Frequency of hypotension during conventional or asymmetric hyperbaric spinal block

Regional Anesthesia and Pain Medicine 24(3): 214-219, 1999

Frequency of Hypotension During Conventional or Asymmetric Hyperbaric Spinal Block Andrea Casati, M.D., Guido Fanelli, M.D., Giorgio Aldegheri, M.D., Eleonora Colnaghi, M.D., Elisabetta Casaletti, M.D., Valeria Cedrati, M.D., and Giorgio Torri, M.D.

Background and Objectives. The purpose of this randomized, double-blind study was to evaluate if use of an asymmetric spinal block affects the incidence of hypotension during spinal anesthesia. Methods. With Ethical Committee approval and patient consent, 120 patients undergoing lower limb surgery were placed in the lateral position with the side to be operated on dependent, and received 8 mg 0.5% hyperbaric bupivacaine through a 25-gauge Whitacre spinal needle. Patients were randomized to one of two groups: (a) local anesthetic was injected with barbotage through a cranially directed needle orifice, then patients were immediately turned to supine (conventional, n = 60); (b) local anesthetic was injected without barbotage with the needle orifice turned toward the dependent side, then the lateral position was maintained for 15 minutes (unilateral, n = 60). A blind observer recorded noninvasive h e m o d y n a m i c variables, as well as loss of cold and pinprick sensation and motor block on both sides. Results. In the unilateral group, 31 patients (52%) showed a unilateral loss of cold sensation and 48 patients (80%) had no motor block on the n o n d e p e n d e n t side for the duration of the study, whereas all conventional patients had bilateral distribution of spinal block (P < .0001). The onset time and two-segment regression of sensory block on the dependent side were more rapid in the conventional group (18 -+ 7 minutes and 60 -+ 18 minutes) than in the unilateral group (22 + 8 minutes and 67 --- 19 minutes) (P < .05 and P < .05, respectively). The incidence of hypotension (SAP decrease > 3 0 % from baseline) was higher in the conventional (22.4%) than unilateral group (5%) (P < .01). The m a x i m u m percentage changes from baseline values of systolic arterial blood pressure and heart rate were greater in conventional group ( - 2 8 % -+ 16% and - 1 9 % + 10%) than in unilateral group ( - 8 % -+ 16% and - 1 2 % -+ 18%) (P < .0001 and P < .01, respectively). Conclusions. Achieving an asymmetric distribution of spinal block by injecting a small dose of 0.5% hyperbaric bupivacaine through a Whitacre spinal needle into patients placed in the lateral position for 15 rain reduces the incidence of hypotension during spinal anesthesia. Reg Anesth Pain Med 1999: 24: 214-219. K e y w o r d s : regional anesthesia, spinal anesthesia, bupivacaine, monitoring, cardiovascular effects, hypotension.

From the Department of Anesthesiology, University of Milan, Milan, Italy. This study was supported in part by a grant from the IRCCS H San Raffaele, Milan, Italy. Accepted for publication November 4, 1998. Reprint requests: Andrea Casati, M.D., Department of Anesthesiology, IRCCS H San Raffaele, Via Olgettina 60, 20132 Milan, Italy. Copyright © 1999 by the American Society of Regional Anesthesia. 0146- 521X/99/2403-000655.00/0

214

Unilateral Spinal Anesthesia

Hypotension is the most frequent side effect of spinal anesthesia, occurring in more than 30% of patients (1). The attempt to achieve an asymmetric distribution of spinal block toward the surgical side only is often used w h e n surgery involves one lower limb (2), even because reducing the extent of spinal block should theoretically protect against undesired cardiovascular effects of neuraxial blockade (2-4). However, until now, no randomized studies have evaluated if unilateral spinal anesthesia allows to reduce the incidence of hypotension compared with conventional bilateral spinal block. The purpose of this investigation was to evaluate in a randomized, double-blind study if the attempt to achieve an asymmetric distribution of spinal block between the operated and nonoperated sides affects the incidence of hypotension during subarachnoid anesthesia.

