The pharmacokinetics and clinical efficacy of ropivacaine (2.5 mg/ml) during a 24-h continuous epidural infusion for postoperative pain relief in 20 patients scheduled for abdominal hysterectomy were characterized using an open-label, increasing-dose design.


Through an epidural catheter inserted at T10-T12, a test dose of 7.5 mg ropivacaine was given 3 min before a bolus dose of 42.5 mg and immediately followed by a 24-h continuous epidural infusion with either 10 or 20 mg/h. Peripheral venous plasma samples were collected up to 48 h after infusion, and urinary excretion was followed up to the end of infusion. Postoperative pain at rest, on coughing, and at mobilization was assessed by means of a visual analog scale 2,4,6,8,12, and 24 h after the end of surgery. Sensory (pinprick) and motor block (modified Bromage scale) were assessed at the same intervals.


The total plasma concentrations of ropivacaine increased markedly and consistently during the 24-h epidural infusion, in contrast to stable unbound concentrations. Both total and unbound plasma concentrations at the end of infusion were proportional to the total dose, although only the latter was proportional to the infusion rate. The total and unbound plasma clearance was independent of dose. Total mean clearance decreased on average by 21% (P < 0.001) during the last 12 h of epidural infusion, i.e., from 539 +/- 191 ml/min to 418 +/- 138 ml/min, indicating time-dependent kinetics. The unbound clearance also varied between estimates after 8 h of infusion and the end of treatment, i.e., a 5.3% decrease from 10.4 +/- 5.3 l/min to 9.5 +/- 3.9 l/min (P < 0.05). The unbound fraction of ropivacaine in plasma decreased during treatment, and this was related to the increase in alpha1-acid glycoprotein concentration. Pain was generally well controlled, and median visual analog scale scores during mobilization were less than 30 mm in patients receiving ropivacaine at 20 mg/h.


The pharmacokinetics of ropivacaine were independent of dose, but total clearance decreased with time over 24 h. The consistent increase in total plasma concentration during the postoperative epidural infusion contrasted to much less variation in the unbound plasma concentrations of ropivacaine.

Key words: Anesthetic techniques: epidural analgesia. Anesthetics, local: ropivacaine. Pharmacokinetics. Postoperative pain.)

ROPIVACAINE, a long-acting local anesthetic, is less prone than bupivacaine to produce central nervous system and cardiovascular changes after intravenous infusion. [1]It has disposition half-lives of 20 and 111 min after intravenous administration, a plasma clearance of 500 ml/min, and a plasma protein-binding of 94%, mainly to alpha1-acid glycoprotein. [2]Data from studies in volunteers and surgical patients suggest that ropivacaine is similar to bupivacaine regarding onset, duration, and extent of sensory block but gives a less intense motor block of shorter duration. [3-6].*

Because postoperative pain is a major indication for a new local anesthetic agent and no pharmacokinetics had been described after continuous or multiple epidural dosing with ropivacaine, the aim of this study was to assess the pharmacokinetics and explore the clinical efficacy (analgesia and degree of motor block) of ropivacaine during continuous 24-h epidural infusion for postoperative pain relief.


The study protocol was approved by the Ethics Committee of the Municipality of Copenhagen.

After giving informed consent, 20 ASA physical status 1 or 2 women were enrolled in the study. They were between 34 and 51 yr old, and their body weights varied from 56 to 94 kg. The demographic characteristics of the two dose groups were similar, with a mean age of 42 +/-4 and 42+/-6 yr and body weights of 69+/-13 and 73+/-9 kg in the 10- and 20-mg/h groups, respectively. No patient had alcohol or drug abuse or received drugs known to interfere with drug metabolism. All patients underwent total abdominal hysterectomy for benign disease. The time between drug administration and subsequent surgery varied between 45 and 185 min (mean 70 min). All patients received a preanesthetic intravenous infusion of sodium chloride or a saline mixture, 500-1400 ml, given 61-190 min before surgery, and all except one were given 10-15 mg diazepam orally.

