Perioperative High-Dose Steroid in Insulin-Treated Patients With Diabetes Undergoing Fast-Track Hip and Knee Arthroplasty: Impact on Length of Stay and Discharge Blood Glucose Levels

BACKGROUND: Perioperative high-dose steroids are widely used to reduce postoperative inflammation, pain, and complications but remain controversial in patients with insulin-treated diabetes due to concerns about hyperglycemia and infections. This study aimed to examine the use of perioperative high-dose steroids and its effect on discharge glucose concentration, length of stay (LOS), and 30-day readmissions in insulin-treated patients with diabetes undergoing fast-track hip or knee arthroplasty.

METHODS: We conducted a prospective observational study in patients with insulin-treated diabetes undergoing fast-track hip or knee arthroplasty across eight Danish centers. Patient demographics, including diabetes treatments, perioperative steroid, discharge blood glucose concentration, LOS, and 30-day readmissions were prospectively extracted from the electronic health records through chart review. Our primary outcome was differences in mean discharge blood glucose concentration between patients with insulin-treated diabetes receiving perioperative high-dose steroid and no/antiemetic steroid. Secondary outcomes included the fraction with a LOS > 2 and 30-day readmissions between the groups. As a sensitivity analysis, discharge blood glucose concentration was compared between the insulin-treated patients and a propensity-score matched cohort of non-insulin-treated patients with diabetes all receiving high-dose steroids.

RESULTS: Of 292 registered patients with insulin-treated diabetes, 64% received high-dose steroids. Median discharge blood glucose was 11.0 mmol/L [IQR 8.9-13.5] in the high-dose steroid group versus 9.7 mmol/L [IQR 7.4-12] in the no/antiemetic steroid group (p = 0.011). LOS > 2 days occurred in 10.1% (95% CI: 6.6-15.3) of the high-dose group and 15.2% (95% CI: 9.6%-23.3%) of the no/antiemetic steroid group (p = 0.200). The 30-day readmission rates were 8% (95% CI: 4.8-13) and 8.3% (95% CI: 4.3-15.6), respectively (p = 0.923). Compared to propensity-matched patients with non-insulin-treated diabetes receiving high-dose steroids, patients with insulin-treated diabetes had a 2.1 mmol/L (95% CI: 1.3-2.8) higher mean discharge blood glucose concentration (p = < 0.001).

CONCLUSIONS: Glucose levels at discharge were slightly, but significantly higher in insulin-treated patients with diabetes receiving high-dose steroids compared to those receiving no/antiemetic doses and to non-insulin treated diabetes patients receiving high-dose steroids. However, there was no increase in fraction with LOS > 2 days or 30-day readmissions in the patients having high-dose steroids. Despite increased discharge glucose concentration, this suggests no evident safety concerns using high-dose steroids in insulin-treated diabetes patients having fast-track arthroplasty.

EDITORIAL COMMENT: This multicenter study addresses the debated use of a high-dose perioperative glucocorticoid in insulin-treated patients with diabetes undergoing fast-track arthroplasty. Despite modestly increased discharge glucose levels, no safety signals were observed regarding length of stay or readmissions. These findings support cautious use of a single high-dose perioperative glucocorticoid in this population. However, larger randomized studies are needed to confirm these findings.