|Home | About | Journals | Submit | Contact Us | Français|
The STandards for Reporting Interventions in Clinical Trials of Acupuncture (STRICTA) were published in five journals in 2001 and 2002. These guidelines, in the form of a checklist and explanations for use by authors and journal editors, were designed to improve reporting of acupuncture trials, particularly the interventions, thereby facilitating their interpretation and replication. Subsequent reviews of the application and impact of STRICTA have highlighted the value of STRICTA as well as scope for improvements and revision.
To manage the revision process a collaboration between the STRICTA Group, the CONSORT Group and the Chinese Cochrane Centre was developed in 2008. An expert panel with 47 participants was convened that provided electronic feedback on a revised draft of the checklist. At a subsequent face-to-face meeting in Freiburg, a group of 21 participants further revised the STRICTA checklist and planned dissemination.
The new STRICTA checklist, which is an official extension of CONSORT, includes 6 items and 17 subitems. These set out reporting guidelines for the acupuncture rationale, the details of needling, the treatment regimen, other components of treatment, the practitioner background and the control or comparator interventions. In addition, and as part of this revision process, the explanations for each item have been elaborated, and examples of good reporting for each item are provided. In addition, the word ‘controlled’ in STRICTA is replaced by ‘clinical’, to indicate that STRICTA is applicable to a broad range of clinical evaluation designs, including uncontrolled outcome studies and case reports. It is intended that the revised STRICTA checklist, in conjunction with both the main CONSORT statement and extension for non-pharmacological treatment, will raise the quality of reporting of clinical trials of acupuncture.
The STandards for Reporting Interventions in Clinical Trials of Acupuncture (STRICTA) reporting guidelines, first published in 2001,1–9 were designed to improve the completeness and transparency of reporting of interventions in controlled trials of acupuncture, in order that such trials may be more accurately interpreted and readily replicated. STRICTA comprised a checklist that expanded the generic content of item 4 of the CONSORT statement,10 11 which relates to the reporting of the intervention.
A survey of authors of clinical trials and systematic reviews was subsequently conducted to determine the usefulness of STRICTA in helping them to write their reports.12 In addition, a survey of 90 acupuncture trials was undertaken to assess whether use of the STRICTA checklist was associated with improved reporting over time.13 The results of these initiatives led to conclusions that most STRICTA items were found to be necessary and easy to use, though some were seen as poorly reported, ambiguous or possibly redundant and a number of suggestions were made for additional items. A revision of STRICTA was therefore proposed.
Meanwhile, extensions to CONSORT have been developed to cover the reporting of non-pharmacological treatments14 15 and pragmatic trials.16 Since there are acupuncture-specific aspects to reporting not covered by these extensions, it was decided that STRICTA should be revised in a manner congruent with CONSORT and its extensions for non-pharmacological treatments and pragmatic trials.
The combination of these developments led to an agreement between the CONSORT Group and the STRICTA Group, in collaboration with the Chinese Cochrane Centre and the Chinese Centre for Evidence-based Medicine, to revise STRICTA as a formal extension to CONSORT. The revision processes have been described in more detail elsewhere.17 This paper describes the outcome and new checklist, updated explanations and published examples of good reporting.
In the summer of 2008, a group of 47 experts from the original STRICTA Group, the CONSORT Group, the World Federation of Acupuncture and Moxibustion Societies, the Acupuncture Trialists' Collaboration,18 the Society for Acupuncture Research19 and clinical trial authors were surveyed.12 The experts were from 15 countries, 41 had academic positions, 31 were acupuncturists, 18 were involved with journals, such as board members, 15 were doctors and 11 had previously helped in developing reporting guidelines. These experts were consulted about a draft of revised STRICTA items that had evolved from previous research.12 13 Feedback was collated and forwarded (with permission) to those invited to a consensus development workshop, the next phase of the revision process.
Twenty-one people attended a workshop in Freiburg, Germany, in October 2008. The attendees included experts in epidemiology, trial methodology, statistics and medical journal editing. Just over half the participants were acupuncturists from a variety of backgrounds, including doctors and non-doctors. All attendees received collated feedback from the 47 experts, together with a draft revised STRICTA checklist for consideration.
The workshop comprised presentations about the history of STRICTA, CONSORT and the then new CONSORT non-pharmacological treatments extension.14 15 The results of two investigations into the utility and acceptability of STRICTA,12 13 and the subsequent consultation with the 47 experts, were also presented. A general discussion and agreement on generic issues relating to STRICTA were followed by a discussion of each nominated checklist item. The aim was to agree, where possible, on the content of the updated draft checklist as well as to develop a revised set of explanations for each included item.
