Preoperative HbA1c and Outcomes following Lower Extremity Vascular Procedures

Abstract

BackgroundLimited data exist evaluating preoperative hemoglobin A1c (HbA1c) in patients undergoing vascular procedures for peripheral arterial disease (PAD). This study evaluated the relationship of preoperative HbA1c on outcomes after open and endovascular lower extremity (LE) vascular procedures for PAD. MethodsWe selected patients with PAD admitted for elective LE procedures between September 2008 and December 2015 from the Cerner Health Facts® database using International Classification of Disease, Ninth Edition, Clinical Modification diagnosis and procedure codes. Bivariable analysis and multivariable logistic models examined the association of patient characteristics, procedure type, and preoperative HbA1c (normal < 6.5%, high ≥ 6.5%) with postsurgical outcomes that included infection, renal failure, respiratory or cardiac complications, length of stay, in-hospital mortality, and readmission. ResultsOf 4087 patients who underwent a LE vascular procedure for PAD, 2462 (60.2%) had a preoperative HbA1c recorded. The cohort was mostly male (60%), white (73%), and underwent endovascular intervention (77%). Patients with high HbA1c levels were more likely of black race (P < 0.02) and had significantly higher comorbidities (P < 0.0001). Elevated HbA1c was associated with diabetes (P < 0.0001) and cellulitis (P = 0.05) on unadjusted analysis. Multivariable logistic regression (adjusting for patient, hospital, comorbidity and procedural characteristics) revealed that elevated HbA1c was significantly associated with 30-day readmission (OR = 1.06, 95% confidence interval = 1.00–1.12), but was not associated with the other outcomes. An independent diagnosis of diabetes was not predictive of complications or readmission. ConclusionsHistoric glucose control, as evidenced by a high preoperative HbA1c level, is not associated with adverse outcome, other than readmission, in patients undergoing LE procedures for PAD. Given the known association of high perioperative glucose levels with poor outcome following vascular procedures, this is suggestive of a more important effect of perioperative, as opposed to chronic, glucose control upon outcome. Thus, we suggest focusing efforts on creating standardized goal-directed guidelines for glucose control in the perioperative period for LE vascular procedures to potentially mitigate complications.