MDRD Study 2 1989.

Methods	Study design: parallel RCT Duration of study duration: 1 January 1989 to 31 December 1992 Study follow‐up period: 2.2 years (mean)
Participants	Country: USA Setting: Renal outpatient clinics; 15 centres Relevant health status: GFR < 30 mL/min/1.73 m2; aged 18 to 70 years; mean arterial BP < 125 mmHg; dietary protein > 0.9 g/kg/d Number: very low protein diet group (126); low protein diet group (129) Mean age: 52 years Sex M/F ratio: 0.6 Exclusion criteria: pregnancy; body weight < 80% or > 160% of standard body weight; DM requiring insulin; kidney transplant; chronic medical conditions
Interventions	Very low protein diet group 0.28 g protein/kg/d with oral keto acid and essential amino acid supplement (nitrogen content 28.8 mg/kg/d) Calculated protein intake: 0.4 g/kg/d (taken from Fig 1) Low protein diet group 0.58 g protein/kg/d Calculated protein intake: 0.7 g/kg/d (taken from Fig 1) Co‐interventions Antihypertensive medications including ACEi and calcium channel blockers, phosphate binders
Outcomes	Slope of GFR decline over time by renal clearance of I‐125 Iothalamate Death (all causes) Number reaching ESKD
Notes	Funding source: National Institute of Diabetes, Digestive and Kidney Disease and Health Care Finance Administration, NIH, USA
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Random permutated blocks to ensure equal balance of participants assigned to each treatment combination" Patients stratified before randomisation according to blood pressure & rate of progression of kidney disease during 3 month baseline period"
Allocation concealment (selection bias)	Low risk	"Centrally administered at data co‐ordination centre through telephone contact"
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not blinded and lack of blinding may influence patient management
Blinding of outcome assessment (detection bias): End or change in GFR End of change in GFR	Low risk	Laboratory measurement (iothalamate clearance) and unlikely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias): Need to start dialysis Need to start dialysis	Unclear risk	No information on criteria for starting dialysis provided
Incomplete outcome data (attrition bias) All outcomes	Low risk	Loss to follow‐up 1.2% (3/255) in study 2
Selective reporting (reporting bias)	High risk	No report of numbers in each group reaching ESKD or death (combined data only). Did report weight & GFR measure
Other bias	Low risk	National Institute of Diabetes, Digestive and Kidney Disease and Health Care Finance Administration, NIH, USA