Defining depth of anesthesia.

In this chapter, drawn largely from the synthesis of material that we first presented in the sixth edition of Miller's Anesthesia, Chap 31 (Stanski and Shafer 2005; used by permission of the publisher), we have defined anesthetic depth as the probability of non-response to stimulation, calibrated against the strength of the stimulus, the difficulty of suppressing the response, and the drug-induced probability of non-responsiveness at defined effect site concentrations. This definition requires measurement of multiple different stimuli and responses at well-defined drug concentrations. There is no one stimulus and response measurement that will capture depth of anesthesia in a clinically or scientifically meaningful manner. The "clinical art" of anesthesia requires calibration of these observations of stimuli and responses (verbal responses, movement, tachycardia) against the dose and concentration of anesthetic drugs used to reduce the probability of response, constantly adjusting the administered dose to achieve the desired anesthetic depth. In our definition of "depth of anesthesia" we define the need for two components to create the anesthetic state: hypnosis created with drugs such as propofol or the inhalational anesthetics and analgesia created with the opioids or nitrous oxide. We demonstrate the scientific evidence that profound degrees of hypnosis in the absence of analgesia will not prevent the hemodynamic responses to profoundly noxious stimuli. Also, profound degrees of analgesia do not guarantee unconsciousness. However, the combination of hypnosis and analgesia suppresses hemodynamic response to noxious stimuli and guarantees unconsciousness.

Decreased protein binding and thiopental kinetics.

Thiopental kinetics and protein binding were determined in seven surgical patients with chronic renal failure and a thiopental free fraction of 28.0 +/- 6.5% (SD) and in seven age- and weight-matched normal surgical patients with a thiopental free fraction of 15.7 +/- 2.4%. Thiopental clearance, based upon total plasma concentrations, rose from 3.2 +/- 0.6 ml/kg/min in the normal group to 4.5 +/- 1.1 ml/kg/min in the chronic renal failure group. Volume of distribution at steady state, also based on total drug concentrations, rose from 1.9 +/- 0.5 l/kg in the normal group to 3.0 +/- 1.0 l/kg in the chronic renal failure group. These changes in clearance and volume of distribution at steady state are secondary to changes in free drug distribution and elimination. When kinetics were calculated from free drug concentrations, the intrinsic clearance, unbound volume of distribution at steady state, and free fraction in tissues in the normal and renal failure groups did not differ substantially. These data suggest that the kinetic changes based on total drug concentrations are secondary to changes in free fraction in plasma. In chronic renal failure patients, the underlying rate and extent of thiopental distribution and elimination are much the same as in normal patients.

Population pharmacodynamic model for ketorolac analgesia.

OBJECTIVE: To derive a population pharmacokinetic-pharmacodynamic model that characterizes the distribution of pain relief scores and remedication times observed in patients receiving intramuscular ketorolac for the treatment of moderate to severe postoperative pain. BACKGROUND: The data analysis approach deals with the complexities of analyzing analgesic trial data: (1) repeated measurements, (2) ordered categorical response variables, and (3) nonrandom censoring because the patients can take a rescue medication if their pain relief is insufficient. METHODS: Patients (n = 522) received a single oral or intramuscular administration of placebo or a single intramuscular dose of 10, 30, 60, or 90 mg ketorolac for postoperative pain relief. Pain relief was measured periodically with use of a five-category ordinal scale up to 6 hours after dosing. In this period, 288 patients received additional medication because of insufficient pain relief. Pharmacokinetic data was available for 85 subjects. Models were fitted to the data with the NONMEM program. RESULTS: The pharmacokinetic data was best described by a two-compartment model with first-order absorption. Pain relief was found to be a function of drug concentration (Emax model), time (waxing and waning of placebo effect), and an individual random effect. The drug concentration at half-maximal effect (EC50) and the first-order rate constant (keo) half-life for pain relief were 0.37 mg/L and 24 minutes. The probability of remedication was found to be a function of the observed level of pain relief and was found to increase with time. Monte Carlo simulations showed that adequate pain relief was achieved in 50% of the patients at 41, 27, 23, and 21 minutes after 10, 30, 60, or 90 mg of intramuscular ketorolac. Adequate pain relief was maintained up to 6 hours in 50%, 70%, 78%, and 81% of patients after these four doses. Only 25% of the patients achieved adequate pain relief with placebo. CONCLUSIONS: A population pharmacokinetic-pharmacodynamic model for the analgesic efficacy of intramuscular ketorolac was derived. The simulated relationship between dose, time, and percentage of patients with adequate pain relief suggested that 30 mg intramuscular ketorolac was the optimal initial dose for postoperative pain relief.

