Sugammadex is increasingly used as the primary antagonist for rocuronium-induced neuromuscular blockade and has reduced, but not eliminated, residual paralysis. A small but continuing series of reports describes a different failure: adequate recovery of the train-of-four ratio followed by a secondary decline in neuromuscular function, sometimes more than an hour later and occasionally after a recommended dose. Sugammadex encapsulates rocuronium in a 1-to-1 ratio, and binding occurs only in plasma. Sustained reversal therefore depends on the molar relationship between the administered dose and the rocuronium remaining in the body, rather than on the milligram-per-kilogram dose. At or above molar equivalence, no uncaptured rocuronium remains, and recurrence is mechanistically excluded while free sugammadex remains in excess; below equivalence, a free fraction persists whose effect is the product of that fraction and the residual burden. A weight-based dose is fixed, whereas the burden varies, concentrating risk where the burden is large or the dosing weight is small: prolonged infusions and repeated boluses, deep block, advanced age, renal impairment, and ideal-body-weight dosing in obesity. Randomized dose-finding trials reported recurrent block almost exclusively in the lowest-dose groups, and a dosing-weight trial reproduced it by reducing the dose alone. Neither 2023 guideline addresses dosing after prolonged infusion or the duration of subsequent observation. Reported events have occurred up to about 100 minutes after reversal, usually without quantitative monitoring. We propose estimating the burden, dosing above equivalence with a margin for uncertainty, and continuing measurements beyond the first confirmed recovery.