Methods To calculate the required study size, we took into account the incidence of hypotension reported during conventional bilateral spinal anesthesia (1). We accepted a one-tailed alpha error of 5% and a beta error of 20% to detect a difference in the incidence of hypotension of 5-10% (5). Based on these figures, the required study size ranged from 48 to 57 patients per group. After the study protocol had been approved by the local Ethical Committee, written informed consent was obtained from 120 ASA I-II patients, aged 25-65 years, receiving spinal anesthesia for elective orthopedic surgery involving one lower limb only (e.g., toes, foot, and ankle surgery). Patients receiving chronic analgesic therapy, as well as patients with marked back arthrosis or scoliosis, diabetes, or peripheral neuropathy were excluded. Standard monitoring was used throughout the study, including noninvasive arterial blood pressure, heart rate, and pulse oximetry. After arrival in the operating theater, a standard crystalloid preload (lactate Ringer's solution, 7 mL/kg) was intravenously infused without premedication and then baseline heart rate and arterial blood pressure were recorded. Patients were placed in the lateral position with the limb to be operated on in the dependent position. The vertebral column was positioned before dural puncture and was maintained as horizontal as possible by tilting the operating table or putting a pillow under the patient's shoulder. Dural puncture was performed at the L3-L4 interspace using a 25gauge Whitacre spinal needle (Becton-Dickinson, NJ) with a midline approach. Using sealed enve-

•

Casati et al.

215

lopes, patients were then randomly allocated to one of two groups. In the first group (unilateral, n = 60), after free flow of cerebrospinal fluid (CSF) had been observed, the needle orifice was turned toward the dependent side and 8 mg (1.6 mL) of 0.5% hyperbaric bupivacaine (Marcaine Spinal Heavy, Astra, Sweden) was injected over 30 seconds without further CSF aspiration (speed of intrathecal injection approximately 0.05 mL/s); the lateral position was then maintained for 15 minutes before patients were turned to the supine position. In the second group (conventional, n = 60), after free CSF flow had been observed, the same dose of 0.5 % hyperbaric bupivacaine was injected over the same interval with three further CSF aspirations (barbotage) through a cranially directed needle orifice; then patients were immediately turned to the supine position. Because the two treatment groups had different positions during the first 15 minutes after dural puncture, measurements were performed by an unblinded observer until patients of the unilateral group were placed in the supine position, and then all measurements were made by a trained independent observer, blinded to the spinal anesthesia technique. Sensory level was evaluated by loss of sensation to both cold (ice test) and pinprick (20-gauge hypodermic needle), whereas motor block was evaluated using a modified Bromage scale (0 = no motor block; 1 = hip blocked; 2 = hip and knee blocked; 3 = hip, knee, and ankle blocked). Hemodynamic variables were measured every 3 minutes during the first 15 minutes after dural puncture, then every 5 minutes until the end of surgery. The evolution of sensory and motor blocks was assessed on both dependent and nondependent sides every 15 minutes until the motor block regressed by one degree on the dependent side. The regression of sensory level by two segments on the dependent side was also recorded (6). Clinically relevant hypotension was defined as a decrease of systolic arterial blood pressure by 30% or more from baseline values, whereas clinically relevant bradycardia was defined as heart rate decrease below 45 beats/min. The occurrence of clinically relevant hypotension was treated with an intravenous bolus of etilephrine (5 mg); bradycardia was treated with 0.5 mg intravenous atropine. During the surgical procedure, a tight pneumatic tourniquet was always inflated 100 m m Hg 1 greater than systolic arterial blood pressure. Oxygen was administered via face mask if required. The adequacy of spinal anesthesia was judged according to the need for supplementary intravenous analgesics and sedation: adequate spinal an-

216

Regional Anesthesia and Pain Medicine Vol. 24 No. 3 May-June 1999

esthesia = neither additional analgesics nor sedation required to complete surgery; inadequate spinal anesthesia = need for additional analgesics (0.1 mg intravenous fentanyl) and sedation (continuous intravenous propofol infusion 2 mg/kg/h) required to complete surgery; or failed spinal anesthesia = general anesthesia required to complete surgery. The occurrence of postoperative headache and pain and/or dysesthesias in the buttocks, thighs, or lower limbs was evaluated 24 hours after dural puncture by the same blind observer and 1 week postoperative during the first routine surgical visit. Statistical analysis was performed using the program Stat-view 3.0 (Abacus Conc., Berkeley, CA). After normal distribution of collected data had been checked, unpaired t-test was used to compare continuous variables in the two treatment groups. Hemodynamic changes from baseline values were analyzed by two-way analysis of variance for repeated measures with Fisher's and Scheffe's tests for multiple comparisons, whereas the chi-square test with the appropriate corrections was used to analyze dichotomous variables. The evolution of sensory and motor blocks was analyzed by Mann-Whitney U test and Wilcoxon test, as appropriate. A value ot P greater than .05 was considered significant. Continuous variables are presented as mean (--- standard deviation (SD)); ordinal data are presented as median (range) or as count (percentage).

