Introduction
The past few decades have seen a surge in the interest in, and utilization of complementary medicine (CM) by the public (Adams et al., Reference Adams, Andrews, Barnes, Magin and Broom2014). CM includes a vast range of self-directed and practitioner-led health practices (eg, meditation, yoga, chiropractic, naturopathy) and products (eg, herbal medicine, nutritional supplements and homeopathy) that are not traditionally considered part of conventional care or the medical curriculum (Adams et al., Reference Adams, Andrews, Barnes, Magin and Broom2014). Large-scale population-based survey findings highlight significant use of CM in Australia and internationally (Harris et al., Reference Harris, Cooper, Relton and Thomas2012; Adams et al., Reference Adams, Sibbritt, Broom, Loxton, Wardle, Pirotta and Lui2013; World Health Organization, 2013; Reid et al., Reference Reid, Steel, Wardle, Trubody and Adams2016). In Australia the role of CM in the contemporary health system is of particular and increasing significance, with CM utilization being among the highest in the developed world (Bodeker and Burford, Reference Bodeker and Burford2007; Reid et al., Reference Reid, Steel, Wardle, Trubody and Adams2016). The estimated numbers of consultations with CM providers in Australia are similar to those of conventional providers (69.2 versus 69.3 million), with reported out-of-pocket expenditure on CM products estimated to be over $4.1 billion annually (US$3.8 billion) (Xue et al., Reference Xue, Zhang, Lin, Da Costa and Story2007). The therapeutic footprint of CM practitioners in Australia is also substantial, with CM practitioners outnumbering conventional primary care providers in some areas (Wardle et al., Reference Wardle, Adams, Magalhaes and Sibbritt2011).
Although general medical practitioners (GPs) often have central primary care gate-keeper roles within health systems, relatively few studies have looked at the practice patterns, attitudes and perceptions of GPs as a specific sample group in relation to CM (Wardle et al., Reference Wardle, Adams, Broom and Sibbritt2013a). High use and support of complementary medicine (CM) by GPs has been observed in Australia (Cohen et al., Reference Cohen, Penman, Pirotta and Da Costa2005; Pirotta et al., Reference Pirotta, Kotsirilos, Brown, Adams, Morgan and Williamson2010; Janamian et al., Reference Janamian, O’rourke, Myers and Eastwood2011) and internationally (Amster et al., Reference Amster, Cogert, Lie and Scherger2000; Corbin et al., Reference Corbin Winslow and Shapiro2002; Hamilton, Reference Hamilton2003; Thomas et al., Reference Thomas, Coleman and Nicholl2003; Cocconi et al., Reference Cocconi, Caminiti, Capriglia, Gennari, Minari, Schianchi and D’aloia2006; Poynton et al., Reference Poynton, Dowell, Dew and Egan2006; Ross et al., Reference Ross, Simpson and McLay2006; Godin et al., Reference Godin, Beaulieu, Touchette, Lambert and Dodin2007; Al Shaar et al., Reference Al Shaar, Ismail, Yousuf and Salama2010). Although these studies appear to show broad support for CM by GPs in most nations – with similar factors underlying this support – it is also important to recognize significant differences across country borders. For example, homeopathy appears to be more popular and publicly accepted in European nations and New Zealand than in other countries, whereas chiropractic and naturopathy seem more popular and publicly accepted in Australia and parts of the United States than in other regions (Bodeker and Burford, Reference Bodeker and Burford2007).
Exploring the practice patterns, attitudes and perceptions of GPs as a specific practitioner group in relation to CM is also particularly important, as studies of Italian and Korean physicians have indicated there do appear to be differences between groups with reference to CM use, attitudes and provision between primary care physicians and specialist or academic physicians (Cocconi et al., Reference Cocconi, Caminiti, Capriglia, Gennari, Minari, Schianchi and D’aloia2006; Kang et al., Reference Kang, Ha, Cho, Lee, Han, Lee and Hahn2011). In a country such as Australia, where GPs as primary care physicians have an official ‘gate-keeper’ capacity to a variety of both medical and non-medical services, the attitudes and perceptions of GPs are likely to have particularly significant practice and policy impact.
