Predicting postoperative analgesia outcomes: NNT league tables or procedure-specific evidence?

Number needed to treat (NNT) values have been recommended and used to assess efficacy of analgesics for acute pain management. However, the data analysed come from a variety of procedures, which may potentially hinder the interpretation of the NNT value for specific procedures. We reanalysed available NNT data with acetaminophen in relation to the magnitude of surgical injury. Acetaminophen was less effective for pain relief after orthopaedic procedures than after dental procedures. The relative risk ratio for more than 50% pain relief, compared with placebo, was only 1.87 compared with 3.77 (P<0.05). Although NNT can give a valuable overview of efficacy, this concept is not necessarily applicable to all types of surgery. We suggest that estimates of NNT should be related to specific surgical procedures.