Sugammadex vs. neostigmine: the train-of-four counts too

In this editorial, we outline the current evidence comparing sugammadex and neostigmine for preventing postoperative pulmonary complications and discuss the strengths and limitations of the relevant evidence. Importantly, we aim to emphasise the importance of quantitative train-of-four monitoring in guiding neuromuscular blockade antagonism; identify barriers to its routine use; and suggest interventions to improve uptake in daily clinical practice.

Using the TriNetX federated real-world data platform [4], Tsai et al. conducted a large retrospective, multicentre cohort study comparing sugammadex with neostigmine for antagonism and rocuronium-induced neuromuscular blockade in 15,730 adults undergoing upper gastrointestinal endoscopy [5]. Sugammadex was associated with a significantly lower rate of 30-day tracheal extubation failure or requirement for mechanical ventilation (3.9% vs. 5.8%, risk ratio 1.49, 95%CI 1.30–1.72) corresponding to a number needed to treat of 52 to prevent one additional case of tracheal extubation failure. Sugammadex was also associated with lower rates of lung atelectasis and unplanned ICU admission, with no significant differences in other secondary outcomes.

The procedure-specific focus of the study by Tsai et al. is a key strength. Upper gastrointestinal endoscopy exposes patients to distinctive respiratory vulnerabilities: non-operating theatre anaesthesia; shared airway constraints; and pharyngo-oesophageal instrumentation [6]. Complete restoration of airway reflexes and respiratory muscle function at tracheal extubation is especially important in this context. This association was confirmed in a prespecified subgroup analysis of patients without chronic lower respiratory disease, suggesting that the procedural insult itself, rather than pre-existing pulmonary disease, may be sufficient to erode the safety margin of inadequate antagonism on neuromuscular blockade.

These findings from Tsai et al. are consistent with a growing body of evidence. The Sugammadex, Neostigmine and Postoperative Pulmonary Complications (SNaPP) trial, which studied patients undergoing major abdominal or thoracic surgery across 44 hospitals in Australia, Aotearoa New Zealand and Hong Kong, allocated 3498 adults randomly to sugammadex or neostigmine for antagonism of neuromuscular block [7]. The adjudicated composite of postoperative pulmonary complications or death was significantly lower with sugammadex (19.0% vs. 21.5%, 95%CI 0.77–1.00), driven predominantly by a reduction in radiologically confirmed atelectasis, with no significant differences in other outcomes including mortality [7]. Together with the STRONGER and STIL-STRONGER matched cohort analyses [8, 9], the cumulative evidence now favours sugammadex for the prevention of postoperative pulmonary complications. The consistency of findings across observational and randomised studies lends additional weight to the suggestion that the observed association is clinically meaningful, but, as we shall argue, drug choice alone is not sufficient to ensure safe recovery from neuromuscular blockade.

The most significant limitation of the study by Tsai et al. is the absence of several intra-operative variables that are mechanistically central to postoperative pulmonary outcomes. Train-of-four data, depth of neuromuscular blockade at antagonism, dose and timing of antagonist drug, anaesthetic duration and ventilatory strategy are not captured within TriNetX and therefore could not be analysed. Propensity matching improves comparability for measured variables, but it cannot account for the unmeasured clinical behaviours that directly determine whether antagonism is timely, appropriately dosed and objectively confirmed. Residual confounding from institutional practice patterns is therefore plausible in both directions: centres using sugammadex more frequently may also have more consistent access to quantitative monitors and greater familiarity with guideline-based reversal, whereas some institutions may reserve sugammadex for higher risk patients, biasing results the other way.

These competing possibilities are precisely why the findings Tsai et al., though compelling, should be interpreted as associations rather than proof of causation. Without knowing the distribution of block depth at antagonism in each group, it is difficult to determine how much of the observed difference reflects an intrinsic advantage of sugammadex, how much reflects inappropriate neostigmine use at moderate-to-deep block and how much reflects broader differences in peri-operative practice. The practical implication is important: antagonist selection and quantitative monitoring are complementary, not competing, strategies and the data from Tsai et al. cannot disentangle the relative contributions of each.

Even in the sugammadex-treated group of both studies, postoperative pulmonary complications still occurred (19.0% in SNaPP [7] and 3.9% in Tsai et al. [5]). Thus, although sugammadex reduces pulmonary complications, it does not eliminate them. Sugammadex is also not pharmacologically infallible; its efficacy as an antagonist is dose-dependent. A standard 200 mg vial may be insufficient for reversal from deep blockade in patients weighing >50 kg, and substantial individual variability in dose requirements has been associated with residual or recurrent paralysis [10]. Overconfidence with sugammadex in the absence of quantitative neuromuscular monitors may, therefore, lead to inadequate recovery of neuromuscular function [11–13].

The quantitative neuromuscular monitoring data from SNaPP are particularly instructive. Although quantitative monitoring was used in approximately 84% of patients in both groups—higher than in many real-world settings—only 64.3% of patients allocated to neostigmine achieved a train-of-four ratio ≥0.9 at tracheal extubation, compared with 87.4% of those allocated to sugammadex. Rescue therapy was required in 17.2% vs. 3.6% of patients allocated to neostigmine and sugammadex, respectively. Among monitored patients, approximately 45% in both groups had a train-of-four count of 0–3 at the time of antagonist administration, representing moderate-to-deep block at which neostigmine is pharmacologically unreliable. These findings support the effectiveness of sugammadex while simultaneously reinforcing the importance of quantitative monitoring in guiding drug selection, dosing and confirmation of recovery. Sugammadex, by encapsulating rocuronium stoichiometrically regardless of block depth, is pharmacologically more forgiving of the failure to monitor. However, a more forgiving drug does not address the underlying failure of not monitoring.

