Pieper 2006.

Methods	Study design: cross‐over study Time frame: 2002 to 2006 Follow‐up period: 20 weeks (2 weeks washout; 8 weeks treatment in each arm) Loss to follow‐up/excluded: efficacy 55%; 22/40. 40 randomised; 34 treated (safety analysis); 30 completed first part; 23 entered cross‐over; 18 completed cross‐over so 22 excluded from efficacy analysis
Participants	Country: Germany Setting: national, multicentre; university teaching hospitals Maintenance PD or HD; eGFR 20 to 60 mL/min/1.7 m2; Ca < 2.75 mmol/L; PTH ≤ 500 pg/mL Number Treatment group: sevelamer first (17); sevelamer second (13), analysed (9) Control group: calcium acetate first (17), calcium acetate second (10), analysed (9) Mean age ± SD: 12.4 ± 4.1 years (efficacy group completing study) Sex (M/F): 13/5 (efficacy group completing study) Exclusion criteria: PTH > 500, Ca > 2.75; cyclosporin use; antiarrhythmic agents; anticonvulsants; pregnant; lactating; difficulty swallowing; intestinal motility disorder or substantial surgery
Interventions	Treatment group Sevelamer to keep P < 2mmol/L (≥ 2 years) or < 2.25 (> 2 years) Starting dose equal to dose administered before study. 8 weeks Control group Calcium acetate to keep phosphate < 2 mmol/L (≥ 2 years) or < 2.25 mmol/L (< 2 years) Starting dose equal to dose administered before study. 8 weeks Co‐interventions Dialysis; vitamin D
Outcomes	Efficacy (18 patients): change in P, Ca, Ca x P, PTH Safety (34 patients): Ca > 2.75
Notes	Data combined for 2 arms of cross‐over for efficacy and safety
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer generated random numbers with prior allocation to each centre
Allocation concealment (selection bias)	Low risk	Computer generated random numbers with prior allocation to each centre
Blinding of participants and personnel (performance bias) All outcomes	High risk	Open label study. Lack of blinding could influence patient management
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Outcomes are laboratory based and unlikely to be influenced by lack of blinding
Incomplete outcome data (attrition bias) All outcomes	High risk	Only 18/40 (45%) completed both parts of the cross‐over study. 30 completed first part. Exclusions could have influenced overall result
Selective reporting (reporting bias)	High risk	Outcomes reported incompletely and cannot be included in meta‐analyses
Other bias	High risk	Chief investigator and study supported by Genzyme Europe