The American Association of Clinical Endocrinologists and the American Diabetes Association have come out with a new consensus statement. The document covers a variety of issues encountered with hospitalized patients and should be a useful reference. The authors seem a little doctrinaire, though, in areas where evidence is lacking.
For critically ill patients:
Insulin therapy should be initiated for treatment of persistent hyperglycemia, starting at a threshold of no greater than 180 mg/dl (10.0 mmol/l).
Intravenous insulin infusions are the preferred method for achieving and maintaining glycemic control in critically ill patients.
That reflects the state of our knowledge after NICE-SUGAR. The IV route is preferred in such patients because of the flexibility it affords in the constantly changing conditions of critical illness as well as the fact poor and fluctuating tissue perfusion makes sub Q absorption unreliable.
The recommendations for noncritical patients seem more problematic:
For the majority of noncritically ill patients treated with insulin, the premeal BG target should generally be less than140 mg/dl (less than 7.8 mmol/l) in conjunction with random BG values less than 180 mg/dl (less than 10.0 mmol/l), provided these targets can be safely achieved.
Although this recommendation seems perfectly reasonable it implies that less ill patients need a stricter glycemic target (which doesn't make a whole lot of sense) and has no high level evidence to back it up.
Concerning the means to get to those targets the statement says:
Scheduled subcutaneous administration of insulin, with basal, nutritional, and correction components, is the preferred method for achieving and maintaining glucose control.
The point about sliding scale insulin here is it shouldn't be the sole regimen. The patient needs some basal coverage to go with that. Note that when diabetologists do recommend sliding scale insulin as part of a regimen they use the more politically correct term correctional insulin.
Noninsulin antihyperglycemic agents are not appropriate in most hospitalized patients who require therapy for hyperglycemia.
That doesn't quite address the question of whether patient's non-insulin diabetes medicines should always be withheld in the hospital although they often are. There's no evidence to support such a rule that I know of and factors such as anticipated NPO status (e.g. in the case of sulfonylureas) and changes in renal function (e.g. for sulfonylureas and metformin) should guide clinical judgment.