Hospital Diabetes Management

Medical Disclaimer

This article is for informational purposes only and does not constitute medical advice. Always consult your physician or a qualified healthcare provider regarding any medical condition or treatment.

Key Takeaways

  • Inpatient glycemic targets per ADA Section 16 are 140 to 180 mg/dL for most hospitalized adults, with 110 to 140 mg/dL appropriate for some critical-care and post-cardiac-surgery contexts — both very high and very low glucose worsen outcomes.
  • Subcutaneous basal-bolus insulin is the preferred regimen for non-critically-ill patients eating; intravenous insulin infusion is preferred for ICU, post-cardiac-surgery, DKA, HHS, and unstable patients — both outperform sliding scale alone.
  • Sliding scale insulin alone (rapid-acting given only when glucose is high) is inferior to scheduled basal-bolus and should not be the only regimen — the RABBIT-2 trial and many subsequent studies show worse glycemic control and more complications with sliding-scale-only protocols.
  • Inpatient hyperglycemia in patients without known diabetes ("stress hyperglycemia") requires A1C testing during admission if not done in the prior 3 months — about 30 percent of these patients have undiagnosed diabetes.
  • Discharge planning starts on admission — medication reconciliation, diabetes self-management education, follow-up scheduling within 2 weeks, and clear written instructions reduce readmission risk substantially.

Hospital diabetes management is a structured approach to glucose control during admission — targeting 140 to 180 mg/dL for most patients, using basal-bolus subcutaneous insulin or IV insulin infusion rather than sliding scale alone, and starting discharge planning on day one. Done well, it reduces wound infections, length of stay, readmissions, and mortality. Done poorly, it produces preventable hypoglycemia, DKA, and discharge medication errors.

Why Inpatient Glycemic Control Matters

  • Hyperglycemia (over 180 mg/dL) is associated with longer hospital stays, more infections, and higher mortality across nearly every condition
  • Hypoglycemia (under 70 mg/dL) also increases mortality and cardiovascular events
  • The relationship is U-shaped — both extremes are harmful
  • Stress hyperglycemia in patients without known diabetes carries similar prognostic weight
  • About 30 percent of hospitalized patients with hyperglycemia have undiagnosed diabetes — A1C during admission identifies them

ADA Inpatient Glycemic Targets (Section 16)

Setting Target Range (mg/dL)
Most non-critically ill 140–180
Critically ill (ICU) 140–180 (110–140 for selected patients)
Post-cardiac surgery 110–140 typical
Pregnancy in hospital Tighter — fasting under 95, post-meal under 140
End-of-life / palliative Liberal — symptom-based
Hypoglycemia threshold Below 70 (severe under 54)
Hyperglycemia threshold Above 180 (sustained above 250 — escalate care)

Inpatient Insulin Strategies

Patient Status Preferred Regimen
Eating, not critically ill Basal insulin (glargine, detemir, degludec) + rapid-acting with meals + correction scale
NPO (not eating) Basal insulin at reduced dose (50–80%) + correction scale every 4–6 h
Receiving enteral nutrition (tube feeding) Basal insulin + scheduled rapid-acting matched to feed schedule
Critically ill / ICU IV insulin infusion with hourly glucose checks
DKA or HHS IV insulin infusion until criteria met for transition to subcutaneous
Post-cardiac surgery IV insulin infusion 24–48 h, then transition
Stable patient on insulin pump (selected) Continue home pump if patient capable, facility allows

Why Sliding Scale Alone Is Inferior

Sliding scale (rapid-acting given for high glucose readings without scheduled basal) is reactive rather than preventive — by the time a fingerstick shows 250 mg/dL, the patient has already been in that range for hours. The landmark RABBIT-2 trial (Umpierrez et al, 2007) randomly assigned hospitalized type 2 diabetes patients to sliding scale alone versus basal-bolus with correction. Basal-bolus produced lower mean glucose, fewer hyperglycemic days, fewer complications, and no increase in severe hypoglycemia. Most major guidelines now recommend basal-bolus as the default; sliding scale should be a correction layer, not the entire regimen.

Managing DKA and HHS Inpatient

  • Initial IV fluid resuscitation — typically isotonic saline at 1 L/h for first 1–2 hours, then adjusted
  • Potassium replacement — start when K+ falls below 5.3 mEq/L, before starting insulin if K+ under 3.3
  • IV regular insulin infusion at 0.05 to 0.1 units/kg/hr (DKA) or 0.025 to 0.05 (HHS)
  • Hourly glucose, electrolytes every 2–4 hours
  • Switch IV fluid to D5-containing when glucose reaches 200 (DKA) or 300 (HHS)
  • Bicarbonate only for severe acidosis (pH under 6.9) — not routine
  • Transition to subcutaneous when anion gap closed, bicarbonate over 18, patient eating — overlap IV with subcutaneous basal for at least 2 hours

Managing Inpatient Hypoglycemia

Glucose Conscious, NPO Conscious, Eating Unconscious
54–70 mg/dL 15 g IV D50 or D10 15 g oral fast carbs IV D50 50 mL
Below 54 mg/dL 25 g IV D50 20–30 g oral fast carbs IV D50 50 mL; consider glucagon if no IV
Recurrent hypoglycemia Reduce basal; investigate cause Reduce next bolus; adjust scheduled basal Stop IV insulin; investigate
Documentation Repeat glucose in 15 min until >70 Same Same

Steroid-Induced Hyperglycemia

  • Corticosteroids raise glucose, especially post-meal and afternoon
  • Most pronounced 4 to 12 hours after a morning prednisone dose
  • NPH insulin given with morning steroid often controls afternoon peak
  • Basal-bolus insulin doses may need 30 to 50 percent increase during steroid course
  • Anticipate dose reduction as steroid tapers
  • Patients without prior diabetes may need insulin during steroid course — most do not need long-term — see our steroid-induced diabetes article

