Effect of perioperative insulin infusion on surgical morbidity and mortality: Systematic review and meta-analysis of randomized trials

Gunjan Y. Gandhi, Mohammad H Murad, David N. Flynn, Patricia J. Erwin, Alexandre B. Cavalcante, Henning Bay Nielsen, Sarah E. Capes, Kristian Thorlund, Victor Manuel Montori, P. J. Devereaux

Research output: Contribution to journalArticle

69 Citations (Scopus)

Abstract

OBJECTIVE: To conduct a systematic review and meta-analysis of randomized controlled trials (RCTs) to evaluate the effect of perioperative insulin infusion on outcomes important to patients. PATIENTS AND METHODS: We used 6 search strategies including an electronic database search of MEDLINE, EMBASE, and Cochrane CENTRAL, from their inception up to May 1, 2006, and included RCTs of perioperative insulin infusion (with or without glucose targets) measuring outcomes in patients undergoing any surgery. Pairs of reviewers working independently assessed the methodological quality and characteristics of included trials and abstracted data on perioperative outcomes (ie, outcomes that occurred during hospitalization or within 30 days of surgery). RESULTS: We identified 34 eligible trials. In the 14 trials that assessed mortality, there were 68 deaths among 2192 patients randomized to insulin infusion compared with 98 deaths among 2163 patients randomized to control therapy (random-effects pooled relative risk, 0.69; 95% confidence interval [CI], 0.51-0.94; 99% CI, 0.46-1.04; I2, 0%; 95% CI, 0.0%-47.4%). Hypoglycemia increased in the intensively treated group (20 trials, 119/1470 patients in insulin infusion vs 48/1476 patients in control group; relative risk, 2.07; 95% CI, 1.29-3.32; 99% CI, 1.09-3.88; I2, 31.5%; 95% CI, 0.0%-59.0%). No significant effect was seen in any other outcomes. The available mortality data represent only 40% of the optimal information size required to reliably detect a plausible treatment effect; potential methodological and reporting biases weaken inferences. CONCLUSION: Perioperative insulin infusion may reduce mortality but increases hypoglycemia in patients who are undergoing surgery; however, mortality results require confirmation in large and rigorous RCTs.

Original languageEnglish (US)
Pages (from-to)418-430
Number of pages13
JournalMayo Clinic Proceedings
Volume83
Issue number4
DOIs
StatePublished - 2008

Fingerprint

Meta-Analysis
Insulin
Morbidity
Confidence Intervals
Mortality
Randomized Controlled Trials
Hypoglycemia
Ambulatory Surgical Procedures
MEDLINE
Hospitalization
Databases
Glucose
Control Groups
Therapeutics

ASJC Scopus subject areas

  • Medicine(all)

Cite this

Effect of perioperative insulin infusion on surgical morbidity and mortality : Systematic review and meta-analysis of randomized trials. / Gandhi, Gunjan Y.; Murad, Mohammad H; Flynn, David N.; Erwin, Patricia J.; Cavalcante, Alexandre B.; Nielsen, Henning Bay; Capes, Sarah E.; Thorlund, Kristian; Montori, Victor Manuel; Devereaux, P. J.

In: Mayo Clinic Proceedings, Vol. 83, No. 4, 2008, p. 418-430.

Research output: Contribution to journalArticle

Gandhi, GY, Murad, MH, Flynn, DN, Erwin, PJ, Cavalcante, AB, Nielsen, HB, Capes, SE, Thorlund, K, Montori, VM & Devereaux, PJ 2008, 'Effect of perioperative insulin infusion on surgical morbidity and mortality: Systematic review and meta-analysis of randomized trials', Mayo Clinic Proceedings, vol. 83, no. 4, pp. 418-430. https://doi.org/10.4065/83.4.418
Gandhi, Gunjan Y. ; Murad, Mohammad H ; Flynn, David N. ; Erwin, Patricia J. ; Cavalcante, Alexandre B. ; Nielsen, Henning Bay ; Capes, Sarah E. ; Thorlund, Kristian ; Montori, Victor Manuel ; Devereaux, P. J. / Effect of perioperative insulin infusion on surgical morbidity and mortality : Systematic review and meta-analysis of randomized trials. In: Mayo Clinic Proceedings. 2008 ; Vol. 83, No. 4. pp. 418-430.
@article{a5e1a173fe18496f98984877e49bdcdd,
title = "Effect of perioperative insulin infusion on surgical morbidity and mortality: Systematic review and meta-analysis of randomized trials",
abstract = "OBJECTIVE: To conduct a systematic review and meta-analysis of randomized controlled trials (RCTs) to evaluate the effect of perioperative insulin infusion on outcomes important to patients. PATIENTS AND METHODS: We used 6 search strategies including an electronic database search of MEDLINE, EMBASE, and Cochrane CENTRAL, from their inception up to May 1, 2006, and included RCTs of perioperative insulin infusion (with or without glucose targets) measuring outcomes in patients undergoing any surgery. Pairs of reviewers working independently assessed the methodological quality and characteristics of included trials and abstracted data on perioperative outcomes (ie, outcomes that occurred during hospitalization or within 30 days of surgery). RESULTS: We identified 34 eligible trials. In the 14 trials that assessed mortality, there were 68 deaths among 2192 patients randomized to insulin infusion compared with 98 deaths among 2163 patients randomized to control therapy (random-effects pooled relative risk, 0.69; 95{\%} confidence interval [CI], 0.51-0.94; 99{\%} CI, 0.46-1.04; I2, 0{\%}; 95{\%} CI, 0.0{\%}-47.4{\%}). Hypoglycemia increased in the intensively treated group (20 trials, 119/1470 patients in insulin infusion vs 48/1476 patients in control group; relative risk, 2.07; 95{\%} CI, 1.29-3.32; 99{\%} CI, 1.09-3.88; I2, 31.5{\%}; 95{\%} CI, 0.0{\%}-59.0{\%}). No significant effect was seen in any other outcomes. The available mortality data represent only 40{\%} of the optimal information size required to reliably detect a plausible treatment effect; potential methodological and reporting biases weaken inferences. CONCLUSION: Perioperative insulin infusion may reduce mortality but increases hypoglycemia in patients who are undergoing surgery; however, mortality results require confirmation in large and rigorous RCTs.",
author = "Gandhi, {Gunjan Y.} and Murad, {Mohammad H} and Flynn, {David N.} and Erwin, {Patricia J.} and Cavalcante, {Alexandre B.} and Nielsen, {Henning Bay} and Capes, {Sarah E.} and Kristian Thorlund and Montori, {Victor Manuel} and Devereaux, {P. J.}",
year = "2008",
doi = "10.4065/83.4.418",
language = "English (US)",
volume = "83",
pages = "418--430",
journal = "Mayo Clinic Proceedings",
issn = "0025-6196",
publisher = "Elsevier Science",
number = "4",

