RESUMO
BACKGROUND: Perioperative hyperglycemia can have an even more detrimental effect on postoperative outcomes in patients without diabetes than in patients with diabetes, but it has not been established if the treatment of patients without diabetes is safe and effective. We hypothesized that sliding-scale insulin for severe postoperative hyperglycemia (glucose ≥180 mg/dL) could lower mean postoperative glucose levels and minimize short-term complications in patients without diabetes undergoing major joint replacement. METHODS: In a prospective study group, 1,398 consecutive patients, with and without diabetes, undergoing joint replacement were monitored and treated for hyperglycemia and were compared with 886 historical, less frequently monitored controls. The primary outcome was the mean glucose level in patients with and without diabetes within 48 hours after the surgical procedure. Two secondary outcomes could be examined only in the prospective study group, which, by design, had much more frequent glucose sampling and insulin use than the historical controls. First, the contribution of comorbidities and procedural factors to postoperative hyperglycemia in patients without diabetes was assessed with multivariable linear regression. Second, the ability of insulin treatment to reduce complications in patients without diabetes who developed hyperglycemia was evaluated. RESULTS: In comparison with 886 historical controls, enhanced glucose management lowered the mean glucose (and standard deviation) from 129 ± 28 mg/dL to 123 ± 23 mg/dL for patients without diabetes (p = 0.041). Multivariable linear regression revealed factors that contributed to elevated mean glucose in patients without diabetes: preoperative fasting glucose (p < 0.001), perioperative steroid use (p < 0.001), general anesthesia (p < 0.001), procedure duration (p = 0.003), and transfusion (p 0.008). Of 968 patients without diabetes, 203 developed severe hyperglycemia. The recommended insulin coverage was given to 129 of these patients, and 74 patients did not receive it for various clinical reasons. Insulin treatment reduced the frequency of positive cultures from any site (p = 0.025) and a composite of positive cultures and readmissions (p = 0.006) in comparison with no insulin treatment. No patient without diabetes who received insulin experienced mild or severe hypoglycemia. CONCLUSIONS: Postoperative hyperglycemia is frequent in patients without diabetes after orthopaedic surgery, but an enhanced glucose management program can lower mean postoperative glucose levels. The treatment of hyperglycemia in patients without diabetes reduced short-term complications and was associated with minimal side effects. LEVEL OF EVIDENCE: Therapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.
