Sheryl Warttig, Phil Alderson, Sharon R Lewis, Andrew F Smith

Inadvertent perioperative hypothermia (a drop in core temperature to below 36°C) occurs because normal temperature regulation is disrupted during surgery, mainly because of the effects of anaesthetic drugs and exposure of the skin for prolonged periods. Many different ways of maintaining body temperature have been proposed, one of which involves administration of intravenous nutrients during the perioperative period that may reduce heat loss by increasing metabolism, thereby increasing heat production. To assess the effectiveness of preoperative or intraoperative intravenous nutrients in preventing perioperative hypothermia and its complications during surgery in adults. We searched the Cochrane Central Register of Controlled Trials (CENTRAL; November 2015) in the Cochrane Library; MEDLINE, Ovid SP (1956 to November 2015); Embase, Ovid SP (1982 to November 2015); the Institute for Scientific Information (ISI) Web of Science (1950 to November 2015); and the Cumulative Index to Nursing and Allied Health Literature (CINAHL, EBSCO host; 1980 to November 2015), as well as the reference lists of identified articles. We also searched the Current Controlled Trials website and ClincalTrials.gov. Randomized controlled trials (RCTs) of intravenous nutrients compared with control or other interventions given to maintain normothermia in adults undergoing surgery. Two review authors extracted data and assessed risk of bias for each included trial, and a third review author checked details if necessary. We contacted some study authors to request additional information. We included 14 trials (n = 565), 13 (n = 525) of which compared intravenous administration of amino acids to a control (usually saline solution or Ringer's lactate). The remaining trial (n = 40) compared intravenous administration of fructose versus a control. We noted much variation in these trials, which used different types of surgery, variable durations of surgery, and different types of participants. Most trials were at high or unclear risk of bias owing to inappropriate or unclear randomization methods, and to unclear participant and assessor blinding. This may have influenced results, but it is unclear how results might have been influenced.No trials reported any of our prespecified primary outcomes, which were risk of hypothermia and major cardiovascular events. Therefore, we decided to analyse data related to core body temperature instead as a primary outcome. It was not possible to conduct meta-analysis of data related to amino acid infusion for the 60-minute and 120-minute time points, as we observed significant statistical heterogeneity in the results. Some trials showed that higher temperatures were associated with amino acids, but not all trials reported statistically significant results, and some trials reported the opposite result, where the amino acid group had a lower core temperature than the control group. It was possible to conduct meta-analysis for six studies (n = 249) that provided data relating to the end of surgery. Amino acids led to a statistically significant increase in core temperature in comparison to those receiving control (MD = 0.46°C 95% CI 0.33 to 0.59; I(2) 0.0%; random-effects; moderate quality evidence).Three trials (n = 155) reported shivering as an outcome. Meta-analysis did not show a clear effect, and so it is uncertain whether amino acids reduce the risk of shivering (RR 0.36, 95% CI 0.13 to 1.00; I(2) = 93%; random-effects model; very low-quality evidence). Intravenous amino acids may keep participants up to a half-degree C warmer than the control. This difference was statistically significant at the end of surgery, but not at other time points. However, the clinical importance of this finding remains unclear. It is also unclear whether amino acids have any effect on the risk of shivering and if intravenous nutrients confer any other benefits or harms, as high-quality data about these outcomes are lacking.