Subsequent to the workshop, a small writing group edited drafts of the revised STRICTA checklist, identifying for each item one or more exemplars of good reporting, and developed text explaining the rationale and discussing relevant evidence. Taking into account further feedback from those attending the Freiburg workshop, the writing group finalised the STRICTA checklist, the explanations and the examples of good reporting.
There was agreement that STRICTA should continue to function as a stand-alone guideline for reporting acupuncture studies and be an official extension of CONSORT for reporting randomised controlled trials. There was also consensus on a minor change of name, in that the word ‘controlled’ in STRICTA should be replaced by ‘clinical’, to indicate that it was applicable for reporting a broad range of clinical evaluation designs, including uncontrolled outcome studies and case reports. The group agreed that the rationale behind reporting should be to provide the information needed to allow replication of a study, reduce ambiguity and enhance transparency. The group recognised that acupuncture trials inevitably differ in the degree of individualisation of care that is permitted and agreed that the reporting guideline should acknowledge this and be applicable across the whole range of designs. The group also suggested that the revised STRICTA statement, when published, should be presented as embedded within the two-group parallel trial CONSORT checklist10 and its non-pharmacological treatment extension checklist.14
The revised STRICTA checklist comprises six items split into 17 subitems (table 1). Table 2 presents details of the way in which the revised STRICTA checklist fits within the CONSORT checklist10 and its extension for non-pharmacological treatments.14 Below we provide the checklist text for each of the six items and their subitems, as well as explanations on the need for their adequate reporting and examples of good reporting from the published literature.
Acupuncture has a long history in many cultures and is characterised by a broad diversity of styles and approaches in both East Asia and the West.20 In order for readers to contextualise the trial within the range of current clinical practices, researchers should state the overall style or approach on which they have based the treatments. If the researcher believes the treatment approach is completely novel, then this should be clearly stated.
The author(s) should provide the reasoning for the chosen treatment, including rationale for diagnosis, point selection and treatment procedures. The ‘rules’ that were used in providing treatments should be described. When treatments were selected that have roots in traditional practice, it is recommended that the historical and cultural context be supplied. This is relevant for interventions within styles such as Traditional Chinese Medicine, where the broad diversity of approaches requires careful identification of where and when the treatment parameters were developed. Where consensus methods, expert clinical panels, practitioner surveys or some combination of sources have been used to define the treatment protocol, it is recommended that full details of the methodology be given. Literature and other sources should be provided where relevant, in order that others can replicate the trial by consulting these source(s) and/or developmental methods on which treatment was based. Authors are encouraged to reference published works that are easily obtainable, such as a book or journal article. If the reference is a thesis, non-published work, written material only available in a different language from the journal article, or a verbal communication, authors are encouraged to present or summarise the information in an appendix or make it otherwise generally available (eg, on a website). For fully individualised trials where the goal is to have representative practitioners who are encouraged to practise as they normally do, it is appropriate to specify the selection process for the practitioners, providing details of criteria for their inclusion. It is important to note that where details of the intended intervention are defined in advance, it is possible that what was actually administered may have differed. In such cases, precise details of the treatments that were provided are also necessary.
The extent to which the treatment was individualised, both between patients and between practitioners, should be described. Trial protocols choose one of the three broad levels of individualisation, ranging from none at all (all patients receiving the same treatment at all sessions), through partially individualised treatments (eg, use of a fixed set of points to be combined with a set of points to be used flexibly), to fully individualised treatment protocols within which each patient receives a unique and evolving diagnosis and treatment. Additionally, the practitioners may have to apply a standardised theoretical framework or may be allowed to apply their own. Many styles of acupuncture, whether based on traditional theories or Westernised concepts such as trigger points, are individualised in routine practice. Trials that are more pragmatic28 in their aim, and designed to replicate routine settings and patient groups, place more of an emphasis on fully individualised treatment. In such cases standardisation may consist of a protocol that instructs practitioners to provide treatments as they normally do. Trials that are more explanatory (mechanistic) in their aim tend to need a tighter definition of specific components in order to minimise variation across treatments.