However, while general practice is one branch of medicine where CM has long made an impact (Adams and Tovey, Reference Adams and Tovey2000; Wardle et al., Reference Wardle, Adams, Broom and Sibbritt2013a), this group has not gained as much research attention as specialist medical professions on this topic. Data from Australian and international studies demonstrate significant levels of interaction – via utilization and referral – between GPs and CM practitioners (Pirotta et al., Reference Pirotta, Farish, Kotsirilos and Cohen2002; Cohen et al., Reference Cohen, Penman, Pirotta and Da Costa2005; Poynton et al., Reference Poynton, Dowell, Dew and Egan2006; Widmer et al., Reference Widmer, Herren, Donges, Marian and Busato2006; Godin et al., Reference Godin, Beaulieu, Touchette, Lambert and Dodin2007; Stange et al., Reference Stange, Amhof and Moebus2008; Pirotta et al., Reference Pirotta, Kotsirilos, Brown, Adams, Morgan and Williamson2010; Perry et al., Reference Perry, Dowrick and Ernst2014). Given these high levels of interaction, the attitudes, perceptions and knowledge of primary care physicians toward CM deserve careful consideration. To provide additional insights into the attitudes, perceptions and use of CM by GPs practicing in rural and regional areas of Australia’s largest State, the free-text open-ended questions relating to attitudes, perceptions and knowledge of CM in a survey of all GPs practicing in non-metropolitan Divisions of General Practice in New South Wales, Australia were analyzed.
Methods
A questionnaire was mailed to all 1486 GPs registered as practicing in rural and regional General Practice Divisions of NSW, with a reminder card sent after two months. The questionnaire was adapted for rural and regional use from previous Australian surveys of GP attitudes, use and practices of CM (Pirotta et al., Reference Pirotta, Farish, Kotsirilos and Cohen2002; Cohen et al., Reference Cohen, Penman, Pirotta and Da Costa2005). The final survey questionnaire contained 27 items and one free-text qualitative question, where respondents were invited to highlight issues associated with CM in their own words. Results from the quantitative parts of the survey have been published elsewhere (Wardle et al., Reference Wardle, Sibbritt and Adams2013c; Reference Wardle, Sibbritt and Adams2013d; Reference Wardle, Sibbritt and Adams2014). This study reports findings from qualitative analysis of the free-text comments.
Results from free-text responses were transcribed and entered into an Excel spreadsheet. The text responses to the open-ended question were then analyzed qualitatively using thematic analysis, with initial codes then being used to create higher order categories in a cumulative manner (Liamputtong and Ezzy, Reference Liamputtong and Ezzy2005). Analysis was conducted by health services researchers in a public health faculty, with backgrounds in nursing, naturopathic medicine, sociology and public health. To reduce potential influence of pre-existing attitudes or experiences of CM, the method of researcher triangulation was also employed with each member of the research team providing independent analysis of selections of the data and comparing results afterward (Patton, Reference Patton2002). Quotations are used to illustrate the themes generated by this analysis, to demonstrate interpretations made. Participant labels were assigned on the order of data entry of the open-ended variable.
Ethical approval for the study was obtained from the School of Population Health Research Ethics Committee of the School of Population Health, University of Queensland and the Human Research Ethics Committee of the University of Newcastle.
Results
A total of 585 questionnaires were returned completed, with 49 questionnaires returned incomplete as ‘no longer at this address;’ giving an adjusted response rate of 40.7% for the entire survey. A total of 152 (26.0%) respondents had filled in the free-text question. The respondent profiles of both the questionnaire as a whole and the free-text question were similar, with no significant differences in respondent demographics. Respondents had an average age between 45 and 54 years and were 53.5% male. Over three-quarters of respondents (77.8%, n=456) had completed their medical training at an Australian university. The respondent profile was broadly representative of the GP community in the study area, apart from a slight over-representation of women (Australian Medical Workforce Advisory Committee, 2005).
Many of those who responded had definite views about CM, which were thematically coded individually with the results listed in Table 1. The central themes which emerged were risk as a primary concern; opposition, resistance and the inappropriateness of complementary therapies; struggles with complexity and ambivalent tolerance. Sub-categories observed within these themes are also listed in Table 1.
