Practice recommendations and referrals, perceptions of efficacy and risk, and self-rated knowledge regarding complementary medicine: a survey of Australian psychologists

The aim of this research was to determine how many types of CM products, CM practices, and CM practitioners are recommended and/or referred by Australian psychologists as part of their clinical practice as well as explore the relationship between psychologists’ perspectives about the efficacy, risks and relevance of CM in psychology, and their self-rated knowledge of CM, with the number of types of CM to which psychologists are recommending or referring their clients.

This study employed an online survey administered exclusively to Australian psychologists who were fully registered and working in a clinical practice setting at the time of survey between February and April of 2021. Email invitations to participate in the study were sent to psychologists whose contact details were collected from their publicly available websites. The recruitment emails contained information about the study, consent forms, and a link to complete the survey online. All participants were sent a reminder email four weeks following the initial invitation email. An advertisement inviting psychologists to participate in the research was also placed on two psychology professional association websites (Australian Association of Psychologists Incorporated and the Australian Psychological Society) and on relevant social media sites including Twitter, LinkedIn and Facebook. Participants accessed the survey via an anonymous link (embedded in the email, website advert or in social media post), which initially directed the psychologist to the participant information and consent form followed by the link to the open survey using Qualtrics software, Version 2021 [42]. The information page at the beginning of the survey included project details such as ethical approval, data protection, and voluntary participation. The information page also served as the participant consent form. Participants indicated their written consent after reading the information and consent page and clicking on the button confirming their agreement to proceed with the survey. Participants who completed the online survey were invited to supply their email address to enter a prize draw to win a gift voucher to the value of $250. Ethical approval was attained from the University of Technology Sydney Human Research Ethics Committee [ETH20-5138].

The survey was distributed to 1,479 Australian psychologists working in clinical practice at the time of recruitment. All psychologists (psychologists with general registration, and those psychologists with general registration plus an area of practice endorsement (AoPE)) were eligible to participate in the study. However, only responses from those psychologists working in an Australian clinical practice setting (e.g., inpatient hospital, private practice) and directly with clients were included in the data analysis for this study.

All psychologists in Australia (n = 34,872) are considered to hold general registration, which enables them to use the title of psychologist (Psychology Board of Australia, 2022). Some psychologists with additional tertiary training in psychology may also hold an AoPE enabling them to use a restricted title (e.g., clinical psychologist). These AoPE titles are clinical neuropsychologist, clinical psychologist, community psychologist, counselling psychologist, educational and developmental psychologist, forensic psychologist, health psychologist, organisational psychologist, and sport and exercise psychologist. A psychologist with an AoPE title has general registration plus an AoPE. To clarify, a psychologist with general registration without an AoPE (i.e., psychologist) can work in clinical practice settings. However, working in clinical practice does not mean a psychologist is a clinical psychologist.

Psychologists self-selected to participate by clicking on the survey link in the email invitation or on the websites listed above. An initial screening question asked participants if they were a psychologist who undertakes the work of a psychologist. Participants who selected “No” were redirected out of the survey. The current study focused exclusively on registered psychologists in clinical practice in Australia. The original sample size was planned to be 400, based achieving a 0.10 confidence interval width on estimates of prevalence of binary questionnaire items. As noted above, we were able to recruit 231 participants, of which 201 passed the inclusion criteria and were used in this study. With this sample size we are able to estimate a confidence interval for the prevalence of a single binary item with a CI width of 0.14; or compare a continuous or binary variable between two equally sized groups with 0.8 power and an effect size of 0.39 (Cohen’s d).

The construction of survey items was informed by previous literature on psychologist engagement with CM to produce survey items that best captured the ways psychologists might be engaging with CM, including the types of CM products and practices they had ever recommended and the types of CM practitioners to which they had ever referred. Participants were also provided with a definition of CM in line with the World Health Organisation [55]. The 79-item questionnaire aimed to examine psychologists’ perspectives relating to how CM is/is not relevant and/or appropriate to (their) psychology practice, their clients, and the treatment of mental health problems. Prior to recruitment, the survey was tested for face validity and functionality by three PhD students from psychology adjacent fields. As part of this process, any identified necessary changes were undertaken to provide clearer definitions and reduce repetitive questions. Based on survey testing feedback, the time required to complete the survey was approximately 15 min. Where relevant this paper adhered to the CHERRIES checklist for reporting results of internet e-surveys [15, 16]. The study reported in this paper focused on the survey data sections examining psychologist demographics, types of CM products and/or practices recommended, and/or CM practitioners referred to, psychologists’ perspectives on CM within psychology, as well as psychologist self-rated knowledge of CM types.

