Utility of β-hydroxybutyrate as a defining criterion for diabetic ketoacidosis resolution in adults
Kevin Fenter, Tyler Bui, Alvin Tomika
Background: The criteria for determining resolution of diabetic ketoacidosis (DKA) have varied over the years. The most recent position statement published in 2024 by the American Diabetes Association (ADA) significantly modified the criteria for DKA resolution to include the use of beta-hydroxybutyrate (BHB). With the development of point-of-care ketone measurement devices it is more feasible to measure serum ketones to define DKA resolution; however, limited data exists on the benefit of serum ketones compared to traditional criteria historically used to define DKA resolution. The objective of this study is to evaluate the utility of BHB as a criterion for DKA resolution to successful transition from intravenous (IV) to subcutaneous (SUBQ) insulin.
Methods: This study was a single center retrospective chart review of electronic health records between November 1, 2023, and October 31, 2025 of patients with DKA treated with intravenous regular insulin as well as additional standard of care therapy. The study included adult patients 18 years of age or older admitted to the hospital with a primary diagnosis of DKA. Patients must have been treated with IV regular insulin and transitioned to SUBQ insulin prior to discharge. Patients were excluded if received insulin infusions related to conditions other than DKA, if treated outside the institutional DKA order set/protocol, received concomitant treatment with IV regular insulin and SUBQ insulin, died, were discharged or enrolled in comfort care prior to insulin transition, pregnant or lactating, or if were missing key clinical data.
Patients were stratified into two groups dependent on if BHB was collected and normalized (Group 1) or not BHB not normalized/collected prior to transition to subcutaneous insulin (Group 2). The primary outcome was post transition treatment failure within 48hrs defined as resumption of IV insulin or meeting two out of the three criteria for DKA diagnosis. Secondary outcomes include duration of ICU stay, duration of hospitalization, time from regular insulin initiation to initiation of SUBQ insulin, duration of IV insulin, hypoglycemia events during admission, IV insulin requirements prior to transition, and post transition insulin requirements. Outcomes were evaluated using statistical methods such as Fisher’s exact test and Mann-Whitney U test as appropriate. Demographics and baseline characteristics were evaluated using descriptive statistics.
Results: A total of 105 patients met criteria for inclusion with 31 patients included within Group 1. The mean age was 44 years with 66% on home insulin prior to admission; of which 61% reported not compliant. Baseline characteristics were similar, however; Group 1 saw 90.3% of patients admitted to the ICU while Group 2 saw 46% admitted to the ICU. For the primary outcome of transition failure there was no statistically significant difference between the two groups (16.1% vs 13.5%, p=0.16). For the secondary outcomes there was no statistically significant difference for hospital length of stay, ICU length of stay, insulin requirements before and after transition as well as hypoglycemic events. In Group 1 there was a non-statistically significant trend toward longer duration of IV insulin therapy and time to transition to subcutaneous insulin.
Conclusion: Among adults admitted for treatment of DKA and transitioned from intravenous insulin to subcutaneous insulin, resolution criteria utilizing normalization of beta-hydroxybutyrate to <0.6 mmol/L were not associated with decreased rates of transition failure.
Contact Information:
[email protected]