Table 1 Main themes and sub-categories of responses to free-text questions
CM=complementary medicine.
Risk as a primary concern
Risk was a major theme in many of the open-text responses, with participants often framing their personal attitudes and perceptions toward CM in relation to risk management rather than other factors such as patient preference or efficacy. Both direct (eg, drug–herb interaction, adverse events) and indirect risks (eg, delayed diagnosis) were identified by respondents, though indirect risks were generally seen as being more influential in clinical practice. CM practitioner (or product) variability or potential monopolization of care by CM provides was highlighted as a major issue that increased potential risks to patients. For many GPs, it was this risk, rather than risk of ineffective therapies, that was of most concern to them as clinicians, as the following quote demonstrates:
‘They all [treatments] work sometimes – it’s just a matter of using them judiciously. I don’t think [CM] is the problem per se, rather that practitioners and patients tend to think their pet therapies will fix everything and anything’.(GP 152)
Concerns around the risks associated with the regulatory vacuum that often surround CM products and services were raised by many respondents, amongst both supporters and detractors of CM. Most respondents held that the risks of CM were maximized due to the variability and heterogeneity in standards, practices and treatments, which in their opinion was related in a large part to this regulatory vacuum.
The risks from integrating or utilizing CM were not seen to be limited to patients, but also to the GPs themselves. For some GPs, resistance to further integration of CM was not necessarily due to their own perceptions on whether integration was appropriate based on patient risk, but rather perceived as a product of external factors, particularly legislative and regulatory concerns that were risky to GPs themselves. As the following response indicates, this could lead to GPs resisting referral, even in instances where they would otherwise be supportive:
‘Fear of litigation over risks associated with cervical manipulation stop me referring to chiropractors or osteopaths on paper. I do believe they have great success with certain conditions’. (GP 80)
Opposition, resistance and the inappropriateness of complementary therapies
Some respondents relayed their concerns about CM purely in evidence-based terms, stating that the only issue they had with most CM was that it did not abide by the principles of evidence-based medicine, and as such, was by default not appropriate for practice or integration into medical practice. The following quote illustrates the attitude presented by many respondents:
‘There is only evidence-based medicine and non-evidence-based medicine. If you can show me the evidence for it I would even start using crystal therapy tomorrow’. (GP 110)
For others, the opposition to CM ran deeper, indicating that even if some CM disciplines were to adopt evidence-based principles of practice, the historical roots of such CM in pseudoscience and unorthodox philosophy should disqualify them from further integration, as demonstrated by the following illustrative quote:
‘Homoeopathy is bunk. Acupuncture is grossly overrated, and is not popular in China. TCM [Traditional Chinese Medicine] is only used by the poor in China. Nutriceuticals [sic] is a scam. Chiropractic is trying to position itself as more scientific but historically it’s based on pseudoscience’. (GP 78)
Respondent opposition to CM sometimes resulted in refusal to discuss CM with patients. Some respondents suggested that discussion of CM in a conventional medical consultation may give CM an undeserved imprimatur. For other GPs, their opposition to CM simply meant that CM was a topic to be ‘kept off the table,’ as illustrated by the following quote:
‘I find if I don’t ask patients about their [CM] they won’t tell me about it. This suits me fine. I don’t want to know what they’re using as it’s all garbage, and they know I’ll just tell them to stop it anyway’. (GP 11)
For some GPs, opposition to CM was absolute and was not solely directed at those that were trained in CM disciplines. Even conventionally trained practitioners drew the ire of some of their colleagues if they dabbled in CM provision, as the following quote demonstrates:
‘Unfortunately CM practitioners aren’t the only charlatans out here. I know of many doctors who practice questionable therapies and I’m never ceased to be amazed by the rubbish that pharmacists have the gall to put on their shelves. It’s the opium for the masses’. (GP 117)
Some stated opposition appeared to be associated not with open resistance to CM, but frustrations among GPs that many CM practitioners were free of the accountability and scrutiny that occurred in general practice. It was felt by many participants that CM therapists practiced in a regulatory vacuum (and were therefore able to practice without restrictions on autonomy), and had a patient-base that was more forgiving and less critical of their practice than was observed in general practice, as evidenced by the following quote:
‘Locally, patients expect orthodox medicine for free but pay for quackery and unproven therapy. They don’t question ridiculous claims that are unproven. Yet question much of what I and other orthodox medicine offer’. (GP 82)
Struggles with complexity
For many GPs, CM use among patients was acknowledged as being a practical modern reality, irrespective of whether they personally perceived merit in the individual CM practiced. CM was perceived as being both too prevalent to ignore, but also too large a field to fully embrace or evaluate. As such, CM use was considered to be a matter for patients and CM practitioners, rather than the respondents themselves:
‘I assume all my patients are using [CM] and pragmatically don’t mind as if they get benefit that’s great. However, I’m not interested in learning more as I already have enough trouble keeping up to do date with orthodox medicine’. (GP 22)
For some GPs, the diversity of CM was said to be overwhelming, and in some instances may result in them not using CM, or feeling uneasy about using CM, even in clinical scenarios where they thought it appropriate, as can be observed in the following quote:
‘I know too little to be confident about prescribing or recommending something’. (GP 29)
Part of this complexity was the fact that there was little distinction among respondents of medicines, therapies and practitioners when constructing personal definitions of CM. However, many GPs did note the fraught nature of referring to CM as a homogenous entity:
‘Some [CM] are good, some are not. There’s too much variability to term everything under the one moniker’. (GP 34)
Some respondents highlighted that the complexities of CM not only affected their perception of individual CM therapies, but also changed their support for a specific CM in differing contexts, where even CM that was generally poorly perceived by GPs could be deemed appropriate in specific circumstances:
‘My opinions on what is effective in CM are hard to generalize into categories. I find some herbal remedies effective, many are rubbish. Likewise I disagree with the general principles of homeopathy but have recommended rescue remedy before. I believe increased communication between practitioners of CM and GPs should be encouraged. It seems slow-going’. (GP 69)
Other GPs were concerned about the practical implications of such broad CM categorization. Many explained the inclusion of well-accepted or perceivably feasible CM with fringe modalities as potentially dangerous, and as a way of giving some CM undeserved legitimacy. This interpretation is demonstrated in the GP quote below:
‘Involving all “alternative medicine” under one umbrella such as [CM] is an attempt to give some sham modalities such as homeopathy credibility by attempting comparison to massage/meditation etc’. (GP 83)
Ambivalent tolerance
Few GPs expressed complete support for inclusion, or further integration of CM. Many GPs did offer supportive statements, though these were often couched with caveats or caution. As such, many respondents exhibited a form of ambivalent – sometimes belligerent – tolerance to CM use, whereby the validity of CM use was still questioned, but respondents adopted a more pragmatic view of CM, one which was neither wholly positive nor negative. Pragmatism (influenced by patient beliefs), rather than explicit support for CM often engendered a tolerant or open attitude toward CM for many GPs, as the following quote illustrates:
‘My community is very alternative. It is important to be open and not judgmental to ensure compliance and openness in patients. They will choose the [CM] practitioner over me if there was any hint of judgment or hostility’. (GP 43)
In some instances some GPs rejected the notion that some aspects of CM were in fact ‘alternative’ or ‘complementary’ at all, but a part of medicine and wanted these differentiated from other CM modalities and practices that they deemed more fringe or unorthodox. Nutritional medicine was often an example given, as indicated in the quote below:
‘Nutritional medicine is part of proper medicine, not complementary. Some [CM] therapists go “too far”, even if they do have useful skills. There needs to be adequate regulation as there is a lot of difference between practitioners and groups’. (GP 35)
Other practitioners relayed the potential for CM to be used as placebo in practuce, particularly in the case of ‘heart-sink’ patients who were unable to be helped by more conventional treatments. As the following quote demonstrates, in this sense they often supported CM use in specific circumstances, even when they had less positive views about CM more generally:
‘[CM] can be useful to generate a placebo response that the GP cannot ethically prescribe – e.g. often useful for anxiety/depression of mild variety. But some of the guff that gets referred back “the naturopath wants you to test/treat my yeast infection” etc. are both aggravating and entertaining’. (GP 6)