Results The two groups of patients were comparable with respect to the demographic variables, duration of surgery, and adequacy of spinal anesthesia (Table 1). Three patients (one in the unilateral group and two in the conventional group) showed failed spinal block (general anesthesia required to complete surgery) and were then excluded from further analysis of both hemodynamic variables and spinal block characteristics. Table 2 shows the maximum sensory level and motor block measured on both the dependent and nondependent sides and the time taken to reach them, as well as the time lasting from local anesthetic injection to both regression of sensory level by two segments and regression of motor block by one degree. Although no differences were observed between the two groups in the m a xi m um sensory level on the dependent side, the unilateral group had a max imum sensory level on the nondependent side markedly lower than that observed in the conventional group (P < .0001), in which all pa-

Table 1. Demographic Variables, Duration of Surgery, and Intraoperative Adequacy of Spinal Anesthesia in Patients Receiving Either an Asymmetric (unilateral, n = 60) or Conventional Bilateral (conventional, n = 60) Spinal Anesthesia

Age (yr) W e i g h t (kg) Height (cm) Sex (Male/Female) ASA physical status Duration of surgery (min) Spinal block a d e q u a c y (Adequate/Inadequate/Failed)

Unilateral (n = 60)

Conventional (n = 60)

48 -+ 14 70 _+ 11 170 _+ 9 26134 I (I-II) 68 -+ 29

47 -+ 16 68 + 10 168 -+ 10 28132 I (I-H) 66 +_ 25

58/1/1

56/2/2

Data are expressed as m e a n s ( - S D ) , w i t h t h e exception ot sex a n d block a d e q u a c y ( n u m b e r ) a n d ASA physical status [ m e d i a n

(range)].

tients showed a symmetric bilateral distribution of spinal block. The maximum sensory level on the dependent side was also reached faster and regressed sooner in the conventional group than in the unilateral group (Table 2). In the unilateral group, 41 patients (68%) showed unilateral loss of pinprick sensation, and 31 patients (52 %) showed a unilateral loss of cold sensation (ice test) for the duration of the study. The maximum motor block on the dependent side was similar in both groups (Table 2). On the contrary, 48 patients in the unilateral group (80%) had no motor block on the nondependent side for the duration of the study, whereas conventional patients had all equal bilateral distribution of motor block (Table 2). Clinically relevant hypotension requiring intravenous etilephrine was more frequent in the conventional group (13 patients, 22.4%) than in the unilateral group (3 patients, 5%) (P < .01). No differences were observed in the incidence of bradycardia [four patients (6.7%) in the unilateral group and five patients (8.6 %) in the conventional group]. Interestingly, all unilateral patients developing clinical hypotension had an equal bilateral distribution of spinal block. The m ean changes from baseline values of arterial blood pressure and heart rate are shown in Fig. 1. When considering the m axi m um percentage change from baseline values of the hemodynamic variables, patients of the conventional group showed a more marked decrease in both systolic arterial blood pressure ( - 2 7 % _+ 16% in conventional group and - 7 % _+ 16% in unilateral group; P < .0001) and heart rate ( - 1 9 % _+ 20% in conventional group and - 1 3 % _+ 18% in unilateral group; P < .01), while no differences in maximum percentage decrease of diastolic arterial blood pressure were observed between the two

Unilateral Spinal Anesthesia

•

Casati et al.

217

Table 2. Maximal Sensory and Motor Block on both Dependent and Nondependent Sides, Time Taken to Reach Them, and Time Lasting from Local Anesthetic Injection to Regression of Sensory Level by Two Segments and to Regression of Motor Block by One Degree in Patients Receiving Either an Asymmetric (unilateral, n = 59) or Conventional Bilateral (conventional, n = 58) Spinal Anesthesia

M a x i m u m s e n s o r y level o n t h e d e p e n d e n t side M a x i m u m s e n s o r y level o n t h e n o n d e p e n d e n t side Time to reach m a x i m u m s e n s o r y level o n t h e d e p e n d e n t side (min) Time to reach m a x i m u m s e n s o r y level o n the n o n d e p e n d e n t side (min) Regression of s e n s o r y level by two s e g m e n t s o n the d e p e n d e n t side (min) M a x i m u m m o t o r block o n t h e d e p e n d e n t side M a x i m u m m o t o r block o n t h e n o n d e p e n d e n t side Time to reach m a x i m u m m o t o r block o n t h e d e p e n d e n t side (min) Time to reach m a x i m u m m o t o r block o n the n o n d e p e n d e n t side (min) Regression of m o t o r block by o n e degree o n the d e p e n d e n t side (min)

