Research methods in complementary and alternative medicine: an integrative review

Journal of Integrative Medicine xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Journal of Integrative Medicine journal homepage: www.jcimjournal.com/jim

Review

Research methods in complementary and alternative medicine: an integrative review Fabiana de Almeida Andrade a,⇑, Caio Fabio Schlechta Portella b a b

Anhembi Morumbi University, São Paulo 01423-001, Brazil Department of Naturology, Anhembi Morumbi University, São Paulo 06730-000, Brazil

a r t i c l e

i n f o

Article history: Received 2 March 2017 Accepted 1 June 2017 Available online xxxx Keywords: Research methodology Complementary medicine Alternative medicine Complementary therapies Comparative effectiveness research Outcome assessment Nonlinear dynamics Evaluation studies

a b s t r a c t The scientific literature presents a modest amount of evidence in the use of complementary and alternative medicine (CAM). On the other hand, in practice, relevant results are common. The debates among CAM practitioners about the quality and execution of scientific research are important. Therefore, the aim of this review is to gather, synthesize and describe the differentiated methodological models that encompass the complexity of therapeutic interventions. The process of bringing evidence-based medicine into clinical practice in CAM is essential for the growth and strengthening of complementary medicines worldwide. De Almeida Andrade F, Schlechta Portella CF. Research methods in complementary and alternative medicine: an integrative 34 review. J Integr Med. 2018; 16(1): –. Ó 2017 Shanghai Changhai Hospital. Published by Elsevier B.V. All rights reserved.

Contents 1. 2. 3.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Important methodological alternatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.1. Whole system research, mixed methods and outcome study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.2. Comparative effectiveness research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.3. Pragmatic trials. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.4. N-of-1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.5. Single-subject research designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.6. Aptitude (or attribute) by treatment interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.7. Observational study or real-world research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.8. Metasynthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.9. Method triangulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Points to be observed in these systems of result analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1. NLDS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2. Analysis triangulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.3. Other possibilities of analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conflicts of interests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

⇑ Corresponding author. E-mail address: [email protected] (F. de Almeida Andrade). https://doi.org/10.1016/j.joim.2017.12.001 2095-4964/Ó 2017 Shanghai Changhai Hospital. Published by Elsevier B.V. All rights reserved.

Please cite this article in press as: de Almeida Andrade F, Schlechta Portella CF. Research methods in complementary and alternative medicine: an integrative review. J Integr Med. (2017), https://doi.org/10.1016/j.joim.2017.12.001

2

F. de Almeida Andrade, C.F. Schlechta Portella / Journal of Integrative Medicine xxx (2017) xxx–xxx

1. Introduction Complementary and alternative medicine (CAM) encompasses several millennia of therapeutic systems such as Ayurvedic medicine and traditional Chinese medicine (TCM), as well as contemporary systems like anthroposophical medicine, naturopathic medicine, chiropractic medicine and homeopathy. CAM is usually associated with specific medical systems that include a form of diagnosis and a treatment within their own specific rationale [1]. These medical systems hold an individualized and essentially not reductionist approach in both diagnosis and in treatment; this is exemplified in a complex nonlinear intervention that operates both in independent and in interdependent forms [2–7]. In the year 2000, with the emergence of evidence-based medicine (EBM), which states that clinical practice should follow an assessment of the best evidence possible, the need for scientific data in the health field has increased [8]. This movement generated debate among CAM practitioners, as the complexity of the interventions leads to difficulty in collecting standardized data [9,10]. This issue was shared among professionals and researchers who concluded that research into these systems was necessary [11]. Alternative methods in the field of clinical research in health were sought, and some differentiated methodological directives were defined, such as whole systems research (WSR) [12] and nonlinear dynamics system (NLDS) [13]. Under these directives, the researchers investigate not only the chain of interactions but also systems dynamics, focusing on change and/or movement in patterns [14] and verifying that in nonlinear systems the macroscopic level has a perspective that derives from multidimensional interactions [15]. Thus, it must be stressed that although conventional research methods have their function and are very well established, the WSR needs further attention. This integrative review of the literature allowed for systematic data collection with the goal of improving the understanding of WSR and similar methodologic models.

care)”[Mesh]) OR ‘‘Nonlinear dynamics”[Mesh]) OR ‘‘Evaluation studies”[Publication Type]) OR ‘‘Whole system research”)) AND ((‘‘Complementary therapies”[Mesh]) AND (‘‘Methods”[Mesh])). All the descriptors used with the three databases are listed below (Table 1). Publications that were included in the review presented: (1) theoretical basis that included investigations, analysis, discussions, proposals or declarations about research in CAM; (2) description of methodology (for clinical trials); (3) priorities or priority definition for research in CAM, in the format of published articles and/or books. All publications included were part of the analysis of a complete text. The eligibility of publications was also examined regarding the exclusion criteria. These criteria were: (1) presenting only one case study or a summary; (2) reporting mainly a specific study project or assessment tool. No article evaluation scale was used to test the quality of the publications that were included. This allowed opinion articles, letters, editorials, and debates to be included in the review whenever they presented a relevant content concerning methodologies in CAM therapies. The document selection was first carried out according to their title. Next, a previous reading of the abstracts was performed to verify the eligibility criteria. Whenever these abstracts were not clarifying, the articles were read in full. The selection of original papers was mainly favored, but opinion articles were also included when a contribution to the theme was found. Then, all the articles meeting the criteria were read, and the descriptive process of the gathered material was started. The data related to this phase are displayed in Table 2. Duplicated articles were excluded. Two researchers were in charge of evaluating the relevance of titles and abstracts obtained through electronic search and the selection criteria were applied independently for each of the potentially relevant studies. Then, the title and abstract of the publication were discussed until an agreement about whether or not to include it was reached.