Kinetics of intravenous and intramuscular morphine.

The disposition of parenteral morphine was assessed in two pharmacokinetic studies. In Study 1, 10 mg of morphine sulfate was administered by intravenous (IV) infusion, intramuscular (IM) injection, or both, to 8 healthy young adult male volunteers. Plasma morphine concentrations were determined by radioimmunoassay in multiple blood samples drawn after each dose. Mean (+/-SE) kinetic parameters following IV morphine were: volume of distribution (Vd), 3.2 (+/- 0.3) L/kg; elimination half-life (t1/2beta), 2.9 (+/- 0.5) hr; clearance, 14.7 (+/- 0.9) ml/min/kg; extraction ratio, 0.70 (+/- 0.04). After IM morphine, peak plasma levels ranged from 51 to 62 ng/ml and were reached within 20 min of injection. The absorption half-life averaged 7.7 (+/- 1.6) min. Systemic availability was 100% complete. In study 2, 4 elderly male patients (61 to 80 yr of age) received 45 to 80 mg of morphine sulfate IV prior to operative repair of an abdominal aortic aneurysm. Morphine pharmacokinetics were determined as described above. Kinetic variables were Vd, 4.7 (+/- 0.2) L/kg; t1/2beta, 4.5 (+/- 0.3) hr; clearance, 12.4 (+/- 1.2) ml/min/kg; extraction ratio, 0.59 (+/- 0.05). Both studies demonstrate that morphine distribution is rapid and extensive and its t1/2beta relatively short. IM morphine is rapidly and completely absorbed.

Pyridostigmine kinetics with and without renal function.

Pyridostigmine kinetics were examined under conditions of clinical use as an antagonist of nondepolarizing neuromuscular blockade in anesthetized patients with and without renal function. Pyridostigmine serum levels were assayed by gas-liquid chromatography, and data were fitted to a 2-compartment kinetic model. Pyridostigmine kinetics following renal transplantation (n = 5) were not different from those in patients with normal renal function. Renal function (n = 5) elimination half-life increased from 112 +/- 12 min (mean +/- SD) to 379 +/- 162 min, and serum clearance decreased from 9 +/- 2 ml/kg/min to 2 +/- 0.6 ml/kg/min in anephric patients (n = 4). We conclude that renal function accounts for 75% of pyridostigmine clearance.

Effects of cardiac surgery with cardiopulmonary bypass on lidocaine disposition.

We investigated lidocaine kinetics after cardiac surgery to elucidate the effects of cardiopulmonary bypass (CPB) and major surgery on its disposition. In five patients lidocaine kinetics were unchanged 15 min and 1 day after CPB, but lidocaine clearance decreased 42% and volume of distribution was reduced by 40% in eight patients 3 days after surgery. Seven days after surgery lidocaine kinetics had returned to baseline in five patients. These changes correlated with a doubling of the alpha 1-acid glycoprotein concentration and a 46% decrease in lidocaine free fraction on day 3, effects that persisted on day 7. Our results suggest that the use of lidocaine need not be altered in the first day after uncomplicated coronary artery surgery, but, in light of our findings the use of lidocaine and the interpretation of measured total lidocaine concentrations 3 to 7 days thereafter should be reexamined.

Electroencephalographic effects of benzodiazepines. I. Choosing an electroencephalographic parameter to measure the effect of midazolam on the central nervous system.

The goal of this investigation was to determine a numerical electroencephalographic parameter that best indicated the degree of the effect of midazolam, administered in hypnotic doses, on the central nervous system. This electroencephalographic parameter could then be used to relate midazolam plasma concentrations and electroencephalographic drug effect (pharmacodynamic modeling). Intravenous doses of midazolam (3.75 to 25 mg) were given to five men at an infusion rate of 5 mg/min. A cortical electroencephalogram was continuously recorded. Two waveform analysis approaches were examined: fast Fourier transformation and aperiodic analysis. From fast Fourier transformation and aperiodic analysis a set of parameters were examined as measures of drug effect. We conclude that the voltage per second from aperiodic analysis provided the electroencephalographic parameter that optimally measured the effect of midazolam on the central nervous system.