Unilateral (n = 59)

Conventional (n = 58)

T9 (T12-T2) L4 (no level-T2) *rl

T9 (T12-T3) T9 (T12-T3)

22 _+ 8 '

18 -+ 7

25 _+ 9 §#

18 + 7

67 -- 1 9 ' 3

60 -- 17 3

(2-3)

(3-3)

n o m o t o r block (no m o t o r block-3} Ir

3 (l-3)

10--4 27 -+ 18 I1~ 120 __ 29¢

9-+5 10-+5 99 -+ 28

C o n t i n u o u s variables are p r e s e n t e d as m e a n s (_+SD); ordinal data are p r e s e n t e d as m e d i a n (range). * P < .0005 c o m p a r e d w i t h t h e d e p e n d e n t side. *P < .05, ¢P < .01, §P < .00l, a n d lip < .0001 c o m p a r e d w i t h t h e c o n v e n t i o n a l group. #Only for t h e 29 patients s h o w i n g loss of cold s e n s a t i o n o n b o t h d e p e n d e n t a n d n o n d e p e n d e n t sides. ~Only for t h e 12 patients s h o w i n g m o t o r block > 1 on t h e n o n d e p e n d e n t side.

groups ( - 1 0 - 13% in conventional group and - 5 + 20% in unilateral group; P < .07). Postoperatively, two cases of urinary retention were reported (one in each group; P = not significant), whereas no cases of postdural puncture headache or other neurologic complications were reported.

Discussion This randomized, double-blind study demonstrates that searching for an asymmetric distribution of spinal block toward the operated side allows one to reduce the hemodynamic effects after spinal anesthesia wh en using a relatively small dose of 0.5 % hyperbaric bupivacaine. The incidence of hypotension observed during conventional bilateral spinal anesthesia is lower than that reported by Carpenter et al. (1) in a previous investigation. This might be explained by both the different arbitrary definition of clinical hypotension used in the present study and the relatively small close of local anesthetic used, because it has been clearly demonstrated that reducing the total amount of local anesthetic limits the cardiovascular response to neuraxial blocks (7). The hemodynamic effects of spinal anesthesia are mainly a result of both the resistance and capaci-

tance changes induced by the block of preganglionic sympathetic fibers (8,9). In the present investigation, no direct measurements of changes in sympathetic activity, such as thermography or skin conductance response, were used. However, the presence of cold sensation strictly correlates with the function of C fibers (10), and only 48% of patients in the unilateral group showed loss of cold sensation on both sides for the duration of the study compared with the 100% symmetric bilateral loss of cold sensation observed in the conventional group. This suggests that the sympathetic block was less profound in the unilateral group than in the conventional group. A further crucial point that might explain these results is the delay in the time taken to achieve the m axi m um sensory level in the unilateral group. The onset time to block of the maxim u m level is related to the time needed for subarachnoid distribution of the local anesthetic: In the conventional group, the local anesthetic injection was made with needle opening in a cephalad direction with three CSF aspiration maneuvers (barbotage), whereas in the unilateral group, the local anesthetic was injected toward the dependent side, and it is know n that the Whitacre spinal needle provides a nonturbulent uniform stream w h e n using relatively low speeds of intrathecal injection (11,12). This probably reduced the mixing of local

218

Regional Anesthesia and Pain Medicine Vol. 24 No. 3 May-June 1999 0

*'~ "r-

-10

E

-20

15

30

45

60

75

90

105

120

-30 '~ -40 II.

-50

,

-60

0

E

*

w Unilateral •Conventional

15

30

45

60

75

90

105

120

-10

-15 -20 -25

*

-30

0 -10 ,4

U Unilateral • Conventional

15

30

45

60

75

90

105

120

-20

-30 -40 '~ -50 -60

[] Ulllldl.~Idl • '~,UII'I~:~IILIUII~I

Fig. 1. M e a n ( + S D ) c h a n g e s of n o n i n v a s i v e arterial b l o o d p r e s s u r e a n d h e a r t rate d u r i n g t h e first 120 m i n u t e s after e i t h e r a n a s y m m e t r i c (unilateral, n = 59) or c o n v e n t i o n a l bilateral ( c o n v e n t i o n a l , n = 58) spinal a n e s t h e sia. SAP, systolic arterial pressure; DAP, diastolic arterial p r e s s u r e ; HR, h e a r t rate. *P < .05 c o m p a r e d w i t h t h e u n i l a t e r a l group. Data are p r e s e n t e d as m e a n _+ SD.