2. Methods 3. Results Structured integrative review of the literature [16] was carried out between June 2015 and July 2016. This research aimed at gathering, describing and synthesizing the differentiated methodological models in the CAM field. Three data bases were used: PubMed, Lilacs and Embase, along with the indexing tool of Academic Google, and articles in English, Spanish, German and Portuguese were included. The search strategy displayed relevant articles published between 1986 and 2016 (up to July 2016). There was no period restriction, so the largest possible number of publications might be included in this review. A specific search strategy based on the concept descriptors was used for each of the databases, as well as a manual complementary search of references. The search strategy used for PubMed was ((((((‘‘Comparative effectiveness research”[Mesh]) OR ‘‘Outcome assessment (health

The literature survey initially identified 2588 documents. After applying eligibility criteria analysis 101 documents were included in the qualitative synthesis (Fig. 1). Two categories of publications, study designs and result analysis systems were found [12–54]. Their characteristics are described in Table 2 since they provide methodological possibilities for research on CAM therapies. 3.1. Important methodological alternatives 3.1.1. Whole system research, mixed methods and outcome study One of the main criticisms of integral treatment systems within conventional research on clinical conduct complete packages has

Table 1 Descriptors used in the databases consulted. Research descriptor

Complementary and alternative medicine descriptor

Method descriptor

Whole system research Nonlinear system Nonlinear dynamics Comparative effectiveness Outcome assessment Outcomes research Evaluation study

Complementary Integrative and complementary practices Medicine, alternative Complementary medicine Medicine, complementary Alternative medicine Alternative therapies Complementary and integrative health practices Integrative and complementary health practices Integrative and complementary medicine

Methodology Methods Procedures, parameters and devices

Please cite this article in press as: de Almeida Andrade F, Schlechta Portella CF. Research methods in complementary and alternative medicine: an integrative review. J Integr Med. (2017), https://doi.org/10.1016/j.joim.2017.12.001

F. de Almeida Andrade, C.F. Schlechta Portella / Journal of Integrative Medicine xxx (2017) xxx–xxx

3

Table 2 Research methods in complementary and alternative medicine. Method

Characteristic

WSR, MM, OS

It attempts to systematically capture the complexity of therapeutic interventions. It seeks to adequately combine study designs and methods in order to measure as many aspects as possible in response to the interventions carried out in a global manner [12–17] They are basically studies about the comparative efficacy between sets of clinical conduct and a control group, where for instance, the intervention in the control group provides the best treatments in conventional medicine [18–20] These studies may waive intervention standardization, allowing an individualized treatment to be carried out and assessed. They also respect the participants preference, thus randomizing only participants with no declared preference [17,20–31] RCT studies with a single patient and multiple crossings. They are randomized and controlled and this randomization is used to determine the order of the treatment. The design of the study displays pairs of treatment periods (control  intervention or placebo  intervention). In between these periods there is a safe ‘elimination’ period for each type of intervention that was used [32–35] Quasi-experimental approach to evaluate the efficacy of a treatment where the participants are assigned as their own controls. The control phase must occur at the beginning of the research, where a target behavior is systematically observed. After a pattern of behavior is established the intervention phase begins [36] The focus of this study is placed on the individual characteristics of each participant in the intervention. There must be at least 2 treatments and 2 aptitudes to be compared, since it is not enough to show that a treatment is different from the other. What has to be shown is the fact that the difference occurs due to the different ‘‘conditions”, ‘‘contexts” or ‘‘characteristics” of the participant [37,38] Study that displays the differences between the practice of the research and the ‘‘real world” practice. It manages to keep the flexibility of protocols in an attempt to achieve the individualization of care and to respect the patient’s preference. The results of these studies tend to be oriented by the patients [27,39–41] Considered as a method directed toward theoretical and conceptual research where the scientific rigor and the highest level of complexity are applied to the investigation of qualitative studies. Aims at classifying concepts, describing and explaining theories and/or developing new theoretical models [42–46] In methodological triangulation there is a choice of two or more methods for the study of a single phenomenon. The aim is to enhance the confidence in the accuracy of the data and its interpretation [47,48] The model evaluates the interactions among observable dynamics by means of biopsychosocial scales, empathy, intrapsychic conflicts, physiological arousal and leukocyte telomerases among others. It also introduces the idea that time dynamics and behavioral patterns are important factors for the understanding of complex systems of intervention [13,14,49] Characterized by the use of one or more methods of data analysis with the goal of increasing the validity of the research [47] MOST; program theory; intervention theory logic model, among others [38,50–54]

Comparative effectiveness research Pragmatic trials

N-of-1

SSRD

ATI

Observational study or RWR

Metasynthesis

Method triangulation Nonlinear dynamical system

Analysis triangulation Other possibilities of analysis

Indentification

been a bias related to the preference of the participants [54–56], since the differences among patients who prefer, believe or accept an integral treatment system may be correlated with characteristics that influence the relief of symptoms [20].

Records identified through database searching (N = 2 588)

Screening

Excluded (n = 2 474) Clearly irrelevant records

Full-text articles assessed for eligibility (n = 114)

Included

Eligibility

Excluded (n = 98) Not meeting the inclusion criteria

Studies included in qualitative synthesis (n = 16)

Articles from hand search (n = 85)

Total studies included in qualitative synthesis (n = 101)

Fig. 1. Flowchart of the literature review process.