Electroencephalographic effects of benzodiazepines. II. Pharmacodynamic modeling of the electroencephalographic effects of midazolam and diazepam.

The comparative pharmacodynamics of midazolam and diazepam were examined by use of the electroencephalogram as a measure of drug effect on the central nervous system. Intravenous doses of 7.5, 15, and 25 mg midazolam and 15, 30, and 50 mg diazepam were given on repeated occasions to three volunteers. Arterial plasma concentration and electroencephalogram voltage were related with nonparameteric and parametric pharmacodynamic models. The peak increases in voltage (maximal effect) and the slopes of the plasma concentration versus effect curve were similar for both drugs. The half-time of blood:brain equilibration was significantly longer for midazolam than diazepam (4.8 minutes versus 1.6 minutes). Midazolam was found to have an intrinsic steady-state potency that was approximately five times greater than that of diazepam (152 ng/ml versus 958 ng/ml).

Plasma and cerebrospinal fluid concentrations of chlordiazepoxide and its metabolites in surgical patients.

Thirty otherwise healthy patients received a 100-mg oral dose of chlordiazepoxide HCl just prior to surgical procedures using spinal anesthesia. Fourteen of these patients had also received 100 mg on the night before surgery. Simultaneous samples of venous blood and cerebrospinal fluid (CSF) were taken immediately prior to injection of spinal anesthesia and were assayed for concentrations of chlordiazepoxide (CDX) and its major metabolite, desmethylchlordiazepoxide. Plasma concentrations of CDX ranged from 2.32 to 13.34 mug/ml. Simultaneous CSF concentrations were considerably lower, ranging from 0.04 to 0.34 mug/ml. Equilibration of CDX between plasma and the lumbar sampling site appeared to be complete within 2 hr of the most recent dose. After attainment of distribution equilibrium, simultaneous plasma and CSF concentrations of CDX were hightly correlated (r = 0.76), with a mean CSF-plasma concentrations ratio of only 0.043 (range; 0.02 to 0.06). The limited passage of CDX into human CSF is probably due to extensive binding to plasma protein. Assuming that transfer of CDX from plasma to CSF is governed by passive diffusion, the extent of plasma protein binding of CDX in healthy individuals averages about 96%.

Kinetics of high-dose intravenous morphine in cardiac surgery patients.

Ten patients received 1.0 mg/kg of morphine sulfate by constant-rate intravenous infusion at 5 mg/min over 9 to 27 min. Multiple arterial blood samples were drawn during the first 30 to 151 min after termination of the infusion, prior to institution of cardiopulmonary bypass. Postinfusion plasma concentrations were fitted by computer to biexponential functions consistent with a 2-compartment open pharmacokinetic model. Mean (+/- SE) pharmacokinetic parameters were: volume of central compartment, 0.09 +/- 0.03 L/kg; total apparent volume of distribution, 1.02 +/- 0.09 L/kg; distribution T 1/2, 0.90 +/- 0.09 min; apparent elimination T 1/2, 137 +/- 14 min; total clearance, 378 +/- 63 ml/min. Thus distribution of morphine is very rapid, but the apparent volume of distribution is only slightly larger than body weight, suggesting limited tissue uptake. Since apparent elimination T 1/2s are similar to those reported after smaller doses, evidence of saturable or capacity-linked elimination is lacking. Total clearances, representing mainly hepatic clearance, averaged about 25% of hepatic blood flow, suggesting clinically important first-pass metabolism of oral morphine.

A comparison of spectral edge, delta power, and bispectral index as EEG measures of alfentanil, propofol, and midazolam drug effect.

BACKGROUND: The effects of anesthetic drugs on electroencephalograms (EEG) have been studied to develop the EEG as a measure of anesthetic depth. Bispectral analysis is a new quantitative technique that measures the consistency of the phase and power relationships and returns a single measure, the bispectral index. The purpose of this study was to compare the performance of the bispectral index, version 1.1, with other spectral analysis EEG measures of drug effect for three commonly used anesthetic drugs. METHODS: The EEG waveforms from 31 adults receiving infusions of alfentanil, propofol, or midazolam were analyzed. The time course of spectral edge (SE95), relative power in delta band, and bispectral index were related to the estimated effect-site concentration with use of a sigmoidal Emax model to estimate the potency (IC50) and the plasma effect-site equilibration rate constant (Ke0) for each measure. The performance of the fitting was assessed by the coefficient of correlation between predicted and observed effect. RESULTS: Alfentanil induced a high-amplitude low-frequency EEG response. Propofol induced a biphasic response. At low concentrations, both frequency and amplitude increased. When the concentration increased, the EEG slowed and the amplitude decreased. High concentration produced burst suppression. Midazolam increased EEG frequency and amplitude. Bispectral index, SE95, and delta power yield similar estimates of IC50 and ke0. Except for alfentanil, the performance of the modeling with the bispectral index was as good that with SE95 or delta power. CONCLUSION: Bispectral analysis can be used as a measure of the EEG effects of anesthetic drugs.

