ACURATE: A guide for reporting sham controls in trials using acupuncture

To promote better reporting quality regarding sham acupuncture in clinical trials for a precise appraisal of the adequacy of the sham acupuncture procedure.

causing stimulating effects rather than remaining inactive. In the case of nonpharmacological interventions, the enigma of a sham intervention becomes more challenging to solve as tactile, or sometimes painful, stimulation is an essential and inevitable component of the treatment process. 1,2,21,22 Acupuncture is a paramount example of invasive, nonpharmacological intervention. Due to its preparation and application procedure, the duration of a single acupuncture treatment session is often longer and more invasive than that of pharmacological treatments. Treatment procedures using sham needles, for example, the Streitberger, 23 the Park, 24 the Takakura, 25 are identical to that of real acupuncture except for the blunt needle that is designed to not penetrate the skin. [26][27][28] Previous studies show that 2,3,22,29 multisensory stimulations or nonspecific effects induced during sham acupuncture procedures may enhance or abate the overall therapeutic effect. 29,30 Furthermore, quantitative components such as dose, duration, and treatment frequency influence stimulation intensity. 31,32 Since sham needling involves minimal stimulation of the skin, 2 it is possible that the quantitative components described above may influence the level of overall effect induced by sham acupuncture. A reassessment of the components of the sham acupuncture procedure and related nonspecific effects would further allow a precise appraisal of the effectiveness of acupuncture. 1,33 The information regarding sham needles in many of the acupuncture studies was shown to be insufficient despite the importance of a thorough appraisal of their components. Our previous study reported the reporting quality of sham acupuncture since 2009, 34 in an approach that is similar to the assessments conducted on clinical trials using acupuncture. 35,36 It showed that the treatment procedure was insufficiently reported in general and also that the reporting quality has not improved over the years. This finding is in contrast to previous studies on the reporting quality of acupuncture clinical trials, which showed that the reporting quality improved, especially since the development of STRICTA, 35 when both real and sham acupuncture needles are used in the study.
Its final objective is a comprehensive reporting of sham acupuncture procedure to allow replicability and, ultimately a thorough evaluation of potential factors which may enhance or moderate the placebo effect of acupuncture.

METHODS
The process of developing ACURATE followed the guidance pro- Informed consent was obtained online from the participants of the Delphi process.

Literature review and development of initial items
Background research was conducted on the reporting quality of previous studies using sham needles. 34 An initial list of potential items was drafted based on this study. Among the 25 items, twenty items were rephrased from relevant checklists including CONSORT, 37,38 STRICTA, 40,41 TIDieR, 42 and TIDieR-Placebo. 43 Additional five items were added based on literature addressing the contextual factors of the placebo effect, that is, the interaction between the patient and the practitioner during applying sham acupuncture. 17,18,44,45 The process of developing the initial items of the checklist is described in detail in our previous study. 34 Due to the low reporting rates of the items and the lack of improvement over the years, all of the items were included in the Delphi survey.

Delphi survey
For a convergence of opinions and consultations from experts in the field of acupuncture research on the checklist items, a three-stage Delphi survey process was conducted. 46 A group of experts from the field of clinical trials using acupuncture, physiotherapy, psychological interventions, mechanisms underlying acupuncture, statistics, methodologies of clinical trials, and placebo effects was convened for this process. An online survey was sent to the experts with a list of items, in which the experts were asked to rate from 1 (not important at all) to 5 (very important). Additional comments were invited in a free-text format for each item as well as for each category and the overall survey. In the following rounds of Delphi survey, the experts were asked to rate in a similar manner the items that reached less than 80% of the whole consensus. Additional rounds were planned in case of which the items did not reach the required consensus. The working group concluded with F I G U R E 1 Detailed timeline of the Delphi survey.
an online meeting to confirm the final checklist. The detailed process and timeline of the Delphi survey is illustrated in Figure 1.

