Hospital-based studies conducted in USA of adults (>18 years old) with hyperglycaemia hospitalised for surgery or medical diagnosis and that used scheduled basal (intermediate or long acting) insulin given subcutaneously were eligible for inclusion. Studies of patients with gestational diabetes, diabetic ketoacidosis or hyperosmolar hyperglycaemia syndrome were excluded.
The included studies were of patients with diabetes or hyperglycaemia. Each study defined hyperglycaemia in different ways: blood glucose levels ranged from more than 120mg/dL to 300mg/dL. Insulin interventions were glargine, glulisine, any basal insulin, basal bolus insulin, routine diabetic medication and SSI; treatment regimens varied between studies. Outcomes reported were whether studies disproved the use of SSI as per current American Diabetes Association (ADA) clinical practice guidelines, glycaemic control, effectiveness in treating and preventing hyperglycaemia and length of hospital stay.
The number of reviewers involved in study selection was not reported.