Study intervention was adopted as standard of care.

Diabetic ketoacidosis (DKA) is a medical emergency that is associated with significant morbidity and mortality for both patients with type I and type II diabetes. By correcting hyperglycemia and inhibiting the release of free fatty acids, insulin administration leads to decreased ketone formation and resolution of acidosis. Short-acting intravenous insulin is often preferred to subcutaneous administration for initial management due to its short half-life and ease of titration, but patients will eventually need to transition to subcutaneous insulin prior to discharge. The timing of initiation or resumption of home long-acting subcutaneous insulin is controversial in the treatment of DKA. It is currently unknown if resuming a portion or all of the patient's home basal regimen during the initial treatment phase of DKA will provide an impact on patient care. The purpose of this study is to evaluate the impact of early glargine administration if the patient was not previously on basal insulin or resuming the patient's home basal insulin regimen within two hours after the start of the intravenous insulin infusion in addition to usual care will improve patient outcomes.

In 2014, there were 168,000 hospitalizations for DKA, and the average length of stay was 3.4 days in 2009. The estimated annual direct medical expense and indirect cost is 2.4 billion dollars per year. DKA is a potentially life-threatening condition due to its ability to cause metabolic acidosis, electrolyte imbalances, and dehydration. While DKA treatment protocols differ from institution to institution, the goals of treatment remain the same --- resolution of acidosis, dehydration, and hyperglycemia. Following fluid resuscitation, correction of hyperglycemia with insulin therapy is the mainstay of treatment. Intravenous (IV) insulin infusions with regular insulin are commonly used due to its short half-life of 30 to 60 minutes and resulting titratability. Treatment is then continued with subcutaneous (SC) insulin once the acidosis and anion gaps have cleared and the patient's hemodynamics and electrolytes abnormalities have resolved.

Traditionally, when SC insulin is initiated, the IV insulin infusion is stopped two hours later. However, the timing of SC initiation has becoming more controversial in recent years. The 2019 American Diabetes Association Standards of Care recommend transition of patients from IV to SC insulin requires administration of basal insulin 2-4 h prior to the intravenous insulin being stopped to prevent recurrence of ketoacidosis and rebound hyperglycemia. However, the 2014 Joint British Diabetes Societies Inpatient Care Group recommend continuing a patient's home basal insulin during the initial management of DKA to prevent rebound hyperglycemia and to avoid extending length of stay. Several recent papers have examined early basal insulin administration as adjunct therapies with IV insulin infusions for the treatment of hyperglycemia and DKA.

In 2012, a single-center prospective, randomized controlled trial by Hsia et al. compared IV insulin infusion with or without early administration of insulin glargine (dose = 0.25 U/kg) in 61 hyperglycemic patients with diabetes. Forty-one percent (25/61) of the patients enrolled in the trial were admitted with DKA. Those patients that received early glargine had lower prevalence of rebound hyperglycemia when the insulin infusion was discontinued (experimental group: 33% vs control group: 93.5%; P < 0.001). The total length of IV insulin infusion was similar in the two groups (control group: 35 ± 13h vs experimental group: 42 ± 24h). The mean blood glucose levels within the 12-hour study period were significantly lower in the intervention group, while there were 3 asymptomatic hypoglycemic measurements in the control group and none in the intervention group.

In 2015, Doshi et al. conducted a two-center prospective, randomized controlled trial of 40 patients admitted for DKA. Patients were randomized to receive early insulin glargine (dose = 0.3 U/kg) or no glargine in addition to regular care. Those in the early insulin glargine arm were to receive the glargine subcutaneous injection within 2 hours of diagnosis. Patients in the control group were given insulin glargine once the anion gap closed and the IV insulin infusion was continued for two additional hours. There was no difference in time to anion gap closure, hospital length of stay, ICU admissions, ICU length of stay, and hypoglycemic events.

In 2015, Houshyar et al. conducted a single-center, prospective, randomized controlled trial of 40 patients admitted for DKA. Patients were randomized to receive early insulin glargine (dose = 0.4 U/kg) or no insulin glargine in addition to usual care. Patients in the control group were given insulin glargine once the anion gap closed. There was no difference in duration of acidosis, hospital length of stay or rates of hypoglycemia. There were higher rates of hyperglycemia following discontinuation of the insulin infusion in those that did not receive early glargine (51% vs. 35% of measurements in the experimental group showed a glucose >8.3 mM or >150 mg/dL; p = 0.046).

Rappaport et al. (2019) conducted a single-center, retrospective cohort study of 106 admissions for DKA comparing early vs late initiation of basal insulin. Patients admitted to the MICU who were treated with home-dose basal insulin prior to DKA resolution and within 24 hours of IV insulin infusion initiation (n = 33) were compared to patients who were initiated on basal insulin within 6 hours before or any time after IV insulin infusion discontinuation (n = 73). There was no difference in transitional failure, rebound hyperglycemia, hypoglycemia events, time to anion gap closure, length of ICU stay, or hospital length of stay. There was a statistically significant decrease for time on IV insulin infusion (early: 13.8h vs late: 17.1h, p = 0.04).

As previously outlined, randomized studies using weight based basal insulin doses have shown that the early initiation of basal insulin can prevent rebound hyperglycemia with no increase in hypoglycemia. One retrospective study demonstrated a decrease of time on IV insulin infusion in those that received early basal insulin when compared to those that did not. These studies have also shown trends toward shorter ICU and hospital length of stay, anion gap closure, and time to resolution of acidosis but none have demonstrated statistical significance. However, no prospective trials have examined resuming the patient's home basal insulin early in the course of DKA and it is currently unknown if resuming a portion or the entirety of the patient's home basal regimen will provide an impact on patient care.

By conducting a prospective study comparing initiation of early glargine administration at the patient's home dose to current standards of care, early administration may lead to shorter ICU and hospital lengths of stay as well as shorter duration of acidosis. This could provide significant cost-savings to the institution and benefit patients in need of treatment for this serious complication of diabetes.

Inclusion Criteria:

1. Presented to Regions Hospital ED for chief complaint of DKA, nausea, vomiting, abdominal pain, hyperglycemia, or similar

2. Meets all below diagnostic criteria for DKA per the American Diabetes Association:

   • Arterial or venous pH </= 7.3
   • Serum Bicarbonate </= 18 mEq/L
   • Ketonuria or ketonemia
   • Anion Gap > 10
   • Blood sugar > 250 mg/dL

3. Receiving IV insulin infusion

4. It is feasible to provide insulin glargine within 2 hours (+/- 30 minutes) of IV infusion start

5. Will be admitted to the ICU for DKA, or already admitted to the ICU for DKA

6. Ability to provide informed consent

Exclusion Criteria:

1. Age < 18
2. End stage renal disease or hepatic disease
3. Hypotension requiring IV vasopressors or inotropes at any point during admission (i.e. norepinephrine, dobutamine, vasopressin, etc.)
4. Need for emergent surgery
5. Pregnant patients
6. Prisoners
7. Indication for insulin therapy other than DKA (hypertriglyceridemia, beta-blocker overdose, hyperglycemia without DKA)
8. Patients receiving prior to admission insulin pump therapy
9. Patients receiving prior to admission combination insulin products (i.e. Novolin® 70/30, Novolog® 70/30, Humalog® 75/25, etc.)
10. Did not consent to study

There is no plan to share IPD with investigators not currently involved in the study.