Acetated Ringer’s solution is one of the most widely used fluids for IV hydration and replacement of blood volume loss during surgery. The rate of infusion is traditionally governed by simplistic means, such as arterial blood pressure and rules-of-thumb. In contrast, the administration of anesthetic drugs is based on a knowledge of their kinetic behavior in the human body. Pharmacokinetic analysis is a tool for dealing with such issues as optimal dosing and dosing intervals, expected residual effects, and quantification of interactions between body physiology and the metabolism of the drug. To allow such considerations in fluid therapy, we have developed a mathematical approach called volume kinetics, which applies pharmacokinetic principles to the distribution and elimination of infusion fluids.
Volume kinetics has been used for the study of acetated and lactated Ringer’s solution in the laboratory setting (1–4) and during the induction of anesthesia (5) but not during surgery. However, the kinetics of crystalloid fluid during surgical procedures is of particular interest because the disposition of infused fluid is likely to be changed by anesthetic drugs and the physiological response to trauma. There is evidence of an increased intravascular retention of Ringer’s solution (6), but otherwise few details are known about the kinetics of the fluid under such circumstances.
In the present study, the kinetics of the volume component of acetated Ringer’s solution is evaluated during laparoscopic cholecystectomy. This operation was chosen because the procedure is well standardized, associated with moderately severe surgical stress, and usually with minimal blood loss despite an operating time of 1 h or more (7,8). Our hypothesis was that kinetic data could be collected during laparoscopy, which allowed us to construct a nomogram to predict how fast this fluid needs to be infused to yield and maintain a predetermined plasma dilution (3,9).
Twelve ASA I–II women between 26 and 69 (mean, 45 yr) yr old, with a body weight of 55–96 kg (mean, 75 kg), undergoing elective laparoscopic cholecystectomy under general anesthesia were studied. The protocol was approved by our ethics committee, and each patient gave her informed consent to participate.
Patients either fasted overnight or had a light morning breakfast without solids depending on whether the surgery was scheduled in the morning or in the afternoon. No premedication was given. With the patient in the supine position, a cannula was placed in the antecubital vein of each arm for sampling blood and infusing fluid.
After breathing oxygen, anesthesia was induced with glycopyrrolate 0.2 mg, propofol 1.5 mg/kg, rocuronium 0.8 mg/kg, and remifentanil 0.8 μg · kg−1 · min−1, followed by endotracheal intubation. Anesthesia was continued with remifentanil 0.3–0.5 μg · kg−1 · min−1) and 1% sevoflurane in oxygen/air. Pneumoperitoneum with CO2 was maintained at a pressure of 12 mm Hg with the patient in the reverse Trendelenburg position and the hips and knees slightly flexed into stirrups to prevent excessive venous pooling to the legs. No IV fluids were given before surgery. Timed with the surgical incision, an IV infusion of 20 mL/kg of acetated Ringer’s solution was given at a constant rate over 60 min (0.33 mL · kg−1 · min−1) via an infusion pump. Acetated Ringer’s solution has the following ionic content (mmol/L): Na 130, K 4, Ca 2, Mg 1, Cl 110, and acetate 30. The osmolality is approximately 270 mOsm/kg.
Acetated and lactated Ringer’s solutions show quite similar distribution and elimination in volunteers (4), but only the acetated solution is marketed in Scandinavia because of better results during lactic acidosis in hypokinetic animals (10).
Monitoring included pulse oximetry and electrocardiography. Noninvasive arterial blood pressure was measured in the arm not used for fluid infusion by an automatic device. The excreted urine was collected and measured every 15 min via an indwelling catheter, which had been inserted into the bladder after the induction of anesthesia but before the infusion started.
Venous blood samples, 3 mL each, were collected every 5 min during the first 90 min and every 10 min during the next 150 min. Furthermore, a 3-mL sample for the measurement of blood glucose was taken immediately before the study, every 15 min for 90 min, and every 30 min during the next 150 min. To preclude any admixture of saline or blood from the previous sampling, a discard volume of 3 mL was drawn before each blood collection. The discard volume was then returned and the cannula flushed with 2 mL of saline to prevent clotting.
