Across the two periods studied, the distribution of patient and procedural characteristics remained relatively unchanged. However, over the last few years, we have observed a trend towards more active reversal with sugammadex. In the period 2015–2017, the sugammadex dose was more often ‘rounded off’ to a full ampule of 200 mg (67.2% compared to 56.0% of the time in 2018–2019). This could be explained by the increased use of NMT monitoring, which facilitates the detection of shallow rNMB where lower doses of sugammadex have been observed to be adequate for reliable reversal (Pongrácz et al. 2013; Schaller et al. 2010).

The POPULAR study, performed in 211 hospitals in Europe, showed the use of any NMT monitoring in only 42.1% of patients receiving a NMBA (Kirmeier et al. 2019). In this study, documented NMT monitoring in spontaneously recovered patients increased from 43 to 57% over the two consecutive periods, in those reversed with sugammadex from 77 to 94%, and overall from 47 to 64% (Table 2). A plausible explanation is the introduction of the integrated NMT module with automatic recording in the anaesthesia record, resulting in a more consistent recording compared to the manual entry of NMT data. Manual documentation can be forgotten, especially since the time around extubation is usually a relatively dynamic moment in the operating room for the anaesthesia team with many possible distractions. Therefore, the NMT monitoring data, especially from 2015 to 2017, might be an underestimation of clinical practice (i.e. what is conducted versus what is documented). Automatic registration is inherently more reliable than manual documentation, thereby minimising the potential for documentation errors. Segregating automatic and manual data for separate analyses would have resulted in overly large tables, compromising readability. Combining all data also made sense to analyse whether any NMT monitoring was used in each case.

Increasing the rate of NMT monitoring in clinical practice has proven to be a formidable challenge. Thomsen et al. (Thomsen et al. 2022) demonstrated that an e-learning module focused on NMT monitoring had no overall impact on the application of NMT monitoring, despite a post-course test suggesting an enhancement in anaesthesiologists’ knowledge within this area. Söderström et al. (Söderström et al. 2017) surveyed 653 Danish anaesthetists and found that while objective NMT monitoring is frequently utilised, it is often associated with technical difficulties. These findings are in line with our own observations and underscore the potential benefits of employing a NMT monitor that features automatic measurement and recording capabilities. Such a system is likely to enhance both the frequency and quality of these recordings, thereby diminishing the likelihood of residual neuromuscular blockade. The meticulous recording of NMT measurements is also crucial both for quality control and for medicolegal considerations. It should be noted that because the period 2015–2017 contains some automatically registered data because the integrated NMT monitor was introduced in 2017. The differences due to the automated registration of neuromuscular transmission monitoring might therefore be underestimated.

The percentage of patients extubated in the OR with neither a documented nor adequate TOFR (> 0.9) decreased from 77% in 2015–2017 to 56% in 2018–2019. This percentage of 56% is lower than the 64.7% (rNMB, defined as TOFR < 0.9) reported in the RECITE-US study years earlier (2012–2013) by Saager et al. (Saager et al. 2019). When comparing our data directly to prospective studies measuring rNMB in the PACU, it is important to realise that not all of the patients in our study, where no (quantitative) NMT monitoring was used or documented, had rNMB. Nevertheless, when (quantitative) NMT monitoring is not performed in patients who receive an NMBA, there is a certain risk of rNMB (Murphy et al. 2011). In some cases, no (further) NMT measurements were carried out after the administration of sugammadex, with only a small chance of rNMB due to the predictable nature of sugammadex-mediated reversal. Therefore, amongst patients reversed with sugammadex, the last TOFR measured prior to extubation may not be distinguishable from the TOFR measured prior to the administration of sugammadex and is therefore not easily compared to rNMB risk amongst spontaneously reversed patients. Although the administration of sugammadex without NMT monitoring does not completely eliminate the risk of rNMB (Kotake et al. 2013; Nemes et al. 2017) it decreases the risk significantly. Patients extubated in the OR, with spontaneous recovery of NMB, with no documented NMT monitoring at all, or an inadequate TOFR at extubation, are obviously most at risk for complications due to rNMB. These percentages decreased from 46% (spontaneous recovery with no documented NMT) and 23% (spontaneous recovery with inadequate TOFR at extubation) during 2015–2017 to 31% and 17% during 2018–2019, respectively. Continuing NMT monitoring until full recovery from NMB after sugammadex reversal could further decrease the risk of rNMB (Kotake et al. 2013).