Development of the initial checklist
The

Delphi survey
Throughout the Delphi survey, a total of 35 participants were invited to complete an online survey, and 66% (n = 23) responded (Table 1).
After the first round, three items did not reach an 80% consensus; after two following rounds of revision and additional Delphi survey, the consensus over two items came close to, but still failed to reach 80%. These two items were: "5b. Report whether the practitioner and the patient had any conversation prior to the treatment" and "5c.
Report if there was any conversation during the treatment" from the initial checklist. The comments from the experts implied that keeping both items might be verbose; furthermore, some experts argued that the range of information garnered from the items might be too heterogeneous as "any conversation" could lead to a number of different scenarios. Based on the comments from the experts and further discussion within the working group, the two items were reduced into one: "Report whether conversations directly linked to the trial design were allowed between practitioner and patient other than scripted instructions and preset information prior to and during the treatment." Additional changes made to the checklist included the term "rationale" to "reason" for item "3c. Report the rationale for the location chosen for placebo acupuncture" from the initial checklist and the term "placebo acupuncture" to "sham acupuncture" throughout the checklist. On the basis of the working group's feedback, the item 6a and 6b in the original checklist were consolidated into a single item. Supplementary Table S2 describes the revision process of the items on the checklist after each Delphi survey.

Scope of ACURATE checklist
The purpose of the ACURATE checklist is to guide authors in describing the use and application of sham controls in clinical acupuncture trials in detail to allow replicability and appropriate assessment of the effectiveness of acupuncture. The scope of the checklist applies to all types of study designs using sham acupuncture, including clinical trials, case-control studies, and experiments using sham needles as their stimulation apparatus.

The ACURATE checklist description and examples
The final checklist includes 23 items in 6 categories ( In electroacupuncture (EA) studies, real acupuncture needles may be used for controls without electric current. In laser acupuncture studies, the same machine may be used in the control group while deactivated.
Some researchers may use their own inventions as sham controls.
While the above examples fall within the scope of sham acupuncture, the level of stimulation may vary across the types of sham devices.
Specifying the type of sham acupuncture and the reason for the chosen sham needle is recommended to enable an understanding of the type and amount of stimulation that study participants will experience and to which the effects experienced by the treatment group will be compared.

ACURATE Category 2. Details of sham acupuncture manipulation
The details of sham acupuncture manipulation may vary from that of real acupuncture due to limitations arising from its physical characteristics. For instance, the number of needles, insertion depth if penetrated, [47][48][49] and collected responses may differ from those of real needles due to difficulties in applying blunt tips. 50 The type of manipulation and retention time may also vary depending on the studies.
Rotation and retention of sham needles may be difficult due to weak adherence to the skin. If the study design is pragmatic, sham acupuncture group may receive other interventions along with the treatment.
Descriptions on the details of sham needle manipulation would help readers recognize the above differences.

ACURATE Category 3. Location of sham acupuncture
Contrary to the location of real acupuncture treatment, which is com-  2b Report the depth of sham acupuncture insertion (if there was no penetration, state this within the paper).
2c Report whether any response was observed during sham acupuncture manipulation (e.g., de qi or muscle twitch response). 2d Report if there was any stimulation using sham acupuncture. 2e Report if there was sham acupuncture retention.
2f Report details of other interventions administered in addition to sham acupuncture during one session.

Location of sham acupuncture 3a
Report the location of sham acupuncture (e.g., acupoint/nonacupoint or the exact location of the sites). 3b Explicitly state in the paper if the points are unilateral or bilateral.
3c Describe the reason for the chosen location of sham acupuncture.

Treatment regimen 4a
Report the number of treatment sessions.
4b Report whether the number of sessions were identical between real and sham acupuncture treatments.
4c Report the frequency and duration of treatment sessions.
4d Report the total trial period.

Practitioner 5a
Report whether the same practitioner is administering both real and control treatments (interventions).
5b Report whether there were conversations between practitioner and patient directly linked to the trial design, other than scripted instructions and preset information, prior to and during the treatment.

Protocol and settings 6a
Report the information regarding sham acupuncture provided to participants.
6b Report whether the information given to patients include the term to openly state that the control is inert (e.g., "fake," "sham," "dummy," "placebo," . . . ). 6c Describe how sham device was blinded from patients, and if done, how the blinding was assessed. 6d If done, report any modification in the sham acupuncture treatment procedure, and reason for the modification.
6e Report any difference in the treatment settings between real and sham acupuncture.
is just as necessary as that of real acupuncture since its details are likely to provide the context of "treatment procedures" to the patient, which will eventually influence the nonspecific effect induced by sham treatment.