The appropriate clinical response to these data is therefore not ‘to switch to sugammadex’ but rather ‘to monitor quantitatively and choose the antagonist drug and dose accordingly’. Quantitative measurement has been shown to be far superior to qualitative or no neuromuscular monitoring for the prevention of postoperative residual curarisation [14]. Both the 2023 ASA guidelines for monitoring and antagonism of neuromuscular blockade and the 2023 ESAIC guidelines on the peri-operative management of neuromuscular blockade are unambiguous: quantitative train-of-four monitoring is the standard of care for all patients receiving neuromuscular blocking drugs, with a confirmed train-of-four ratio ≥0.9 required before tracheal extubation [15, 16]. Neither guideline recommends sugammadex as a blanket policy; drug choice and dose should be determined by the depth of neuromuscular block at the time of antagonist administration.

Neostigmine remains appropriate at shallow block (train-of-four count of 4 with a train-of-four ratio 0.4–0.9) under quantitative guidance. Sugammadex is generally preferred for deeper aminosteroid-induced blockade [15, 16]. Importantly, quantitative monitoring also identifies patients who have recovered spontaneously to a train-of-four ratio ≥0.9, in whom pharmacological antagonism can be safely omitted altogether, thereby reducing drug costs, avoiding adverse effects of antagonist drugs and eliminating the rare (but recognised) risk of sugammadex hypersensitivity.

The central challenge may be less about which antagonist drug is chosen and more about whether guideline-recommended quantitative monitoring is performed routinely. Despite guidelines published in 2023 [15, 16] and evidence linking residual block to postoperative pulmonary complications dating back nearly three decades to a landmark study by Berg et al. [17], uptake of quantitative neuromuscular monitoring remains inconsistent. In a cluster randomised crossover trial, quantitative neuromuscular monitoring use increased from only 41%–46% following an educational intervention, even as antagonist use approached 100%, implying that clinicians may administer neuromuscular block antagonists without confirming that recovery is adequate [18].

In a qualitative study exploring the barriers and aids to routine neuromuscular monitoring and consistent reversal practice [19], identified barriers included the inconvenience or unavailability of monitoring equipment; perceived unreliability of the measurements; time pressure at case turnover; difficulty placing monitors; lack of familiarity with monitor use; and attitudinal barriers, including the belief that minor residual neuromuscular blockade is clinically irrelevant. Departmental culture and the absence of standardised local guidelines were also cited as a source of frustration [19].

These barriers are surmountable. Equipment availability is an infrastructure issue: quantitative neuromuscular monitors should be standard at every anaesthetic workstation and treated as patient safety equipment analogous to pulse oximetry or capnography. Where acceleromyography using the hand is impractical because the arms are tucked in or inaccessible, alternatives such as facial nerve–corrugator supercilii monitoring or electromyography-based devices should be considered. Time pressure is a systems issue, not a legitimate trade-off against confirming adequate neuromuscular block recovery before tracheal extubation. Knowledge and attitudinal gaps call for targeted educational responses. Simulation-based training that confronts clinicians with the limitations of clinical and qualitative assessment, combined with peer-led audit and feedback in which anaesthetists are shown data from their own patients, is more likely to shift entrenched beliefs than didactic teaching alone. Departments should formalise quantitative monitoring within standard operating procedures rather than leaving its use to individual discretion.

Future work should therefore focus not only on further drug comparisons but also on interventions that improve adherence to quantitative monitoring guidelines and on the cost-effectiveness of different antagonist strategies in real-world systems. Universal substition with sugammadex should await formal economic analysis. Enthusiasm is tempered by Tsai et al’s number needed to treat of 52 [5] and the modest absolute risk reduction in SNaPP [7], which was driven largely by atelectasis – an outcome where broader clinical significance remains debated. This is especially important as the acquisition cost of sugammadex falls. A cheap sugammadex strategy without neuromuscular monitoring may still be unsafe; an expensive monitoring programme that does not change clinician behaviour may also fail to deliver benefit. Beyond cost, the key policy question is how best to combine drug availability, objective measurement, education and local governance into a reliable system of care. A well-powered trial comparing guideline-adherent, quantitatively monitored reversal with usual care, with drug choice and dosing as prespecified variables, would be more actionable than further registry analyses alone.

In conclusion, Tsai et al. [5] and the SNaPP trial [7] provide the strongest evidence to date that sugammadex is associated with fewer postoperative pulmonary complications than neostigmine across a range of surgical and procedural settings. This evidence is welcome and clinically important. However, the lesson is not ‘use sugammadex routinely and monitoring becomes optional’. The lessons are instead that residual neuromuscular blockade remains prevalent, preventable and clinically consequential; that neostigmine is less reliable when used at inappropriate depths of block; and that sugammadex, although more effective across a wider range of block depth, still requires appropriate dosing and objective confirmation of recovery. Safer emergence from anaesthesia depends on universal quantitative train-of-four monitoring; guideline-adherent drug and dose selection; and a professional culture in which tracheal extubation is not considered safe until the train-of-four ratio of at least 0.9 has been confirmed. Until that standard is achieved, we risk attributing to drug choice what is in fact a failure of monitoring, and our patients will continue to bear the consequences.