Discussion
This study provides interesting contextualization of previous descriptive survey findings around medical practitioner utilization and integration of CM practices and offers new insights to complement other work regarding GPs’ perceptions of CM and the interface with primary health care. Such work has focused largely on quantitative reporting of figures support for, referral to and practice of individual CMs (Wardle et al., Reference Wardle, Adams and Sibbritt2013b; Reference Wardle, Sibbritt and Adams2013c; Reference Wardle, Sibbritt and Adams2013d; Reference Wardle, Sibbritt and Adams2014). Findings from this study suggest that there are few default positions of GPs relating to CM. GP perceptions toward CM appear to be accompanied by a number of caveats and qualifying statements, and there is significant diversity of view among GPs on most issues relating to CM. The only group of GPs in this study which appeared to hold definitive views on CM was the minority of GPs who ideologically opposed CM on all levels, who suggested CM was pseudoscientific and incompatible with medical practice. GPs who supported CM, or GPs who stated they could support it in specific contexts, usually only did so conditionally (eg, using it as a benign placebo, rather than being convinced of its effectiveness). For most respondents their attitudes and perceptions of CM in specific circumstances was framed in a discourse of risk, with risk minimization being seen as more important than maximizing benefits of CM use, possibly due to an overarching skepticism of CM efficacy. Efficacy seems to be only one aspect of CM critically evaluated by GPs when gauging their support for individual CM therapies – risk, alignment with medical principles and openness to exploring new avenues of therapies where others have failed, all appear to be equally important considerations. However, most contemporary commentary around the issues surrounding CM and primary practice appears to focus on whether CM is effective enough to integrate into medical practice (Ernst, Reference Ernst2016). Findings from this study suggest that more nuanced discussion relating to the policy and practice interface between CM and primary care is necessary to ensure it adequately reflects the issues experienced at the coalface of primary care practice.
The findings from this study seem broadly consistent with other studies of Australian GPs, which show generally high levels of support for CM, yet also show a significant minority of GPs harboring negative attitudes or heightened skepticism toward CM (Easthope et al., Reference Easthope, Tranter and Gill2000; Hall and Giles-Corti, Reference Hall and Giles-Corti2000; Cohen et al., Reference Cohen, Penman, Pirotta and Da Costa2005). The prominence of risk as a factor in attitudes and perceptions of CM in practice in our study is also supported by analysis of Australian biomedical journal content on CM, which showed that most CM content published in those journals was framed under a discourse of risk (Lewis, Reference Lewis2013). Our work adds clarity and context to these previous studies, by offering insights into the factors that underpin varying levels of support for CM. The qualifications and caveats around issues such as efficacy, for example, the support for CM use by skeptical GPs in ‘heart-sink’ patients where other options have failed, or that efficacy may not be the primary concern in relation to CM, also mirrors earlier international qualitative work of CM in general practice, which uncovers a similarly nuanced approach to CM integration (Adams, Reference Adams2003; Brien et al., Reference Brien, Howells, Leydon and Lewith2008; Jarvis et al., Reference Jarvis, Perry, Smith, Terry and Peters2015). Just as there may be no such thing as a ‘typical’ GP, these findings would also suggest that there appears to be no such thing a ‘typical’ GP view of CM. These findings, and the reasons that underlie such differing views, suggest that information tools and strategies to encourage GP discussion around CM may need to be tailored to a wide variety of views, rather than assuming a ‘one-size fits all’ strategy. This would necessitate a variety of communication and dissemination strategies for GP CM education, rather than the development of resources aimed at a single GP market. Such development should also account for the significant heterogeneity of CM. Although often defined as a discrete and homogenous group, there is significant heterogeneity in CM approaches, which according to participants in our study included therapies identified as ‘fringe’ or ‘pseudoscientific’ as well as therapies and practices that are considered relatively mainstream in Australia (eg, chiropractic is covered by the Australian public health insurer Medicare upon referral by a GP). There may also be significant regional differences in CM utilization and prevalence that influence GP perceptions. However, relatively few respondents in our study acknowledged this heterogeneity, often discussing CM as a homogenous entity.