The relationship between patient and practitioner, and the patient’s expectations are important points in these methodologies; methodologies that develop positive relationships and expectations lead to better treatment outcomes than methodologies that maintain impersonal, formal and distant conduct [3,55,57–60]. This binomial (professional-patient) becomes one of the main focuses in studies trying to understand the rebalancing of the relationship; in a therapeutic relation, which is closer and more detailoriented, both the professional and the patient are changed with the therapeutic encounter [12,60]. Furthermore, Sidani et al. [61] advocated minimizing exclusion criteria and maintaining a wide set of inclusion criteria in order to represent real-world patient populations, increasing the chances of generalizing the intervention results and their clinical relevance. Analyzing the overall response to interventions is also important in the evaluation of complex care systems. The use of validated measurement tools that attempt to address the complexity of the interventions is recommended [28,62]. These methods should be prioritized in research on CAM practices [2,27,63,64] in addition to traditional treatment outcomes [38,65,66]. They are also recommended for clinical research in biomedicine [67]. 3.1.2. Comparative effectiveness research They are designed to observe the ‘‘real world” context of care, comparing the benefits and harms of diagnosis, treatment and evaluation of the treatment of a clinical condition between CAM and conventional medicine, in most cases. It uses patientcentered and less standardized treatment protocols, which are also seen in WSR or pragmatic trials, where the patient-professional binomial is highly relevant. The comparison in this case is not limited to randomized controlled trials (RCTs). Among other options, we can use data from epidemiological studies and/or observational records [64].

Please cite this article in press as: de Almeida Andrade F, Schlechta Portella CF. Research methods in complementary and alternative medicine: an integrative review. J Integr Med. (2017), https://doi.org/10.1016/j.joim.2017.12.001

4

F. de Almeida Andrade, C.F. Schlechta Portella / Journal of Integrative Medicine xxx (2017) xxx–xxx

3.1.3. Pragmatic trials The main challenges in the application of RCT projects in the study of CAM approaches occur due to difficulties determining adequate placebos [22,60,63,68], the impossibility of doubleblind in some interventions and the individualized and complex approach of many complementary practices [55,69–73]. Pragmatic RCTs attempt to overcome these obstacles so that analyses of diagnosis subgroups in CAM therapies (such as the syndromes in TCM) can be used in these studies, improving the analytical relationship between qualitative data and quantitative components [10,74]. A double diagnostic classification can be performed by first establishing a conventional biomedical diagnosis and then a diagnosis in the CAM system under study [27,75–77]. To Hotopf [23], this is a method that represents the middleground between observational and conventional RCT. 3.1.4. N-of-1 This method determines the study of one patient based on objective oriented criteria [32]. It can be used for severe chronic conditions when conventional treatment is no longer functional [34]. Even with the use of randomization, double-blind, treatment pair replication and quantification of results, the qualitative observations by the therapist result in a far more complete assessment of the treatment efficacy when it is followed by visual inspection [35]. If various N-of-1 treatments investigating the same type of intervention were initiated concomitantly, the study of the sets of data that were generated would be possible. These analyses might explore trends in the data, such as side effect profiles, as well as common characteristics among patients, allowing the researcher to correlate these characteristics with better and worse responses to treatment. Under this approach, large population-based studies could be leveraged for that purpose [32,78]. Thus, the use of combined results from N-of-1 studies could be a technique for developing a stronger evidence base in the CAM approach. One of the reasons why this method has not been incorporated in practice yet may be the lack of emphasis from higher education institutions. With increased emphasis on this approach, coupled with assistance in developing this type of study, these approaches would be more widely spread and would help develop a more rich understanding of treatment outcomes [32,79]. 3.1.5. Single-subject research designs There are several types of single-subject research designs (SSRDs) but all of them comprise 2 research phases: an initial phase and an intervention phase. This method differs from the N-of-1 because the SSRDs are also performed with small groups of participants and also because there is no randomization in the control: the control is always the initial intervention of the research, in which there is a systematic observation of the ‘‘target behavior”. Taking more measurements in shorter periods of time increases the validity of the study, and the probability that some results come up by chance is decreased [36]. 3.1.6. Aptitude (or attribute) by treatment interaction This study suggests that knowledge of a participant’s individual characteristics may be useful to understand variability in response to treatment among individuals. This technique could help understand problems caused by studying group averages, where the range of responses is wide or poorly distributed: that is, different patients respond differently to treatments, and therefore, the ‘‘average effect” may misrepresent treatment effects [38]. This may lead to favoring more trustworthy forecasts about treatments according to patients’ characteristics [37].

The participants may be randomly assigned and the research must be leveraged by a clinical hypothesis in which there is a relation between studied groups that receive the same treatment after having been selected and divided according to their specific characteristics. Data are gathered pre- and post-testing and the analysis generally seeks associations among aptitudes, treatments and results [37]. 3.1.7. Observational study or real-world research If this method is well-outlined, it may produce results that are comparable to those in RCT, thus reducing operating costs. The studies must be long, to allow for the distinction between the condition remission time and the time of response to treatment; they must be very well documented, clearly displaying in the reports which interventions were performed and how they were carried out, and several branches of study can be included, so the report validity is increased [27]. An example of this kind of study is anthroposophic medicine outcome study, a prospective observational multicenter study, which has been consistently cited by studies of this nature in recent years [40]. 3.1.8. Metasynthesis The contributions of this approach are generally used to check for gaps in knowledge that emerge from comparing quantitative and qualitative information present in other studies. The works of Sousa et al. [46] and Bastos et al. [80] present an overview of the possibilities provided by this method. Different methodologic options may be used, depending on the specific objectives of the work, ranging from integrative interpretation to meta-analytic approaches [46]. 3.1.9. Method triangulation There are 6 types of triangulation: data triangulation, investigator triangulation, methodological triangulation, theory triangulation, interdisciplinary triangulation and analytical triangulation. Analytical triangulation is addressed in the next topic of this article. Methodological triangulation gathers data derived from different approaches, thus constructing a more complex panorama of results including a range of observed phenomenon. A simple example is the qualitative-quantitative approach, which generates potentially complementary data. When conducting a research project in which you want to use triangulation methods, it is recommended that these are defined according to the question and the objective of your research. Based on this, the appropriate outcome measures will be also defined [48]. The researcher must choose structured tools balancing their weak and strong points to minimize the ‘‘threat” to the internal and external validity of the study [47,48]. 3.2. Points to be observed in these systems of result analysis 3.2.1. NLDS NLDS are investigative tools that go beyond the quantitative and qualitative methods traditionally used [49,14]. In this method, emergent effects are considered, such as those that present nonlinear outcomes, promote local and global change or are the result of synergy [81]. The data that feed this system may stem from a range of sources; these include self-reporting, clinical notes, physiological readings, as well as data derived from other statistical analyses [13]. One of the main software tools that can be used for the development of this kind of work is known as GridWare (Version 1.1, http://statespacegrids.org) [82]. This can offer useful possibilities to test the mechanism of action hypotheses for clinical phenomena in CAM [81]. Addition-