The duration of surgery varied between 60 and 120 min (median 85 min) for patients receiving 10 mg/h ropivacaine and between 65 and 140 min (median 98 min) in those receiving 20 mg/h ropivacaine.

Drug Treatment

The epidural administration was performed with 2.5 mg/ml ropivacaine (batch 470-5-1 and 470-15-1) at T10 to T12. All patients received a test dose of 3 ml (7.5 mg) at a rate of 1 mg/s. Three minutes after the test dose, 17 ml (42.5 mg) was injected epidurally as a bolus dose over 1-5 min, followed immediately by a continuous epidural infusion for 24 h (22.3-24.4 h). The epidural infusion rate was increased from 4 ml (10 mg)/h to 8 ml (20 mg)/h in patients 11 to 20. This decision was based on tolerability after the ropivacaine infusion in the first ten patients, the apparent safety as judged by the total plasma concentrations of ropivacaine achieved, and a desire to improve postoperative pain relief. After the end of infusion, a 5-ml sample of the infusion fluid was collected for analysis of its actual content of ropivacaine hydrochloride (from each bottle used). The actual content of ropivacaine hydrochloride in the infusion fluid varied between 2.4 and 2.7 mg/ml according to assay. The nominal content of 2.5 mg/ml has been used in all calculations.

Induction of general anesthesia was performed with thiopental sodium, pancuronium, and succinylcholine and anesthesia maintained using enflurane.

Blood Sampling and Drug Assay

Peripheral blood was collected from the cubital vein via an indwelling catheter from the arm not used for drug and/or fluid injection/infusion. Blood samples were taken before *symbol* the test dose, 15 and 30 *symbol* min and 1, 2, 4 *symbol*, 8 *symbol*, and 12 h after the start of the epidural bolus dose and 0 *symbol* (end of infusion), 5, 15, 30, and 45 min and 1, 2, 4 *symbol*, 8, 12 *symbol*, 24 *symbol*, and 48 *symbol* h after the end of infusion. At times indicated by a bullet (*symbol*), 7 ml of blood was collected for assay of total and unbound concentrations as well as alpha1-acid glycoprotein; otherwise, 4 ml was collected for assay of total concentrations. The samples were taken in heparinized tubes (Venoject, MEDA, Gladsaxe, Denmark), and plasma was separated by centrifugation at room temperature within 30 min of collection. The plasma samples were stored at -20 degrees Celsius until drug assay.

One urine sample was collected before epidural drug administration (before infiltration of the skin and the epidural test dose of ropivacaine). The total urine output was collected from administration of the test dose up to 4 h after the start of the epidural bolus dose of ropivacaine. Thereafter, samples were collected during the 4-8-, 8-12-, and 12-24-h periods after the start of the epidural bolus dose. The volume (ml) of the total urine collected during each interval was recorded, and 10-ml portions were rapidly frozen and stored at -20 degrees C until drug assay.

Assay of plasma and urine for ropivacaine base was performed by gas chromatography with a nitrogen-sensitive detector after liquid-liquid extraction. [7]The limit of quantification in plasma and urine was 10 micro gram/l and the between-day precision about 5%. The recovery of spiked control samples was close to 100%. The unbound concentration of ropivacaine base was determined by direct injection of plasma ultrafiltrate into a coupled column liquid chromatographic system. Ropivacaine was detected by ultraviolet light at 210 nm. [8]The limit of quantification was 3 micro gram/l and the between-day precision about 7%. The concentration of ropivacaine is given in mg/l. However, in the evaluation of drug protein-binding, a molar unit is more appropriate. To obtain molar units, the conversion factor for ropivacaine is 3.64 (ropivacaine base Mr274.4).

The concentration of alpha1-acid glycoprotein in plasma was determined in the samples taken before drug administration and in the samples selected for analysis of unbound concentration by using an immunodiffusion technique with commercially prepared kits (NOR-Partigen-acid-alpha1-Glycoprotein, Behringwerke, Marburg, Germany). The limit of determination was 2 micro mol/l and the between-day precision about 9%.