The hemoglobin (Hb) concentration in whole blood, the red blood cell count, and mean corpuscular volume were measured by a Sysmex SF 3000 (Tillquist Instrument AB, Sundbyberg, Sweden) with a coefficient of variation of 2%. Plasma glucose was analyzed on a Vitros device (Ortho Clinical Diagnostics, Sollentuna, Sweden). Samples were drawn in duplicate and the mean values used in the calculations.
Volume kinetics is a pharmacokinetic method for use with infusion fluids (1–5), which was adapted here to account for the fact that awakening from anesthesia affected the plasma dilution to the extent that an analysis of the postoperative phase could not be made. The volume kinetic model fitted to the Hb and urine data (Fig. 1) implied that fluid infused at a rate ki expands a central body fluid space v, which strives to return to its baseline volume V (to regain body fluid steady-state) by three mechanisms: first, translocation to a much larger peripheral body fluid space at a rate proportional by a constant kt to the dilution of V, i.e., the rate of translocation at any time was kt × (v-V)/V; second, elimination of fluid by urinary excretion at a rate proportional by a constant kr to the dilution of V, i.e., the rate of elimination at any time was kr × (v-V)/V; and third, a baseline loss (kb) fixed at 0.5 mL/min to account for evaporation and basal diuresis, which occur at a rate of approximately 700 mL/d in adults.
From a kinetic point of view, fluid infused into the central body fluid space readily equilibrates with antecubital venous blood, whereas the peripheral fluid space becomes expanded after a delay. The volume change in the central body fluid space (v) was expressed as:
Input data obtained at baseline (time 0) and at any later time (t) comprised Hb, red blood cells (RBC), and hematocrit. They were used to calculate the plasma dilution, which also represented the dilution of V:
A correction for variations in cell size was made using the mean corpuscular volume. A further correction of the plasma dilution for iatrogenic dilution resulting from the blood sampling was made based on the baseline blood volume estimated from a regression equation using the height and weight of the patients as input data (4). However, iatrogenic dilution has little influence on the estimates of the kinetic variables (3).
The unknown variables V and kt in the model (primary variables) were estimated by fitting the mathematical solution to the differential equation (Equation 1) to the data on plasma dilution (Equation 2). For this purpose, a nonlinear least-squares regression routine, based on a modified Gauss-Newton method, was repeated until no variable changed by more than 0.001 (0.1%) in each iteration.
The variable kr was calculated from the urinary excretion (Equation 3). The integral of the urinary excretion rate from 0 to 90 min is represented by the total urine volume collected during the studied period of time T. Therefore, the variable kr could be calculated as the urinary excretion divided by the area under the curve for the entire dilution-time profile. After considering that approximately half of the basal fluid losses consisted of excretion of urine, this relationship at any time t was expressed as:
Secondary variables were calculated based on the fact that kt and kr are clearance constants, and, therefore, the exponent for the elimination function is −kr/V. Because fluid also leaves V by distribution to peripheral tissues, the slope resulting from both distribution and elimination is the sum of both clearance constants divided by V. The inverted expressions corresponding to these situations are the elimination half-life and the context-sensitive half-life, the latter representing the time required for the plasma dilution to be reduced by 50% after infusion with regard to both distribution and elimination (11). All secondary variables ignore the minimal contribution of basal fluid losses to the dilution-time curve.
where ln is the natural logarithm. The present adaptation of the volume kinetic analysis for a short postinfusion observation period with a known urinary excretion has recently been used in another study (12).
Because of skewed distribution, data are reported as the median (interquartile range). The Wilcoxon ranked sum test was used to compare the groups. Selected variables were compared by simple linear regression analysis. P < 0.05 was considered significant.
The induction of anesthesia reduced the systolic and diastolic arterial blood pressures by 30% and the heart rate by approximately 12% (each P < 0.002). These variables had been restored when the patients recovered from anesthesia at 80 min (75–103 min) (Fig. 2).