Please cite this article in press as: de Almeida Andrade F, Schlechta Portella CF. Research methods in complementary and alternative medicine: an integrative review. J Integr Med. (2017), https://doi.org/10.1016/j.joim.2017.12.001

F. de Almeida Andrade, C.F. Schlechta Portella / Journal of Integrative Medicine xxx (2017) xxx–xxx

ally, it may help document the indirect, dynamic and emerging effects in the interventions. Each analysis parameter may serve as an independent variable including multivariate statistical models [13]. 3.2.2. Analysis triangulation Analysis triangulation can be thought of as observing the same set of results from a different angle, to form a more complete understanding of the phenomenon under study. It also provides a broader perspective on the relationships among treatments or on any other link established in the study objectives [47]. 3.2.3. Other possibilities of analyses Some data analysis strategies, such as multiphase optimization strategy, display the main effects of interactions among individual components in a multicomponent intervention [50]. The program theory helps to assess results of complex interventions in which the process takes place throughout the therapy. The intervention theory logic modeling has been developed to investigate and evaluate the social policies; it is a tool that helps to understand and observe what works, for whom it works and under what circumstances [52]. There are also possibilities of data analyses used in clinical research, including variable analysis modeling [53], cluster design or grouping analysis [22,52,83], covariance analysis [30], stratified analysis, multivariate adjustment and sensitivity analysis [54], regression lines and discontinuity regression [38]. There are also analytical methods, including time series, correlation dimensions, detrended fluctuation analysis and multitiered entropy [55]. 4. Discussion An increase in the search for CAM therapies and the emergence of EBM have boosted research in this area, since there is a need for more evaluation and substantiation of these treatment systems both for patients and for practitioners. In addition, as national health systems are showing interest in offering this type of care, the need for research is even more present. It is not easy to produce evidence in CAM therapies that meets the standards of RCTs; the initial difficulties lie in their practice presuppositions and complex interventions performed. Firstly, a decision has to be made between focusing on the ‘‘gold standard” of clinical evidence, or on methodological pathways that elucidate ‘‘evidence in the real-world practice”. These paths may be poorly described and infrequently practiced within clinical research. From the clinical research perspective, it is clear that good randomized clinical trials have, for example, made a fundamental contribution to conventional medicine in certain fields, such as describing active pharmacological constituents. However, some authors, such as Verhoef et al. [27] and Walach et al., [65] reinforce the idea that CAM presuppositions are so fundamentally different from those of conventional medicine that they may require specific research. CAM is not the only family of treatment that has faced this difficulty in the research of complex interventions. All the complex treatments within conventional medicine, such as those used in oncology, gerontology, or palliative care, for instance, have experienced this difficulty. According to Bell et al. [49] and Ahn et al., [55] the conventional research models that use an inferential statistical analysis and investigation of the standardized interventions efficacy to evaluate specific results may be fundamentally incompatible with approaches where complex clinical conducts are considered. Thus, we may suppose that research in the CAM therapies should be directed toward multidisciplinary research, mixedmethod research, and multicenter studies [20,22,71,84–87], to improve research validity and rigor. However, this will not solve the issue.

5

From a broad point of view, issues of a methodological nature in CAM could be resolved if circular and nonhierarchal approaches to analysis were adopted [84,88,89]; the gap between scientific evidence and clinical evidence needs to be rethought [90]. According to El-Hani and Pereira [91], including a wider range of concepts in research permits broader approaches; greater perspective can be gained when we include properties of the whole as well as constituent parts when explaining properties of observed phenomena. This perspective suggests that several approaches produce a set of different and potentially complementary information. Hence, when the qualitative and quantitative methods are appropriately combined and undergo differentiated data analysis, there is a tendency toward a more comprehensive picture that reflects the complex, patient-centered, relevant and sensitive approach employed by complex treatment systems with different philosophical presuppositions [7,89,92–94]. The openness of EBM to the complexity, through pragmatic RCTs [20,23,27,95–97], and the incorporation of the concepts of respecting the ‘‘patient’s value” and respecting the professional’s ‘‘clinical experience”, can provide a greater integration/interaction between conventional methods of research on clinical practice and methods that WRS directs us to research in the CAM field. It was observed, however, that this openness could require more actions, and perhaps the acceptance of evidence from other hierarchical levels [98]. Verhoef et al. [60] and Borgerson [98] suggest a few ways to deal with EBM within the field of CAM. The first could be to accept the hierarchy of evidence and carry out RCT or pragmatic RCT whenever possible. Another way would be to argue that there are special characteristics for research in CAM, thus justifying the acceptance of the proof even if they do not stem from RCT. A third manner would be to work to develop research projects that may demonstrate new patterns of evidence that are consistent with this complex approach [60,98]. We would like to highlight that it is important to maintain the ‘‘basic” types of research, for without them the possibility of finding the ‘‘fundamental causes” of the unbalances would be restricted, and the focus of the complex interventions [83,98] would be shifted, since these types of research form the basis for clinical investigation [87]. The possibilities of differentiated statistical analyses are another key factor which can support a good study design, leading to a result assessment in a complex manner. Within this issue the choice of statistical tools is critical, since they provide, along with the intervention and the correct data collection, scientific evidence of good quality in CAM. All these factors may help keep the complexity of interventions, and may also avoid losing sight of the presuppositions of such interventions in the research. Nevertheless, besides the difficulties that are inherent in the methods, there are some others that arise from a social/cultural/ educational system that culminates in the academic education of most current researchers. The publication bias would be an example, since only articles that display positive results tend to be published. Other examples would be (1) the heterogeneity of vocabulary among the researchers in CAM, which complicated database searches in many topics; (2) the peer review bias, which is formed because conventional statistical methods taught in school produce researchers, when faced with new methodologies, suspect that they are devoid of rigor and scientific quality; (3) the indexation bias, in which the database search strategy is not enough to uncover comprehensive material on a subject, requiring manual searches for references that are relevant to the research. Such difficulties are intensified by the costs of the research process and available funding. The limited budget is a recurring problem in the execution of the research [20,99,100], and it does not match the evidence shown by large studies in the CAM area, which,

