| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the *Department of Anesthesia, Edith Wolfson Medical Center, Holon and Sackler School of Medicine, Tel Aviv University, Israel; Outcomes Research Institute and Department of Anesthesiology and Perioperative Medicine, University of Louisville, Louisville, Kentucky; Department of Outcomes Research, Cleveland Clinic, Cleveland, Ohio; Departments of Obstetrics and Gynecology and ¶Epidemiology, Edith Wolfson Medical Center, Holon and Sackler School of Medicine, Tel Aviv University, Israel.
Address correspondence and reprint requests to Tiberiu Ezri, MD, Department of Anesthesia, Edith Wolfson Medical Center, Holon 58100, Israel. Address e-mail to tezri{at}netvision.net.il. On the World Wide Web: www.or.org.
Abstract
BACKGROUND: Accidental cannulation of an epidural vein is a common complication associated with epidural anesthesia or analgesia. On the basis of a pilot study and previous reports, we tested the hypothesis that predistention of the epidural space with saline before epidural catheterization would ease catheter insertion and decrease the incidence of this complication.
METHODS: Two-hundred-three laboring women were randomly assigned to receive an epidural with loss of resistance technique with 2 mL (nondistention) or 5 mL saline (distention). In the distention group, the syringe plunger was held closed before epidural catheter insertion. Then in both groups, a test dose of 3 mL of 1.5% lidocaine was injected through the epidural catheter.
RESULTS: There were fewer accidental intravascular catheter placements (2% vs 16%, P = 0.0001) in the distention group, and 91% of patients in this group did not have any unblocked segments versus 67% in the nondistension group (P = 0.0001). The difference in onset time of analgesia was small (5.0 ± 2 min vs 6 ± 3 min, P = 0.0001) and not clinically important. The quality of analgesia (visual analog scores and ropivacaine consumption) was similar between groups.
CONCLUSIONS: Distention of the epidural space with 5 mL saline before epidural catheter insertion decreased the incidence of accidental venous cannulation and the number of unblocked segments.
Identification of the epidural space by loss of resistance (LOR) with normal saline (NS) or lidocaine is superior to the use of air (1–6). But once the space is identified, some clinicians inject additional NS into it before catheter insertion (7–9) whereas others do not (4,5). Large volumes of NS in the epidural space may impair the quality of analgesia (10,11).
Accidental epidural vein cannulation is common when epidural catheters are inserted, with an incidence that depends on multiple factors including patient position, the angle of the epidural needle, and the flexibility of the epidural catheter tip (1,5,12–16). Our primary hypothesis, based on a pilot study, was that injecting a low volume of NS before catheter insertion and holding the catheter plunger closed would distend the epidural space and reduce the propensity for accidental venous cannulation without diminishing the speed of onset of analgesia or increasing the number of unblocked segments.
METHODS
After obtaining institutional ethics committee approval and informed consent, we enrolled 210, ASA class I-II nulliparous women with singleton cephalic presentation at term, who were in active spontaneous labor and requested epidural analgesia. We excluded patients with preeclampsia; morbid obesity (body mass index >35 kg/m2); history of drug or alcohol abuse; heavy smoking; or abnormal hepatic, renal, or hematological test results. Patients in whom cesarean delivery was anticipated were also excluded.
Protocol
Standardized, active management of labor protocol was used in all patients. Parturients were randomly allocated to receive epidural analgesia with or without distention of the epidural space before placement of the epidural catheter. Randomization was simple, and was based on computer-generated codes that were maintained in sequentially numbered opaque envelopes until just before use. The randomization envelopes were opened and the designated treatment started when visual analog scale (VAS) scores for pain reached 30 mm (VAS is a 100-mm scale where 0 mm = no pain and 100 mm = worst pain imaginable). Patients were blinded to treatment, as was the anesthesiologist who recorded or evaluated outcomes of analgesia and block levels.
