An integrated model for the interaction of muscle relaxants with their antagonists.

An integrated model describing the interaction of nondepolarizing neuromuscular blocking agents with reversible anticholinesterase agents is derived and compared with a naive model using experimental data obtained from four anesthetized dogs. Three consecutive but separate steady-state d-tubocurarine blocks (approximately 50, 70, and 90%) were induced in each of the four dogs and reversed by short edrophonium infusions. Edrophonium arterial concentrations and twitch tension of the anterior tibialis muscle were measured. Both the integrated and the naive model were fit to the twitch tension data using a model with a hypothetical "effect" compartment. The integrated model consistently fit the twitch tension data better than the naive model; the sum of squared deviations was lower by 46, 45, 87, and 69%, respectively, with the integrated model than with the naive model. Also, in contrast to the naive model, the integrated model is capable of describing the interaction of the anticholinesterase agent and the neuromuscular blocking agent when the concentration of either varies with time.