Please cite this article in press as: de Almeida Andrade F, Schlechta Portella CF. Research methods in complementary and alternative medicine: an integrative review. J Integr Med. (2017), https://doi.org/10.1016/j.joim.2017.12.001

6

F. de Almeida Andrade, C.F. Schlechta Portella / Journal of Integrative Medicine xxx (2017) xxx–xxx

despite not being extensive if compared with other knowledge areas, show that the treatment costs with CAM in some interventions are lower than the conventional treatments costs [79]. The study and application of research methods that try to encompass complex interventions can be considered a priority for the future [10]. A possible solution could be carrying out not only individual projects aiming to use methodologies that focus on spanning the complexity of the interventions, but also multimethod, multicenter and multiprofessional research programs capable of producing consistent evidence in CAM [27,88]. Another priority is to devote more attention to the teaching of EBM in the CAM schools, including for example, mandatory EBM courses, development of critical reading in research, lectures, workshops and study groups, that will produce a stronger research basis and an increase in the publications in this field [101]. For the issues mentioned above to be adequately dealt with, institutions of higher education need a ‘‘standardization” of language, laws, curricular framework and instruction in evidence-based research, so that practitioners are more prepared to effectively participate in research and clinical practices that can be communicated well with the more conventional Western medicine system. Moreover, practicing professionals also need to receive education in the methods and standards of EBM, improving the intellectual commerce among fields. Producing evidence in CAM is a social concern, since it provides possible health solutions, which may be at much lower costs than conventional alternative, for contemporary problems. The same applies to complex research methodologies; although CAM is an area of knowledge that is immediately and positively impacted by this approach, the adoption of these methods remains limited. Issues regarding safety and risk of interventions were not dealt with here, but the topic should not be ignored, since investigation about it is paramount [88]. Researchers in CAM must face this methodological challenge, discuss it and study it. If they merely conform to the conventional molds, they run the risk of missing a critical part of the presuppositions concerning the nature of the therapeutic system under study. Engaging in the production of consistent and coherent data is a challenge for the whole scientific community, and this is not different for CAM practitioners. Therefore, there is a clear need for research methods in CAM to be differentiated and/or adapted. Within this perspective, we cannot dispense with rigor nor quit evaluating clinical conduct best practices. We must manage to express what CAM really does, by testing, introducing and divulging other methodological possibilities. The challenge, when adopting a practical perspective, is to acknowledge the importance of multidimensional and multifactor models in the evaluation of the participants in the research, to observe the effects of the complex systems of treatment and their interrelations, taking the diversity of theories that give support to complex/integral care systems into consideration, without disregarding practical difficulties such as heterogeneity of systems and vocabulary, use of several diagnoses (conventional medical diagnoses and diagnoses of unbalances of traditional medicine) as well as comorbidities and specific laws from each location of practices. Conflicts of interests The authors have no conflict of interests. References [1] Tesser CD, Luz MT. Medical and therapeutic alternative rationalities. Sci Public Health 2008;13(1):195–206. Portuguese.

