Comparison of short and intermediate-acting insulin protocols for perioperative glycaemic management in diabetic dogs undergoing phacoemulsification: a randomised clinical trial.

OBJECTIVE

To evaluate whether a perioperative insulin protocol using short-acting insulin is noninferior to a biphasic, intermediate-acting insulin protocol for maintaining blood glucose (BG) within a target range of 8.3-11.1 mM in diabetic dogs undergoing phacoemulsification.

STUDY

DESIGN

Prospective, randomised, double-blinded, noninferiority clinical trial.

ANIMALS: A group of 52 client-owned diabetic dogs scheduled for phacoemulsification surgery.

METHODS

Dogs were randomly allocated to receive either short-acting insulin intramuscularly (NEU; n = 25) or intermediate-acting insulin subcutaneously (CAN; n = 27) on the morning of surgery. Intravenous glucose supplementation was provided based on a standardised protocol. BG was monitored for 600 minutes postinsulin injection. The primary end point was the ratio of geometric mean BG concentrations (NEU: CAN). The prespecified noninferiority margin was 1.10 (an allowable 10% increase in BG). Secondary outcomes included proportions of readings within the target range and incidence of adverse events.

RESULTS

Noninferiority of the short-acting insulin protocol was not identified. The estimated ratio of geometric mean BG (NEU: CAN) was 1.03, with a 95% confidence interval (0.86-1.23) exceeding the noninferiority margin. High inter-individual variability was observed in both groups. The median proportion of readings within the target range was 0.11 [interquartile range (IQR) 0.08-0.23] for NEU and 0.17 (IQR: 0.08-0.29) for CAN (p = 0.413). Incidence of hypoglycaemia (BG < 3.3 mM) was low (4.0% NEU versus 11.1% CAN; p = 0.362), whereas hyperglycaemia (BG > 11.1 mM) occurred in > 96% of all dogs. Nonglycaemic complications, including hypotension (52.0% NEU versus 37.0% CAN; p = 0.402), were comparable between groups.

CONCLUSIONS AND CLINICAL RELEVANCE: Both protocols showed poor perioperative glycaemic control with high BG variability. These results illustrate the importance of vigilant BG monitoring with appropriate insulin or glucose supplementation during the perioperative period irrespective of insulin type.