well as Anna for allowing us to present our disclosures during the app. So while diabetes has been identified as an independent risk factor for post-operative infection, inferior clinical outcomes, and failure of bone tendon healing following arthroscopic rotator cuff repair, limited studies to date have evaluated the effect of perioperative glycemic control in diabetics undergoing rotator cuff repair. Hemoglobin A1C is commonly used as a measure of glycemic index, and while lower extremity total joint arthroplasty has used these values to study hemoglobin A1C levels and their association with adverse events, no similar studies have used this value to determine if similar associations exist following arthroscopic rotator cuff repair. Therefore, the objective of the current study was to determine the association between glycemic control and adverse events, namely infection, revision surgery, and post-operative stiffness following arthroscopic rotator cuff repair. To accomplish this, we identified all patients undergoing arthroscopic rotator cuff repair with the following CPT code in the Pearl Diver database. We excluded any patients undergoing more open or complex procedures and any procedures being done for infection. Of these patients, we identified all of those diabetics with a hemoglobin A1C within three months of surgery, and we subsequently grouped these into six mutually exclusive groups based on hemoglobin A1C level by increments of one milligram per deciliter from less than six to greater than 10. Our primary outcome of interest was infection within six months, and this was identified using CPT and ICD-9 codes for post-operative infection or septic arthritis or procedures for these indications. Subsequently, rates of deep infection were plotted versus hemoglobin A1C level group to construct a receiver-operator characteristic curve to determine a threshold level to be tested. This level was then tested using a multivariable binomial logistic regression analysis comparing patients below the threshold with those above for the following outcomes of infection, ipsilateral revision shoulder surgery, and lysis of adhesions and manipulation under anesthesia. So in total, we identified more than 3,700 patients with a history of diabetes who underwent rotator cuff repair with a hemoglobin A1C recorded. The overall incidence of infection was 0.4% and was less than 0.3% with those with a hemoglobin A1C of less than six and 1.14 for those with a hemoglobin A1C of greater than 10. The inflection point on the ROC curve was eight milligrams per deciliter, and after controlling for demographics and comorbidities, patients with a hemoglobin A1C of greater than eight has significantly higher risk of deep infection following rotator cuff repair. In addition, there was no significant increased risk of revision surgery or lysis of adhesions with patients above this threshold level. This graph represents infection risk stratified by hemoglobin A1C, and that is our threshold value that was tested. And this represents our ROC curve with the inflection point at eight, which was significant. However, the specificity and sensitivity were 61% and 59% with an area under the curve of 0.64. So while our study does provide evidence that an elevated hemoglobin A1C is associated with infection following rotator cuff repair, our AUC of 0.64 indicates that the accuracy of this threshold as an independent predictor for infection after rotator cuff repair is poor and of limited clinical utility, and there are likely other factors that influence the risk of infection that are unable to be accounted for in this analysis. And so while we believe that a hemoglobin A1C level of greater than eight is helpful and should be referenced when stratifying a patient's infection risk, it should be interpreted in the context of other risk factors that also affect infection. Thank you.

glycemic control

adverse events

infection

revision surgery

post-operative stiffness