The induction of anesthesia diluted the plasma by 4.2% (1.0%–5.3%), although no fluid was infused until surgery started 45 min (35–60 min) later. During surgery, the plasma dilution transiently increased by an additional 17.5% (10.1%–19.3%) in response to the infusion of 1500 mL (1375–1645 mL) of fluid (Fig. 3). The blood loss, as obtained by suction, was only a few milliliters. The patients excreted 19.9% (8.1%–23.6%) of the infused fluid during the 4-h experiment. For individual patients, the urine flow rate during anesthesia decreased with the mean arterial blood pressure (r = 0.84; P < 0.001) and rarely exceeded 1 mL/min (Table 1; Fig. 3). There was a marginal increase in plasma glucose (Fig. 3).
Fitting the kinetic model to the individual dilution-time profiles was limited to the period of anesthesia only (Fig. 4) because the termination of anesthesia clearly affected the curve, resulting in abrupt concentration or dilution (Fig. 4). The size of V averaged 3.2 L, the clearance constant for distribution of fluid (kt) was 115 mL/min, and the clearance constant for urinary excretion (kr) was 6.8 mL/min (Table 2).
Computer simulations based on the kinetic data in Table 2 were used to predict the plasma dilution expected to result from infusing Ringer’s solution at various rates and at various periods of time (Fig. 5). Simulation based on a single kinetic analysis of the pooled data from all patients extending throughout the 240-min experiment yielded a dilution-time curve that was practically identical to a simulated curve based on the kinetic variables shown in Table 2 (Fig. 3).
The results illustrate how acetated Ringer’s solution is distributed in patients during laparoscopic cholecystectomy performed under general anesthesia. Slow urinary excretion is the most apparent difference from laboratory experiments in which the same fluid was infused. Healthy female volunteers excreted between 50% and 73% of the fluid within 30 minutes after ending an infusion (2), and males eliminated approximately 50% of infused lactated or acetated Ringer’s solution within three to four hours (3,4). In the present study, the excretion rate was much slower and amounted only to 5.3% during surgery and to 19.9% for the entire four-hour study. There are probably several reasons for the limited diuretic response to intravascular fluid administration during laparoscopy, including anesthesia-induced hypotension and surgical stress, as well as the induced pneumoperitoneum (13). However, the excretion is still twice as much as for isotonic glucose 2.5% with electrolytes infused during laparoscopy (14). The even more limited diuretic response to glucose 2.5% can probably be explained by the osmotic translocation of fluid into the intracellular space during the metabolism of the exogenous glucose (15).
A small diuretic response should increase the plasma dilution resulting from crystalloid administration. Comparisons with female volunteers (9) show that the rates and infusion times of Ringer’s solution required to yield any predetermined plasma dilution are almost identical to the present ones. However, the women undergoing cholecystectomy had a body weight 20% larger than that of the volunteers, which implies that they, in fact, had a more pronounced plasma dilution response.
The fate of the Ringer’s solution was further evaluated using pharmacokinetics adapted for infusion fluids. The volume kinetics of acetated and lactated Ringer’s solution have been studied in the laboratory setting where it was shown that the fluids expand a central fluid space (here denoted as V) of approximately 3.5 L in size and that distribution to a peripheral fluid space of approximately 7 L requires 25–30 min for completion (1–4). These are functional body fluid spaces that may or may not correspond to the physiological terms plasma volume and interstitial fluid space. Sometimes the peripheral fluid space becomes difficult to identify because of prompt elimination of infused fluid from V, which mainly occurs by voiding. Such exceptions are unlikely during surgery because of the poor diuretic response to IV fluid.
