Inappropriate conclusion from meta-analysis of randomized controlled trials.

To the Editor: Paul et al. should be congratulated for examining the body of evidence on the comparison of patient-controlled analgesia opioid or epidural analgesia with femoral nerve blockade for analgesia after a common operation, total knee arthroplasty. A meta-analysis, by pooling and analyzing statistically the results from several combinable studies, can increase the statistical power to detect small but clinically important treatment harms and effects. Paul et al. clearly demonstrated that femoral nerve blockade improved analgesia and reduced opioid consumption compared with intravenous patient-controlled analgesia techniques. The quality of the systematic review performed by Paul et al. is high, as evidenced by the comprehensive and systematic search of the literature and explanation of statistical tests and models used for their analyses. Furthermore, the authors (and the accompanying editorial) clearly emphasize the limitations of their meta-analysis regarding comparison of single shot femoral nerve blockade (SSFNB) with continuous femoral nerve block, including that: