Duloxetine for the treatment acute postoperative pain in adult patients: A systematic review with meta-analysis.

BACKGROUND

Duloxetine administered during the acute perioperative period has been associated with lesser postoperative pain and analgesic consumption.

STUDY OBJECTIVES

The study aimed to quantify the pooled effects of duloxetine on postoperative pain, analgesic consumption, and side-effects in the first 48 postoperative (PO) hours.

DESIGN

Systematic review with meta-analysis.

SETTING

Postoperative pain management.

PATIENTS

Adult patients undergoing elective surgery. Search strategy and study selection. Medline, Cochrane, EMBASE, CENTRAL, and Web of Science were searched without language restrictions for prospective, parallel randomized controlled trials comparing duloxetine to placebo for the management of postoperative pain in adult patients.

MEASUREMENTS

Pain scores (11-point scales), opioid consumption (i.v. morphine equivalents), and frequency of side-effects were compared between duloxetine and placebo. Effect sizes were summarized as mean differences (MD), standardized mean differences (SMD) or risk ratios (RR) with the respective 95% confidence intervals (95% CI). Grading of Recommendations Assessment, Development, and Evaluation (GRADE) criteria were used to classify the quality of evidence.

RESULTS

Thirteen studies were included. Duloxetine decreased pain at 24 h (MD = -0.66 points; 95% CI = -1.14 to -0.19 points; SMD = -0.59; 95% CI = -1.06 to -0.12; p = 0.01; I2 = 88%), and at 48 PO hours (MD = -0.90 points; 95% CI = -1.54 to -0.26 points; SMD = -0.66; 95% CI = -0.94 to -0.38; p = 0.01; I2 = 93%); and opioid consumption at 24 PO hours (MD = -8.21 mg; 95% CI = -13.32 mg to -3.10 mg; SMD = -2.17; 95% CI = -3.10 to -1.24; p < 0.001; I2 = 95%), and at 48 PO hours (MD = 7.71 mg; 95% CI = -13.86 mg to -1.56 mg; SMD = -2.13; 95% CI = -3.51 to -0.75; p = 0.02; I2 = 97%). Duloxetine did not affect the prevalence of postoperative nausea and/or vomiting (PONV) pruritus, headache or dizziness. High inter-study heterogeneity and within-study bias resulted in very-low quality of evidence for the primary outcomes.

CONCLUSIONS

Although statistically significant effects of duloxetine were found on postoperative pain and opioid consumption during the first 48 postoperative hours, the effect sizes were below the expected minimal clinically relevant differences. Also, high risk-of-bias and inter-study heterogeneity caused the very-low quality of evidence (GRADE). We conclude that the currently available evidence does not support the clinical use of duloxetine for the management of acute postoperative pain.