Simultaneous modeling of pharmacokinetics and pharmacodynamics: application to d-tubocurarine.

We propose a model of drug pharmacodynamic response that when integrated with a pharmacokinetic model allows characterization of the temporal aspects of pharmacodynamics as well as the time-independent sensitivity component. The total model can accommodate extremes of effect. It allows fitting of simultaneous plasma concentration (Cp) and effect data from the initial distribution phase of drug administration, or from any non-equilibrium phase. The model postulates a hypothetical effect compartment, the dynamics of which are adjusted to reflect the temporal dynamics of drug effect. The effect compartment is modeled as an additional compartment linked to the plasma compartment by a first-order process, but whose exponential does not enter into the pharmacokinetic solution for the mass of drug in the body. The hypothetical amount of drug in the effect compartment is then related to the observed effect by the Hill equation, a nonlinear sigmoid form. Nonlinear least-squares data fitting is used for parameter estimation. The model is demonstrated on two different sets of Cp and effect data for the drug d-tubocurarine (dTC). In 7 normal subjects, the (mean +/- SD) rate constant for equilibration of dTC effect (paralysis) and Cp is 0.13 +/- 0.04 min-1 and the (mean +/- SD) steady-state Cp required to produce 50% paralysis is 0.37 +/- 0.05 microgram/ml.

Pharmacokinetic-pharmacodynamic modeling in drug development: application to the investigational opioid trefentanil.

OBJECTIVE: We determined the possible benefits of a new opioid, trefentanil, relative to fentanyl and alfentanil using high-resolution pharmacokinetic-pharmacodynamic modeling and computer simulations of clinical dosing scenarios. METHODS: First, we determined in nine volunteers the electroencephalographic (EEG) effects and the trefentanil infusion rate that gave maximal EEG changes in 3 to 10 minutes. Then, in a crossover fashion in five volunteers, we compared the pharmacokinetics and EEG pharmacodynamics of trefentanil with fentanyl and alfentanil. Finally, we used computer simulations to predict offset of opioid effects of trefentanil, fentanyl, and alfentanil when given in different dosing schemes. RESULTS: The pharmacokinetic-pharmacodynamic profile of trefentanil was similar to alfentanil, except for a higher elimination clearance. Trefentanil versus alfentanil pharmacokinetic parameters were as follows: Elimination clearance, 0.444 +/- 0.073 versus 0.184 +/- 0.031 L/min; steady-state distribution volume, 37 +/- 7 versus 23 +/- 3 L; and elimination half-life, 127 +/- 24 versus 114 +/- 19 minutes. Trefentanil versus alfentanil pharmacodynamics were as follows: the equilibration half-time between EEG effect and arterial drug concentration, 1.2 +/- 0.5 versus 0.6 +/- 0.4 minutes; and the concentration resulting in 50% of maximal EEG effect, 429 +/- 313 versus 577 +/- 273 ng/ml. The pharmacokinetic-pharmacodynamic profile of fentanyl was significantly different from trefentanil and alfentanil. Simulation of effect compartment concentration decay curves after variable-length infusions predicted more rapid recovery from trefentanil than from alfentanil or fentanyl. CONCLUSION: We suggest that high-resolution pharmacokinetic-pharmacodynamic studies and computer simulations of clinical dosing scenarios may have significant usefulness in appreciating differences between new and established drugs in early phase I studies.

Haloperidol kinetics after oral and intravenous doses.

