First author, year, country
Study funding source
Treatment group
Control group
Core inclusion criteria
Number of patients randomized/analyzed
Mean age (years)
Females (%)
Primary outcomes
Key secondary outcomes
Behavioral approaches
Büttner, 1991, Germany [6]
NR
Biofeedback three times per week for 5 weeks
Hand exercises three times per week for 5 weeks (placebo intervention)
RP
Tx: NR/10
35–59
NR
Number attacks pre-post
Duration of attacks
Tx = 4.8–3.4
Tx = reduction from 17 to 15 min
Control: NR/10
Control 3.9–3.1
Control = reduction from 28 to 21 min
P = NS
P = < 0.05 for difference between groups before and after intervention
Freedman, 1983, USA [18]
NHLBI
1. Temperature biofeedback in 10 biweekly sessions
Sham frontalis EMG bio-feedback in 10 biweekly sessions
Primary RP
Tx (1): NR/8
20–65
Overall: 88
Decrease in attack frequency
None
Tx (2): NR/8
Tx (3): NR/8
Control: NR/8
Tx (1): 93 %
2. Temperature biofeedback under cold stress in 10 biweekly sessions
Tx (2): 67 %
Tx (3): 33 %
3. Autogenic training in ten biweekly sessions
Control: 17 %
P < 0.05 for 1 and 2, but not 3, versus control
Freedman, 1984, USA [17]
NHLBI
1. Temperature biofeedback in ten biweekly sessions
Sham frontalis EMG biofeedback in ten biweekly sessions
RP secondary to systemic sclerosis
Tx (1): NR/8
NR
Overall: 92
Decrease in attack frequency
None
Tx (2): NR/8
No significant differences
Control: NR/8
2. Autogenic training in ten biweekly sessions
Guglielmi, 1982, USA [26]
Rehabilitation Services Administration
Skin temperature biofeedback
1. Sham biofeedback (EMG relaxation of forehead muscles)
RP, not taking vasoactive medication or medication for Raynaud’s
Tx: 12/12
Tx: 33
NR
Total attacks in 5 months
None
Tx: 225
Control (1): 204
Control (1): 34
Control (2): 254
Control (1): 12/12
2. No treatment
P = NS
Control (2): 34
Control (2): 15/12
Raynaud’s Treatment Study Investigators, 2000, USA [57]
NIH
Temperature biofeedback with suggestions for warmth imagery and passive volition delivered in ten 1-h sessions over 5–10 weeksa
Sham (frontalis muscle
Surface EMG biofeedback) delivered in ten 1-h sessions over 5-10 weeksa
Primary RP with ≥2 attacks per day during previous cold season
Tx: 81/81
Tx: 44
Overall: 70b
Daily attacks at 1 year c
Daily attacks at 2 months c
Tx = 0.23
Tx = 0.12
Control: 74/74
Control: 46
Control = 0.16
Control = 0.15
P = 0.38
P = 0.57
Sporbeck, 2012, Germany [63]
Physiomed Elektromedizin AG
Biofeedback three times per week for 4 weeks
Waiting list
RP secondary to systemic sclerosis
Tx: NR/8
Tx: 50
Tx: 88
Scleroderma VAS RP item at 4 weeks
Scleroderma VAS RP item at 12 weeks
Tx = NR
Tx = NR
Control: NR/10
Control: 59
Control: 90
Control = NR
Control = NR
P = 0.021 (favors Tx)
P = 0.093 (favors Tx)
Surwit, 1978, USA [67]
NIMH
Autogenic relaxation training or autogenic relaxation training with biofeedback during six biweekly sessions
Waiting list
Primary RP
Tx: NRd/15
23–54
100
Number attacks pre-post
Attack intensity
Control: NRd/15
Greater reduction in Tx group
Tx = 32 % reduction
P = NS
Control = 10 % reduction
P = NS
Other approaches
Al-Awami, 2004, Austria [1]
NR
Low level laser therapy every other day for ten sessions total
Sham level laser therapy every other day for ten sessions total
RP for 2 years or more with average of ≥4 episodes per week, age 18–65
Tx: NR/24
Median age:
Tx: 67
Daily attacks at 6 weeks
Daily attacks at 3 months
Control: 91
Control: NR/23
Tx: 45
Tx = 3
Tx = 3
Control: 46
Control = 5
Control = 6
P = 0.007
P = 0.02
Severity attacks (0–10) at 6 weeks
Severity attacks (0–10) at 3 months
Tx = 1
Tx = 0
Control = 4
Control = 4
P = 0.02
P = 0.04
Appiah, 1997, Germany [2]
NR
Acupuncture with seven treatments over 2 weeks
No treatment
Age 18–60 years, primary RP
Tx: NR/17
Tx: 46
Tx: 71
Daily attacks e
None
Tx = reduction from 1.4 to 0.6
Control: NR/16
Control: 42
Control: 69
Control = reduction from 1.6 to 1.2
P = NS
Hahn, 2004, Germany [27]
NR
Acupuncture weekly for 8 weeks
Sham acupuncture weekly for 8 weeks
Secondary RP
Tx: NR/11
Tx: 47
Tx: 91
Daily attacks
Duration attacks
Tx = reduction from 15 to 12 min
Tx = reduction from 1.9 to 1.4
Control = reduction from 31 to 16 min
P = NS
Control: NR/8
Control: 41
Control: 75
Control = reduction from 2.8 to 1.8
Severity attacks (0–5)
Tx = reduction from 2.6 to 2.1
P = NS
Control = reduction from 3.0 to 3.0
P = NS
Hirschl, 2002, Austria [33]
NR
Low level laser therapy for 30–40 min in five sessions per week for 3 weeks
Sham laser therapy for 30–40 min in five sessions per week for 3 weeks
Primary RP
Crossover design: 18 (15)
53
80
Daily attacks (1–5)f
None
Tx = 0.67
Control = 0.72
P = 0.520g
Hirschl, 2004, Austria [34]
NR
low level laser therapy for 30–40 min in five sessions per week for 3 weeks