It is recommended that the reporting of this item should include a total of needle insertions per subject per session. This item is relevant to all designs of randomised controlled trials, from pragmatic to explanatory. For more explanatory designs where a formula of points is prescribed, the number of needle insertions should be reported as a simple total. For more pragmatic designs, with individualised treatments, the mean and range should be reported. Clearly, full details of individualised treatment cannot be reported in every section of item 2 below. However, each item should be considered and as much information given as possible.
The point descriptions in the seminal classic texts, such as the Huangdi Neijing (Inner Canon of the Yellow Emperor) are rare and vague. The depiction of acupuncture points in relation to precise anatomical structures dates back only 100 years. Since the mid-1950s a process of standardisation has been taking place, and the acupuncture point descriptions based on anatomical locations and proportional cun measurement systems have served as a blueprint for many Western translations. It should be noted that these locations have not been universally adopted. Given this historical context, it remains important to know which acupuncture points have been used in clinical trials, with as accurate descriptions as possible of the location of these points, and, where relevant, the method used to identify the points.
The specific point locations used in treatments where standardised should be described by an accepted nomenclature (eg, GB21)35 or by an anatomical location where there is no accepted name. Whether the needles are inserted unilaterally or bilaterally should be stated. For protocols with partially individualised prescriptions, list any prescribed essential or optional points, and describe (in the Results section) both the points used at every visit, and all the points used on an ad hoc basis. If the list is extensive, the most commonly used points (with percentages) should be reported. Where protocols specify using fully individualised point prescriptions, authors should consider the best way to report the points used—for example, by listing all points across all subjects or by identifying the most commonly used points if the list is extensive.
The depth of insertion should be expressed using the Chinese measurement of the cun; as anatomical depth—for example, of subcutaneous tissue, fascia, muscle or periosteum; or in millimetres. For some trials, the protocol might specify the angle and direction of insertion together with depth of insertion, in which case these should also be reported.
If the study protocol requires that specific responses to needling be elicited—for example, the de qi sensation in traditional Chinese acupuncture, the muscle twitch in trigger point treatment or muscle contraction in electroacupuncture, these elicited responses should be reported. Where relevant, the authors should differentiate between the responses required in the protocol and those actually obtained (which should be reported in the Results section).
Needle stimulation techniques, where used, should be clearly described for all points. For manual stimulation, such techniques include lifting, thrusting or rotating the needle to manipulate the de qi sensation. For electrical stimulation, the current, amplitude and frequency settings should be recorded.
Needle retention times should be reported as either a standard or a mean and range. Authors should make it clear that they are reporting the time elapsed between the insertion and removal of needles (retention time) and distinguish it from treatment time, which may include other procedures such as history taking, discussion and preparation for treatment.
Details should be given of the types of needles used, including the diameter and length as well as the manufacturer and/or the material. This information is of importance since the effect of different metals or needle sizes on the body is not known. For trials using a variety of different types of needles, the ranges of diameters and lengths as well as types of material should be reported.
The planned number of sessions and frequency of treatment should be clearly documented. The actual number of treatments received by participants should be reported in the Results section. If there is variation between patients, then the mean and range should be reported.
The frequency and duration of sessions should be documented, with mean and range to be reported where there is variation across patients. Any variation in frequency of treatment (eg, if subjects are to be treated twice weekly in the first 2 weeks then once a week for the next 6 weeks) should be clearly reported.
Additional components of treatment refer to the auxiliary techniques, prescribed self-treatment and lifestyle advice provided by the practitioner. All additional components, whether carried out by the practitioner or patient and whether integral or adjunctive to the acupuncture needling, should be described clearly. For acupuncture-related interventions, such as moxibustion or cupping, detail should be provided equivalent to that recommended for acupuncture needling. If the protocol specifies the options of prescribed self-help treatments such as Qigong or muscle stretching exercises and/or lifestyle advice such as dietary changes based on acupuncture-related diagnostic criteria, then these too must be reported. The frequency with which the advice was given, and participants' compliance with this advice, should be reported. ‘Other components of treatment’ should be distinguished from ‘co-interventions’—that is, interventions that are provided additionally to both groups, which should be fully reported as described in STRICTA item (6b) below.
The setting and context of treatment can also provide important additional components to treatment.47 Context includes instructions to practitioners that might modify their normal practice—for example, prescribing or proscribing explanations to patients about their diagnosis. For patients, the context includes the information they have been given about the trial that might be expected to modify outcomes. Therefore, the information that the patient receives regarding the treatment and control intervention should be reported, including any relevant wording on consent forms and information leaflets designed to influence beliefs or expectations. For example, describing a sham acupuncture control as ‘a type of acupuncture’ may have a different effect on outcome than saying it is ‘not acupuncture, but will involve a similar experience to acupuncture’.
