Salusky 2005a.

Methods	Study titles Salusky 2005, Salusky 2005a and Salusky 2005b represent the same RCT. The study title (Salusky 2005a) has been used to allow the presentation of data for vitamin D groups + sevelamer (treatment groups 2 and 4) Study design: parallel RCT, 2 x 2 longitudinal factorial study design. Time frame: 2003 to 2005 Follow‐up period: 8 months Loss to follow‐up/excluded: 15%; 9/60 did not complete study; transplant (5), non‐compliance (4)
Participants	Country: USA Setting: University teaching hospital CCPD, PTH > 400 pg/mL; bone histomorphometry of secondary hyperparathyroidism Number (randomised/analysed): treatment group 2 (14/12); treatment group 4 (14/13) Mean age ± SD (years): treatment group 2 (14.5 ± 3.7); treatment group 4 (15.0 ± 2.2) Sex (M/F): treatment group 2 (7/7); treatment group 4 (4/10) Exclusion criteria: previous history of poor compliance; parathyroidectomy in last 12 months; immunosuppressive agents; rhGH
Interventions	Treatment group 2 Sevelamer: initial dose extrapolated from previous calcium carbonate doses; then titrated to keep P at 4 for 6 mg/dL Doxercalciferol: 3 times/week. Initial dose depended on PTH level, then titrated to keep PTH at 300 to 400 pg/mL and Ca 8.4 to 10.2 mg/dL Treatment duration: 8 months Treatment group 4 Sevelamer: initial dose extrapolated from previous calcium carbonate doses; then titrated to keep P at 4 to 6 mg/dL Calcitriol: 3 times/week. Initial dose depended on PTH level, then titrated to keep PTH at 300 to 400 pg/mL and Ca 8.4 to 10.2 mg/dL Treatment duration: 8 months Co‐interventions 1000 mg oral calcium in sevelamer group if Ca < 8.2 mg/dL
Outcomes	Outcomes for comparison between vitamin D preparations Bone formation rate Other bone histomorphometric parameters Final levels of PTH, Ca, P, FGF 23
Notes	Factorial analysis provided no evidence of treatment interaction between two sterols so comparisons reported for phosphate binders irrespective of D sterol given 2011 report included 60 allocated patients with data on 51 provided for 4 groups
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomised longitudinal factorial study. Computer generated randomisation
Allocation concealment (selection bias)	Low risk	Computer generated, allocated by statistician
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants, care givers, investigators not blinded to interventions
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Outcomes laboratory based and unlikely to be influenced by lack of blinding
Incomplete outcome data (attrition bias) All outcomes	High risk	9/60 (15%) did not complete study. Lack of data on these patients could have influenced results
Selective reporting (reporting bias)	High risk	Primary outcome was bone histology; secondary outcomes only reported graphically
Other bias	Low risk	USPH grants and Casey Lee Ball Foundation but Bone Care International provided medications and other unrestricted support for the study