}

TY - JOUR

T1 - Effect of perioperative insulin infusion on surgical morbidity and mortality

T2 - Systematic review and meta-analysis of randomized trials

AU - Gandhi, Gunjan Y.

AU - Murad, Mohammad H

AU - Flynn, David N.

AU - Erwin, Patricia J.

AU - Cavalcante, Alexandre B.

AU - Nielsen, Henning Bay

AU - Capes, Sarah E.

AU - Thorlund, Kristian

AU - Montori, Victor Manuel

AU - Devereaux, P. J.

PY - 2008

Y1 - 2008

N2 - OBJECTIVE: To conduct a systematic review and meta-analysis of randomized controlled trials (RCTs) to evaluate the effect of perioperative insulin infusion on outcomes important to patients. PATIENTS AND METHODS: We used 6 search strategies including an electronic database search of MEDLINE, EMBASE, and Cochrane CENTRAL, from their inception up to May 1, 2006, and included RCTs of perioperative insulin infusion (with or without glucose targets) measuring outcomes in patients undergoing any surgery. Pairs of reviewers working independently assessed the methodological quality and characteristics of included trials and abstracted data on perioperative outcomes (ie, outcomes that occurred during hospitalization or within 30 days of surgery). RESULTS: We identified 34 eligible trials. In the 14 trials that assessed mortality, there were 68 deaths among 2192 patients randomized to insulin infusion compared with 98 deaths among 2163 patients randomized to control therapy (random-effects pooled relative risk, 0.69; 95% confidence interval [CI], 0.51-0.94; 99% CI, 0.46-1.04; I2, 0%; 95% CI, 0.0%-47.4%). Hypoglycemia increased in the intensively treated group (20 trials, 119/1470 patients in insulin infusion vs 48/1476 patients in control group; relative risk, 2.07; 95% CI, 1.29-3.32; 99% CI, 1.09-3.88; I2, 31.5%; 95% CI, 0.0%-59.0%). No significant effect was seen in any other outcomes. The available mortality data represent only 40% of the optimal information size required to reliably detect a plausible treatment effect; potential methodological and reporting biases weaken inferences. CONCLUSION: Perioperative insulin infusion may reduce mortality but increases hypoglycemia in patients who are undergoing surgery; however, mortality results require confirmation in large and rigorous RCTs.

AB - OBJECTIVE: To conduct a systematic review and meta-analysis of randomized controlled trials (RCTs) to evaluate the effect of perioperative insulin infusion on outcomes important to patients. PATIENTS AND METHODS: We used 6 search strategies including an electronic database search of MEDLINE, EMBASE, and Cochrane CENTRAL, from their inception up to May 1, 2006, and included RCTs of perioperative insulin infusion (with or without glucose targets) measuring outcomes in patients undergoing any surgery. Pairs of reviewers working independently assessed the methodological quality and characteristics of included trials and abstracted data on perioperative outcomes (ie, outcomes that occurred during hospitalization or within 30 days of surgery). RESULTS: We identified 34 eligible trials. In the 14 trials that assessed mortality, there were 68 deaths among 2192 patients randomized to insulin infusion compared with 98 deaths among 2163 patients randomized to control therapy (random-effects pooled relative risk, 0.69; 95% confidence interval [CI], 0.51-0.94; 99% CI, 0.46-1.04; I2, 0%; 95% CI, 0.0%-47.4%). Hypoglycemia increased in the intensively treated group (20 trials, 119/1470 patients in insulin infusion vs 48/1476 patients in control group; relative risk, 2.07; 95% CI, 1.29-3.32; 99% CI, 1.09-3.88; I2, 31.5%; 95% CI, 0.0%-59.0%). No significant effect was seen in any other outcomes. The available mortality data represent only 40% of the optimal information size required to reliably detect a plausible treatment effect; potential methodological and reporting biases weaken inferences. CONCLUSION: Perioperative insulin infusion may reduce mortality but increases hypoglycemia in patients who are undergoing surgery; however, mortality results require confirmation in large and rigorous RCTs.

UR - http://www.scopus.com/inward/record.url?scp=42149174900&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=42149174900&partnerID=8YFLogxK

U2 - 10.4065/83.4.418

DO - 10.4065/83.4.418

M3 - Article

C2 - 18380987

AN - SCOPUS:42149174900

VL - 83

SP - 418

EP - 430

JO - Mayo Clinic Proceedings

JF - Mayo Clinic Proceedings

SN - 0025-6196

IS - 4

ER -