A 69-year-old woman with type 2 diabetes, hypertension, dyslipidemia, and chronic kidney disease was admitted following a fall at home and sustained a right hip fracture. She underwent a right hip arthroplasty on the day of admission. Her blood glucose levels were noted to be in the range of 200-300 mg/dL after her surgery. The orthopedic team ordered a preset aspart sliding scale with meals. Due to persistent hyperglycemia over the following 24 h, an endocrine consult was requested.

The patient has had a history of diabetes for 20 years prior to this admission. She takes glyburide 5 mg daily, sitagliptin 100 mg daily, and a combination of metformin 850 mg/pioglitizone 15 mg twice daily. She denies taking insulin in the past. She does not check her blood glucose at home and denies hypoglycemic events. She reports a history of retinopathy for which she has received laser treatment in the past. She complains of right hip pain, but otherwise feels well. She denies nausea, vomiting, polydipsia, and polyuria. On physical examination, her weight is 80 kilograms (kg) and her height is 152 cm with a body mass index of 34.6 kg/m². She has acanthosis. Dorsalis pedis pulses were 2+ bilaterally and she had no lesions on her feet.

Laboratory tests showed a creatinine level of 1.77 mg/dL, GFR 29 mL/min/m², and a fasting serum glucose of 325 mg/dL. Her liver function tests were normal. Her hemoglobin A1c (HbA1c) was 10.6 %. Review of prior laboratory tests shows a GFR of 47 mL/min/m² approximately 2 years prior to this admission.

Controversy surrounds glycemic targets for inpatients. A critical care versus a noncritical care setting is important in defining glycemic goals in the hospital. Although poor blood glucose control has been associated with negative outcomes (infections, longer length of stay, mortality), further evidence-based data is needed to determine optimal glycemic goals for hospitalized patients in a noncritical care setting [1]. One of the main factors limiting tight glycemic control is the risk of hypoglycemia, which itself may result in increased mortality. The American Diabetes Association (ADA) and the American Association of Clinical Endocrinologists (AACE) published a Consensus Statement on inpatient glycemic control in 2009 and the ADA includes recommendations for inpatient glycemic goals in the Standards of Medical Care in Diabetes yearly [2, 3]. The Endocrine Society published a Clinical Practice Guideline for managing hyperglycemia in a noncritical care setting in 2012 [4]. According to guidelines, the glycemic goals for noncritically ill patients are: fasting or premeal glucose levels ≤140 mg/dL and a random blood glucose level ≤180 mg/dL. In our clinical vignette, the patient’s blood glucose levels are above goal and therefore her insulin regimen should be adjusted.

Most oral antihyperglycemic medications are either ineffective in the inpatient setting, contraindicated, or prone to causing hypoglycemia in the unstable patient. Our patient’s oral medications were discontinued. Metformin is contraindicated in women with creatinine levels >1.4. The drug is also not an ideal choice given the potential for patient exposure to contrast agents, procedures, or hemodynamic instability while she is acutely sick in the hospital. Pioglitizone is known to increase fracture risk in some patients; given that the patient recently sustained a fragility fracture, it would not be prudent to continue this oral antihyperglycemic agent. Glyburide, a secretogogue, carries a high risk of causing hypoglycemia in elderly patients, particularly in the context of renal dysfunction; Hypoglycemia would compromise safety in a patient already at high risk for falls. The dose of sitagliptin could be adjusted according to her renal dysfunction; however, the use of sitagliptin alone will be unlikely to control this patient’s hyperglycemia. While a DPP-4 inhibitor could be considered at discharge, combining a DPP-4 inhibitor and insulin would also increase her risk of hypoglycemia. She will need insulin at discharge.

There are three main factors that contribute to determining the most appropriate insulin regimen for the hospitalized patient:

For patients who are prescribed a diet, it is important to provide both basal and prandial insulin coverage. The RABBIT-2 medicine [5] and surgery [6] trials demonstrated that basal/bolus insulin is superior to sliding scale insulin in the treatment of noncritically ill patients with type 2 diabetes. Unfortunately, it is very common to find patients only receiving “sliding scale” insulin prior to meals for several days in the hospital setting. It is important to educate healthcare professionals about the importance of basal/bolus insulin treatment.

The main component to establishing an inpatient insulin regimen is to first calculate the patient’s estimated total daily dose (TDD) of insulin. On average, patients with type 2 diabetes require a TDD of 0.4–1 units/kg. For elderly patients and those with renal or hepatic dysfunction, a TDD of 0.3 units/kg is a reasonable starting dose. For obese patients or those receiving steroids, 0.5–0.6 units/kg provides additional insulin to counter the increased insulin resistance associated with these states. Generally, basal insulin requirements make up 50 % of the patient’s TDD. Initial insulin requirements in the hospital may often be higher than the patient’s insulin needs at home due to stress, infection, and altered eating patterns.

In our patient with renal dysfunction, we selected a formula of 0.3 units/kg to determine her total daily insulin requirements: for an 80 kg woman, TDD = 0.3 units × 80 kg, or 24 units. Her basal dose should be 50 % of her estimated TDD requirements or 12 units; therefore, she received 12 units of glargine insulin. The remaining 50 % of her TDD approximates the patient’s prandial rapid-acting insulin dose, which is further divided by 3 to give 4 units rapid acting insulin with each meal. Her prandial dose may be supplemented by a correctional insulin scale for hyperglycemia as well.

For patients who are already receiving an established insulin regimen at home, it is common practice to initiate basal insulin at 80–100 % of their home dose in the inpatient setting. Lower insulin doses should be given to patients with decreased dietary intake, especially if the home basal insulin dose is inappropriately high and appears to be covering prandial needs as well. Patients who understand carbohydrate counting prior to their admission and are reliable at practicing diabetes self-management may be allowed (and often encouraged) to use their home insulin ratios to determine their premeal insulin dose in the hospital.