Patients were prehydrated with 10 mL/kg Ringer’s lactate solution. Anesthesiologists with at least 4-yr experience in obstetric anesthesia and familiar with the LOR technique were responsible for initiating epidural analgesia. In both groups, patients were placed in sitting position and an 18-G, Tuohy epidural needle (B.Braun Inc., Melsungen, Germany) was inserted at the L3–4 or L2–3 intervertebral space, using the LOR technique with either 2 or 5 mL NS. Similarly to that found by Beilin et al. (5), we found that 2 mL of NS was sufficient for detecting the epidural space with LOR. This was also the minimal volume required in the control group in order not to distend the epidural space. Once the needle was positioned properly, in the nondistention group a 20-G, multiorifice epidural catheter (B.Braun Inc., Melsungen, Germany) was threaded 2–3 cm into the epidural space. In the distention group, 5 mL NS was injected into the epidural space while the syringe plunger was held closed for 20 s (the time period we found appropriate to avoid backflow). We choose to add 3 mL of NS to the 2 mL of NS used to detect the epidural space using the LOR technique, considering that larger volumes of NS might impair the quality of analgesia (10,11). Therefore, epidural space distention was achieved with a smaller volume of NS, while holding the syringe plunger closed. Immediately thereafter, a 20-G, multiorifice epidural catheter was treaded 2–3 cm into the epidural space.
After confirming that no blood or cerebrospinal fluid returned through the catheter upon aspiration or by free flow to gravity, 3 mL of 1.5% lidocaine was injected through the epidural catheter in both groups. Parturients were placed in the left lateral position, and 5 min later, if there were no signs of subarachnoid or intravascular injection, 5 mL of 0.2% ropivacaine was injected epidurally and another 5 mL was injected 5 min later. Adequate analgesia was confirmed by lack of bilateral response to pinprick to T8; if necessary, additional anesthetic was given, the catheter was repositioned (by pulling up to 2 cm depth), and the patient’s position was modified to a sitting position toward and during the second stage or labor, or to a side position when one side was less anesthetized (unilateral block).
Analgesia was maintained throughout labor and delivery with patient-controlled epidural analgesia (PCAM syringe pump model P500, IVAC Medical System, Basingstoke, NH), using a continuous epidural infusion with 5 mL/hr basal infusion rate and 5 mL patient-controlled boluses of plain ropivacaine 0.2% with 20 min lockout interval (20 mL/hr limit). Patients requesting additional analgesia received a 10 mL bolus of the same ropivacaine solution in 5 mL aliquots, using the patient-controlled epidural analgesia device. This bolus was included in the 20 mL hourly limit. Parturients remained in bed until delivery.
Measurements
Maternal monitoring consisted of noninvasive arterial blood pressure and heart rate measurements and tocodynamometry. External fetal heart rate was measured continuously. Invasive monitoring was provided as clinically indicated.
Routine demographic and morphometric characteristics were assessed before beginning the epidural analgesia. Duration of ruptured membranes, duration of labor and other routine obstetric variables were also recorded. Vaginal examinations were performed every hour during the first stage and every 30 min during the second stage of labor.
An anesthesiologist blinded to group assignment (not providing the epidural analgesia) assessed the highest sensory level of the epidural block bilaterally, pain score (VAS), modified Bromage score from 0 to 3 (0 = able to fully flex knees and feet, 1 = able to flex knees only, 2 = able to move feet only and 3 = unable to move either feet and knees) and total dose of ropivacaine administered, postdural puncture headaches, and transient neurological symptoms. All these variables were evaluated before the epidural, 1 hr after epidural catheter insertion, at the end of first stage of labor and in the second stage of labor, before delivery.
Total ropivacaine consumption and the number of unblocked segments in the range of zero to three were also recorded. Each side was considered a separate segment; thus, a bilateral block failure at a single dermatome was considered two failed segments.
The anesthesiologist who performed the epidural procedure recorded the occurrence of inadvertent intravascular or subarachnoid insertion of the catheter as indicated by aspiration of blood or free-flow clear fluid. Time to the first painless (VAS <5) contraction (i.e., speed of onset of analgesia) was tested at 1-min intervals.