[2] Ribeaux P, Spence M. CAM evaluation: what are the research questions? Complement Ther Med 2001;9(3):188–93. [3] Di Blasi Z, Harkness E, Ernst E, Georgiou A, Kleijnen J. Influence of context effects on health outcomes: a systematic review. Lancet 2001;357 (9258):757–62. [4] Moerman DE, Jonas WB. Deconstructing the placebo effect and finding the meaning response. Ann Intern Med 2002;136(6):471–6. [5] Harlan Jr WR. New opportunities and proven approaches in complementary and alternative medicine research at the National Institutes of Health. J Altern Complement Med 2001;7(1):53–9. [6] Campbell M, Fitzpatrick R, Haines A, Kinmonth A, Sandercock P, Spiegelhalter D, et al. Framework for design and evaluation of complex interventions to improve health. BMJ 2000;321(7262):694–6. [7] Craig P, Dieppe P, Macintyre S, Michie S, Nazareth I, Petticrew M. Medical Research Council Guidance. Developing and evaluating complex interventions: the new Medical Research Council guidance. BMJ 2008;337: a1655. [8] Leach MJ, Gillham D. Are complementary medicine practitioners implementing evidence based practice? Complement Ther Med 2011;19 (3):128–36. [9] Polich G, Dole C, Kaptchuk TJ. The need to act a little more ‘scientific’: biomedical researchers investigating complementary and alternative medicine. Sociol Health Illn 2010;32(1):106–22. [10] Coulter ID, Lewith G, Khorsan R, Kirk R, Mittman B. Research methodology: choices, logistics, and challenges. Evid Based Complement Alternat Med 2014;2014:780520. [11] National Center for Complementary and Alternative Medicine. NCCAM third strategic plan: 2011–2015. Exploring the science of complementary and alternative medicine. (2017-9-24) [2017-11-20]. https://nccih.nih.gov/ about/plans/2011. [12] Ritenbaugh C, Verhoef M, Fleishman S, Boon H, Leis A. Whole systems research: a discipline for studying complementary and alternative medicine. Altern Ther Health Med 2003;9(4):32–6. [13] Howerter A, Hollenstein T, Boon H, Niemeyer K, Brule D. State-space grid analysis: applications for clinical whole systems complementary and alternative medicine research. Forsch Komplementarmed 2012;19(S1):30–5. [14] Bell IR, Koithan M. Models for the study of whole systems. Integr Cancer Ther 2006;5(4):293–307. [15] Argibay P. Advanced research methodology: introduction to the study of complex systems and their applications part III: complex dynamics. Rev Hosp Ital B Aires 2012;32(3):140–7. Spanish. [16] Whittemore R, Knafl K. The integrative review: updated methodology. J Adv Nurs 2005;52(5):546–53. [17] Leis AM, Weeks LC, Verhoef MJ. Principles to guide integrative oncology and the development of an evidence base. Curr Oncol 2008;15(Suppl 2):S83–7. [18] Witt C, Brinkhaus B, Jena S, Linde K, Streng A, Wagenpfeil S, et al. Acupuncture in patients with osteoarthritis of the knee: a randomised trial. Lancet 2005;366(9480):136–43. [19] Witt CM, Jena S, Brinkhaus B, Liecker B, Wegscheider K, Willich SN. Acupuncture for patients with chronic neck pain. Pain 2006;125(1– 2):98–106. [20] Macpherson H. Pragmatic clinical trials. Complement Ther Med 2004;12(2– 3):136–40. [21] Cardini F, Wade C, Regalia AL, Gui S, Li W, Raschetti R, et al. Clinical research in traditional medicine: priorities and methods. Complement Ther Med 2006;14(4):282–7. [22] Fischer F, Lewith G, Witt CM, Linde K, von Ammon K, Cardini F, et al. A research roadmap for complementary and alternative medicine: what we need to know by 2020. Forsch Komplementarmed 2014;21(2):e1–e16. [23] Hotopf M. The pragmatic randomised controlled trial. Adv Psychiatr Treat 2002;8(5):326–33. [24] Linde K. Invited commentary on Hyland ME: methodology for the scientific evaluation of complementary and alternative medicine. Complement Ther Med 2003;11(3):154–5. [25] Paterson C, Dieppe P. Characteristic and incidental (placebo) effects in complex interventions such as acupuncture. BMJ 2005;330(7501):1202–5. [26] Torgerson DJ, Klaber-Moffett J, Russell IT. Patient preferences in randomised trials: threat or opportunity? J Health Serv Res Policy 1996;1(4):194–7. [27] Verhoef MJ, Lewith G, Ritenbaugh C, Boon H, Fleishman S, Leis A. Complementary and alternative medicine whole systems research: beyond identification of inadequacies of the RCT. Complement Ther Med 2005;13 (3):206–12. [28] Vickers A. Methodological issues in complementary and alternative medicine research: a personal reflection on 10 years of debate in the United Kingdom. J Altern Complement Med 1996;2:515–24. [29] Preference Collaborative Review Group, Adamson SJ, Bland JM, Hay EM, Johnson RE, Jones GT, et al. Patients’ preferences within randomised trials: systematic review and patient level meta-analysis. BMJ 2008;337:a1864. [30] Walker L, Anderson J. Testing complementary and alternative therapies within a research protocol. Eur J Cancer 1999;35(11):1614–8. [31] Jena S, Witt CM, Brinkhaus B, Wegscheider K, Willich SN. Acupuncture in patients with headache. Cephalalgia 2008;28(9):969–79. [32] Lillie EO, Patay B, Diamant J, Issell B, Topol EJ, Schork NJ. The n-of-1 clinical trial: the ultimate strategy for individualizing medicine? Per Med 2011;8 (2):161–73.

Please cite this article in press as: de Almeida Andrade F, Schlechta Portella CF. Research methods in complementary and alternative medicine: an integrative review. J Integr Med. (2017), https://doi.org/10.1016/j.joim.2017.12.001