Haloperidol kinetics were determined after oral and intravenous drug doses in 15 men. Mean elimination t1/2 for the subjects was 17.9 +/- 6.4 (SD) hr. After 0.125 mg/kg IV, mean distribution t1/2s in six subjects were 0.19 +/- 0.07 and 2 +/- 1 hr, and in 12 subjects mean clearance was 11.8 +/- 2.9 ml/kg/min and mean steady-state volume of distribution was 17.8 +/- 6.5 l/kg. After 0.50-mg/kg oral doses in eight subjects, mean lag time before drug absorption was 0.82 +/- 0.25 hr. Mean absorption t1/2 was 0.37 +/- 0.18 hr and mean distribution t1/2 was 0.96 +/- 0.20 hr. Bioavailability was 0.65 +/- 0.14 after oral doses. In 14 kinetic studies in nine subjects, data was analyzed by both model-dependent (open two- and three-compartment models using nonlinear regression) and model-independent (AUC and first moment curve) approaches. Results of the two were found to be in close agreement. The long elimination t1/2 of haloperidol is explained by the drug's extensive tissue distribution.

Pharmacodynamic modeling of the electroencephalographic effects of flumazenil in healthy volunteers sedated with midazolam.

The purpose of this study was to model pharmacodynamically the reversal of midazolam sedation with flumazenil. Ten human volunteers underwent four different sessions. In session 1, individual midazolam pharmacokinetics and electroencephalographic pharmacodynamics were determined. In sessions 2 and 3, a computer-controlled infusion of midazolam with individual volunteer pharmacokinetic data was administered, targeting a plasma concentration corresponding to a light or deep level of sedation (20% or 80% of the maximal midazolam electroencephalographic effect) for a period of 210 minutes. After obtaining a stable electroencephalographic effect and constant midazolam plasma concentrations, a zero-order infusion of flumazenil was started until complete reversal of midazolam electroencephalographic effect was obtained. The flumazenil infusion was then stopped and the volunteer was allowed to resedate because of the constant midazolam drug effect. The electroencephalographic response was measured during a 180-minute period and analyzed by aperiodic analysis and fast-Fourier transforms. In session 4, a midazolam plasma concentration corresponding to a deep level of sedation was targeted for 210 minutes to examine for the possible development of acute tolerance. No flumazenil was given in session 4. For a light sedation level, with a mean midazolam plasma concentration of 160 +/- 64 ng/ml, the mean half-life of the equilibration rate constant of flumazenil reversal is 5.0 +/- 2.5 minutes, and the mean effect site concentration causing 50% of Emax is 13.7 +/- 5.8 ng/ml. For a deep level of sedation, with a mean midazolam plasma concentration of 551 +/- 196 ng/ml, the mean half-life of the equilibration rate constant is 3.9 +/- 1.5 minutes, and the mean effect site concentration causing 50% of Emax is 20.6 +/- 6.8 ng/ml. This study provides an estimate of the magnitude of the blood/central nervous system equilibration delay for flumazenil antagonism of midazolam sedation and further defines the usefulness of the electroencephalogram as a measure of midazolam pharmacodynamic effect.

Effect of cardiopulmonary bypass on the pharmacokinetics of drugs.

The cardiopulmonary bypass apparatus must temporarily substitute for the cardiac and pulmonary function of the patient undergoing heart surgery. In order to meet the metabolic needs of the patient and the technical demands of the surgeon, within the limits of engineering technology, a number of major alterations are made in normal physiology. The patient is typically cooled to 27 degrees C and perfused with a non-pulsatile flow of blood which has been diluted with saline to a haematocrit in the mid-20s. Blood flow and pressure are often considerably less than normal. Blood coagulation is prevented by administration of a massive dose of heparin. Central redistribution of blood flow, elaboration of stress-reactant hormones, and fluid and electrolyte shifts occur in response to these changes. In the postoperative period, these alterations are reversed, and normal physiology is restored. Effects upon the pharmacokinetics of drugs are anticipated. The clearance of many drugs may be reduced. Protein binding is diminished by haemodilution, but may rise above normal in the postoperative period for basic drugs which bind to alpha 1-acid glycoprotein. Changes in volume of distribution depend upon the opposing influences of protein binding and reduced peripheral perfusion. Previous studies on the pharmacokinetics of drugs during and after cardiopulmonary bypass illustrate many of these effects. The clearance of digoxin, fentanyl, and the cephalosporins is reduced after cardiopulmonary bypass, and the volume of distribution of cefazolin is increased during cardiopulmonary bypass. Studies of digitoxin and propranolol are also reviewed. Many of the investigations in this area of study have been limited by logistical and methodological factors. Thus, the effects of cardiopulmonary bypass on the pharmacokinetics of drugs are incompletely understood, and the subject merits further attention.