anesthetic molecules with the CSF, leading to a significant delay in the cephalic spread of spinal block in the unilateral group compared with the conventional group. The slower onset time coupled with the reduced extent of sympathetic block probably allowed unilateral group patients to activate more effective homeostatic vascular mechanisms than those patients receiving conventional bilateral spinal anesthesia (13). In fact, epidural anesthesia is known to produce neuraxial block with a slower onset time and less hemodynamic effects than subarachnoid anesthesia (14), whereas incremental injections of bupivacaine through a subarachnoid catheter produce an equally effective block with less cardiovascular changes than a single injection of the same dose of local anesthetic solution (15,16). Conventional group patients also showed a maximum decrease in heart rate greater than that observed in those patients receiving a unilateral spinal block. Neuraxial blocks are known to produce a vagotonic baroceptor response, which can be re-

stored by increasing the cardiac filling pressures (8,17). The less marked negative effect on heart rate observed in the unilateral group patients could be related to decreased venous pooling compared with the conventional group patients, thus producing less vagotonic shift of the baroceptor response (1719). In spite of the lower effect on maximum percentage decrease of heart rate, our results failed to demonstrate a significant reduction in the incidence of bradycardia in the unilateral group compared with the conventional group. However, it should be considered that when estimating the study size, we considered the incidence of hypotension during conventional spinal anesthesia. Since the incidence of bradycardia is lower than that of hypotension (1), we cannot exclude a type 2 error in the evaluation of the incidence of bradycardia. In clinical practice, both intravenous fluid preload and prophylactic vasopressors are used for decreasing hypotension during spinal anesthesia (20). In the present study, all patients received a moderate-volume intravenous infusion of crystalloid solution before placing spinal anesthesia without prophylactic vasopressors. The association of prophylactic vasopressors with crystalloid preload could probably reduce the incidence of hypotension in both conventional and asymmetric spinal block, further minimizing the risk for undesired cardiovascular side effects. In agreement with previous studies (21,22), our results also demonstrated that the regression of sensory level by two segments after intrathecal injection of 8 mg 0.5 % hyperbaric bupivacaine required 60-67 minutes, suggesting the usefulness of low doses of hyperbaric bupivacaine for surgical procedures of short duration. Interestingly, the larger spread of spinal block observed in the conventional group was also associated with a faster regression of both sensory and motor blocks compared with patients receiving unilateral spinal anesthesia. Similar findings have been reported by Kooger-Infante et al. (23), who demonstrated that with identical doses of local anesthetic, the duration of spinal anesthesia was longer in patients with lower anesthetic levels, probably because of the reduced surface available for anesthetic molecule absorption and elimination from the subarachnoid space. The hemodynamic advantages of an asymmetric distribution of spinal block may be more relevant in elderly, chronically ill patients, and further controlled studies should be performed to evaluate the usefulness of a one-sided spinal anesthesia in these higher-risk patients. Overall, we conclude that searching for an asymmetric distribution of spinal block by injecting 8 mg 0.5% hyperbaric bupiva-

Unilateral Spinal Anesthesia

caine t h r o u g h a Whitacre spinal needle into patients placed in the lateral position for 15 minutes reduces the incidence of h y p o t e n s i o n during spinal anesthesia.

Acknowledgment The authors w o u l d like to t h a n k the staff of anesthesia nurses (University D e p a r t m e n t of Anesthesiology, IRCCS San Raffaele Hospital), w i t h o u t w h o s e help and cooperation this study w o u l d not have been possible.

