This multimodal perioperative analgesia protocol that included infiltration of a local anesthetic offered improved pain control and minimal side effects to patients undergoing total knee arthroplasty. Our study also confirmed the safety of the protocol.

IntroductionLiver microcirculation in the perfused rat liver was assessed by the multiple indicator dilution technique. Comparative studies were carried out in noncirrhotic rats and in rats with cirrhosis secondary to chronic exposure to phenobarbital and carbon tetrachloride. Chronic administration of carbon tetrachloride (CC14) to phenobarbital-treated rats was chosen as the experimental model because it has been shown (7) to yield morphologic alterations resembling human cirrhosis. In this model, we defined the hepatic venous outflow patterns resulting from portal vein injection of labeled erythrocytes (RBCs)' and plasma-dissolved labeled substances not metabolized by the liver (albumin, sucrose, and water). In normal liver, RBCs are confined to the vascular space (vascular reference), while plasma-dissolved substances gain access to the extravascular space (the space ofDisse and hepatocytes) through fenestrae in and between the endothelial cells: albumin (69,000 mol wt) and sucrose (342 mol wt) diffuse in extracellular spaces that are inversely related to their molecular *eights, and water diffuses in the cellular space (5).-This produces a major delay in, and a decrease in the magnitude of, the labeled albumin, sucrose, and water curves with respect to that of labeled RBCs, which can be analyzed using the flow-limited linear two-compartment model described by Goresky (5). This method yields estimates of sinusoidal blood volume and of the extravascular volume of distribution of diffusible substances (peak-time method). Extravascular volumes can also be calculated in a more direct manner from the transit times (transit-time method) (8), and will always be correct as long as all of the label is recovered at the outflow, even in a system with reduced permeability.In cirrhotic liver, ifconditions for a flow-limited distribution are no longer fulfilled, the extravascular volume measured by the peak-time method may depart widely from the volume measured by the transit-time method. This disparity can result either from shunting, in which both reference and diffusible labels pass through the vasculature with little difference in their transit times, or from collagenization and capillanization, in which the devel-

Using the multiple indicator dilution approach, events occurring in the microvascular bed can be characterized in experimental animals with different types of cirrhosis and in man. Intrahepatic shunts can be found shunting blood away from sinusoids in both cirrhotic patients and cirrhotic animals. Such shunts were present in about one-third of cirrhotic patients with portal hypertension, and occurred mainly between the portal vein and hepatic veins. In cirrhotics, portohepatic anastomoses are usually large in diameter (more than 20 micron in diameter). Collagenization of the space of Disse and the progressive transformation of sinusoids into capillary-like channels decrease the extravascular space accessible to albumin and probably to other large molecules and protein-bound substances. However, unlike findings obtained in well-capillarized organs, these sinusoidal changes do not appear to limit the diffusion of sucrose, water, and lipophilic substances, such as lidocaine in the extravascular and intracellular spaces. The pattern observed for labeled sucrose curves following hepatic artery injection in cirrhotic patients could be secondary to the passage through the dense peribiliary capillary plexus originating from the enlarged arterial bed in cirrhosis. The difference in the perfusion of cirrhotic nodules with regard to the portal venous and hepatic artery routes introduces important new concepts in the overall mechanism of the elimination of endogenous and exogenous substances by the cirrhotic liver: blood entering the liver by the two afferent vessels will not flow through the same vascular bed before reaching the efferent hepatic veins.

The basic characteristics of NMBAs, namely, molecular weight, lipid solubility and protein binding, are strongly associated with the kinetics of the drug response.

The pharmacokinetics and pharmacodynamics of atracurium, a nondepolarizing neuromuscular blocking agent, were compared between morbidly obese patients and nonobese patients. Atracurium besylate (0.2 mg/kg) was administered intravenously as a bolus to patients who had received anesthesia. The force of contraction of the adductor pollicis was measured and plasma samples were collected for a 2-hour period. The concentrations of atracurium and its major end product, laudanosine, were determined by use of a chromatographic method. The pharmacokinetic-pharmacodynamic relationship was characterized by use of several models. No difference was observed between obese patients and nonobese patients in atracurium elimination half-life (19.8 +/- 0.7 versus 19.7 +/- 0.7 minutes), volume of distribution at steady state (8.6 +/- 0.7 versus 8.5 +/- 0.7 L), and total clearance (444 +/- 29 versus 404 +/- 25 ml/min). However, if values were expressed on a total body weight basis, there was a difference between obese and nonobese patients in the volume of distribution at steady state (0.067 versus 0.141 L/kg) and total clearance (3.5 +/- 0.2 versus 6.6 +/- 0.5 ml/min/kg). Although atracurium concentrations were consistently higher in obese patients than in nonobese patients, there was no difference in the time of recovery from neuromuscular blockade between the two groups. Consequently, the median effective concentration was higher in obese than in nonobese patients (470 +/- 46 versus 312 +/- 33 ng/ml).