Combined Spinal-Epidural Anesthesia for Outpatient Surgery: Dose-Response Characteristics of Intrathecal Isobaric Lidocaine Using a 27-Gauge Whitacre Spinal Needle

Data were collected prospectively from 90 patients using a randomized, double-blinded study design. Patients were ASA physical status 1-3 who were scheduled for ambulatory knee arthroscopy. Patients recruited for the study were 18-60 yr of age and were without contraindication to spinal or epidural anesthesia. After Institutional Review Board approval and written informed consent, patients were randomly assigned using a computer-generated random sequence to receive CSE with one of three initial intrathecal injection doses of 2% isobaric plain lidocaine (Abbott, North Chicago, IL): 40, 60, or 80 mg (2, 3, or 4 ml).

For the original study design, we devised a randomization table for 100 patients randomly assigned to four study groups: 80, 60, 40, and 30 mg 2% lidocaine as the initial intrathecal CSE dose. A few patients were studied from this table, including two patients who were randomly assigned to receive 30 mg 2% lidocaine. Each patient who received 30 mg lidocaine had inadequate spinal anesthesia. As a result, we eliminated the 30-mg dose group. We created a new randomization table limited to 40-, 60-, or 80-mg lidocaine 2% initial intrathecal doses and began the study again.

Patients received no premedication and sedation was limited to a maximum of 10 mg intravenous midazolam. No opioids or other analgesics were administered intraoperatively. All patients had the following monitors: electrocardiogram, automated blood pressure cuff (Critikon, Tampa, FL) every 3 min, and finger pulse oximeter. Heart rate and blood pressure trends were recorded for the duration of the operative procedure. Bradycardia was defined before data collection as a heart rate of less than 50 beats/min and hypotension as a systolic blood pressure of less than 80 mmHg at any point in time.

After placement of an intravenous catheter, patients received 500 ml lactated Ringer's solution. Combined spinal-epidural anesthesia was performed and initial intrathecal injection was administered at the L3-L4 interspace with patients in the right lateral decubitus position using a midline approach. The L3-L4 interspace was operationally defined as the first palpable interspace that was cephalad to the iliac crest. A 17-G Weiss needle was inserted into the epidural space using loss of resistance. A 27-G, 4 11/16-inch Whitacre point spinal needle (Becton-Dickinson, Franklin Lakes, NJ) was then placed through the 17-G, 3 1/2-inch (patients 1-49) or 3-inch (patients 50-90) Weiss epidural needle. The Whitacre aperture was directed laterally, toward the operative knee, for the entire injection. Ability to aspirate CSF was assessed after each 1-ml increment of the intrathecal injection. The spinal needle was then removed and a 20-G (B. Braun Medical, Bethlehem, PA) epidural catheter was inserted 3 cm into the epidural space.

If the dural puncture attempt with the Whitacre needle through the Weiss needle at L3-L4 was unsuccessful (defined as no spontaneous CSF flow on removal of stylet), the Whitacre needle was removed and an epidural catheter was advanced through the Weiss needle. After this, spinal anesthesia using a separate 27-G Whitacre needle spinal was performed in the same interspace using a 20-G spinal introducer inserted just lateral to the epidural catheter.

Supplementation of the original spinal anesthetic by use of the epidural catheter was done by a separate anesthesiologist (blinded to the original intrathecal dose) according to strict criteria. Criteria for epidural supplementation were: (1) patient movement of the lower extremities, (2) surgeon complaint of inadequate muscle relaxation, or (3) patient discomfort. Reinforcement was with 5-ml increments of 2% lidocaine after a record of the patient's pain (0 = no pain to 10 = worst imaginable pain). This was repeated every 3 min, if required, to a maximum reinforcement dose of 15 ml.

After injection of intrathecal lidocaine, a single, blinded observer measured onset and regression of thoracic and lumbar sensory anesthesia by pinprick and motor blockade by use of the Bromage scale. [9]For motor testing, a score of 0 = no paresis, full movement; 1 = partial paresis, ability to flex knee only; 2 = partial paresis, ability to flex foot only; and 3 = full paresis, no movement. Assessment of perirectal sensation was not performed. Measurements were made at 5, 10, 15, 20, 30, 40, 50, and 60 min and every 15 min thereafter until complete resolution of anesthesia. Measurements began preoperatively were continued intraoperatively and into the postoperative period.