References I. Carpenter RL, Caplan RA, Brown DL, Stephenson C, Wu R. Incidence and risk factors for side effects of spinal anesthesia. Anesthesiology 1992: 76: 906916. 2. Tanasichuk MA, Shultz EA, Matthews JH, Van Bergen F. Spinal hemianalgesia: an evaluation of a method, its applicability and influence on the incidence of hypotension. Anesthesiology 196h 22: 74-85. 3. Casati A, Fanelli G, Berti M, Beccaria P, Agostoni M, Aldegheri G, Torri G. Cardiac performance during unilateral lumbar spinal block after crystalloid preload. Can J Anaesth 1997: 44: 623-628. 4. Casati A, Fanelli G, Beccaria P, Leoni A, Berti M, Senatore R, Torri G. Block distribution and cardiovascular effects of unilateral spinal anaesthesia by 0.5 % hyperbaric bupivacaine. A clinical comparison with bilateral spinal block. Minerva Anestesiol 1998: 64: 307-312. 5. Browner WAS, Black D, Newman B, Hulley SB. Estimating sample size and power. In: Hulley SB, Cummings SR, eds. Designing Clinical Research--An Epidemiologic Approach. Baltimore, Williams & Wilkins, 1988:139-150. 6. Stienstra R, Gielen M, Van Poorten F, Kroon JW. Spinal anesthesia with plain bupivacaine 0.5%: Regression of sensory and motor blockade with different temperatures of the anesthetic solution. Anesth Analg I989: 69: 593-597. 7. Klimscha W, Weinstabl C, Ilias W, Mayer N, Kashanupour A, Schneider B, Hammerle A. Continuous spinal anesthesia with a microatheter and low-dose bupivacaine decreases the hemodynamic effects of centroneuraxis blocks in elderly patients. Anesth Analg 1993: 77: 275-280. 8. Hogan Q. Cardiovascular response to sympathetic block by regional anesthesia. Reg Anesth 1996:2 l: 26-34. 9. McCrae AF, Wildsmith JAW. Prevention and treatment of hypotension during central neural block. Br J Anaesth 1993: 70: 672-680.

•

Casati et al.

219

10. Liu S, Kopacz DJ, Carpenter RL. Quantitative assessment of differential sensory nerve block after lidocaine anesthesia. Anesthesiology 1995: 82: 60-63. 11. Serpell MG, Gray WM. Flow dynamics through spinal needles. Anaesthesia i997: 52: 229-236. 12. Holman SJ, Robinson RA, Beardsley D, Stewart SFC, Klein L, Sevens RA. Hyperbaric dye solution characteristics after pencil-point needle injection in a spinal cord model. Anesthesiology 1997: 86: 966973. 13. Bridenbaugh PO, Greene NM. Spinal (subarachnoid) neural blockade. In: Cousins MJ, Bridenbaugh PO, eds. Neural Blockade in Clinical Anesthesia and Management of Pain. Philadelphia, JB Lippincott, 1988: 213-251. 14. Casati A, Fanelli G, Beccaria P, Aldegheri G, Berti M, Agostoni M, Torri G. Hemodynamic monitoring during alkalinized lidocaine epidural block for lower limb surgery: A comparison with subarachnoid anesthesia. Eur J Anaesth 1997: 14: 300-306. 15. Casati A, Zangrillo A, Fanelli G, Torri G. Comparison between haemodynamic changes after single-dose and incremental subarachnoid anesthesia. Reg Anesth 1996: 21: 298-303. 16. Favarel-Garrigues JF, Sztark F, Petijean ME, Thicoipe M, Lassie P, Dabadie P. Hemodynamic effects of spinal anesthesia in the elderly: Single dose versus titration through a catheter. Anesth Analg 1996: 82: 312-316. 17. Baron JF, Decaux-Jacolot A, Edouard A, Bordeaux A, Samii K. Influence of venous return on baroreflex control of heart rate during lumbar epidural anesthesia in humans. Anesthesiology 1986: 64: i88-193. 18. Oberg B, Thoren P. Increased activity in left ventricular receptors during hemorrhage or occlusion of caval veins in the cat: A possible cause of vasovagal reaction. Acta Physiol Scand 1972: 85: 164-173. 19. Jacobsen J, Sofelt S, Brocks V, Fernandes A, Warberg J, Secher NH. Reduced left ventricular diameters at onset of bradycardia during epidural anaesthesia. Acta Anaesthesio] Scand 1992: 36: 831-836. 20. Arndt JO, Bomer W, Krauth J, Marqardt B. Incidence and time course of cardiovascular side effects during spinal anesthesia after prophylactic administration of intravenous fluids or vasoconstrictors. Anesth Analg 1998: 87: 347-354. 2i. Liu SS, Ware PD, Allen HW, Neal JM, Pollock JE. Dose-response characteristics of spinal bupivacaine in volunteers. Clinical implications for ambulatory anesthesia. Anesthesiology 1996: 85: 729-736. 22. Ben-David B, Levin H, Solomon E, Admoni H, Vaida S. Spinal bupivacaine in ambulatory surgery: The effect of saline dilution. Anesth Analg i996: 83: 716-720.