Baldwin 2012.

Methods	Study design: parallel RCT Study time frame: June 2009 to June 2011 Duration of follow‐up: 6 days
Participants	Country: USA Setting: multicentre (3 sites) Inclusion criteria: patients with type 2 DM > 1 year of duration; > 18 years; eGFR ≤ 45 mL/min/1.73 m2; at least one hospital blood glucose level > 10 mmol/L; if on insulin, outpatient dose ≥ 0.5 U/kg Number: treatment group (57); control group (50) Mean age ± SD (years): treatment group (63.7 ± 13.0); control group (65.3 ± 10.6) Sex (M/F): treatment group (28/29); control group (20/30) Exclusion criteria: type 1 DM; pregnancy; chronic dialysis; solid‐organ transplant within the past 12 months; steroid therapy > 7.5 mg/d of prednisolone or equivalent medication; known hypopituitarism or adrenal insufficiency; known hypoglycaemia unawareness; length of stay < 48 h; severe liver disease
Interventions	Both groups All oral antidiabetic agents were stopped on hospital admission Treatment group SC insulin: 0.25 U/kg, half the dose given as glargine and half the dose as glulisine 3 times/d equally Control group SC insulin: 0.5 U/Kg, half the dose glargine and half the dose as glulisine 3 times/d equally
Outcomes	Percentage of BGL within the range of 5.6 to 10 mmol/L, and the percentage of subjects experiencing a hypoglycaemic event defined as a blood glucose < 3.9 mmol/L Hypoglycaemic events were further separated into moderate hypoglycaemia (2.8 to 3.8 mmol/L) and severe hypoglycaemia (< 2.8 mmol/L)
Notes	Funding source: "This study was sponsored by an investigator‐initiated grant from sanofi‐aventis."
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	The method of random sequence generation was not reported. All the paper reports was: “Eligible patients gave informed consent and were randomised 1:1 into two protocol groups by a research pharmacist”.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Insufficient information to permit judgement
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Insufficient information to permit judgement
Incomplete outcome data (attrition bias) All outcomes	Low risk	There were no drop‐outs in each group, and all subjects were analysed in the groups to which they were randomised
Selective reporting (reporting bias)	Low risk	The prespecified outcomes were available on a clinical trials database and all outcomes were reported
Other bias	Low risk	No conflicts of interest were reported