Clinical Assessments

The upper and lower spread of analgesia on the right and left sides of the patient (loss of sharp sensation in response to pinprick) was determined with a short bevelled needle. The testing times were 5, 10, 15, 20, and 30 min after the end of the bolus injection and 2, 4, 8, 12, and 24 h after the end of surgery.

Motor block was assessed by recording the motor function of the lower limbs immediately after each determination of sensory block: 0 = no paralysis (full flexion of knees and feet); 1 = inability to raise extended legs (just able to move knees); 2 = inability to flex knees (able to flex ankle joint); 3 = inability to flex ankle joints (unable to flex ankle joint and knee).

A 100-mm visual analog scale was used for the assessment of pain (0 mm = no pain, 100 mm = worst imaginable pain.) Pain was assessed when the patient was recumbent, during mobilization to a sitting position, and during coughing, when the patient was in a sitting position. The assessments were performed by the same anesthesiologist (C-J. E.) 2, 4, 8, 12, and 24 h after the end of the surgery.

Heart rate and systolic and diastolic blood pressure were measured before premedication, before administration of the test dose, and immediately after each determination of motor block. If hypotension occurred, the foot of the bed was raised and an electrolyte solution given by rapid intravenous infusion. If hypotension persisted, ephedrine (30 mg) was given intramuscularly or intravenously.

Before the study, a routine preoperative medical examination was performed, including clinical chemistry screening with the following tests: B-hemoglobin; B-thrombocytes; B-leukocytes, including differential count; S-albumin; alpha1-acid glycoprotein; S-creatinine; and aspartate aminotransferase.

All adverse experiences of a local or systemic nature were recorded.

Pharmacokinetic Analysis

Pharmacokinetic calculations were performed in each patient using a validated pharmacokinetic program written in the RS/1 command language as an underlying structure. ** Release 6.08 of the SAS system under VMS also has been used in the analysis and presentation of pharmacokinetic parameters using established programs within SAS. *** ****.

Cstopis the plasma concentration at the end of infusion and Cnthe concentration at time tn. The area under the total (AUCinfinity) as well as the unbound (AUCinfinity, u) plasma concentration-time curve was calculated by numeric integration using the linear trapezoidal rule. The estimated area from the last sampling time to infinity was calculated from the plasma concentration at the last sampling time divided by the terminal disposition rate constant (lambdaz), which was estimated by the method of least squares from the terminal linear part of the log plasma concentration versus time curve. The total (AUCinfinity(n)) and unbound (AUCinfinityu(n)) area under the plasma concentration-time curve up to time tnwas calculated in a similar way using the disposition rate constant (lambdaz) from the terminal linear part of the plasma concentration-time curve to estimate the residual area.

Total (CL) and unbound (CLu) plasma clearance were calculated, assuming total (100%) absorption after epidural administration, as: Equation 1, Equation 2. where the dose is expressed as ropivacaine base. The corresponding clearances after a cumulative dose up to time tnwere calculated as: Equation 3, Equation 4.

The fraction of unbound drug in plasma (fu) was calculated as the unbound divided by the total plasma concentration at each sampling time. The terminal elimination half-life was estimated as ln2/lambdaz.

Statistical Analysis

To compare results of variables measured on two occasions, the individual difference was calculated, and Wilcoxon's signed-rank test was used to determine whether the individual differences [9]varied from zero. For the half-life, a comparison between dosage was done using a two-sample Wilcoxon's test. Normal approximation was used in both cases. Hodges-Lehmann estimates and 95% confidence limits were calculated for the differences in clearance and the shift in location between the groups. [9]A simple linear regression was used to study the relationship between AUC and dose for both total and unbound plasma concentration. Ninety-five-percent confidence intervals for the intercepts were calculated. The relationship of protein concentration, total dose, time, and patient to fraction of unbound plasma concentration was studied by fitting a multiple regression line to the data. The model was chosen by Mallow's Cp. [10]The model chosen was considered the most relevant of several possible, according to the selection criteria.

