Powell 2005.

Study characteristics
Methods	Randomised controlled three‐arm parallel trial
Participants	48 children with JIA, and persistent lower limb pain were enrolled (sample size calculation not stated) Age: 5 to 19 years (mean 12 years, 7 months, SD 3.7) Gender: 30 F, 10 M Setting; outpatients, Southern Californian children's hospitals, Location: San Diego, USA Inclusion criteria: diagnosed juvenile chronic arthritis; over 5 years old, active foot disease, history of foot pain over 1 month, able to walk 50 metres, stable medications Exclusion criteria: previous use of shoe inserts, joint injections in last 6 months, osseous anomaly Baseline characteristics: fewer males in CFO group (13.3% versus 33.3% in PFO, 30.8% in shoes); more polyarthritis in CFO group (73.3% versus 50% in PFOs, and 30.8% in shoes); CFO group had more pain on VAS (5.23) versus PFO (VAS 3.5), versus shoe (VAS 4.7)
Interventions	1: custom‐made semi‐rigid foot orthoses (CFO) 2: prefabricated foot orthoses (PFO) 3: new athletic footwear with a medial longitudinal arch and shock‐absorbing sole
Outcomes	Outcomes measured at baseline and 3 months Paediatric Pain questionnaire: 0 to 10‐point VAS scale (0 = no pain) Timed walking speed Foot Function index: 3 subscales ‐ activity limitation, foot pain, disability, 0 to 100‐point scale (0 = best function) Physical functioning subscale of PedsQL, 0 to 100‐point scale (0 = worst function)
Source of funding	Foot orthoses supplied by Langer Biomechanics, Neoprene insoles supplied by Spenco Medical
Notes	Pain reported in the trial using two instruments (Pediatric pain questionnaire and foot pain subscale of the Foot Function index) For this review, we extracted data using the Pediatric pain questionnaire for the analysis of pain
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Method of random sequence generation unclear. ‘Once accepted into the study, each subject was randomly placed…’ Insufficient information about the sequence generation process to permit judgement of low risk or high risk.
Allocation concealment (selection bias)	Low risk	‘Sealed envelope containing a predetermined numbered placement card into one of the 3 intervention groups’
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding and the outcome may be influenced by lack of blinding Physical therapist 1 who administered the interventions was not blinded Physical therapist 2 who performed baseline and follow‐up measures was blinded
Blinding of outcome assessment (detection bias) ‐ self‐reported outcomes (e.g., pain, function)	High risk	Pain, level of disability, and activity limitation measured by participants
Blinding of outcome assessment (detection bias) ‐ objective outcomes	High risk	Speed of ambulation influenced by participant and measured by personnel The outcome measurement may be influenced by lack of blinding
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Withdrawals few and similar between groups No child withdrew from the study because of discomfort or lack of efficacy. Children who withdrew were not different from children who completed the study in terms of parental education level, family income, race/ethnicity, or child’s age, gender, or type of arthritis.’
Selective reporting (reporting bias)	Unclear risk	No reference to study protocol. Insufficient information to permit judgement of low risk or high risk.
Other bias	Unclear risk	Orthoses and shoe inserts reported to be donated by medical products companies. No assurance was made of their independence from the research design, conduct, or report.