“…Eight trials, 52,56,63,64,70 (one low risk of bias, 52 two trials included patients with diabetes, 56,78 four trials used dexamethasone equivalent doses ≥16 mg, 52,70,78,81 1220 patients) reported blood glucose levels (Figure, SDC 5.7, http://links.lww.com/EJA/A780). No significant difference in blood glucose was found between groups on POD 1 or 2 (POD 1: 0.2 mmol l −1 , 95% CI 0.0 to 0.5, I 2 = 56%, P = 0.08; POD 2: -0.1 mmol l −1 , 95% CI -0.3 to 0.2, I 2 = 45%, P = 0.66).…”