How to Design the Control Group in Randomized Controlled Trials of Acupuncture?
Table 1
Acupuncture study control group designs.
Number
Categories
Method
Purpose
Advantage
Limitation
(1)
Nontreatment contrast
CG: not receiving any treatment; TG: acupuncture [40]
To assess the effects of acupuncture
Observe the progression of the condition and patient recovery
No blinding; placebo effects of acupuncture are not eliminated. This treatment would be contrary to medical ethics when treatment involves acute or severe conditions
(2)
Complementary contrast
CG: standard western medicine (with [41] or without [42] placebo acupuncture); TG: standard western medicine and acupuncture
To compare the effects of western medicine and western medicine plus acupuncture
Clarify the costs and side effects of acupuncture as the complementary therapy
Blinding is inadequate unless placebo acupuncture is used. Comparisons with standard treatments cannot be made, and the error may not be correct
(3)
Alternative contrast
CG: standard western medicine; TG: acupuncture [43]
To compare the effects of acupuncture and western medicine
Demonstrate the effectiveness of acupuncture as the alternative therapy and assess the costs and side effects
No blinding; placebo effects of acupuncture are not eliminated, resulting in the error
(4)
Nonacupuncture contrast
CG: nonpenetrating intervention, for example, TENS or laser acupuncture; TG: acupuncture [44, 45]
To assess the different effects of acupuncture, TENS, and laser acupuncture
Similar amounts of time and attention are spent on each group, thereby helping to eliminate some of the placebo effect
It is impossible to ensure blinding because of the substantial differences between the CG intervention and acupuncture. The CG intervention may have therapeutic effects. The effects of acupuncture are underestimated
(5)
No effects of nonacupuncture contrast
CG: mock nonpenetrating intervention, for example mock-TENS or mock laser acupuncture; TG: acupuncture [46]
To observe whether the therapeutic effects of acupuncture are greater than those of a nonpenetrating placebo intervention
If the TG receives TENS or laser acupuncture, elimination of the placebo effect and blinding can be assured. The CG intervention has no therapeutic effects and allows the observation of the progression of the condition and patients’ recovery
It is impossible to ensure blinding because of the substantial differences between the CG intervention and acupuncture
(6)
Sham acupuncture contrast
CG: insertion points are not acupoints or meridians; TG: acupuncture at acupoints [47, 48]
To compare the effects of acupoints with sham points
The needling methods are the same in the CG and TG, resulting in an optimal elimination of placebo effect, and blinding can be performed
The effects of acupuncture on pain are underestimated. Having no uniform protocol for the sham points precludes accurate comparison of the results and conditions. Participants with previous acupuncture experience may be conscious of the difference in acupoint site(s), which influences the blinding
(7)
Nonspecific sites contrast
CG: acupuncture at acupoints which are considered to produce no or only minor effects; TG: acupuncture at specific acupoints related to the objective illness [49, 50]
To compare the specific effects of acupuncture points
Needling methods are the same in the CG and TG, resulting in an optimal elimination of placebo effect, and blinding can be performed
The effects of acupuncture on pain are underestimated. Selecting an unsuitable acupoint in the CG would render the result invalid
(8)
Minimal acupuncture contrast*
CG: insertion points are not acupoints, and needle penetration is to a depth of less than 4 mm; TG: standard acupuncture [48, 51]
To assess the effects of acupuncture at acupoints with manipulating stimulation
The elimination of placebo effect, blinding is ensured; the procedure is easily manipulated
Potential therapeutic effects in the control group. To increase the efficacy of blinding, acupuncturists decreased the manipulation in the TG which then reduced the effects and confounded the results, and analysis. “De-chi” was not attained
(9)
Superficial acupuncture contrast
CG: dull needles or other tools (e.g., needle tube, toothpicks) are slapped on acupoints and tapped on them; dull needles do not prick the skin; TG: standard acupuncture [52]
To examine whether acupuncture is more effective than placebo acupuncture
The physiological reaction in the control group is minimal; the procedure is easily manipulated and suitably applied without requiring a novel study design; the blinding is effective
The operating locations were restricted to areas patients could not see, such as the neck, upper back, and dorsal side of limbs. The sensation was minimal and “de-chi” was not attained, which influenced the blinding for patients with previous experience of acupuncture; thus, the protocol cannot be used in long-term research
(10)
Placebo needle contrast
CG: Streitberger’s needle, Park Sham Device, or Takakura’s needle with a blunt tip was tapped onto the skin; TG: Streitberger’s needle, Park Sham Device, or Takakura’s needle with a real needle [17–20]
To observe whether real acupuncture is more effective than placebo acupuncture
The placebo effect and blinding are regarded as optimal, and, thus, the protocol has been widely used with good confidence
The major limitation of placebo needles is the associated lack of the “de-chi” sensation. Certain body sites cannot undergo acupuncture with these devices, such as the fingers, toes, and scalp, as well as sites that require transverse insertion or oblique insertion. These devices do not overcome the problem of double-blinding [33]. The needling methods in the TG had to be limited, thereby reducing the effects; thus, this procedure cannot be used in long-term research
(11)
Combined multiple methods contrast
(a)
CG: medicine and acupuncture placebo; TG: acupuncture and placebo medicine [53]
To reduce the psychological influence
The blinding and elimination of placebo effects were enforced to contrast the specific therapeutic effects between acupuncture and medicine
The procedure is difficult to use in long-term research
(b)
CG: placebo needles inserted into acupoints and real needles inserted into sham points; TG: real needles inserted into acupoints and placebo needles inserted into sham points [54]
To produce similar therapeutic experience in the 2 groups; promote blinding and eliminate placebo effects
Blinding was effective, and the nonspecific effect of placebo needles was reduced
Participants with previous acupuncture experience may be conscious of the difference in acupoint site(s), which affects the blinding
CG: Control group; TG: Test group; TENS: Transcutaneous electrical nerve stimulation; *Minimal acupuncture was termed superficial acupuncture in some studies [51].