F. de Almeida Andrade, C.F. Schlechta Portella / Journal of Integrative Medicine xxx (2017) xxx–xxx [33] Mahon J, Laupacis A, Donner A, Wood T. Randomised study of n of 1 trials versus standard practice. BMJ 1996;312(7038):1069–74. [34] Guyatt G, Sackett D, Taylor DW, Chong J, Roberts R, Pugsley S. Determining optimal therapy randomized trials in individual patients. N Engl J Med 1986;314(14):889–92. [35] Guyatt G, Sackett D, Adachi J, Roberts R, Chong J, Rosenbloom D, et al. A clinician’s guide for conducting randomized trials in individual patients. CMAJ 1988;139(6):497–503. [36] Backman CL, Harris SR. Case studies, single-subject research, and N of 1 randomized trials: comparisons and contrasts. Am J Phys Med Rehabil 1999;78(2):170–6. [37] Ridenour TA, Treloar JH, Biner PM, Henriksen LW, Dean RS. Attributetreatment interaction research using nonexperimental data and the allocation average. Int J Neurosci 1999;99(1–4):19–47. [38] Caspi O, Bell IR. One size does not fit all: aptitude chi treatment interaction (ATI) as a conceptual framework for complementary and alternative medicine outcome research. Part II—research designs and their applications. J Altern Complement Med 2004;10(4):698–705. [39] Luo J, Xu H, Liu B. Real world research: a complementary method to establish the effectiveness of acupuncture. BMC Complement Altern Med 2015;15:153. [40] Hamre HJ, Kiene H, Ziegler R, Tröger W, Meinecke C, Schnürer C, et al. Overview of the publications from the anthroposophic medicine outcomes study (AMOS): a whole system evaluation study. Glob Adv Heal Med 2014;3 (1):54–70. [41] Ernst E. Back to basics? How important is basic research in CAM? Altern Ther Health Med 2001;7(5):30–1. [42] Carrillo-González GM, Gómez-Ramírez OJ, Vargas-Rosero E. Metasynthesis: a research methodology. Rev Sal Púb 2007;9(4):609–17. Spanish. [43] Gálvez A. Evidence-based clinical practice: a bibliographic approach. Nurs Sci Dev 2000;8(4):119–21. Spanish. [44] Galvez A. Contemporary methods of scientific synthesis: a proposal for homogeneity, classification of research by design and utility. Index de Enferm 2003;12(43):45–9. Spanish. [45] Jifa G. Meta-synthesis knowledge system: basics and practice. Int J Knowl Syst Sci 2010;1(1):58–72. [46] Sousa CF, Branco MZPC. Meta-synthesis: a review of literature—contributions to knowledge and nursing care. Enferm Foco 2014;4(2):1–10. Portuguese. [47] Hamilton D, Bechtel GA. Research implications for alternative health therapies. Nurs Forum 1996;31(1):6–10. [48] Breitmayer BJ, Ayres L, Knafl KA. Triangulation in qualitative research: evaluation of completeness and confirmation purposes. Image J Nurs Sch 1993;25(3):237–43. [49] Bell IR, Koithan M, Pincus D. Methodological implications of nonlinear dynamical systems models for whole systems of complementary and alternative medicine. Forsch Komplementmed 2012;19(S1):15–21. [50] Collins LM, Murphy SA, Nair VN, Strecher VJ. A strategy for optimizing and evaluating behavioral interventions. Ann Behav Med 2005;30(1):65–73. [51] Pawson R, Greenhalgh T, Harvey G, Walshe K. Realist review: a new method of systematic review designed for complex policy interventions. J Health Serv Res Policy 2005;10(S1):21–34. [52] Paterson C, Baarts C, Launso L, Verhoef M. Evaluating complex health interventions: a critical analysis of the ‘outcomes’ concept. BMC Complement Altern Med 2009;9:18. [53] Katz MH. Multivariable analysis: a primer for readers of medical research. Ann Intern Med 2003;138(8):644–50. [54] Ziegler R. Elements of therapeutic research structure. Forsch Komplementamed Klass Naturheilkd 2004;11(S1):5–12. [55] Ahn AC, Nahin RL, Calabrese C, Folkman S, Kimbrough E, Shoham J, et al. Applying principles from complex systems to studying the efficacy of CAM therapies. J Altern Complement Med 2010;16(9):1015–22. [56] Wolff N. Randomised trials of socially complex interventions: promise or peril? J Health Serv Res Policy 2001;6(2):123–6. [57] Verhoef MJ, Vanderheyden LC, Dryden T, Mallory D, Ware MA. Evaluating complementary and alternative medicine interventions: in search of appropriate patient-centered outcome measures. BMC Complement Altern Med 2006;6:6–38. [58] Scriba PC. Placebo and the relationship between doctors and patients: overview. Bundesgesundheitsblatt Gesundheitsforsch Gesundheitsschutz 2012;55(9):1113–7. German with abstract in English. [59] Crow R, Gage H, Hampson S, Hart J, Kimber A, Thomas H. The role of expectancies in the placebo effect and their use in the delivery of health care: a systematic review. Health Technol Assess 1999;3(3):1–96. [60] Verhoef MJ, Casebeer AL, Hilsden RJ. Assessing efficacy of complementary medicine: adding qualitative research methods to the ‘‘Gold Standard”. J Altern Complement Med 2002;8(3):275–81. [61] Sidani S, Epstein DR, Moritz P. An alternative paradigm for clinical nursing research: an exemplar. Res Nurs Heal 2003;26(3):244–55. [62] Eton DT, Bauer BA, Sood A, Yost KJ, Sloan JA. Patient-reported outcomes in studies of complementary and alternative medicine: problems, solutions, and future directions. Explore (NY) 2011;7(5):314–9. [63] Zick SM, Schwabl H, Flower A, Chakraborty B, Hirschkorn K. Unique aspects of herbal whole system research. Explore (NY) 2009;5(2):97–103. [64] Witt CM, Chesney M, Gliklich R, Green L, Lewith G, Luce B, et al. Building a strategic framework for comparative effectiveness research in

[65]

[66]

[67]

[68]

[69] [70] [71]

[72] [73] [74]

[75] [76]

[77]

[78]

[79]

[80] [81]

[82] [83]

[84]

[85]

[86]

[87] [88]

[89] [90] [91] [92] [93] [94] [95] [96]

