Perioperative hemoglobin A1c as a predictor of deep infection following single-level lumbar decompression in patients with diabetes

Background Context Although multiple studies have cited that diabetes mellitus as a risk factor decreased functional outcomes, increased infectious complications, and overall increased reoperation rate following degenerative lumbar spinal surgery, few have investigated how perioperative glycemic control influences such complications. Purpose The primary goal of the present study was to use a national database to evaluate the association of perioperative glycemic control as demonstrated by hemoglobin A1c (HbA1c) levels in patients with diabetes undergoing primary, single-level decompression without concomitant fusion with the incidence of deep postoperative infection requiring operative irrigation and debridement. Our secondary objective was to calculate a threshold level of HbA1c above which the risk of postoperative infection after lumbar decompression increases significantly in patients with diabetes. Study Design/Setting This is a retrospective case control database study, with Level III evidence. Patient Sample This study comprised private-payer patients with diabetes mellitus undergoing single-level lumbar decompression with an HbA1c laboratory value recorded in the database within 3 months of surgery. Outcome Measures The outcome examined in this study was deep infection following primary, single-level lumbar decompression requiring surgical intervention. Postoperative infection within 1 year of the index primary, single-level lumbar decompression was assessed using Current Procedural Terminology (CPT) procedure codes and the International Classification of Diseases, 9th Revision (ICD-9) diagnostic codes. Methods The Humana private-payer dataset from the PearlDiver database was used for this study. The database was queried for patients with diabetes mellitus undergoing primary, single-level lumbar decompression surgery using CPT codes. Patients with a diagnosis of diabetes mellitus who had an HbA1c level drawn within 3 months before or after their surgical date were then selected to form the study group using the ICD-9 diagnostic codes. Patients were then divided into groups based on their HbA1c level by increments of 0.5 mg/dL. The incidence of deep infection requiring operative intervention within 1 year for each HbA1c group was then identified using CPT and ICD-9 codes. A receiver operating characteristic (ROC) and area under the curve (AUC) analysis was performed to determine an optimal threshold value of the HbA1c above which the risk of postoperativeinfection was significantly increased. The threshold value was tested using a multivariable binomial logistic regression analysis. Results A total of 5,194 patients who underwent primary, single-level lumbar decompression with diabetes and a perioperative HbA1c recorded within 3 months of surgery were included in the study. The rate of infection ranged from a low of 0.5% up to 3.5% for patients with an HbA1c level >11.0 mg/dL (p=.012). The inflection point of the ROC curve corresponded to an HbA1c level above 7.5 mg/dL (p=.01, AUC=0.71, specificity=70%, sensitivity=53%). After controlling for patient demographics and medical comorbidities, patients with an HbA1c level of 7.5 mg/dL or above had a significantly higher risk for deep infection compared with patients below this threshold (odds ratio: 2.9, 95% confidence interval: 1.8–4.9, p<.0001). Conclusions The risk of deep postoperative infection requiring surgical intervention following single-level lumbar decompression in patients with diabetes mellitus increases as the perioperative HbA1c increases. The ROC and multivariable regression analyses determined that a perioperative HbA1c above 7.5 mg/dL could serve as a threshold for a significantly increased risk of deep postoperative infection following lumbar decompression.