Meta-analysis of lower perioperative blood glucose target levels for reduction of surgical-site infection

Abstract Background There is a clear association between hyperglycaemia and surgical-site infection (SSI). Intensive glucose control may involve a risk of hypoglycaemia, which in turn results in potentially severe complications. A systematic review was undertaken of studies comparing intensive versus conventional glucose control protocols in relation to reduction of SSI and other outcomes, including hypoglycaemia, mortality and stroke. Methods PubMed, Embase, CENTRAL, CINAHL and WHO databases from 1 January 1990 to 1 August 2015 were searched. Inclusion criteria were RCTs comparing intensive with conventional glucose control protocols, and reporting on the incidence of SSI. Meta-analyses were performed with a random-effects model, and meta-regression was subsequently undertaken. Targeted blood glucose levels, achieved blood glucose levels, and important adverse events were summarized. Results Fifteen RCTs were included. The summary estimate showed a significant benefit for an intensive compared with a conventional glucose control protocol in reducing SSI (odds ratio (OR) 0·43, 95 per cent c.i. 0·29 to 0·64; P < 0·001). A significantly higher risk of hypoglycaemic events was found for the intensive group compared with the conventional group (OR 5·55, 2·58 to 11·96), with no increased risk of death (OR 0·74, 0·45 to 1·23) or stroke (OR 1·37, 0·26 to 7·20). These results were consistent both in patients with and those without diabetes, and in studies with moderately strict and very strict glucose control. Conclusion Stricter and lower blood glucose target levels of less than 150 mg/dl (8·3 mmol/l), using an intensive protocol in the perioperative period, reduce SSI with an inherent risk of hypoglycaemic events but without a significant increase in serious adverse events.