7

complementary and integrative medicine. Evid Based Complement Alternat Med 2012;2012:531096. Walach H, Jonas WB, Lewith GT. The role of outcomes research in evaluating complementary and alternative medicine. Altern Ther Health Med 2002;8 (3):29–45. Bell IR, Caspi O, Schwartz GER, Grant KL, Gaudet TW, Rychener D, et al. Integrative medicine and systemic outcomes research issues in the emergence of a new model for primary health care. Arch Intern Med 2002;162(2):133–40. Verhoef MJ, Lewith G, Ritenbaugh C, Thomas K, Boon H, Fønnebø V. Whole systems research: moving forward. Focus Altern Complement Ther 2004;9 (2):87–90. Bensoussan A, Talley NJ, Hing M, Menzies R, Guo A, Ngu M. Treatment of irritable bowel syndrome with Chinese herbal medicine: a randomized controlled trial. JAMA 1998;280(18):1585–9. Hróbjartsson A. What are the main methodological problems in the estimation of placebo effects? J Clin Epidemiol 2002;55(5):430–5. Mehling WE, DiBlasi Z, Hecht F. Bias control in trials of bodywork: a review of methodological issues. J Altern Complement Med 2005;11(2):333–42. Verhoef MJ, Mulkins A, Kania A, Findlay-Reece B, Mior S. Identifying the barriers to conducting outcomes research in integrative health care clinic settings: a qualitative study. BMC Health Serv Res 2010;10:14. Tamayo C, Boon H, Ghishan F, Trinh K. Research methodology: evaluating complementary and alternative therapies. Drug Inf J 2002;36(3):535–48. Sagar SM. Integrative oncology in North America. J Soc Integr Oncol 2006;4 (1):27–39. Liu JP, Han M, Li XX, Mu YJ, Lewith G, Wang YY, et al. Prospective registration, bias risk and outcome-reporting bias in randomised clinical trials of traditional Chinese medicine: an empirical methodological study. BMJ Open 2013;3(7). pii: e002968. Bansal D. Research methodological issues in evaluating herbal interventions. Medicine 2010;10:11. Guyatt GH, Haynes RB, Jaeschke RZ, Cook DJ, Green L, Naylor CD, et al. Users’ guides to the medical literature: XXV. Evidence-based medicine: principles for applying the users’ guides to patient care. Evidence-based medicine working group. JAMA 2000;284(10):1290–6. Eskinazi D. Methodologic considerations for research in traditional (alternative) medicine. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998;86(6):678–81. Sung L, Feldman BM. N-of-1 trials: innovative methods to evaluate complementary and alternative medicines in pediatric cancer. J Pediatr Hematol Oncol 2006;28(4):263–6. Johnston BC, Mills E. N-of-1 randomized controlled trials: an opportunity for complementary and alternative medicine evaluation. J Altern Complement Med 2004;10(6):979–84. Bastos JDA, de Oliveira AAS, Trancoso AER, Canuto LT. Metasynthesis. Res Health 2015;1:147–52. Spanish. Koithan M, Bell IR, Niemeyer K, Pincus D. A complex systems science perspective for whole systems of complementary and alternative medicine research. Forsch Komplementmed 2012;19(S1):7–14. Lamey A, Hollenstein T, Lewis MD, Granic I. GridWare (Version 1.1). (2004) [2017-11-20]. http://statespacegrids.org. Elder C, Aickin M, Bell IR, Fønnebø V, Lewith GT, Ritenbaugh C, et al. Methodological challenges in whole systems research. J Altern Complement Med 2006;12(9):843–50. Walach H, Falkenberg T, Fønnebø V, Lewith G, Jonas WB. Circular instead of hierarchical: methodological principles for the evaluation of complex interventions. BMC Med Res Methodol 2006;6:29. Jonas WB. Building an evidence house: challenges and solutions to research in complementary and alternative medicine. Forsch Komplementarmend Klass Naturheilkd 2005;12(3):159–67. Verhoef MJ, Vanderheyden LC, Fønnebø V. A whole systems research approach to cancer care: why do we need it and how do we get started? Integr Cancer Ther 2006;5(4):287–92. Harlan Jr WR. Research on complementary and alternative medicine using randomized controlled trials. J Altern Complement Med 2001;7(S1):S45–52. Verhoef MJ, Leis A. From studying patient treatment to studying patient care: arriving at methodologic crossroads. Hematol Oncol Clin North Am 2008;22 (4):671–82. Hyland ME. Methodology for the scientific evaluation of complementary and alternative medicine. Complement Ther Med 2003;11(3):146–53. Duggal R, Menkes DB. Evidence-based medicine in practice. Int J Clin Pract 2011;65(6):639–44. El-Hani CN, Pereira AM. Reductionism or holism: wandering the question. Ideation 1999;3:73. Portuguese. Carter B. Methodological issues and complementary therapies: researching intangibles? Complement Ther Nurs Midwifery 2003;9(3):133–9. Walach H. The efficacy paradox in randomized controlled trials. J Altern Complement Med 2001;7(3):213–8. Riley D, Berman B. Complementary and alternative medicine in outcomes research. Altern Ther Health Med 2002;8(3):36–7. Ernst E, Resch K. Pragmatic trials: propaganda or research tool? Focus Altern Complement Ther 2005;10(3):181–3. Witt CM. Clinical research on acupuncture: concepts and guidance on efficacy and effectiveness research. Chin J Integr Med 2011;17(3):166–72.

Please cite this article in press as: de Almeida Andrade F, Schlechta Portella CF. Research methods in complementary and alternative medicine: an integrative review. J Integr Med. (2017), https://doi.org/10.1016/j.joim.2017.12.001

8

F. de Almeida Andrade, C.F. Schlechta Portella / Journal of Integrative Medicine xxx (2017) xxx–xxx

[97] Roland M, Torgerson DJ. What are pragmatic trials? BMJ 1998;316 (7127):285. [98] Borgerson K. Evidence-based alternative medicine? Perspect Biol Med 2005;48(4):502–15. [99] Fischer FH, Lewith G, Witt CM, Linde K, von Ammon K, Cardini F, et al. High prevalence but limited evidence in complementary and alternative medicine: guidelines for future research. BMC Complement Altern Med 2014;14–46.

[100] Harlan Jr WR. New opportunities and proven approaches in complementary and alternative medicine research at the National Institutes of Health. J Altern Complement Med 2001;7(S1):S53–9. [101] Mills EJ, Hollyer T, Guyatt G, Ross CP, Saranchuk R, Wilson K, et al. Teaching evidence-based complementary and alternative medicine: 1. A learning structure for clinical decision changes. J Altern Complement Med 2002;8 (2):207–14.

Please cite this article in press as: de Almeida Andrade F, Schlechta Portella CF. Research methods in complementary and alternative medicine: an integrative review. J Integr Med. (2017), https://doi.org/10.1016/j.joim.2017.12.001