Times were measured and recorded by postanesthesia care unit nurses (who were unaware of the original intrathecal dose) for each of the following milestones: ability to sit unassisted in a chair, first spontaneous urination, initiation of oral intake, and ambulatory unit discharge. Strict criteria were used by these nurses in judging patients ready for discharge. [3]These included all of the following requirements: complete resolution of thoracic and lumbar sensory anesthesia and motor blockade, return of vital signs and mental status to preanesthetic levels, ability to urinate, ability to ambulate, adequate analgesia, and absence of nausea and vomiting. Follow-up phone calls were done by ambulatory surgical nurses (who were blinded to treatment) to elicit symptoms consistent with postdural puncture headache or other complications in all patients on postoperative days 1 and 3, using a standardized questionnaire.

Statistical analysis was by ANOVA, chi-squared, or Kruskal-Wallis nonparametric analysis. To define differences in resolution of thoracic sensory, lumbar sensory, and lower extremity motor blockade, the time to full resolution of each of these was measured and survival analysis was performed with between-group comparisons using a Mantel Cox test. To account for multiple comparisons, alpha was set at 0.01.

Ninety-nine patients were randomized; two of these were excluded because of surgical reasons (changes in procedure from preoperative diagnosis). An additional seven patients who consented to be studied were excluded because it was difficult to perform the procedure while they were in the lateral decubitus position because of obesity or scoliosis. Subjectively, all 90 patients had high-quality spinal anesthesia, and epidural reinforcement was successful in each case in which it was required. Demographic data are displayed for each of the study groups in Table 1. There were no dural punctures with the Weiss epidural needle and no patients had postdural puncture headaches at follow up.

The original study design included a 30-mg lidocaine 2% group (see Materials and Methods section). Two patients received 30 mg lidocaine, and each had inadequate motor blockade and complained of pain in the knee. One had a duration of 38 min before complaining of pain at the surgical site. The second complained of pain at less than 35 min and it was our impression that this patient's spinal was not complete before the 35-min point. While operating on both of these patients, the surgeon, who was blinded to dosage, complained of inadequate motor blockade that, he said, made arthroscopic examination of the knee very difficult. For this reason, the fourth group (30-mg dosage) was eliminated from the study design.

(Table 2) shows mean spinal anesthetic durations for each of the three groups as well as those patients who required epidural reinforcement. Times to two-dermatome regression, resolution of thoracic sensory anesthesia, resolution of lumbar sensory anesthesia, and full return of motor blockade by Bromage scale were all approximately 30 min shorter for the 40-mg group than for the 60- or 80-mg groups (P < 0.01 for all variables). Thoracolumbar sensory level of anesthesia as a function of time from injection is shown in Figure 1. Median Bromage score as a function of time from spinal anesthetic injection is plotted for each of the three groups in Figure 2.

The results of survival analysis for resolution of sensory and motor blockade are shown in Figure 3. This included patients who received epidural supplementation. Duration of thoracic sensory, lumbar sensory, and motor blockade were significantly longer in the 60- than in the 40-mg group (P < 0.0002) and longer in the 80- than in the 40-mg group (P < 0.0001).

(Table 3) shows mean times from intrathecal injection to fulfillment of ambulatory milestones and discharge. The 40-mg dose significantly decreased all times compared with the 60- and 80-mg dose groups (P < 0.01). There were no significant differences, however, between the 60- and 80-mg groups.

Seven patients (7.7%) required epidural supplementation intraoperatively, all for discomfort in the knee: three in the 40-mg lidocaine group (10.3%), three in the 60-mg group (9.4%), and one in the 80-mg group (3.4%). The mean time to reinforcement was 49 plus/minus 11 min and was similar in all groups. The mean volume of 2% lidocaine administered epidurally was 9.3 ml (range 5-15 ml). Patients who received epidural supplementation tended to have a longer duration of sensory anesthesia, motor blockade, and delayed postoperative recovery in the postanesthesia care unit (Table 2and Table 3).

There were no statistically significant differences in the incidence of bradycardia (heart rate < 50 beats/min), hypotension (systolic < 80 mmHg), or the need for intravenous atropine or ephedrine between groups (Table 4). One patient developed asystole 15 min after onset of spinal anesthesia with 80 mg of 2% lidocaine. This patient was quickly and successfully treated with immediate precordial thump-pacing and concurrent administration of 8 micro gram epinephrine and 1 mg atropine by intravenous route. This patient never lost consciousness and communicated with us throughout the treatment. She had no subsequent or residual complications.

Using the 3 1/2-inch Weiss epidural needle (patients 1-49), 24.5% of patients required separate spinal injection because of inadequate protrusion of the tip of the Whitacre spinal needle, which was measured at up to 12 mm. Most often, the dura was believed to be encountered and pushed, but the needle had inadequate length to penetrate. This caused a visible recoiling of the Whitacre needle out of the Weiss needle hub.