sham laser therapy for 30–40 min in five sessions per week for 3 weeks
primary RP and not taking vasoactive medication
Crossover design: 50 (48)
46
79
Intensity of attacks (1–5)f
Daily attacks f
Tx = 2.3
Tx = 1.6
Control = 2.8
Control = 2.0
P < 0.001g
P = 0.001g
Ko, 2002, Canada [43]
Thermoflow
Ceramic-impregnated gloves for 3 months
Placebo gloves for 3 months
Age ≥ 18 years, “Pal” questionnaire criteria for RP
Tx: 49/30
Tx: 52
Tx: 67
Pain VAS e
None
Tx = 50.8
Control: 44/30h
Control = 57.9
Control: 54
Control: 87
Sporbeck, 2012, Germany [63]
Physiomed Elektromedizin
Deep oscillation three times per week for 4 weeks
Waiting list
RP secondary to systemic sclerosis
Tx: NR/10
Tx: 53
Tx: 80
Scleroderma VAS RP item at 4 weeks
Scleroderma VAS RP item at 12 weeks
Tx = NR
Tx = NR
Control: NR/10
Control: 59
Control: 90
Control = NR
Control = NR
P = 0.055 (favors Tx)
P = 0.081 (favors Tx)
Table 19.2
Assessment of risk of bias in randomized controlled trials of non-drug treatments for Raynaud’s phenomenon
Trial, year, country | Cochrane risk of bias tool domains | |||||||
---|---|---|---|---|---|---|---|---|
Random sequence generation | Allocation concealment | Blinding of participants and personnel | Blinding of outcome assessmenta | Incomplete outcome data | Selective outcome reporting | Study funding and author–industry financial tiesa | Other sources of bias | |
Behavioral approaches | ||||||||
Büttner, 1991, Germany [6] | Unclear risk | Unclear risk | Low risk | Low risk | Unclear risk | Unclear risk | Low risk | Low risk |
Freedman, 1983, USA [18] | Unclear risk | Unclear risk | Low risk | Low risk | Unclear risk | Unclear risk | Low risk | Low risk |
Freedman, 1984, USA [17] | Unclear risk | Unclear risk | Low risk | Low risk | Unclear risk | Unclear risk | Low risk | Low risk |
Guglielmi, 1982, USA [26] | Unclear risk | Unclear risk | Low risk | Low risk | Low risk | Low risk | Low risk | Low risk |
Raynaud’s Treatment Study Investigators 2000, USA [57] | Low risk | Low risk | Unclear risk | Unclear risk | Low risk | Low risk | Low risk | Low risk |
Sporbeck, 2012, Germany [63] | Unclear risk | Unclear risk | High risk | High risk | Unclear risk | Unclear risk | Low risk | Low risk |
Surwit, 1978, USA [67] | Unclear risk | Unclear risk | High risk | High risk | Low risk | Unclear risk | Low risk | Low risk |
Other approaches | ||||||||
Al-Awami, 2004, Austria [1] | Low risk | Low risk | Low risk | Low risk | Unclear risk | Unclear risk | Unclear risk | Low risk |
Appiah, 1997, Germany [2] | Unclear risk | Unclear risk | High risk | High Risk | Unclear risk | Unclear risk | Low risk | Low risk |
Hahn, 2004, Germany [27] | Unclear risk | Unclear risk | Unclear risk | Low risk | Unclear risk | Unclear risk | Low risk | Low risk |
Hirschl, 2002, Austria [33] | Unclear risk | Unclear risk | Low risk | Low risk | Low risk | Unclear risk | Unclear risk | Low risk |
Hirschl, 2004, Austria [34] | Unclear risk | Unclear risk | Low risk | Low risk | Low risk | Unclear risk | Unclear risk | Low risk |
Ko, 2002, Canada [43] | Low risk | Unclear risk | Low risk | Low risk | Unclear risk | Unclear risk | High risk | Low risk |
Sporbeck, 2012, Germany [63] | Unclear risk | Unclear risk | High risk | High risk | Unclear risk | Unclear risk | High risk | Low risk |
Our descriptions of RCTs are focused on outcomes directly relevant to RP attacks, and risk of bias ratings done using the Cochrane Risk of Bias tool, a standard evaluation system for RCTs [32]. The Cochrane Risk of Bias tool is used to evaluate the design and execution of individual RCTs and identify factors that could influence the validity of published results from RCTs. Domains of the Risk of Bias tool include randomization sequence generation; allocation concealment; blinding of participants, personnel, and outcome assessors; handling of incomplete outcome data; selective outcome reporting, and other possible sources of bias. Based on evidence that study sponsorship and author–industry financial ties are associated with outcomes [59, 60], we evaluated the risk of bias from this factor as well. See Table 19.3 for descriptions of Risk of Bias domains. In addition to bias, in evaluating RCTs, we considered the sample size of the RCTs. It is increasingly well documented that results from small, substantially underpowered trials are highly imprecise. Due to publication biases, very small trials that are published often report large and statistically significant effects that are false positives and do not replicate in larger, well-conducted studies, whereas very small studies with negative findings are typically not published [38, 39].
Table 19.3
The Cochrane tool for assessing risk of bias