The kinetic analyses were limited by the fact that recovering from anesthesia often distorts the dilution-time curve (14). To overcome this problem, only the period of anesthesia was subjected to kinetic analysis, and the measured urinary excretion was used as an input variable to yield the clearance constant for elimination kr (16). Approximately 20 minutes of postinfusion data were analyzed in addition to the 60-minute infusion period, which is too short to obtain useful estimates of the size of the peripheral space. However, partial derivatives from the curve-fitting of volunteer experiments reveal that the size of the peripheral fluid space contributes little to the dilution-time profile before 20–25 minutes of the postinfusion. By setting its size to high values, a robust kinetic analysis of the infusion period and immediate postinfusion period could still be obtained, yielding an approach having the advantage of allowing direct comparison between distribution and elimination from the estimates of kt and kr (12). The drawback of this approach is a blurring of comparisons between the estimates of V and kt obtained in this study with those in which a postinfusion period of normal length, usually three hours, was also used to estimate the peripheral fluid space.
The kinetic variables obtained during the curve-fitting procedure show that V expanded by the infused fluid had a volume close to the expected plasma volume (42 mL/kg). The clearance constant for distribution, kt, was high compared to healthy nonpregnant women to whom the same adapted kinetic model was applied (12). As could be expected from the limited diuretic response, the value of kr was low and ended up between 5% and 10% of that obtained in the volunteers (1–4). The secondary (derived) variables illustrate the importance of distribution to peripheral parts of the body for the short persistence of Ringer’s solution in the blood. The half-life for the removal of fluid from V (context-sensitive half-life) was 16 minutes, whereas the half-life of the fluid in the whole body was 17 times longer. Such differences promote peripheral edema, which is therefore a more significant issue during surgery than in laboratory experiments. However, the difference will be reduced with time because the larger peripheral fluid volume gradually becomes diluted, a fact not accounted for in this simplified version of the kinetic model.
The nomogram created from the kinetic variables is based on the experience that similar kinetic variables are usually derived regardless of infusion rate and the infused volume of isotonic or nearly isotonic fluid (2,17,18). The results of other studies have been illustrated in the same way (3,9,14). The rationale behind the nomogram is that most favorable effects of Ringer’s solution are attributable to its intravascular persistence. An additional issue during surgery is the high ratio of kt:kr, which promotes long-standing peripheral edema. Without the use of diuretics, such edema can be prevented only by maintaining a stable plasma dilution over time at the lowest acceptable level.
A possible strategy to achieve these goals would be to rapidly hydrate the patient to a desired plasma dilution and then decrease the infusion rate to maintain this dilution. Infusing the fluid at the same rate throughout surgery would result in progressively increasing dilution, which does not benefit the patient. The degree of plasma dilution that should be recommended reflects clinical judgments relating to the need for a safety margin to ensure stable hemodynamics with respect to the anesthesia and the risk of sudden blood loss. This issue will be the target for further studies, but we speculate that a plasma dilution between 5% and 10% is sufficient. Providing more fluid would result in the formation of much more edema than required.
An additional issue in the present study is the slight plasma dilution that was caused by the induction of anesthesia alone. The fluid volumes injected just to complete the actual induction amounted to approximately 30 mL, whereas the 4.2% dilution corresponded to a volume expansion that is the product of V (3.24 L) and 4.2%, i.e., 136 mL. The volume expansion that cannot be accounted for by IV fluid is small but still of theoretical interest. A similar expansion occurs when epidural anesthesia is induced without IV fluid (19) and might be due to the vasodilation associated with anesthesia, which allows peripheral fluid to enter the plasma volume. The pronounced plasma dilution that develops when crystalloid is infused during the induction of spinal or general anesthesia probably has the same background. We have interpreted this effect to indicate that infused fluid circulates in a V that is smaller than the plasma volume (probably the central circulation) (5). The present and other results (14) suggest that it represents an upward shift of the baseline level, similar to that occurring during hemorrhage. Further support of this interpretation is given by the urinary flow rate that, in the volume kinetic model, increases with the plasma dilution. A comparison between the two upper graphs of Figure 3 illustrates that such a relationship existed during the surgery, whereas after awakening, the diuretic response became stronger. The increased baseline seems to return to the preanesthetic level when the anesthesia-associated vasodilation is resolved. Interestingly, the excessive plasma dilution is contra-intuitive—one could expect that vasodilation would reduce the dilution after IV fluid, but the opposite occurs. From a kinetic point of view, the fluid becomes distributed in a smaller volume than usual, yet one might assume that it would be distributed in a larger one (19).
