C. J. Goodenough1, M. T. Nguyen1, D. H. Nguyen1, J. S. Roth2, C. J. Wray1, L. S. Kao1, M. K. Liang1 1University Of Texas Health Sciences Center At Houston,Surgery,Houston, TX, USA 2University Of Kentucky,Surgery,Lexington, KY, USA
Introduction:
Glycosylated hemoglobin(HbA1C) is diagnostic of and a measure of chronic hyperglycemia. Both HbA1c and perioperative hyperglycemia have been targeted as modifiable risk factors for post-operative complications. The aim of this study was to determine whether HbA1C or peak 24-hour post-operative peak glucose has a stronger association with major complications and to identify the peri-operative goals for both to minimize complications.
Methods:
A prospective study of all abdominal surgeries from a single institution from 2007-2010 was performed. All patients with a HbA1C within 3-months prior to surgery and at least one 24-hour post-operative glucose were included; only the peak glucose level was used. During the study period, standardized order sets for blood glucose control were routinely utilized. Variables were collected according to criteria established by the National Surgical Quality Improvement Project. The primary outcome was major complication defined as any Dindo-Clavien 3-5 complication within 30-days of surgery. Stepwise, multivariable analysis was performed including clinically relevant variables chosen a priori. Predicted probabilities for major complications were calculated at fixed HbA1C and peak glucose levels.
Results:
Among 374(out of 1017) patients who had both a HbA1c and post-operative glucose level, 92(24.6%) experienced a major complication. One-third(n=129) of these patients had a diagnosis of diabetes mellitus. On univariate analysis, both HbA1c and peak glucose level were predictive of complications. When HbA1c and peak glucose were considered independently, HbA1c≤6% and peak glucose 80-119 mg/dL had the lowest rates of major complications. When both HbA1c and peak glucose were included in the multivariable analysis, glucose was no longer statistically significant. HbA1C>6% predicted major complications (OR 2.12, 95% CI 1.19-3.76). The predicted probabilities of major complications demonstrated that within each HbA1c stratum, there were fewer major complications at lower peak glucose levels excluding <80 mg/dL (Table). However, the higher the HbA1c, the more likely the patient was to have elevated peak glucose levels.
Conclusion:
Both HbA1c and peak glucose are predictive of major complications after abdominal surgery, although HbA1c is a better predictor. Nonetheless, both should be considered together in stratifying patients’ risks for major complications. Whether these are both modifiable risk factors depends upon the urgency of the procedure and feasibility of achieving the target levels given the patient severity of disease and compliance. Further studies are necessary to determine safe, effective methods for optimizing both acute and chronic glycemic control around surgery.
