clinicians and PCT managers in the present study. But the discourse of these NHS professionals may be misleading, and their pool of influential sources may be more closely aligned with service users than reported. ⁷ , ⁴⁰ , ⁴¹ For example, in a study that brought together managers from Health and Local Authorities and the voluntary sector to explore and observe their use of research evidence in decisions about service provision, researchers found that the group members relied on anecdotes from friends and relatives to guide their development of services, rather than research evidence. This was the case even when the relevant literature was available during meetings where such decisions were made. ⁴²

Evidence based decision‐making has much to recommend it. As Lambert notes, before 1985 many medical interventions were based on tradition or the preference of the administering practitioner. ⁴³ But what was once put forward to promote patient benefit now appears to be used to thwart those same service users from accessing the treatments they believe can improve their health. In the event, it appears that some of the clinicians and PCT managers in this study, and perhaps more widely, have fallen into a ‘naïve rationalist’ trap. ⁴⁴ They contended that decisions pertaining to complementary therapies were based on the selection, evaluation and implementation of research evidence, when clearly, they were not. Thus, ‘evidence rhetoric’, rather than evidence based medicine or evidence based policy making, appears to hold sway. If a patient‐led NHS is to become a reality, NHS professionals need to cede the power that they wield with evidence rhetoric and acknowledge the legitimacy of patient preferences, views and alternative sources of evidence.