Patient-safety rubric: What is Popular on outcomes after administration of muscle relaxants?

Patient-safety rubric: What is Popular on outcomes after administration of muscle relaxants?

  • Issue 76

Marc Giménez-Milà MD,
A Sabaté MD, PhD

Although it is in observational nature, its large sample combined with the paucity of previous studies – many of them retrospective – It is highly foreseeable that the POPULAR1 study will become one of the most commented, discussed, and analysed trials providing evidence on outcomes after use of neuromuscular blocking agents (NMBA).

To test the author’s hypothesis that NMBA are associated with postoperative pulmonary complications, 211 hospitals from 28 European countries prospectively collected data from 22,803 patients. Following procedures under general anaesthesia (GA), data on patients that were admitted to the hospital > 12 h after surgery and did not require postoperative treatment in the Intensive Care Unit were gathered. Patients were stratified in 5 cohorts – Cohort 1 GA (21,694 patients), Cohort 2 NMBA (17,150 patients), Cohort 3 any NM monitoring (6868 patients), Cohort 4 quantitative NM monitoring (4182 patients), Cohort 5 use of reversal (8795 patients); only 4001 patients (18.4%) did not receive NMBA. Primary outcome was postoperative pulmonary complications up to postoperative day 28. Definition of postoperative pulmonary complications were the same as in the PERISCOPE study,2collected on physical examination performed during the postoperative 3 days and through patients’ charts after hospital discharge.

Incidence of any postoperative pulmonary complications in each cohort were 7.6%, 8.4%, 10.7%, 10.5%, and 8.9%, respectively, with an increased risk when NMBA were administered (OR adjusted 1.86, 95% CI 1.53–2.26). Use of NM monitoring did not show reduced risk of pulmonary complications, neither the use of reversal nor the use of sugammadex instead of neostigmine.

From the propensity score matching, when NMBA administration is unlikely (first six deciles compared with four second deciles) the mean risk for a postoperative pulmonary complication is low (4.4%). This may well reinforce the idea of avoidance of NMBA for lower-risk patients and peripheral surgical procedure using supraglottic devices.

Kirmeier et al.1describe very thoroughly the limitations of the study with the exception of how pulmonary complications data are obtained from day of hospital discharge to day 28. While this data may be in patient electronic records of many other patients, this may have been missed, undermining the incidence of pulmonary complications. In similar lines, we missed more information on those patients who did not receive NMBA and how neuromuscular monitoring and reversal was performed in those patients in whom a single dose of NMBA was used.

Once more, this study reflects the complexity of perioperative care research, where many interventions are performed at the same time and may impact on the main outcome. In this case impact on incidence of pulmonary complications of intraoperative ventilation management, intravenous fluid therapy, postoperative nurse ward management, and physiotherapy approach is not analysed. We also do encourage similar studies for ambulatory surgery to analyse the impact of NMBA on pulmonary complications.

Finally, lack of reduction on incidence of pulmonary complications in cohorts 4 and 5 require careful reflection. In our opinion, the fact this this study does not show statistical difference could not be translated into changes in monitoring and reversal of current practice until more evidence is obtained from future trials. Until then, we do now have the POPULAR study, which will serve to promote judicious usage of muscle relaxants, balancing their benefits such as patient immobility, and to facilitate mechanical ventilation and its inherent risks such as development of postoperative pulmonary complications.


1. Kirmeier E, Eriksson LI, Lewald H, et al. Lancet Respir Med2018. pii: S2213-2600(18)30294-7.
2. Canet J, Gallart L, Gomar C, et al. Anesthesiology2010;113(6):1338–50.