Non-disclosure of CM use in general practice was identified as a significant issue in our study, which may pose potential direct and indirect risks to patient care (Wardle and Adams, Reference Wardle and Adams2014). An international review of CM disclosure to conventional medical practitioners found that, although risks of non-disclosure are known to be significant, patients are generally hesitant to disclose CM use to medical practitioners, even when prompted, and medical practitioners routinely do not initiate discussion around CM use in their patients (Robinson and McGrail, Reference Robinson and McGrail2004). Our study suggests simply highlighting well-known risks of non-disclosure to those GPs opposing CM appears not to encourage discussion of CM use with patients, even when those GPs may acknowledge the importance of understanding CM use to ensure optimal patient care, and as such offers insights into why addressing CM non-disclosure between GPs and patients remains difficult. Our study suggests CM use by patients may be perceived by some GPs to be an issue best ignored due to it being ‘too complex’ or not being as important as other general practice issues. This suggests that not only does CM information for GPs need to be tailored for a wide variety of views, but it also needs to be presented in an accessible and clinician-friendly manner. Development of tailored and flexible CM information and clinical tools may assist GPs to better navigate the interface between CM and general practice, but current tools do not appear to be fulfilling this role. Additionally, highlighting the ways in which encouraging CM discussion can be important in improving patient care beyond reducing direct risks (eg, using CM discussion to establish trust), may also help to reduce the rate of non-disclosure of CM use in general practice.
The large and varied study area was chosen to be broadly representative of Australian general practice demographics. Nevertheless, the location of the GPs in this study should be considered when generalizing the study’s results to the broader Australian or international general practice population, or to other medical specialties. The non-metropolitan nature of respondents may impact the findings of this study, given that CM use is observed to be highest in rural and regional areas in Australia (Wardle et al., Reference Wardle, Lui and Adams2012). Further research in other settings is warranted, particularly given the continued significance of CM in primary care practice, and the relative paucity of recent studies in this area.
Other limitations of the study include self-selection which may have resulted in a response bias, due to CM being a controversial issue in general practice, and those with particular strong views may have been more likely to participate. This effect may be particularly pronounced in the free-text response section which required additional respondent effort beyond other questions in the survey (Rich et al., Reference Rich, Chojenta and Loxton2013). The response rate is typical for large-scale GP surveys on CM conducted in Australia over the past decade, which have reported response rates of between 29 and 58% (Cohen et al., Reference Cohen, Penman, Pirotta and Da Costa2005; Pirotta et al., Reference Pirotta, Kotsirilos, Brown, Adams, Morgan and Williamson2010; Janamian et al., Reference Janamian, O’rourke, Myers and Eastwood2011), and compares well with general surveys of Australian GPs, which routinely have difficulty receiving response rates of over 30% (Bonevski et al., Reference Bonevski, Magin, Horton, Foster and Girgis2011). While the free-text nature of this survey cannot replicate depth of more-focused qualitative exploration, the broad range of perspectives gathered from this open-ended question nested within a larger quantitative survey represent an important series of insights which may otherwise have not been exposed.
Conclusion
GPs in Australia have a wide variety of perceptions toward CM. Only a subset of those GPs who oppose CM have absolute views toward the topic, with most GPs having numerous caveats and qualifications of individual CM. Efficacy seems to be only one aspect of CM critically evaluated by GPs when they gauge their support for individual therapies – risk, alignment with medical principles, and an openness to exploring new avenues of treatment where others have failed, all appear to be equally important considerations when GPs form their views around CM.
Acknowledgments
The authors would like to acknowledge the Divisions of General Practice for their support in recruitment for this study.
Financial Support
This work was funded by the National Health and Medical Research Council and the University of Queensland Research Excellence Foundation.
Ethical Standards
Ethical approval for the study was obtained from the School of Population Health Research Ethics Committee of the School of Population Health, University of Queensland (JW130508) and the Human Research Ethics Committee of the University of Newcastle (H0080344).
Conflicts of Interest
None.