Logistic regression analysis (Table 3) revealed independent factors associated with the pharmacological reversal of rocuronium-induced neuromuscular blockade with sugammadex. Patient-related variables included high BMI, higher ASA classification, and age > 60 years. This may be attributed to a more careful approach of anaesthesiologists regarding the prevention of rNMB in patients with a higher risk of perioperative complications, including those with a higher ASA classification, age, or BMI. This is in concordance with data from the Multicentre Perioperative Outcomes Group in the United States published by Dubovoy et al. (Dubovoy et al. 2020) and data from Leiden University Medical Centre (another large teaching hospital in the Netherlands) presented by Martini et al. (Martini et al. 2021).

Surgery-related factors associated with the use of sugammadex are the duration of surgery less than 120 min, and emergency, laparoscopic, and open abdominal/thoracic surgery. Martini et al. also found an association between reversal with sugammadex and shorter duration of surgery (Martini et al. 2021). In both open abdominal or thoracic surgery and laparo/thoracoscopic procedures, NMBAs are often repeatedly administered to improve surgical conditions until the end of surgery. Additionally, open abdominal or thoracic surgery is inherently associated with a higher risk of pulmonary complications (Canet et al. 2015) and rNMB might therefore be treated more rigorously. Patients who underwent emergency surgery were more often reversed with sugammadex. This can be readily explained by the fact that many of these patients were intubated following rapid sequence induction of anaesthesia with a high dose of rocuronium. A higher dose of rocuronium resulted in a higher chance of rNMB and a need for active reversal with sugammadex. To illustrate, in 30.3% of emergency cases where rocuronium was used, the total dosage of rocuronium was at least 3× ED95 vs 20.6% for non-emergency surgery. One thousand seventy patients with an emergency case received succinylcholine as the only NMBA, but these were excluded from the final analysis.

Anaesthesia-related factors associated with the use of sugammadex included the use of NMT monitoring and the total dose of rocuronium corrected for ABW. When NMT monitoring is used, there is a higher chance of diagnosing rNMB. However, the observed relationship with sugammadex administration may also be explained by our local practice, as in our hospital, sugammadex is rarely administered without a NMT measurement. NMT monitoring will also be used more often when there is a high chance of rNMB due to a recent rocuronium dose or high clinical suspicion of rNMB. This is in concordance with the fact that NMT monitoring is highly correlated with sugammadex-mediated reversal.

When a higher intraoperative total dose of rocuronium is used, there may also be a higher risk of rNMB. This was reflected in the results. We used a variable that was corrected for ABW (ED95) to better compare patients.

Both magnesium chloride (Fuchs-Buder et al. 1995) and inhalational anaesthetics (Jellish et al. 2000; Wulf et al. 1998) are known to increase the potency of NMBA. Sevoflurane and isoflurane are the only inhalational anaesthetics used at our hospital. There was a slightly lower chance of sugammadex reversal in patients receiving inhalation anaesthetics for maintenance of anaesthesia (odds ratio 0.883, 95% CI 0.785–0.992). A possible explanation for this observation could be that, in general, lower doses of rocuronium are required to achieve the same level of muscle relaxation during inhalation anaesthesia. Magnesium chloride is mainly used as a multimodal analgesic in our hospital because of its N-methyl-D-aspartate (NMDA) receptor-blocking properties. Our analysis did not reveal a significant association between sugammadex reversal and magnesium chloride use.

A strength of this study is that the combination of automatic extraction from the EMR and manual checks ensures high-quality data and a large dataset with sufficient power for analysis of NMBA and sugammadex dosing and NMT monitoring. Because this is a single-centre study in a tertiary teaching hospital, the results are likely more generalizable to other local tertiary care centres than to community or smaller nonteaching hospitals. These hospitals often have a higher turnover of shorter procedures, and pharmacological reversal may be used more often. A study by Dubovoy et al. (Dubovoy et al. 2020) showed a wide variation in sugammadex use amongst different hospitals in the US. However, in the absence of global guidelines on NMB management, there is a large variation in NMB management, specifically reversal trends between regions, reflecting variability in practice, training, and exposure to new technologies.