Minimum efficacy criteria for comparisons between treatments using individual patient meta-analysis of acute pain trials: examples of etoricoxib, paracetamol, ibuprofen, and ibuprofen/paracetamol combinations after third molar extraction.

We defined response in acute pain trials according to percentage of maximum possible efficacy. Minimum efficacy criteria (MEC) of 0%, or at least 15%, 30%, 50%, and 70% pain relief were used to examine stability over time using total pain relief and summed pain intensity difference (SPID), sex differences, and sensitivity. We used individual patient data from placebo-controlled third molar extraction trials: 4 with single-dose oral etoricoxib 120 mg, and 2 with paracetamol, ibuprofen, and ibuprofen plus paracetamol combinations. With etoricoxib, numbers needed to treat (NNTs) were stable between response levels of at least 15% (MEC15) and 50% pain relief (MEC50), and similar for total pain relief and SPID. NNTs were higher (worse) at extremes of MEC, especially with SPID. Results for women and men were similar. NNTs of lower efficacy treatments (paracetamol 500 and 1000 mg) rose rapidly at higher MEC. NNTs of high efficacy treatments (ibuprofen plus paracetamol combinations) showed greater separation at higher MEC. The highest degree of discrimination between treatments was with MEC50 and MEC70. Etoricoxib 120 mg (NNT for ≥50% maximum 6-hour pain relief 1.7) and ibuprofen 200/400 mg plus paracetamol 500/1000 mg (NNTs 1.5 and 1.6, respectively) produced the lowest (best) NNTs in the dental pain model. Timing of patient request for additional analgesia is an alternative analgesic efficacy outcome measure