Table 4.

Quality Assessment of RCT of MSM and DMSO randomized clinical trials.

Study, year (reference)	Baseline characteristics presented?	Concealement of allocation	Reported to be double blind	Adequate blinding of patients	Adequate blinding of researchers	Consort Diagram reported	Withdrawal rate in the intervention group	Withdrawal rate in the control group	Power Calculation/ Statistical analysis	Intention to Treat analysis performed	Method to handle missing data	JADAD Score
Bookman et al., 2004 [36]	YesNo significant group difference for demographics, baseline knee pain, radiographic status or compliance	Yes	No, but clear from methodology this was a double blind study	Yes	Yes	Yes	18% (N = 14/80)	18% (N = 15/84)	N = 40 per group plus 10 for drop outs, to detect a difference of 3, 4, or 5 units in WOMAC pain 80% power and α = 0.05 ITT analysis. Primary outcomes: analysis of covariance	Yes	LOCF Except for missing baseline data which were replaced with day 1 scores	Score = 5 Blinding = 2 Randomization = 2 Withdrawal = 1
Eberhardt et al., 1995 [56]	Yes No significant group differences at baseline for symptom duration, age, height, weight, length of activation of current symptoms	No details reported	Yes	Yes	No details reported	No	5% (N = 3/56)	2% (N = 1/56)	No ITT Primary outcome: ANOVA, other outcomes: contingency board method and t-test	Yes	LOCF	Score = 4 Blinding = 2 Randomization = 1 Withdrawal = 1
Kim et al., 2006 [43]	Yes No significant group differences at baseline for sex, age, height, wgt, and symptom duration, NSAID use, MSM or DMSO use, ACR classification, baseline VAS scores, PGA or PhGA scores, and radiological stage	Yes	Yes	Yes	Yes	Yes	16%  (N = 4/25)	24%   (N = 6/25)	Yes ITT analysis Group differences in baseline to week 12 by T test	Yes	Not reported	Score = 5 Blinding = 2 Randomization = 2 Withdrawal = 1

LOCF, last observation carried forward.