Data Analysis
Inadvertent intravascular insertion of the epidural catheter was defined as the primary outcome. Secondary outcomes were inadvertent subarachnoid insertion of the catheter, the number of unblocked segments, the time to first painless contraction, the highest sensory level, and the modified Bromage score.
Sample-size estimate indicated that 200 patients (100/group) would provide an 80% power to detect a reduction in the incidence of inadvertent intravascular catheter insertion from 10% to 3% with a two-sided 
of 0.05. We enrolled 210 patients to allow for dropouts.
Analysis of data was performed using the SPSS 900 statistical analysis software (SPSS, Chicago, IL; 1999). For continuous variables descriptive statistics were reported as mean ± sd. Normalcy of distribution of continuous variables was assessed using the Kolmogorov– Smirnov test (cut-off P = 0.01). Continuous variables that could not be normalized via log transformation were compared by treatment group using the Mann– Whitney U-test. Normally distributed continuous variables were compared by treatment group using t-test for independent samples. Categorical variables were described using frequency distributions and are presented as frequency (%). The 
2 test was used to detect differences in categorical variables by treatment group. All tests were two-sided and considered significant with a P < 0.05.
RESULTS
The study was completed by 203 of 210 parturients. The remaining seven parturients were excluded from the study because they developed preeclampsia.
Among the 203 parturients who completed the study, 101 were allocated to the nondistention group and 102 to the distention group. There were no epidural failures. Patients’ demographic and labor characteristics were similar in the two groups (Table 1), except for subjects’ age in the distention group that was significantly older than the other patients.
|
The onset of analgesia was 1 min faster in the distention group (5.0 ± 2 min vs 6.3 ± 3 min, P < 0.0001). Other characteristics of the epidural blockade were similar in the two groups (Table 2).
|
There were fewer accidental intravascular catheter insertions (2 vs 16, P = 0.0001) in the distention group (Table 3). Ninety-one percent of the patients in the distention group did not have any unblocked segments compared with 67% in the nondistention group (P = 0.0001). The incidence of other side effects and complications attributable to epidural analgesia was similar in each treatment group (Table 3). There were no intrathecal catheter placements.
|
DISCUSSION
We demonstrated that distention of the epidural space with 5 vs 2 mL NS before epidural catheter insertion decreased the incidence of accidental intravascular placement, the incidence of unblocked segments and shortened the time to onset of analgesia. The theory of our study was that once the epidural space is identified, further distention of the epidural space with an incompressible fluid (i.e., NS) would provide lubrication and push blood vessels and other anatomical components away from the catheter. We hypothesized that this distention would, in turn, facilitate catheter insertion and reduce the incidence of accidental intravascular catheterizations. Our results were, therefore, consistent with this theory.
The amount of NS used in previous studies of LOR varied. Beilin et al. (5) used 2 mL while Valentine et al. (4) used 4 mL NS; however, both groups compared the speed of onset of analgesia with that achieved by using the same volume of air for LOR. Others injected additional volumes of NS into the epidural space, before or after catheter insertion. Sarna et al. (12), for example, injected 10 mL of NS or air, during or after the search for LOR and found no difference in analgesia and complication rate. Iwama (11) compared the quality of analgesia and incidence of complications using 2 or 10 mL NS for LOR. Epidural anesthesia with 10 mL NS was more extensive; however, 2 of 15 patients had unsatisfactory analgesia. It may thus be that higher volumes of NS dilute local anesthetics thereby reducing the potency of analgesia. Consistent with this theory, Okutomi and Hoka (10) using 1–10 mL of NS for the LOR technique demonstrated a decreased spread of the block to pinprick with larger volumes of NS injected.