Demographic data included year of birth, gender identity, and the predominant state/territory in which they practice. Psychologists were also asked to provide practice characteristics, such as AoPE, their work setting (solo or group setting), and years in practice as a psychologist.

This section asked psychologists to self-rate their knowledge across eleven types of CM: Aboriginal and Torres Strait Islander Traditional Medicine /Healing practices, acupuncture, dietary intervention, exercise/movement interventions, herbal medicine, hypnotherapy, massage, meditation, nutrition supplements, probiotic supplements, and yoga. Psychologists could rate their knowledge on a four-point Likert scale from excellent, good, fair, and poor.

IBM SPSS Statistics Premium Edition Version 27 (Armonk, New York, IBM Corp) was used to analyse the data. Prior to conducting the analyses, raw survey data were screened for any missing or incomplete responses. During this process, nine cases were removed as the data (responses) were incomplete. After removal of the nine cases, 222 cases were included in the initial analyses which identified significant outliers. On review the outlier responses were mostly from cases who did not work in clinical practice settings. These cases were removed resulting in 201 participants in the final data set. Descriptive statistics were used to determine the percentages and frequencies. A Poisson regression model was used to estimate rate ratios between demographic and practice covariates and the outcomes of CM engagement l (number of types of recommending CM products and practices and referring to CM practitioner types, agreement with statements about risks and relevance of CM, self-rated knowledge of types of CM).

Variables and categories were created to best capture psychologist responses to survey items. For analysis the types of CM products and practices were categorised into six groups that were informed by previous literature (Ng et al., 2016; Wieland et al., 2011) and consistent with the definition of CM used in this study: mind/body approaches (hypnotherapy, meditation, yoga), movement approaches (exercise and movement-based activities, such as walking), ingestive therapies (herbal medicine, probiotics, vitamin and nutrition supplements), dietary changes and manual approaches (acupuncture and massage). The sixth category cultural and spiritual approaches included participant’s free text responses indicating recommending or referring to music, creative arts, prayer and spirituality, and Aboriginal and Torres Strait Islander traditional healing. The number of types of CM products and/or practices psychologists recommend to their client was collapsed into three categories: recommended none; recommended one to three types of CM products and/or practices; and recommended four plus types of CM products and/or practices.

Psychologists could also select from six categories of CM practitioners to which they have ever referred their clients. Practitioner types included mind/body practitioners (e.g., hypnotherapists and yoga teachers), movement practitioners (e.g., exercise and movement trainers and/or coaches), practitioners who predominantly prescribe ingestibles category (e.g., naturopaths, herbalists, and traditional Chinese practitioners), prescribes nutrition (e.g., nutritionists), manual practitioners (e.g., acupuncturists, massage therapists), and cultural/spiritual practitioners (e.g., Aboriginal and Torres Strait Islander traditional healers). The number of types of CM practitioner to which a psychologist referred their clients was combined into three categories: refers to none, refers to one to three types of CM practitioner, and referred to four plus types of CM practitioner.

Variables relating to perspectives about the efficacy and relevance of CM, and self-rated knowledge of CM, were also adjusted for the purpose of analysis. Due to the small number of responses in some categories the statement responses were collapsed into fewer categories. Both categories relating to agreement with perspectives about CM (efficacy and relevance) were collapsed into two responses of either agree or disagree. Finally, the self-rated knowledge category was also collapsed into two responses of either excellent/good or fair/poor.

Count variables were also created for the Poisson regression analyses. A variable representing the total number of types of CM products and practices that psychologists would refer patients to was created, and similarly one for the number of types of CM practitioners that psychologists refer to as part of their clinical practice. For the additional summary data tables, the count variables were converted to ordinal variables (for grouping into three categories) for both recommend or referred; none, one to three types, and four or more types of CM.

The study sample (N = 202) was comprised of 165 women (81.6%), 36 men (17.8%) and one person who identified as other (0.5%). The mean age of psychologist participants was 48 years (M = 48.260, SD = 26.53). All of the Australian states and territories were represented within the sample, with highest representation from New South Wales (n = 65) and Queensland (n = 64), and the lowest from Northern Territory (n = 1). Most psychologists in the study identified as having the AoPE as a clinical psychologist (n = 79). Psychologists with general registration (n = 76) were also represented. There were also psychologists who identified as Other AoPE including counselling psychologists (n = 25), forensic psychologists (n = 8), health psychologists (n = 7), educational and developmental psychologist (n = 6) and one community psychologist. Solo private practice was the most common work setting reported among participants (n = 137). The highest proportion of participants had 11 to 20 years of clinical experience (n = 72) and the lowest proportion had 31 plus years of clinical experience (n = 31). Table 1 provides a summary of demographic and professional characteristics.