Subsequently, a 3-inch Weiss needle was used (patients 50-90), which allowed greater protrusion of the Whitacre needle tip. In spite of this, it was still necessary to give a separate spinal in 12.5% of these patients because of an inability to access cerebrospinal fluid (CSF; the Whitacre needle encountered bone or there was no free flow of CSF).

In the 33 patients in whom successful CSE was performed, the protrusion of the Whitacre needle beyond the tip of the Weiss needle was measured. This ranged from 5 to 15 mm, with a mean value of 8.5 mm.

Spontaneous flow of CSF through the Whitacre spinal needle was seen in all 90 patients before injection. However, it was not always possible to aspirate CSF via the Whitacre needle. In seven patients, CSF could not be aspirated at all, and in an additional four patients, aspiration of CSF was not possible after at least one of the 1-ml incremental doses of the spinal anesthetic.

A reliable initial dose of isobaric lidocaine 2% to administer intrathecally when using CSE for ambulatory knee arthroscopy is 40 mg. This dose provided excellent anesthesia, with a more rapid resolution of sensory and motor blockade than either 60 or 80 mg. We considered 30 mg as a potentially better initial dose, but because of our preliminary experience with two patients who received 30 mg and had inadequate anesthesia, we abandoned the 30-mg dose as being ineffective. The combined technique of CSE, using an initial dose of 40 mg 2% lidocaine, had qualities of rapid onset, short duration, and offered flexibility as provided by an epidural catheter.

Not only was the duration of sensory and motor blockade shortened in the 40-mg group, but all indices of recovery in the postanesthesia care unit were improved. Patients in the 40-mg group were able to sit in a chair, drink, and urinate before the 60- and 80-mg groups and were discharged on average almost 1 h earlier. Three patients of 29 in the 40-mg group developed pain in the surgical site intraoperatively, which was successfully treated using 2% lidocaine via the epidural catheter. Even when the three patients who required supplementation in the 40-mg group were considered, this had little impact on the recovery time.

Increasing the dose of anesthetic has been shown to prolong the duration of epidural anesthesia [8,10-12]and spinal anesthesia with 0.5% bupivacaine. [13,14]This is the first report showing a dose-response relationship between the dose of intrathecally administered isobaric lidocaine and duration of spinal anesthesia.

The CSE anesthetic that we chose to evaluate for ambulatory knee arthroscopy used a 27-G Whitacre needle-through-needle technique. This was deliberately chosen because: (1) it is technically easy to perform, (2) fine-gauge Whitacre-point needle are associated with low incidence of postdural puncture headache, [15]and (3) it involves minimal patient discomfort. A recent study of 25-G Whitacre needle spinal anesthetics in more than 3,000 obstetric patients by Hurley et al. [16]found an incidence of postdural puncture headache of 1.1% in patients who underwent cesarean section, a group known to be at high risk for this syndrome. In addition, headaches that occurred in these patients were characterized as being milder, and responded to conservative treatment without the necessity to resort to epidural blood patch. Thus, although the 27-G needle has not been well studied to date, one would expect an incidence of postdural puncture headache at least as low as, and possibly less than, that associated with the 25-G needle. An incidence in this range would certainly be acceptable in the ambulatory surgical center. We observed no headaches in the ninety patients, but a study of much greater numbers would be necessary to define a postdural puncture headache incidence.

After changing to a 3-inch Weiss needle to enable greater protrusion of the tip of the 27-G Whitacre needle, there were still patients (12.5%) in whom it was not possible to perform a successful dural puncture using the CSE technique at the L3-L4 interspace that satisfied all study criteria. These criteria included a single midline insertion and spontaneous CSF flow without the aid of aspiration with the Whitacre needle aperture oriented toward the operative knee. In some patients, the Whitacre needle appeared to strike bone; in others, no CSF could be detected. Although this may appear to be a limitation of the 27-G Whitacre needle-through-needle technique, our clinical experience supports an approximate 100% success rate for CSE anesthesia in general.

In conclusion, CSE with a 40-mg initial dose of isobaric lidocaine provided reliable and significantly shorter-duration anesthesia for knee arthroscopy, allowing faster discharge than was possible with CSE using 60- or 80-mg initial doses. Duration of spinal anesthesia was indeed dose related. However, this dose dependence was not linear; rather, it presented as a step function with the 40-mg dose characterized by shorter duration. In contrast, the 60- and 80-mg doses were of approximately equal and significantly longer duration.

The authors thank Dr. Donald Lambert, for his review of the manuscript.