In conclusion, the plasma dilution resulting from intravascular fluid administration with acetated Ringer’s solution in women undergoing laparoscopic cholecystectomy is slightly stronger than expected owing to a limited diuretic response. The distribution of the infused fluid to peripheral tissues occurs relatively fast, which promotes edema. Future studies are required to clarify whether these results are applicable to surgical operations performed without pneumoperitoneum.
1. Svensén C, Hahn RG. Volume kinetics of Ringer solution, dextran 70 and hypertonic saline in male volunteers. Anesthesiology 1997;87:204–12.
2. Hahn RG, Drobin D, Ståhle L. Volume kinetics of Ringer’s solution in female volunteers. Br J Anaesth 1997;78:144–8.
3. Drobin D, Hahn RG. Volume kinetics of Ringer’s solution in hypovolemic volunteers. Anesthesiology 1999;90:81–91.
4. Drobin D, Hahn RG. Kinetics of isotonic and hypertonic plasma volume expanders. Anesthesiology 2002;96:1371–80.
5. Ewaldsson C-A, Hahn RG. Volume kinetics during induction of spinal and general anaesthesia. Br J Anaesth 2001;87:406–14.
6. Hahn RG. Volume effect of Ringer solution in the blood during general anaesthesia. Eur J Anaesth 1998;15:427–32.
7. Schauer PR, Sirinek KR. The laparoscopic approach reduces the endocrine response to elective cholecystectomy. Am Surg 1995;61:106–11.
8. Thorell A, Nygren J, Essén P, et al. The metabolic response to cholecystectomy: insulin resistance after open compared with laparoscopic operation. Eur J Surg 1996;162:187–91.
9. Hahn RG, Svensén C. Plasma dilution and the rate of infusion of Ringer’s solution. Br J Anaesth 1997;79:64–7.
10. Kveim M, Nesbakken R. Utilization of exogenous acetate during canine hemorrhagic shock. Scand J Clin Lab Invest 1979;39:653–8.
11. Bailey JM. Context-sensitive half-times: what are they and how valuable are they in anaesthesiology? Clin Pharmcokinet 2002;41:793–9.
12. Drobin RG, Hahn RG. Distribution and elimination of crystalloid fluid in pre-eclampsia. Clin Sci 2004;106:307–13.
13. Odeberg S, Ljungqvist O, Svenberg T, Sollevi A. Lack of neurohumoral response to pneumoperitoneum for laparoscopic cholecystectomy. Surg Endosc 1998;12:1217–23.
14. Sjöstrand F, Hahn RG. Volume kinetics of 2.5% glucose solution during laparoscopic cholecystectomy. Br J Anaesth 2004;92:485–92.
15. Hahn RG, Edsberg L, Sjöstrand F. Volume kinetic analysis of fluid shifts accompanying intravenous infusions of glucose solution (review). Cell Biochem Biophys 2003;39:211–22.
16. Hahn RG, Drobin D. Urinary excretion as an input variable in volume kinetic analysis of Ringer’s solution. Br J Anaesth 1998;80:183–8.
17. Svensén C, Drobin D, Olsson J, Hahn RG. Stability of the interstitial matrix after crystalloid fluid loading studied by volume kinetic analysis. Br J Anaesth 1999;82:496–502.
18. Brauer L, Svensén C, Hahn RG, et al. Influence of rate and volume of infusion on the kinetics of 0.9% saline and 7.5% saline/6% dextran 70 in sheep. Anesth Analg 2002;95:1547–56.
© 2004 International Anesthesia Research Society
19. Holte K, Foss NB, Svensén C, et al. Epidural anesthesia, hypotension and changes in intravascular volume. Anesthesiology 2004;100:281–6.