Using NS to distend the epidural space is but one method of reducing accidental venous puncture. For example, puncture risk can be reduced by cephalad needle angulation (15), by using softer epidural catheter tips (16), or by having epidural injection performed by an experienced anesthesiologist (16). Behar et al. (13,14) were able to decrease the rate of this complication from 11% to 2% in both nonobese and obese parturients by performing the epidural with the patient in the head-down lateral recumbent position rather than the sitting position. Gadalla et al. (7) used an unknown volume of air with LOR for combined spinal–epidural labor analgesia. While injecting 10 mL NS before epidural catheter placement, they were able to reduce the incidence of accidental venous cannulation from 20% to 2% when compared to the non-NS injection patient group. Mannion et al. (8) assessed the effect of patient position, diameter of epidural catheter, and the injection of 10 mL NS before epidural catheterization, on the incidence of accidental epidural venous cannulation. The epidural space was identified with LOR with air. The injection of 10 mL of NS through an 18-G epidural needle before epidural catheterization with the patients in sitting position decreased the incidence of venous cannulation from 16% to 0%. Patient position had no significant effect on this complication. In another study (9), the injection of smaller volumes of NS or lidocaine (3 mL) after identification of the epidural space with LOR with 2 mL of air did not reduce the incidence of accidental epidural venous cannulation (9). The differences in the results may stem from the different techniques used, patient position, the volume of fluid injected, or catheter type inserted. Many studies found no differences in the incidence of intravascular cannulation when LOR to air was compared with LOR to NS. The role of fluid volume injected before catheter insertion has now become clearer. 10 mL of NS decreases the incidence of venous cannulation, but at the expense of possible impairment of the quality of analgesia (10–11). Using our technique, we achieved fewer accidental intravascular cannulations with a smaller volume of NS, but with the syringe plunger held closed to maintain the NS within the epidural space.
Despite facilitating catheter insertion (1,5,6,17), the LOR technique with NS or lidocaine is reportedly associated with unblocked segments in 3.2%–19% of patients, even when volumes as small as 2–4 mL are injected (2,4,5). This incidence is still lower when compared with that (6.6%–36%) which follows LOR with air (2,4,5). Interestingly, 91% of the patients in our predistention group did not have any unblocked segments compared to 67% of those in the nondistended group. In contrast to previous work, our results thus suggest that injection of 5 mL NS and holding the syringe plunger closed for 20 s before catheter insertion reduces the risk of unblocked segments.
These differing results may be explained by methodological details, including the volume of NS, concentration of local anesthetics, different time lag between the injection of the NS and the local anesthetic, and prolonging the distention time as we did by holding the syringe plunger closed for 20 s after NS injection. Delaying the injection of local anesthetic after injecting the NS results in the same beneficial effect (18). Another explanation for this discrepancy might be differences in the definition of onset of analgesia: the time to the first painless contraction (the definition used in our study), the time taken from the start of epidural to the peak of first "comfortable" contraction (19), or the time measured by loss of sensation to pinprick. Although there was a statistically significant difference in the onset time of analgesia between the two groups (Table 2), this difference (5 vs 6 min) does not seem to be clinically relevant.
In conclusion, predistention of the epidural space with 5 mL of NS was associated with a lower incidence of intravascular catheter insertion and fewer unblocked segments. This simple technique thus appears to reduce the risk of an important complication and to improve several aspects of epidural analgesia.
Footnotes
Accepted for publication April 4, 2007.
Supported by NIH Grant GM 061655 (Bethesda, MD), the Gheens Foundation (Louisville, KY), the Joseph Drown Foundation (Los Angeles, CA), and the Commonwealth of KY Research Challenge Trust Fund (Louisville, KY).
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  S. M. Siddik-Sayyid and A. S. Baraka Predistention of the Epidural Space with Saline Before Catheter Insertion Anesth. Analg., February 1, 2008; 106(2): 672 - 672. [Full Text] [PDF]
|
|
|
|
|
|
S. Evron and T. Ezri Predistention of the Epidural Space with Saline Before Catheter Insertion Anesth. Analg., February 1, 2008; 106(2): 672 - 673. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
