Perioperative Risk of and Monitoring for Hyperglycemia in Spine Surgery

Abstract

Perioperative hyperglycemia occurs frequently among patients with and without diabetes and is associated with increased morbidity and mortality. The purpose of this study was to describe perioperative hyperglycemia and to evaluate potential factors associated with postoperative hyperglycemia in a retrospective cohort of patients admitted following planned spine surgery between April 2010 and October 2013. Methods: Data were collected from electronic medical records for variables including age, gender, ethnicity, insurance status, diabetes status, HbA1c, ASA class, type and duration of surgery, and estimated blood loss. Glucose data were retrieved for morning of surgery, intraoperatively and postoperatively. Log-linear regression was used to assess for associations with postoperative hyperglycemia. Results: Preoperative glucose screening occurred in 71% of those with diabetes and 7% of those without. Intraoperative glucose testing occurred in 81% of those with diabetes and 57% of those without. Postoperative monitoring occurred in 80% of those with diabetes and 74% of those without. Early postoperative hyperglycemia was identified in 65% of cases with diabetes and 6% of those without diabetes. Most cases (95%) of postoperative hyperglycemia were identified by postoperative day two. Early postoperative hyperglycemia was independently associated with preoperative glucose ≥140mg/ml (RR 1.34) and A1c > 8 (RR 1.26) among those with diabetes. Among those without diabetes, early postoperative hyperglycemia was independently associated with procedure duration ≥6 hours (RR 6.20) and intraoperative glucose ≥180mg/dL (RR 3.53). Median length of stay for those with early postoperative hyperglycemia was 6.8 days compared to 4.2 days for those without hyperglycemia. Linear regression revealed that early postoperative hyperglycemia was independently associated with longer length of stay in this study sample (p<.01). Conclusion: In current practice, glucose is inconsistently and insufficiently monitored. Determining the appropriate frequency of monitoring and delineating risk predictors will be essential to improving identification of perioperative hyperglycemia and for targeting interventions. These results suggest that among those with diabetes, poor preoperative glucose control represents a target for intervention to reduce the risk of postoperative hyperglycemia and subsequent potential complications. Among those without diabetes, more thorough data is needed to inform best practices. However, these results suggest that procedure duration and intraoperative hyperglycemia are risk factors for patients without diabetes. Protocols that include preoperative screening and postoperative monitoring for the first 48 hours will identify the vast majority of patients who develop postoperative hyperglycemia.