Increased postoperative glucose variability is associated with adverse outcomes following orthopaedic surgery

Aims: The aim of this study was to examine the association between postoperative glycaemic variability and adverse outcomes following orthopaedic surgery. Patients and Methods: This retrospective study analyzed data on 12 978 patients (1361 with two operations) who underwent orthopaedic surgery at a single institution between 2001 and 2017. Patients with a minimum of either two postoperative measurements of blood glucose levels per day, or more than three measurements overall, were included in the study. Glycaemic variability was assessed using a coefficient of variation (CV). The length of stay (LOS), in-hospital complications, and 90-day readmission and mortality rates were examined. Data were analyzed with linear and generalized linear mixed models for linear and binary outcomes, adjusting for various covariates. Results: The cohort included 14 339 admissions, of which 3302 (23.0%) involved diabetic patients. Patients with CV values in the upper tertile were twice as likely to have an in-hospital complication compared with patients in the lowest tertile (19.4% versus 9.0%, p < 0.001), and almost five times more likely to die compared with those in the lowest tertile (2.8% versus 0.6%, p < 0.001). Results of the adjusted analyses indicated that the mean LOS was 1.28 days longer in the highest versus the lowest CV tertile (p < 0.001), and the odds of an in-hospital complication and 90-day mortality in the highest CV tertile were respectively 1.91 (p < 0.001) and 2.10 (p = 0.001) times larger than the odds of these events in the lowest CV tertile. These associations were significant even for non-diabetic patients. After adjusting for hypoglycaemia, the relationships remained significant, except that the CV tertile no longer predicted mortality in diabetics. Conclusion: These results indicate that higher glycaemic variability is associated with longer LOS and in-hospital complications. Glycaemic variability also predicted death, although that primarily held for non-diabetic patients in the highest CV tertile following orthopaedic surgery. Prospective studies should examine whether ensuring low postoperative glycaemic variability may reduce complication rates and mortality. Cite this article: Bone Joint J 2018;100-B:1125-32.