To the Editor,

In a small randomized controlled trial which only included 26 patients undergoing total shoulder arthroplasty, Sun et al.1 compared the occurrence of hemidiaphragmatic paresis between patients receiving a continuous high-thoracic erector spinae plane block (HT–ESPB) and a continuous interscalene block (CIB). They reported that the incidence of hemidiaphragmatic paresis was 100% with the CIB compared with 0% with the HT–ESPB. Given that application of a multimodal analgesic protocol including a nerve block can improve postoperative analgesia and decrease opioid consumption as emphasized in current Enhanced Recovery After Surgery protocols,2 this study has potential implications. Other than the limitations described in the discussion, however, there were several issues in the article that we wish to bring to the readers’ attention and invite the authors to comment on.

First, when patients arrived in the postanesthetic care unit (PACU), 5 mL of 0.5% ropivacaine was given via the indwelling catheter in the two groups. Then, both automated bolus and demand bolus doses of 0.2% ropivacaine were administered. The available evidence indicates that the postoperative analgesic efficacy of interscalene blockade for shoulder surgery is not significantly different between 0.25%, 0.5%, and 0.75% ropivacaine solutions.2 Furthermore, continuous erector spinae plane block with 0.2% ropivacaine can provide adequate postoperative pain relief.3 As sensory nerves are more sensitive to the blocking effect of local anesthetics than motor nerve, and the intensity of motor blockade increases with increased local anesthetic concentrations,2 we argue that a decreased incidence of hemidiaphragmatic paresis would have been achieved in the CIB group had a lower concentration of ropivacaine been used for initial postoperative pain control in the PACU.

Second, the CIB resulted in a 100% incidence of hemidiaphragmatic paresis, with a larger decrease in incentive spirometry volumes from baseline and greater reduction in motor and sensory function. Nevertheless, the incidence of adverse events, duration of PACU stay, activity domain scores, time to discharge, and patient satisfaction were not significantly different between the two groups. On this basis, it is difficult for readers to determine whether the hemidiaphragmatic paresis associated with the CIB for pain control after total shoulder arthroplasty should be considered as being a clinically important event.

Finally, the mean (standard deviation) cumulative opioid consumption (oral morphine milligram equivalent, mg) at postoperative day 0 was significantly lower in the interscalene group than in the HT–ESPB group (16.6 [21.9] mg vs 45.3 [39.9] mg; P = 0.04). As 3 mg oral morphine is equivalent to 1 mg intravenous morphine, the net between-group difference in the mean cumulative opioid consumption at postoperative day 0 is about 9.6 mg intravenous morphine, which does not exceed the recommended minimal clinically important difference of cumulative opioid consumption for postoperative pain control in the literature, i.e., an absolute reduction of 10 mg intravenous morphine in the morphine milligram equivalent of 24 hr opioid consumption.4 That is, the between-group difference in mean cumulative opioid consumption at postoperative day 0 is statistically significant, but its clinical relevance is debatable.