Managing Inpatient Diabetes: Practical Pearls and Complex Scenarios

In this high-yield session, Lily Ackermann, MD, a clinical associate professor of medicine at Thomas Jefferson University Hospital in Philadelphia, addressed the complexities of inpatient diabetes management, emphasizing real-world decision making and individualized care. Drawing from current evidence and large trials, Dr. Ackermann offered practical frameworks to guide hospitalists through nuanced insulin regimens, glycemic targets, and transitions of care.

She began by challenging the default use of sliding scale insulin, or SSI, citing data from the RABBIT-2 trial, which demonstrated the superiority of basal-bolus regimens in hospitalized patients with type 2 diabetes. However, she cautioned against a one-size-fits-all approach. Patients’ oral intake, renal function, frailty, and risk of hypoglycemia must inform decisions. For example, a well-nourished patient with a hemoglobin A1C over 9% and glucose over 300 mg/dL would benefit from a full basal-bolus regimen, whereas a frail, elderly patient with limited intake might fare better on a basal-plus or correction-only strategy.

Premixed insulin (e.g., 70/30) was discussed as an option only at discharge in select patients due to a higher risk of nocturnal hypoglycemia when used during acute hospitalization. Neutral Protamine Hagedorn (NPH) insulin was highlighted for its utility in managing steroid-induced hyperglycemia, particularly in patients on prednisone, given its temporal glycemic-lowering overlap. In contrast, rapid-acting analogs were preferred for their closer match to postprandial glucose excursions, though cost remains a barrier for many.

Dr. Ackermann outlined the use of the “basal plus” strategy, basal insulin with correctional doses for patients with moderate hyperglycemia (glucose 200 to 300 mg/dL), especially when oral intake is adequate and hemoglobin A1C is over 9%. Trials like BASAL PLUS and SITA-HOSP support the efficacy and safety of this approach, particularly when combined with dipeptidyl peptidase-4 inhibitors. These agents are preferred due to their low hypoglycemia risk and renal safety profile, even in dialysis patients.

For patients with severe hyperglycemia driven by steroid use, insulin regimens must be tailored to the pharmacokinetics of the steroid. For example, for hydrocortisone-induced hyperglycemia, rapidly acting insulins can be used at 0.1 units/kg. For prednisone-induced hyperglycemia, which responds well to NPH given its 12-hour peak, clinicians can use NPH at 0.2 units/kg for doses less than 40 mg, and 0.3 units/kg for doses more than 40 mg. For dexamethasone, clinicians may require longer-acting insulin. She encouraged dosing adjustments based on steroid dose and duration.

Nutritional support presented additional challenges. Patients on continuous enteral feeds should receive regular insulin every six hours or NPH at night, while those on total parenteral nutrition (TPN) benefit from insulin mixed directly into the TPN bag, with dose adjustments based on prior 24-hour correctional insulin use. This approach avoids severe hypoglycemia if TPN is abruptly stopped.

Dr. Ackermann also reviewed glycemic targets in hospitalized patients. For most non-critically ill patients, 140 to 180 mg/dL is the target. However, in terminally ill, frail, or dialysis-dependent individuals, a relaxed goal of 180 to 250 mg/dL may be safer. She emphasized that glycemic goals must be individualized and adjusted for risk of hypoglycemia, which is now a reportable quality metric if glucose drops below 40 mg/dL.

Device use in the hospital was addressed next. Continuous glucose monitors, CGMs, and insulin pumps are increasingly encountered, though not U.S. Food and Drug Administration-approved for inpatient use. Many institutions allow continued use under supervision, provided nursing staff can cross-check readings. This is particularly helpful for overnight monitoring in high-risk patients.

A key theme of the session was the importance of discharge planning. Dr. Ackermann noted that many patients, especially those uninsured or underinsured, rely on the hospital for diabetes care. She stressed the need for early education, identification of affordable insulin options (e.g., Walmart’s $25 NPH and regular insulin), and simplification of regimens for safe outpatient use. Human insulins, while associated with more glycemic variability, remain lifesaving when analogs are cost-prohibitive.

For newly diagnosed or uncontrolled patients with poor access to medications, Dr. Ackermann recommended simplified 70/30 regimens, dosed as 60% in the morning and 40% in the evening, provided food intake is consistent. If meal patterns are irregular, individualized dosing with NPH and regular insulin remains an option. She encouraged clinicians to initiate assistance applications early, citing programs that cap insulin costs at $35/month.

Perioperative management included holding sodium-glucose co-transporter 2 inhibitors 72 hours before surgery due to the risk of ketoacidosis, and the evolving debate around Glucagon-Like Peptide-1 receptor agonists. Dr. Ackermann noted limited evidence around gastric emptying delay and aspiration risk, encouraging shared decision-making with anesthesiologists and adherence to local protocols.

The session closed with a practical case involving the discharge of a young patient with diabetic ketoacidosis and no insurance. Dr. Ackermann recommended a pragmatic approach using premixed insulin or human insulin formulations, with careful counseling on hypoglycemia symptoms, dose timing, and follow-up care.

Dr. Alqawasmi is an internal medicine resident at the University of New Mexico.