For the millions of people living with diabetes, a hospitalization for diabetic ketoacidosis is both dangerous and costly — and the metric clinicians use to decide when to transition off intravenous insulin turns out to matter far more than previously appreciated. This finding directly challenges the direction of updated clinical guidelines, which have been steering hospitals toward a newer, more expensive monitoring approach.
In this two-site retrospective study of 178 patients with comparable DKA severity, those whose resolution was assessed using anion gap transitioned to subcutaneous long-acting insulin in a median of 13.3 hours, compared to 23.5 hours in patients monitored with serum β-hydroxybutyrate — a difference of more than 10 hours. Critically, this accelerated transition did not come at the cost of safety: rates of DKA recurrence following the switch to subcutaneous insulin were statistically equivalent between groups. Both cohorts followed identical continuous intravenous insulin protocols, isolating the monitoring metric as the key variable.
The broader context here is important. β-hydroxybutyrate is the predominant ketone body in DKA and is physiologically more direct than anion gap, which is a calculated surrogate affected by multiple variables including albumin levels, chloride shifts from saline infusion, and renal function. Updated guidelines favoring β-hydroxybutyrate rest on that biochemical logic. However, this study suggests that anion gap may normalize faster in clinical practice — or that β-hydroxybutyrate thresholds used for resolution may be set too conservatively — leading to unnecessarily prolonged insulin infusions. Longer infusions mean more ICU or step-down time, greater nursing burden, and higher costs. Key limitations include the retrospective design, modest sample size, and the possibility of unmeasured site-level confounders. Prospective randomized data are needed before guidelines reverse course, but this is a clinically meaningful signal that warrants urgent scrutiny.