In the description of analgesia over time, the horizontal axis is the time from end of surgery, and it is truncated at the time for the last planned measurement. In the graphs of group medians, the median values are plotted at the target times planned in the protocol. The median values are based on individual values, which are aligned with these target times, i.e., a complete set of values at the target times was determined for each subject. Essentially, aligned values are set to original values recorded sufficiently close to the target times. Moreover, the "last value carried forward" principle is used in case an original value is not close enough in time to a target time. Thus, each subject was assigned a complete sequence of values at the target times.

All tests were two-sided, and the statistical significance level was 0.05. The results are presented as the mean+/-SD or the median.

Pharmacokinetic Evaluation

In all calculations, time zero is the time of the start of the test dose.

The total plasma concentrations of ropivacaine did not reach a plateau but increased during the whole 24-h epidural infusion in almost all patients. The largest increase was seen in patients given 20 mg/h (Figure 1). The unbound plasma concentrations were more stable during the infusion period after the plateau had been reached, although increasing concentrations were seen in a few patients (Figure 2). The mean total plasma concentration increased 36% from 0.39+/-0.16 to 0.53+/-0.22 mg/l during the last 12 h of epidural infusion with 10 mg/h and 38% from 0.64+/-0.23 to 0.88+/-0.24 mg/l during epidural infusion with 20 mg/h. The corresponding differences in mean unbound plasma concentration between 8- and 24-h epidural infusion were much smaller, 0.019+/-0.0084 to 0.017+/-0.0059 mg/l and 0.032+/-0.016 to 0.035+/-0.015 mg/l, respectively. In 10 of the 20 patients, the highest total plasma concentration was seen 5-25 min after the end of infusion.

The mean total plasma concentrations at the end of infusion (C sub stop, 24 h) were 0.53 and 0.88 mg/l after ropivacaine infusion rates of 10 and 20 mg/h, corresponding to mean total doses of 290 (288-293) and 511 (487-528) mg ropivacaine. This increase was proportional to the total dose. The corresponding mean unbound plasma concentrations (Cu, stop, 24 h) were 0.017 and 0.035 mg/l after 10 and 20 mg/h, which were proportional to the infusion rate and the total dose.

The apparent terminal half-life (t1/2) was 2.4+/- 0.9 h in the 10-mg/h group and 2.7+/-0.9 h in the 20-mg/h group and estimated from the last 3-11 data points on the terminal linear phase of the total plasma concentration-time curve. The extrapolated residual area under the total plasma concentration-time curve was small, < 2%. The increase in mean total AUC (AUCinfinity) after the epidural infusion of 10 to 20 mg/h was 58% (12.2+/-5.4 to 19.3+/-6.5 mg *symbol* h sup -1 *symbol* 1 sup -1). AUC was roughly proportional to the total dose, and the intercept of the linear regression line did not differ significantly from zero (95% confidence limits -7.7 to 13.3). The half-life was independent of dose and varied from 1.5 to 4.4 h.

The mean increase in total area under the unbound plasma concentration-time curve (AUCinfinity,u) after the epidural infusion of 10 to 20 mg/h was 68% (0.55+/-0.16 to 0.89+/-0.43 mg *symbol* h sup -1 *symbol* l sup -1). AUCinfinity,u was approximately proportional to the total dose, and the intercept of the linear regression line did not differ significantly from zero (95% confidence limits -0.39 to 0.93).

The total plasma clearance (418+/-138 ml/min) was independent of dose (Figure 3) but decreased significantly (P < 0.001), on average by -21% (-7 to -36%), when estimated after 12 h of infusion compared with the end of treatment. The 95% confidence limits for the difference in percent were -16.1 to -25.5. The unbound plasma clearance (CLu) was similar regardless of dose (Figure 3) and significantly different (P = 0.04) between estimates after 8 h of infusion and the end of treatment, although the differences were less pronounced, -5.3% (10 to -28%). The 95% confidence limits for the difference in percent were -0.04 to -12.4.

The alpha1-acid glycoprotein concentration showed a slight decrease during the first 6 h of epidural infusion. Thereafter, there was a steady increase during the whole 72-h follow-up period (Figure 4).

