| Methods |
Randomized double‐blind (blinding of patient, counselor and physician for phase 1 and 2, as well as nurse for phase 1) controlled trial.
Phase 3: no blinding, randomization by procedure day.
Phase 4: no blinding.
Free‐standing urban abortion clinic in British Columbia, Canada. |
| Participants |
Total of 480 pregnant patients for phase 1, 2 and 4. Phase 3 with 139 pregnant women. First trimester.
Exclusion criteria: unable to understand or unwilling to sign the consent form. Allergic to lidocaine, bupivacaine or ibuprofen. |
| Interventions |
Phase 1 Group 1: 600g ibuprofen 30 minutes prior to procedure
Phase 1 Group 2: placebo 30 minutes prior to procedure
All participants in phase 1: paracervical block (PCB) with 20ml 1% lidocaine (10ml injected in 4 to 6 sites around the cervix and 5ml each between 3 and 4 o'clock and between 8 and 9 o'clock, about 1 inch deep from the reflection of the vagina into the lower uterine segment, and as decsribed in their included study from 1992).
Phase 2 Group 1: PCB with 20ml plain 1% lidocaine
Phase 2 Group 2: PCB with 20ml 1% lidocaine buffered with 2ml 8.4% sodium bicarbonate
Phase 2 Group 3: PCB wih 20 ml 0.25% bupivacaine
Phase 3: comparison of waiting time 1, 3, 10 or 20 minutes after administrating PCB with bupivacaine
Phase 4: fast (30 sec) versus slow (60 sec) injection of buffered lidocaine for PCB on right or left side of cervix (randomized which to speed and side which was first injected). No waiting time. between injection and procedure.
All participants: premedication with lorazepam 0.5‐1mg SL per patient request 30 minutes prior to procedure. Vacuum aspiration followed by sharp curettage. |
| Outcomes |
Phase 1: pain with procedure, measured at end of procedure. Pain 30 minutes after the procedure.
Phase 2: pain with injection and with procedure.
Phase 3: pain with procedure.
Phase 4: pain with injection of each side.
Pain measured on a verbal 11‐point pain scale (0 = no pain to 10 = worst pain you can imagine) in phase 1 to 3, and with a 10cm long visual anlog pain scale in phase 4. |
| Notes |
Per power analysis 65 patients per group required to detect mean pain score difference of 1 with a standard deviation of 2 at the 0.05 significance level. Not all groups had that many patients.
Per e‐mail communication with Dr. Wiebe: randomization using tables of computer‐generated random numbers, allocation concealment with opaque numbered envelopes.
Phase 3 was randomized by procedure day, which is not adequate randomization.
No major complicaions reported. |
| Risk of bias |
| Bias |
Authors' judgement |
Support for judgement |
| Allocation concealment? |
Low risk |
A ‐ Adequate |
