We thank Dr. Pace for selecting our publication highlighting the lack of difference between liposomal and plain bupivacaine, and presenting alternative findings that are based on the differences between the frequentist and Bayesian approaches for meta-analyses. We present in this reply a few of our thoughts on the different analyses.

First, while we have limited awareness of the use of the Bayesian method as a mainstream approach in meta-analyses, we are familiar with the nascent  Bayesian statistics that are slowly gaining popularity as an alternative analytic method.

Second, for acute pain meta-analyses comparing two different interventions where the mean of a group of data (i.e., randomized trials) is most important, a frequentist analysis may be better suited for the clinician. In contrast, a Bayesian model may be more applicable for those clinicians wishing to evaluate the probability that a certain intervention will be better for an outcome. On a broad scale, clinicians assessing the results of a single outcome (e.g., rest pain at 24 h) may be more interested in knowing whether one treatment is “better” than another treatment, rather than the “probability” of being better; this is more consistent with a frequentist approach.

Third, we would caution readers against an over-reliance on P values for interpreting clinical research. While the threshold for statistical significance does provide some insights into the results of a clinical research question, there should always be interpretation based on clinical importance and the minimal clinically important difference. For even if we utilize the Bayesian model to accomplish a statically significant 95% credible interval of –0.43 cm.h to –0.01 cm.h, this alternative result is still not clinically important, especially when the clinical importance threshold is 3.0 cm.h. Indeed, the differences between the results of Bayesian and frequentist approaches were trivial, and the effect size of liposomal bupivacaine was very small.

Finally, interpretation of research findings is contextual. Using the frequentist method, we have already demonstrated that liposomal bupivacaine is not superior to plain bupivacaine when used in perineural nerve blocks, periarticular infiltration and surgical field infiltration.  Our current findings in the setting of fascial plane blocks are consistent, and the alternative Bayesian analysis does not alter this finding in a clinically meaningful way.