Stress Hyperglycemia in Non-Diabetic Patients

  • Glucose over 140 mg/dL in a hospitalized patient without known diabetes
  • Causes: critical illness, infection, infarction, surgery, steroids, parenteral nutrition
  • Treat to standard inpatient targets (140 to 180)
  • Check A1C during admission if not done in past 3 months
  • A1C above 6.5 percent — likely undiagnosed diabetes; refer for outpatient follow-up
  • A1C 5.7 to 6.4 percent — prediabetes; lifestyle counseling and rescreening in 6 months
  • A1C below 5.7 percent — pure stress hyperglycemia; usually resolves with recovery but still carries 50 percent risk of future diabetes within 5 years

Medication Reconciliation

Step Action
Admission Compare current home medications against admission orders; identify holds and substitutions
Transitions (ICU to floor) Re-verify; transition from IV to subcutaneous insulin requires overlap
Pre-discharge Side-by-side comparison: home meds, inpatient meds, discharge meds — flag changes
Discharge education Written list in patient’s language; teach-back to confirm understanding
Post-discharge Pharmacist follow-up call within 72 hours; clinician visit within 2 weeks

Discharge Planning

  • Medication reconciliation as above
  • Written sick day rules — see our sick day rules guide
  • Hypoglycemia recognition and treatment instructions
  • Glucagon prescription if on insulin or sulfonylureas — teach a household member
  • Glucose meter and strips at home — confirm patient has them
  • Diabetes self-management education (DSME) — inpatient consult or outpatient referral
  • Nutrition consultation if dietary changes are needed
  • Follow-up appointment within 1 to 2 weeks (sooner for new insulin start or DKA/HHS admission)
  • Specialist referral for complex cases — endocrinology, podiatry, ophthalmology, nephrology
  • Communication with primary care clinician before discharge

Inpatient Diabetes Educator and Specialist Consults

  • Most academic hospitals have inpatient diabetes management teams or consult services
  • Indications include: DKA, HHS, persistent hyperglycemia despite basic management, new insulin start, severe hypoglycemia, pregnancy with diabetes, complex regimens, insulin pump or CGM use
  • Early consultation reduces length of stay and improves discharge outcomes
  • Inpatient diabetes self-management education improves post-discharge adherence and reduces readmission

For pre- and post-hospital care, see our guides on surgery and diabetes, sick day rules, and the broader treatment options overview.

The Bottom Line

Hospital diabetes management targets 140 to 180 mg/dL using basal-bolus subcutaneous insulin or IV insulin infusion — not sliding scale alone. ICU and post-cardiac surgery patients tolerate IV insulin with tighter targets; ward patients do better on scheduled basal plus mealtime plus correction. Steroid-induced and stress hyperglycemia require active management even in patients without prior diabetes diagnoses, with A1C testing during admission to catch undiagnosed disease. Discharge planning is the make-or-break step — medication reconciliation, written instructions, diabetes education, and follow-up within 2 weeks prevent the readmissions that account for a large fraction of avoidable diabetes admissions.

Frequently Asked Questions

What blood sugar level is the goal in the hospital?

The American Diabetes Association recommends 140 to 180 mg/dL for most hospitalized patients, both critically and non-critically ill. Some critical-care contexts (post-cardiac surgery, neurosurgical patients with stroke) target 110 to 140 mg/dL. Below 70 mg/dL is hypoglycemia requiring immediate treatment. Above 180 mg/dL increases length of stay, infection, and mortality. Tighter targets (under 110 mg/dL) are not recommended for most patients because of hypoglycemia risk.

Why does the hospital use insulin instead of my usual diabetes pills?

Most non-insulin diabetes medications are held in hospital because of safety, dosing, and predictability concerns. Metformin is held if kidney function is unstable or contrast is planned. SGLT2 inhibitors are held because of euglycemic DKA risk. Sulfonylureas are held while NPO. GLP-1 receptor agonists slow gastric emptying and complicate perioperative care. Insulin (basal plus correction or IV infusion) is dose-titratable to current glucose, kidney function, and intake — which is why it dominates inpatient regimens.

What is sliding scale insulin and why is it not enough?

Sliding scale gives rapid-acting insulin only when a glucose check shows hyperglycemia. It reacts to high glucose after it has already happened, rather than preventing it. Multiple trials including RABBIT-2 showed sliding scale alone produces worse glucose control, more hyperglycemia, and more complications than scheduled basal insulin plus mealtime bolus plus correction. Sliding scale can be a useful correction layer on top of scheduled basal-bolus, but it should not be the entire regimen for most inpatients.

What happens when I am discharged from the hospital?

A good discharge plan includes: medication reconciliation comparing pre-admission, in-hospital, and discharge medications side-by-side; written instructions in plain language; diabetes self-management education or referral; follow-up appointment within 1 to 2 weeks for medication titration; A1C result and what it means; instructions for sick days, hypoglycemia, and when to seek care; and connection to a diabetes educator or pharmacist. Patients with new diabetes diagnoses, new insulin starts, or admissions for DKA or HHS need particularly close follow-up.

Sources

  1. American Diabetes Association. Standards of Care in Diabetes 2024. Section 16 Diabetes Care in the Hospital. Diabetes Care 47(Suppl 1).
  2. Umpierrez GE et al. RABBIT-2 Trial. Randomized study of basal-bolus insulin therapy in the inpatient management of patients with type 2 diabetes. Diabetes Care.