The unbound fraction (fu) decreased during treatment. The fuvalues calculated from individual samples taken between 7 and 8.7 h after the start of treatment were 5.6+/-1.2% (n = 9) and 5.2 +/-1.4% (n = 10) in patients with infusion rates of 10 and 20 mg/h, respectively. The corresponding fuvalues calculated from samples taken between 22 and 25 h after start of treatment were 3.3 +/-0.8% (n = 10) and 3.7+/-1.2% (n = 9), respectively. The decrease in fuwithin each treatment group was statistically significant (P < 0.001). There was a linear relationship between the decrease in fuand the increase in alpha1-acid glycoprotein (Figure 5). The relationship of alpha1-acid glycoprotein concentration, ropivacaine total plasma concentration, total dose, time, and patient to the unbound fraction of ropivacaine in plasma was analyzed by multiple linear regression. Only the concentration of alpha1-acid glycoprotein gave a statistically significant response (P < 0.0001) on the unbound fraction. About 50% of the variation in the unbound fraction was explained by variation in the alpha1-acid glycoprotein over time (r sup 2 = 0.49).

Up to the time when the epidural infusion was stopped (24 h), the cumulative amount of unchanged ropivacaine recovered from the urine varied from 1.1 mg (0.4% of dose) to 12.9 mg (3.0% of dose).

Clinical Evaluation

(Figure 6) presents median visual analog scale scores at rest, during coughing, and during mobilization. The patients receiving the highest infusion rate generally experienced less pain than those receiving the lowest infusion rate. The patients who received 20 mg/h showed marked pain relief (visual analog score < 30) during mobilization in the postoperative period.

The maximum upper spread achieved by any patient was T3 in the 10-mg/h group and T2 in the 20-mg/h group, both occurring 4 h after the start of the epidural infusion. The maximum lower spread was S5 in both groups. The maximum median upper spread was T4 in the 10-mg/h groups and T3 in the 20-mg/h group. The corresponding maximum median lower spread was L4 and L4-L5.

The median upper sensory block in the 10-mg/h group decreased from T4 at the start of infusion to T8 at 24 h. The block in the 20-mg/h group decreased from T3 at 4 h to T5-T6 at 24 h. The median lower spread of block decreased from L4 at 2 h to L1 at 24 h in the 10-mg/h group and from L4-L5 at 4 h to L2 at 24 h in the 20-mg/h group.

Only four patients, two in each group, experienced motor block (degree 1).

Seven and four patients in the 10- and 20-mg/h groups, respectively, received additional morphine for postoperative pain. In three patients, there was insufficient information on the amount of morphine given. Among the others, large morphine doses were used in the group given 10 mg/h (median 28 vs. 5 mg in the 20-mg/h group).

Except for two patients requiring ephedrine for intraoperative hypotension, no clinically important changes in blood pressure or heart rate were observed in the course of the study.

The principal findings of this first study of pharmacokinetics of ropivacaine during continuous epidural infusion in surgical patients are (1) a marked time-dependent decrease in total but not unbound clearance and (2) a stable unbound but increasing total plasma concentration. Furthermore, 20 mg/h continuous epidural ropivacaine was more effective for postoperative pain than 10 mg/h and without adverse effects.

Ropivacaine is predominantly eliminated by metabolism in the liver with an intermediate to low extraction ratio. In volunteers, low to medium individual hepatic extraction ratios of between 0.20 and 0.57 have been reported.***** Thus, the rate of elimination should depend on the unbound plasma concentration of ropivacaine. According to pharmacokinetic theory, the plasma clearance of ropivacaine (CL) is then expected to vary proportionally with changes in the unbound fraction (fu) according to Equation 1and when the extraction ratio is low, the intrinsic clearance (CLint) remains unchanged. [11,12]This is supported by the current results, in which the plasma clearance of ropivacaine showed a marked and consistent decrease with a decrease in the unbound fraction, whereas the unbound clearance (approximately CLint) of ropivacaine showed smaller and less consistent changes. Deviations from this in individual patients probably is explained by additional variation in the intrinsic metabolic clearance. The lack of multicompartment characteristics in the total ropivacaine plasma concentrations after infusion (not shown) implies that reasonable estimates of the half-life of unbound ropivacaine are expected although based on a small number of samples in each patient. An error in the clearance estimates is, therefore, an unlikely explanation of the consistent decrease in plasma clearance of ropivacaine in all patients and the corresponding smaller and less consistent changes in unbound plasma clearance. Furthermore, bupivacaine clearance has been reported to be inversely related to serum alpha1-acid glycoprotein concentrations. [13]In the calculations of clearance after epidural administration, complete absorption was assumed. This assumption is supported by reports of 95% bioavailability of epidural ropivacaine in the rhesus monkey. [14].

