STATISTICS Formal meta-analysis pooling of the efficacy results was not performed because: 1. the IV acetaminophen dosing regimen varied among studies (eg, a single 1-g dose given either before surgery or during surgery, 1 g administered every 6 hours for 24 hours, or 1 g given once a day every day for 7 days). 2. the studies had different primary end points:pain score at rest or with movement (11 studies),
opioid reduction(5studies),need for rescue anal-gesic (3 studies), pain relief on a 5-point categorical scale (2 studies), or difference between mean pain scores(1study).Also,duration of follow-up ranged from less than 1 hour to 3 days, with median of 1 day. Not enough articles were available to separate surgery types into those associated with mild, moderate, and severe pain. Also, it was not possible to assess statistically whether IV acetaminophen affected the incidence of opioid side effects, because the end point was either not reported (eg, postoperative nausea and vomiting [PONV] not reported in n = 5 studies and pruritis not reported in n = 13 studies), definitions differed (eg, PONV on visual analog scale, nausea only, vomiting only, or nausea and vomiting combined), or end points occurred infrequently when measured such that the trials were not appropriately powered to assess side effect differences.