Not All Patients With Diabetes Have the Same Risks: Perioperative Glycemic Control Is Associated With Postoperative Infection Following Knee Arthroscopy

Abstract

To determine the association between glycemic control and infection following knee arthroscopy, and to determine the clinical utility of a threshold HbA1c level. A national database identified patients who underwent knee arthroscopy from 2007 to 2016. Patients with concomitant open portions, more complex knee procedures, procedures performed for infection and patients with prior septic knee arthritis were excluded. Patients with an HbA1c level checked within 3months of surgery were compared to control groups of nondiabetics and diabetics undergoing knee arthroscopy without a perioperative HbA1c. The study group was then stratified based on their HbA1c. The incidence of deep infection within 6months was identified and compared to those in the control groups. A subgroup analysis was performed to investigate any trend in the timing of postoperative infection. A receiver operating characteristic (ROC) analysis was performed to determine and test a threshold value of HbA1c. 13,470 study patients were included, with an overall rate of infection of 0.33%. The rate of infection ranged from 0.25% to 1.03%. The rate of infection in study patients was greater than the nondiabetes control group's (P < .0001) and not significantly different from that in the nonstudy diabetes control (P= .765). The inflection point of the ROC curve corresponded to an HbA1c level of 8.0mg/dL (P= .006, specificity= 76%, sensitivity= 44%, area under curve [AUC]= 0.619). The risk of infection following knee arthroscopy increases as the perioperative HbA1c increases. ROC analysis determined that an HbA1c above 8.0mg/dL could serve as a threshold level; however, the AUC and low sensitivity reflected the poor utility of this test as an independent predictor for infection. Level III, retrospective comparative study.