Of the 201 psychologist responses included in the analysis, 5% (n = 11) reported having not recommended any type of CM, 32% (n = 65) reported having recommended one to three types, and 63% (n = 126) reported having recommended four or more types of CM products and/or practices (see Appendix Table 2). The rate ratios of psychologist response to statements about efficacy, risk, relevance and knowledge of CM with recommending and referring to a number of types of CM are reported in Table 3. The lowest rates of CM product and practice recommendations were from psychologists who agreed “CM is not scientifically valid” (RR = 0.77 [0.65; 0.93]), and that “referring to CM practitioners puts client safety at risk” (RR = 0.76 [0.63; 0.91. Of the 23% (n = 47) of participating psychologists who agreed with the statement that “CM as not scientifically valid”, some (n = 6) did not recommend any types of CM. However, most of those psychologists who agreed with this statement still recommended CM to their clients (n = 41).

In contrast, psychologists who agreed with statements describing CM training and integration within psychology as beneficial, had the highest rates of recommending CM. For example, psychologists who agreed “CM practitioners can play a valuable role in assisting clients with their mental health problems” were more likely to recommend CM (RR = 1.55 [1.24; 1.94]).

The following CM types had the highest proportion of self-ratings as excellent/good among participating psychologists; meditation 88% (n = 178), dietary interventions 65% (n = 131), yoga 64% (n = 129), and exercise/movement interventions 54% (n = 110). With regards to the remaining seven CM types, more than half of the psychologists self-rated their knowledge as fair/poor for each type. Aboriginal and Torres Strait Islander Traditional Medicine /Healing practices attracted the lowest self-rated knowledge amongst the psychologists with only 6% (n = 12) of participants rating their knowledge as excellent/good for this medicine/healing practices. Psychologists who self-rated their knowledge of meditation as excellent/good had the highest rate of recommending multiple types of CM (RR = 1.51 [1.15; 1.96]).

Of the participants, 25% (n = 50) had not referred to any type of CM practitioner, 42% (n = 84) had referred to one to three types of CM practitioner, and 33% (n = 68) had referred to four or more types of CM practitioner. The lowest reported rates of referral to CM practitioners came from psychologists who agreed with the statement “referring clients to CM practitioners or services puts client safety at risk” (RR = 0.39 [0.29; 0.52). Psychologists who agreed with the statement “CM is not a good match with psychology” also reported low rates of referral to CM practitioners (RR = 0.48 [0.33; 0.73]). Of those psychologists who agreed with the statement “referring clients to CM practitioners or services puts client safety at risk” (n = 48), half (n = 24) reported referring to one or more types of CM practitioner. Meanwhile, another 10% (n = 21) of the study sample who also agreed with the statement “CM is not a good match with psychology”, half (n = 11) still reported referring to one or more CM practitioners. Across the demographic and practice characteristics psychologists aged 51 to 65, and those with 31 years plus experience, were more likely to refer to CM practitioners (Table 2). Clinical psychologists were the least likely to refer to any CM practitioner types. (See Appendix Table 1).

Psychologists who agreed with the statement “it is important for psychologists to understand and engage with their client’s preference for CM as part of their mental health treatment” had the highest reported rate of CM referral (RR = 4.08 [2.03; 8.21]). Similarly, psychologists who agreed with the statement “there is potential to improve mental health outcomes with the integration of evidence-based CM within psychology practice” also reported high rates of referral to CM practitioners (RR = 4.06 [2.17; 7.60). Similarly, psychologists who self-rated their knowledge of dietary intervention as excellent/good reported the highest rate of referring to multiple types of CM practitioner (RR = 1.77 [1.43; 2.19]).

This study is the first to focus on the rates of recommending and referring to types of CM amongst a diverse range of psychologists in clinical practice (e.g., in terms of years of experience, AoPE) in the context of psychologist perceptions of the risks and relevance of CM, and their self-rated knowledge of some types of CM. Although the number of participants in the study was small, it is representative of the Australian psychology workforce according to current workforce demographics provided by the Psychology Board of Australia (2022). Although the study participants were from some AoPE, not all psychology AoPEs were represented. Further, when interpreting the results, it is important to be mindful that most participants were either psychologists with general registration or psychologists with an AoPE in clinical psychology. We suggest caution when interpreting the results as there is potential for bias in our research due to participants being self-selecting. Those who have strong opinions on the relationship between psychology and CM are potentially more likely to respond.