Like other local anesthetics, the plasma protein-binding of ropivacaine mainly is due to association with alpha1-acid glycoprotein, a so called acute-phase or stress protein that increases postoperatively. The molar concentration of alpha1-acid glycoprotein was approximately 10 times higher than the molar concentration of ropivacaine, i.e., a linear drug protein-binding is expected. Therefore, an increase in the alpha1-acid glycoprotein concentration, and consequently its binding capacity, is expected to result in a decrease in fu. The consistent and steady increase in alpha1-acid glycoprotein during the 72-h study period is in accordance with previous reports. [15,16]The initial decrease in alpha1-acid glycoprotein during the first 6 h may have been influenced by fluid replacement and circulatory effects during surgery. CLintwill be independent of protein-binding, and the total plasma concentration will vary inversely with fu. Significant changes in both total and unbound plasma concentration would be expected at intermediate values of CLint. [12]The postoperative increase in plasma concentrations of alpha1-acid glycoprotein, associated with a marked increase in plasma binding and total plasma concentrations of ropivacaine with minor changes in the unbound plasma concentration, is in accordance with previous results in patients receiving intravenous infusions of bupivacaine before and after surgery. [17]Although total bupivacaine plasma concentrations were doubled postoperatively, the unbound concentrations were similar to those measured preoperatively.

Thus, an excessive increase in total plasma concentration during postoperative epidural infusion with ropivacaine requires cautious interpretation and is not necessarily an indication that the plasma concentration and, consequently, dose rate should be decreased. The safe limits should be based on the relatively more stable unbound (i.e., pharmacologically active) plasma concentrations of ropivacaine. The mean threshold unbound plasma concentration for central nervous system toxicity has been estimated to be 0.6 mg/l after intravenous administration of ropivacaine in humans,****** illustrating the safety of the reported unbound plasma concentrations after epidural infusion of ropivacaine in this study.

The epidural bolus dose in this study, 42.5 mg, was estimated as a pharmacokinetic loading dose for the postoperative epidural infusion. Hence, it was not expected to result in adequate surgical anesthesia, which is why all patients were also to receive general anesthesia. The patients who received the larger infusion rate, 20 mg/h, generally experienced less postoperative pain than those who received 10 mg/h. However, this dose must be considered suboptimal as such, because several patients needed addition of morphine.

The relative analgesic insufficiency was probably not caused by tachyphylaxis, because the level of sensory block only decreased from T3 at 4 h to T3-T6 at 24 h in the 20-mg/h group.

In conclusion, unbound plasma concentrations are stable, in contrast to increasing total concentrations during continuous postoperative epidural infusion of ropivacaine.

*Zaric D: Quantitative methodological studies of motor and sensory blockade in epidural analgesia with special reference to early mobilization and a new local anesthetic--ropivacaine. Doctoral thesis, Uppsala University, Uppsala, Sweden, 1995.

**BBN Software Products Corporation: RS/1 Release 4, Cambridge, BBN Software Products Corporation, 1988.

***SAS Institute Inc: SAS Language: Reference. Version 6. 1st edition. Cary, SAS Institute Inc. 1990.

****SAS Institute Inc: SAS/STAT user's guide. Version 6. 4th edition. Cary, SAS Institute Inc, 1989.

*****Emanuelsson B-M: Personal communication. 1995.

******Knudsen K, Beckman M, Blomberg S, Sjovall J, Edvardsson N: Central nervous and cardiovascular effects during intravenous infusions of ropivacaine, bupivacaine and placebo in healthy volunteers (abstract), XIV Annual ESRA Congress. Prague, International Monitor of Regional Anesthesia, Medicom Europe, 1995, p 15.

