Adults with DKA have achieved early hyperglycaemia correction and ketoacidosis reversal with low hypoglycaemia risk after undergoing variable rate intravenous insulin infusions, according to an Australian team.

They say the protocol appears to have numerous benefits over fixed rate infusions, which are currently recommended by international guidelines, although head-to-head trials are needed to determine the best option.

The finding follows a study involving 97 patients who received insulin infusions for DJA at St Vincent’s Hospital, Melbourne, between 2012 and 2017.

Insulin dosages varied, based on hourly blood glucose level (BGL) readings, with rates up to six units per hour in patients with BGL >15.0 mmol/L down to 0.5 units per hour in patients with 3.5-5 mmol/L, under the hospital’s protocol.

Findings were positive, with 58% of patients reaching a BGL <10 mmol/L within 12 hours and 88% within 24 hours, the team reported in Journal of Diabetes and Its Complications (link here).

Moreover, hypoglycaemia rates during 12 and 24 hours of treatment were only 6.2% and 8.2%, respectively.

This was notable as it came despite such a large proportion of patient experiencing an early, intense, correction of the hyperglycaemia, the researchers said.

“Furthermore, rates of hypoglycaemia, hypokalaemia or any other adverse outcomes were not related to the rate of correction of BGLs,” they wrote.

They stressed the findings should be interpreted in the context of lower insulin infusion rates and consequential rates of BGL and ketone lowering compared with those recommended by current international guidelines.

Beyond that, the data contrasted with less positive outcomes seen in some international studies, according to the authors.

This may have reflected the graduated step-downs in insulin infusion rates in the hospital’s DKA protocol, which likely influenced the lower hypoglycaemia rates, they said.

One concern was the high rate of hypokalaemia, which occurred in 40% of patients and was associated with a lower admission pH.

This finding supported the hospital’s recent change in its institutional protocol for a more a aggressive approach to potassium replacement, particularly for DKA patients with marked acidosis, the team said.

Additionally, hypokalaemia was not associated with the correction rate of hyperglycaemia to <10 mmol/L, they noted.

“Prospective randomised control trials would be beneficial to establish the most effective, efficient, and safe way to treat patients with DKA,” they concluded.