Optimal timing for checking glucose level after total knee arthroplasty in patients with or without diabetes

Abstract

Introduction: Perioperative hyperglycaemia is an important risk factor for periprosthetic joint infection. However, controversies exist in the hyperglycaemic threshold and the optimal timing of glucose measurements. The primary aim was to determine the best timing of glucose measurement to detect hyperglycaemia. The secondary aim was to characterise the postoperative glucose profile and whether preoperative glucose control predicts postoperative hyperglycaemia. Methods: Primary total knee arthroplasty (TKA) from January to May 2019 were reviewed. Postoperative glucose was measured on day (D)0 at 5 pm and 10 pm and on D1 and D2 at 7 am and 5 pm. Hyperglycaemia was defined using strict (7 mmol/L), intermediate (7.6 mmol/L), and lenient (10.8 mmol/L) thresholds. Results: Seventy-six primary TKA patients reviewed. Hyperglycaemia on D0 at 10 pm for 7.6 mmol/L and 10.8 mmol/L thresholds was 100% and 78%, respectively and were significantly higher than other time points (p=0.009; p=0.001). Mean postoperative glucose peaked on D0 at 10 pm, then gradually tail down for diabetes, pre-diabetes, and non-diabetes. On D1, glucose values were significantly higher in diabetes and pre-diabetes than non-diabetes (p=0.003). On the night of surgery (D0 at 5 pm and 10 pm), glucose levels were not significantly different between diabetes, pre-diabetes, and non-diabetes. The haemoglobin A1c level was linearly correlated with both maximum and mean postoperative glucose (p<0.0001) with an R2 value of 0.304 and 0.325, respectively. Conclusion: Checking glucose at 10 pm on D0 was most sensitive to detect hyperglycaemia. Although patients with diabetes were at higher risk of postoperative hyperglycaemia, patients without diabetes were also susceptible to postoperative hyperglycaemia at this time. Optimising preoperative haemoglobin A1c is associated with lower postoperative glucose levels.