Scott DB, Lee A, Fagan D, Bowler GMR, Bloomfield P, Lundh R: Acute toxicity of ropivacaine compared with that of bupivacaine. Anesth Analg 1989; 69:563-9.
Lee A, Fagan D, Lamont M, Tucker GT, Halldin M, Scott DB: Disposition kinetics of ropivacaine in humans. Anesth Analg 1989; 69:736-8.
Brown DL, Carpenter RL, Thompson GE: Comparison of 0.5% ropivacaine and 0.5% bupivacaine for epidural anesthesia in patients undergoing lower-extremity surgery. ANESTHESIOLOGY 1990; 72:633-6.
Brockway MS, Bannister J, McClure JH, McKeown D, Wildsmith JAW: Comparison of extradural ropivacaine and bupivacaine. Br J Anaesth 1991; 66:31-7.
Morrison LMM, Emanuelsson BM, McClure JH, Pollok AJ, McKeown DW, Brockway M, Jozwiak H, Wildsmith JAW: Efficacy and kinetics of extradural ropivacaine: Comparison with bupivacaine. Br J Anaesth 1994; 72:164-9.
Wolff AP, Hasselstrom L, Kerkkamp HE, Gielen MJ: Extradural ropivacaine and bupivacaine in hip surgery. Br J Anaesth 1995; 74:458-60.
Bjork M, Pettersson K-J, Osterlof G: Capillary gas chromatographic method for the simultaneous determination of local anaesthetics in plasma samples. J Chromatogr 1990; 533:229-34.
Arvidsson T, Eklund E: Determination of free concentration of ropivacaine and bupivacaine in blood plasma by ultrafiltration and coupled column liquid chromatography. J Chromatogr 1995; 688:91-8.
Lehmann EL: Nonparametrics: Statistical Methods Based on Ranks. San Francisco, Holden-Day, 1975, pp 5-31,123-32.
Weisberg S: Applied Linear Regression. 2nd edition. New York, John Wiley & Sons, 1985, pp 216-8,285-93.
Rowland M: Protein binding and drug clearance. Clin Pharmacokin 1984; 9(suppl 1):10-7.
Smallwood RH, Mihaly GW, Smallwood RA, Morgan DJ: Effect of a protein binding change on unbound and total plasma concentrations for drugs of intermediate hepatic extraction. J Parmacokinet Biopharm 1988; 16:529-42.
Denson DD, Raj PP, Saldahna F, Finnsson RA, Ritschel WA, Joyce TH III, Tucker JL: Continuous perineural infusion of bupivacaine for prolonged analgesia: Pharmacokinetic considerations. Int J Clin Pharmacol Ther Toxicol 1983; 21:591-7.
Katz JA, Sehlhorst CS, Thompson GA, Denson DD, Coyle D, Bridenbaugh PO: Pharmacokinetics of intravenous and epidural ropivacaine in the rhesus monkey. Biopharm Drug Dispos 1993; 14:579-88.
Aronsen K-F, Ekelund G, Kindmark C-O, Laurell C-B: Sequential changes of plasma proteins after surgical trauma. Scan J Clin Lab Invest 1972; 29(suppl 124):127-36.
Wulf H, Winkler K, Maier CH, Heinzow B: Pharmacokinetics and protein binding of bupivacaine in postoperative epidural analgesia. Acta Anaesthesiol Scand 1988; 32:530-4.
Tucker GT, Mather LE: Properties, absorption, and disposition of local anesthetic agents, Neural Blockade in Clinical Anesthesia and Management of Pain. Edited by Cousins MJ, Bridenbaugh PO. Philadelphia, JB Lippincott, 1988, pp 81-2.