ACURATE Category 5. Practitioner
Clinical trials often involve multiple practitioners or physicians admin-

ACURATE Category 6. Protocol and settings
The category of protocol and settings covers a substantial amount of the "treatment context" induced by sham acupuncture treatment: trial information, use of specific terms, blinding, and modifications in the treatment protocol or other noticeable changes in the treatment settings. Each component creates cues that patients construe throughout the trial, some of which imply similarities of real and sham acupuncture (e.g., modification to further mimic real acupuncture) and the others imply uncertainty in the expected sensation (e.g., use of the phrase "may or may not feel the sensation from acupuncture"). Consequently, the potential effect of context on trial outcomes may change due to the expectation and expectancy of the patient formulated thereafter.
Each procedure described above supports participant blinding, implying that the participant will likely have difficulties in identifying what they received. Sufficient blinding will ensure similarities in the context between real and sham groups, and subsequently higher credibility of the measured outcomes. Reporting blinding index is recommended from this perspective to help demonstrate the level of which patients found the needles indistinguishable. 16,26,51 It is noteworthy that the quality of the trial does not necessarily depend on the success of blinding, especially in trials where the main objective is to measure the specific therapeutic effect of real acupuncture. Blinding index, if done, is a useful tool in understanding the trial but not always necessary in understanding the effectiveness of the targeted treatment.

DISCUSSION
The ACURATE guideline was developed to promote better reporting of the aspects of clinical trials related to sham acupuncture. The final 22 items developed cover the application procedure of sham acupuncture needling in clinical trials, using the following six categories: types of sham acupuncture, details of sham acupuncture manipulation, location of sham acupuncture, treatment regimen, practitioner, and protocol and settings. The intention is that this guideline will have a positive impact on reporting the quality of trials using sham acupuncture in ways similar to CONSORT 52 and STRICTA. 40 It is hoped that the rigor in reporting the components of sham acupuncture is not overlooked through this guideline. From this effort, a clearer demonstration of sham acupuncture application and an understanding of the scope of diversity in sham acupuncture procedures will be made possible.
Literature over the recent years discusses the impact of the placebo effect on treatment outcomes, much of which focuses on components of a sham intervention or an inert pill that are not "active" or "specific" but nonetheless have an influence on the overall therapeutic effect. 5,11,12,18,33 For instance, the effectiveness of acupuncture may be partially attributable to components also present in sham needles. 1,2,27,28,30 While discussing the potentially "active" components of nonpharmacological interventions is beyond the scope of this study, 53 the Delphi results and the final items included in the ACU-RATE suggest that a number of aspects related to sham acupuncture procedure need a thorough consideration prior to interpretation of its "placebo" effect.
The findings of this study and our previous study 34 suggest that a broader understanding of the definition of sham acupuncture is necessary. Sham acupuncture can be defined as a minimally invasive intervention which allows accumulation or reduction of its duration and intensity. It is through this minimal effect that signaling the patient as having received real acupuncture is possible. Although further research is required for sham acupuncture using devices such as transcutaneous electrical nerve stimulation (TENS) or laser acupuncture devices, the potential specific and nonspecific effects of any sham acupuncture involving needles can be understood from this perspective.
A number of limitations exist in this study. Although a relatively high response rate has been reached in the Delphi survey, an external review will help ensure its usability; therefore, further studies are being planned on the implementation and dissemination of ACURATE, which will include external reviews and subsequent revision of the checklist. In addition, the current Delphi included an expert group of which most members were from universities. This may need to be more representative of the overall opinions, and further studies are required to include professionals from other types of institutions such as hospitals and private medical clinics.
Using the ACURATE in not only clinical trials but also in systematic reviews may help readers understand the diversity of sham acupuncture procedure included in the review and consequently the possibility of different placebo effect evoked in each study. Providing information on the components of sham acupuncture in individual trials which remained otherwise vague would lead to a more precise appraisal and analysis of its effect. When evaluating the quality of reporting of sham acupuncture, we recommend that at least two independent reviewers use this checklist to assess the publications, with the help of a third reviewer, if consensus is not reached.
Lastly, ad hoc modifications of the guideline are not recommended as with the CONSORT and the STRICTA. Instead, we invite the researchers to provide feedback and comments on the ACURATE, as this guideline will be updated based on the reported publications and direct feedback to the authors. The contributions will be reflected in the following meetings and the updates of the ACURATE.

CONCLUSION
This paper presents the ACURATE checklist, an extension of the CON-SORT and to be used along the STRICTA when sham acupuncture needles are used in the study. Previous study indicated low reporting quality of sham acupuncture procedures from a literature review. 34 This checklist focuses on a clear depiction of sham needling procedures to endorse a transparent description of its components in the future studies and to enable a precise assessment of the effects. Doing so will allow a deeper understanding on the level of stimulation from sham needling and the effects of context from its treatment regimen.