In Reply

Our study¹  was appropriately powered for the primary comparison among randomized groups. Even in moderately large randomized trials, there can be important differences in baseline patient characteristics. In our case, for example, there were 19 of 58 patients in the placebo group who took opioids chronically, whereas only 9 of 57 patients in the lidocaine group did. Fry and Davis ask whether this difference might have influenced our results.

In a post hoc analysis, we therefore first assessed both the relation between chronic opioid use and postoperative morphine equivalent dose. The ratio (95% CI) of mean postoperative IV morphine equivalent dose comparing chronic opioid users with nonusers was estimated at 1.31 (0.76 to 2.24). We then assessed the differential treatment effect among chronic opioid users and among nonusers. The ratio (95% CI) of mean postoperative IV morphine equivalent dose comparing chronic opioid users randomized to lidocaine with chronic opioid users randomized to placebo was 0.69 (0.28 to 1.67). For nonusers, this ratio was 0.84 (0.47 to 1.51).

In our main analysis, we did not adjust for chronic opioid use. A separate post hoc analysis, which adjusts for chronic opioid use, reveals an estimated ratio of means (lidocaine vs. placebo) of 0.79 (0.49 to 1.28) (the estimate from the main analysis was 0.75 [0.47, 1.20]).

There is thus no compelling indication that the chance imbalance on chronic opioid use substantively influenced our conclusion that IV lidocaine is analgesic for complex spine surgery and noninferior on postoperative opioid consumption.