Source: Picard & Tramer, Anesth Analg 2000; 90:963-9

Reviewer:  R. Prasad, MD

Comments:

• Not a new study, but many of the residents and CRNAs still seem surprised I do this.
• Best intervention: lidocaine 0.5mg/kg IV with tourniquet, 30-120sec before injection
• Number Needed to Treat (NNT) to prevent pain in one person who would have had pain if they had received placebo = 1.6 (60mg dose)
• i.e., reduces incidence from ~70% to ~10-12%
• Meperidine and metoclopramide with tourniquet also work, but not as well; probably b/c of their local anesthetic properties
• Speed on injection made no difference in incidence of pain
• There are a LOT of bad studies, with a variety of drugs

• Literature review, meta-analysis
• Studies reviewed, graded for quality (1-5, 5 best)
• Used RCTs looking at dichotomous data only (pain vs. no pain, not VAS of pain)
• Calculated NNT based on means from studies weighted by sample size

• 56 RCTs, 6264 pts
• Avg trial size 111pts. Median quality score 2
• Studies looked at
• lido before prop
• lido mixed with prop
• lido as Bier's block
• opioids (fent, alfent, meperidine before and with tourniquet)
• metoclopramide before, or with tourniquet
• cold, warm, or room temp prop
• other interventions (ondansetron, droperidol, nafamostat, mesilate, ketamine, ASA, ketorolac, prilocaine, MSO4, lido iontophoresis, dilution with blood or dextrose, long-chain TG, injection site) only tested in single trial: no conclusions drawn