Table 1

Characteristics of index meta-analyses (n=31)

Characteristicsn (%)*
Review protocol status
 Published in the Cochrane Database of Systematic Reviews12 (39)
 Unavailable or no mention that a review protocol was used19 (61)
Clinical condition
 Rheumatoid arthritis or osteoarthritis15 (48)
 Depressive or anxiety disorder16 (52)
 Number of trials in the meta-analysis, median (IQR)6 (3, 10)
Outcome domain
 Depression12 (39)
 Pain8 (26)
 Function2 (6)
 Swollen joint count2 (6)
 Other (anxiety, obsessive compulsive symptoms, aerobic capacity, fatigue, physical activity, quality of life, range of motion)7 (23)
Outcome label
 Primary16 (52)
 Non-primary (secondary or not labelled)15 (48)
Effect measure
 Mean difference4 (13)
 Standardised mean difference27 (87)
Meta-analysis model
 Fixed-effects5 (16)
 Random-effects24 (77)
 Not reported†2 (6)
Type of comparison
 Placebo/no intervention controlled comparison29 (94)
 Head-to-head comparison2 (6)
Type of active intervention
 Pharmacological12 (39)
 Non-pharmacological19 (61)
Source of funding for systematic review
 Pharmaceutical industry2 (6)
 Non-industry (governmental agency or other not-for-profit organisation)11 (35)
 No funding9 (29)
 Not reported9 (29)
Meta-analysis specification
 RA/OA meta-analysesn=15
 Defined by scale
   Yes (eg, Health Assessment Questionnaire score)3 (20)
   No (eg, disability)12 (80)
  Defined by time point
   Yes (eg, pain at 6 weeks)4 (27)
   No (eg, pain)11 (73)
 Depressive/anxiety disorder meta-analysesn=16
  Defined by scale
   Yes (eg, Beck Depression Inventory score)4 (25)
   No (eg, depression score)12 (75)
  Defined by time point
   Yes (eg, anxiety at 3 months)9 (56)
   No (eg, anxiety)7 (44)
  • *All values given as n (%) except where indicated.

  • †Meta-analysis model not stated and unclear because both a fixed-effects and random-effects model produced the same SMD and 95% CI.

  • OA, osteoarthritis; RA, rheumatoid arthritis; SMD, standardised mean difference.