Characteristics of the acupuncturists providing treatment should be reported, including qualification or affiliation, years in acupuncture practice, as well as any other experience that may be relevant to the trial. Relevant differences (if any) in the qualification, training and experience of the participating acupuncturists should be highlighted. The recent survey of authors of acupuncture trials and reviews reinforced the need for these characteristics to be reported well,12 especially since the actual level of reporting has historically been poor.13 In trials where different acupuncturists provide treatment to different treatment arms, the background of both groups should be reported. The eligibility criteria for acupuncturists should be explained, as these will influence generalisability of the trial results. Where there are known to be potential variations between practitioners, selecting a random sample of practitioners will reduce expertise bias and help improve the applicability of the results.49
The rationale for choice of control or comparator should be presented and justified in relation to the research question and the methodology. In studies in which a group receiving acupuncture is compared with another group, the control or comparator can be sham acupuncture, usual care, an active treatment, a waiting list or no treatment. Whereas ‘control’ is sometimes used for a group that receives no intervention, the term ‘comparator’ may be more appropriate for an active intervention, such as physiotherapy, for which the intended action of the comparator is expected to be therapeutic. If using an acupuncture-like control in a participant-blinded trial then one of the following terms: active acupuncture control, penetrating needle control or non-penetrating sham needling control might be helpful descriptors. Control procedures involving invasive or non-invasive sham needling techniques may be therapeutically active, evoking neurophysiological and/or localised immune and circulatory responses. The extent to which sham acupuncture needling, whether penetrating or not, might elicit acupuncture-specific physiological mechanisms is not known, and is in part a consequence of our lack of knowledge of the mechanism(s) of true acupuncture. There are also variations in assumptions about the precision required for point location, as for some clinicians and investigators acupuncture points are considered as areas of reactivity rather than points of action. Such assumptions have a bearing on the integrity of the sham as an appropriate control. Some non-needling control procedures can be assumed to be physiologically inert, such as an inactivated transcutaneous electrical nerve stimulation machine; however, these procedures may not have the same total psychophysiological credibility as acupuncture, thereby compromising the interpretation of the results. Sources that led to the choice of control, such as literature or expert opinion, should also be reported and referenced. The author should reference prior work that supports the use of the selected comparator, such as from the conclusion of a systematic review or from another randomised controlled trial.
A precise description of the components of the control or comparator should be presented. If the control treatment is an acupuncture-like intervention, such as a form of sham acupuncture, then it should be specified whether the sham is invasive (penetrating the skin) or non-invasive (non-penetrating). The theoretical basis, needling details and regimen of an acupuncture-like control need to be reported in the same way as is set out in STRICTA items 1–3 above. The lack of a worldwide consensus on the location and size of acupuncture points reinforces the importance of accurate documentation of the sham points actually used, their precise location and the method used to locate them. If usual care or another active treatment is the comparator, all the components should be reported in full detail. This will enable readers to compare usual care as provided in the trial with what is usually provided to participants in another setting. Where usual care is also provided to those receiving acupuncture, these data will also allow readers to compare the intensity of usual care in the comparator arm with that of the experimental arm. If it is a waiting list arm, then the period of waiting needs to be specified. While precise description of the control or comparator is fairly straightforward in principle, the more complex the components, the more care is required to specify them precisely.
This revised STRICTA statement has been designed to help improve the reporting of interventions in clinical trials of acupuncture, with the intention that it will help authors of acupuncture trials provide readers with a clear, accurate and transparent account of their acupuncture protocols as well as their control and/or comparator procedures. In addition to revising the STRICTA checklist, we have improved the explanations of each item and provided examples of good reporting. To enhance awareness, endorsement and adherence, the revised STRICTA statement has been developed as an extension to CONSORT. Authors of clinical trials of acupuncture should use the STRICTA recommendations for the acupuncture intervention (item 5 in the CONSORT 2010 statement) in conjunction with the other 25 items of the checklist in the main CONSORT guidelines.10 11 The extension to CONSORT for non-pharmacological interventions is also highly relevant to acupuncture trials.14 15 There are other extensions to CONSORT that may be relevant, depending on the type of trial design, including extensions for cluster trials, equivalence and non-inferiority trials and pragmatic trials, and the reporting of abstracts and of harms (eg, adverse events) associated with the intervention. The most recent versions of all CONSORT guidance documents can be found on the CONSORT website (http://www.consort-statement.org, accessed 16 April 2010).
