Fifty-six studies with a total of 6264 participants were included.
Quality of the included trials.
Most of the trials were of rather a poor quality. Blinding was often inadequate (no further details of study quality such as range of scores was provided).
Effectiveness of the interventions.
On average, 70% of the patients reported pain on injection. Fifteen drugs, 12 physical measurements, and combinations were tested. With IV lidocaine 40 mg, given with a tourniquet 30 to 120 s before the injection of propofol, the number of patients needed to be treated (NNT) to prevent pain in one who would have had pain had they received placebo was 1.6. The closest to this came meperidine 40 mg with tourniquet (NNT 1.9) and metoclopramide 10 mg with tourniquet (NNT 2.2). With lidocaine mixed with propofol, the best NNT was 2.4; with IV alfentanil or fentanyl, it was 3 to 4. IV lidocaine before the injection of propofol was less analgesic. Temperature had no significant effect. There was a lack of data for all other interventions to allow meaningful conclusions. The diameter of venous catheters and speed of injection had no impact on pain.