No effect of age on the dose requirement of thiopental in the rat.

With increasing human age (20-80 years), the electroencephalogram (EEG) dose requirement for the intravenous anesthetic thiopental decreases approximately 10% per decade of life. The goal of this study was to compare the dose required to attain isoelectric EEG in young (4-5 month) vs. aged (24-25-month) Fischer 344 rats. One second isoelectricity was found to be an endpoint where minimal cardiorespiratory depression occurred. The effects of age, infusion rate, and repeated administration were examined in nine young and nine old rodents. Thiopental dose requirement increased with increasing infusion rates. Repeated administration at two-day intervals did not demonstrate tolerance to thiopental. No difference in thiopental dose requirement was detected in the young vs. elderly rats. In a separate group of five young and five old rats, thiopental plasma, brain, heart, and CSF concentrations were measured when five seconds of EEG isoelectricity was achieved: no consistent differences were noted. The rat may not be an appropriate model to investigate acute age-related anesthetic effects in humans, because cardiovascular changes with age are dissimilar between species.

Factors influencing blood concentrations of chlordiazepoxide: a use of multiple regression analysis.

Three groups of male and female subjects aged 24-74 years received 25, 100, or 200 mg of chlordiazepoxide hydrochloride by mouth as a single dose or as two divided doses. The relation of plasma or whole blood concentrations for chlordiazepoxide (CDX) and its metabolite, desmethylchlordiazepoxide (DMCDX), to time since the last dose, weight, age, and sex were determined by simple and multiple regression analyses. Both CDX and DMCDX levels were negatively correlated with weight. Concentrations of CDX decreased, while those of DMCDX increased, with the time since the last dose. Lower levels of both drugs were associated with female sex, and lower levels of DMCDX were noted with increasing age. In the largest sample group, age and weight were more important variables than sex in accounting for CDX and DMCDX. Sex was of significance, and more important than time or age in explaining the variance of CDX in one series of observations. Multiple regression analysis is a useful approach to assessing interrelated factors influencing blood levels of drugs, especially when combined with a consideration of the interactive components of variance. Age and sex, in addition to weight and time, may be important factors that deserve further attention.

Efficacy and safety of single doses of intramuscular ketorolac tromethamine compared with meperidine for postoperative pain.

Ketorolac tromethamine, a potent nonnarcotic prostaglandin synthetase-inhibiting analgesic, was compared with meperidine for relief of moderate to severe postoperative pain. In a double-blind, randomized study, 125 patients received single intramuscular doses of ketorolac 30 or 90 mg or meperidine 50 or 100 mg. The degree of pain and pain relief were quantified verbally and with visual analog scales at baseline and 30 minutes, then hourly for 6 hours. Ketorolac 30 and 90 mg were significantly superior to meperidine 50 mg in six of nine efficacy measures. The onset of and peak analgesic effect of both doses of ketorolac and of meperidine were equivalent. Compared with both doses of meperidine, the two doses of ketorolac exhibited significantly longer duration of analgesic effect, as measured by the percentage of patients who terminated the study because of inadequate pain relief. The frequency of side effects was not significantly different between the drugs. The prolonged efficacy of intramuscular ketorolac combined with the reduced risk of respiratory depression suggest an important use of this drug for the relief of postoperative pain.

Factors affecting the measurement of lidocaine protein binding by equilibrium dialysis in human serum.

A systematic study was undertaken to assess in vitro factors that influence the value of the lidocaine free fraction obtained by equilibrium dialysis in human serum. These factors include pH readjustment to 7.40 after serum storage; choice of buffers for dialysis; the effect of phosphate buffer ionic strength; temperature of storage for serum samples; the use of untreated versus silanized glassware for storage; and age of serum. It was concluded that the pH of serum that contains lidocaine must be brought back to the original whole blood pH found in the patient before equilibrium dialysis because the protein binding of lidocaine is critically dependent on pH. It was also found that Krebs-Ringer bicarbonate buffer, when used with room air atmosphere in the dialysis cell, is not adequate to control pH even when serum pH is readjusted to the physiological pH of the patient. Isotonic phosphate buffer and 0.10 M phosphate buffer are effective for pH control and give identical values of lidocaine free fraction when the original serum sample is first pH-adjusted. If the pH of the serum is correct and the pH of the buffer remains constant, then freezing, the choice of container, or the age of serum are not important variables affecting the measurement of the lidocaine free fraction.