A complete, accurate and transparent trial report facilitates dissemination, interpretation, translation and replicability. There continues to be a need for better reporting generally, as has been highlighted in a recent study of what is missing from descriptions of treatments in trials and reviews.55 The authors found that elements of the intervention were missing in half of the published articles that they reviewed, giving insufficient detail—for example, with practitioners unable to use the treatments as described and researchers unable to replicate studies. This finding is similar to that from a review of authors of acupuncture trials.12 Improved reporting reduces reader ambiguity in interpretation, is likely to increase credibility and application of the results by providing better evidence on which to base decisions on patient care.
Reporting guidelines do help to improve the quality of reporting randomised trials,56 although it is difficult to observe their maximal benefit because too few journals endorse reporting guidelines57 and fewer adhere to them.57 To maximise this potential, we encourage journals to unambiguously endorse the revised STRICTA reporting guidelines. This can be most readily achieved by updating journal Instructions to Authors, thereby alerting prospective authors. In addition, we encourage journals to implement strategies to improve author adherence to reporting guidelines. These efforts might also help peer reviewers and journal editors in deliberating the merits of such trials.
The initial consultation with experts, which was piloted with the help of Mark Bovey, Val Hopwood and AW, involved a panel comprising Joyce Anastasi, Stephen Birch, Joao Bosco, Claudia Citkovitz, Remy Coeytaux, Misha Cohen, Agatha Colbert, Helen Elden, Reginaldo de Carvalho Silva Filho, Alastair Forbes, Nadine Foster, Joel Gagnier, Mark Goldby, Marita Gronlund, Richard Harris, Dominik Irnich, Helene Langevin, Lixing Lao, Anna Lee, Hyangsook Lee, Myeongsoo Lee, Sanghoon Lee, George Lewith, Klaus Linde, Jianping Liu, Ryan Milley, Scott Mist, Dieter Melchart, Albrecht Molsberger, Vitaly Napadow, Richard Niemtzow, Jongbae Park, Mahmood Saghaei, Koosnadi Saputra, Rosa Schnyer, Charles Shang, Karen Sherman, Byung-Cheul Shin, Caroline Smith, Elisabet Stener-Victorin, Kien Trinh, Jorge Vas, Andrew Vickers, Peter White, Claudia Witt, Hitoshi Yamashita and Christopher Zaslawski.
Support and administrative help in managing the process of revising STRICTA has come from Anne Burton, Ann Hopton, Suzanne Jenna, Stephanie Prady and Tracy Stuardi.
Members of the STRICTA Revision and Steering Groups The Steering Group comprised DA and DM (CONSORT), HM and RH (STRICTA) and YL and TW (Chinese Cochrane Centre). The STRICTA Revision Group, who participated in the consensus-building workshop in Freiburg, comprised the six members of the Steering Group and Stephen Birch, Isabelle Boutron, Mark Bovey, Yutong Fei, Joel Gagnier, Sally Hopewell, Val Hopwood, Susanne Jena, Klaus Linde, Jianping Liu, Kien Trinh, Emma Veitch, AW and Hitoshi Yamashita.
Funding The workshop in Freiburg was supported by the White Rose Health Innovation Partnership, Enterprise and Innovation Office, Charles Thackrah Building, 101 Clarendon Road, Leeds LS2 9LJ, Leeds, UK. HM is supported by a Career Scientist Award from the UK National Institute for Health Research. DGA is supported by Cancer Research UK. DM is supported by a University of Ottawa Research Chair.
Competing interests AW is employed by the British Medical Acupuncture Society as Editor of Acupuncture in Medicine.
Provenance and peer review Not commissioned; externally peer reviewed.
In order to encourage dissemination of the STRICTA criteria, this article is freely accessible on aim.bmj.com and will also be published in the Australian Journal of Chinese Medicine, Journal of Alternative and Complementary Medicine, Journal of Evidence Based Medicine, Medical Acupuncture, and Public Library of Science. The authors jointly hold the copyright of this article. For details on further use, see the CONSORT website (http://www.consort-statement.org/consort-statement/)
iIn the Examples that follow, the embedded references to figures, tables and sources that are presented are where the authors have reported them in the published studies, and therefore details of these data are not provided in this article.