Secondary Logo

Journal Logo

Ambulatory Anesthesia: Research Report

A Comparison of Selective Spinal Anesthesia with Hyperbaric Bupivacaine and General Anesthesia with Desflurane for Outpatient Knee Arthroscopy

Section Editor(s): White, Paul F.Korhonen, Anna-Maija MD*†; Valanne, Jukka V. MD, PhD*; Jokela, Ritva M. MD, PhD; Ravaska, Pirjo MD*; Korttila, Kari T. MD, PhD, FRCA

Author Information
doi: 10.1213/01.ANE.0000139351.40608.05

Controversial results concerning recovery times have been found in studies comparing ambulatory spinal anesthesia and general anesthesia (GA). Spinal anesthesia with 10 mg of bupivacaine resulted in prolonged discharge home compared with IV anesthesia with propofol and remifentanil (1), but 6 mg of bupivacaine, together with 15 μg of fentanyl intrathecally, resulted in equal recovery times with propofol-maintained GA (2). Patients receiving selective spinal anesthesia (SSA) produced by a combination of lidocaine 10 mg and sufentanil 10 μg for ambulatory gynecological laparoscopy were able to walk but could not be discharged earlier than patients receiving GA with desflurane (3). Compared with propofol, desflurane has been shown to produce better fast-tracking possibilities (4) and faster early stage recovery (5,6) in outpatients. In ambulatory knee arthroscopy, SSA with bupivacaine has not been compared to GA with desflurane.

The goal of this study was to assess whether SSA with small-dose hyperbaric bupivacaine provides an equal possibility of fast-tracking, a shorter stay in the postanesthesia care unit (PACU), and earlier discharge home compared with GA with desflurane.


After approvals of the Ethics Committee of Lapland Central Hospital, we obtained the written, informed consent of 64 ambulatory adult American Society of Anesthesiologists physical status I–III knee arthroscopy patients, with a body mass index <32. Patients with allergy to any of the study drugs or contraindication either to spinal or GA were excluded.

We used a sealed envelope technique with computer-generated numbers to randomize the patients prospectively into two groups: the SSA group and the GA group. The patients were given 600 mg of ibuprofen or 1000 mg of paracetamol PO preoperatively. An IV infusion was started before anesthesia and standard monitoring (electrocardiograph, noninvasive arterial blood pressure, and pulse oximetry) was used. Bradycardia <50 bpm was treated with glycopyrrolate and bradycardia <40 bpm was treated with atropine, and hypotension <90 mm Hg or systolic arterial blood pressure decreasing more than 50 mm Hg from the baseline were treated with etilefrine. Postoperative pain was evaluated with a visual analog scale (VAS) from 0 = no pain to 10 = the worst pain imaginable. Severe postoperative pain (VAS >7) was treated with fentanyl IV (up to 0.1 mg), moderate pain (3<VAS<8) with a combination of paracetamol and codeine (0.5 g of paracetamol and 30 mg of codeine × 1–2) PO, or if contraindicated, with tramadol (100 mg) IV/PO. If the pain relief was still insufficient, the patients received IV oxycodone (3–5 mg IV repeatedly until pain relief) and were admitted to the hospital. The Bispectral index (BIS) and the adductor pollicis train-of-four (TOF) ratio (by using kinemyographic measurement with Datex-Ohmeda MechanoSensor®; Datex-Ohmeda, Helsinki, Finland) were monitored in the GA group. The inspired and end-tidal concentrations of oxygen, CO2, desflurane, and N2O were also monitored.

SSA Group

The patients in the SSA group received 4 mg (0.8 mL) of hyperbaric bupivacaine injected through a 27-gauge cutting needle (Becton-Dickinson Yale Spinal; Becton-Dickinson, Franklin Lakes, NJ). A very low-flow (i.e., 0.4 mL/min) steady injection, at the lumbar 2/3 interspace with the bevel of the needle directed laterally towards the nerves involved, was used. The patient was maintained in a lateral decubitus position, the operative side dependent, for 10 min. Before the spinal injection, a level was used to ensure that the vertebral column was horizontal. Perioperatively, the patients were given midazolam up to 2 mg IV and/or alfentanil up to 0.5 mg IV when needed. If the dermatomes L1–5 were not blocked, the spinal anesthesia was considered to have failed. The sensory block to cold stimulus was tested by using thermal stimuli (acetone drop) (7,8). The motor block was assessed according to a modified Bromage scale (7,8). After the release of the tourniquet, and every 20 min thereafter, the sensory and motor block were evaluated until the transfer to the ambulatory surgery unit (ASU). The time in the PACU was recorded as the time from admission to the PACU until transfer to ASU. The PACU discharge was evaluated by using both the new fast-tracking scoring system by White and Song (9) and also by complete recovery of motor block, sensory block not above Th12, and the ability to sit up. The patient was fast-tracked to ASU if the latter criteria were fulfilled at the end of surgery and if the new fast-tracking criteria were also fulfilled.

GA Group

The risk for postoperative nausea and vomiting (PONV) was assessed before GA (10). The patients with more than two risk factors (female, nonsmoker, history of PONV, or motion sickness) were given prophylactic dexamethasone 5 mg IV after induction and 5-HT3 antagonist (ondansetron 4 mg) IV at the end of surgery. Rescue medication for PONV was provided by metoclopramide 10 mg IV for the patients with the prophylaxis and ondansetron 4 mg IV for the patients without prophylaxis. GA was induced with propofol 2–3 mg/kg and 0.1 mg fentanyl. To facilitate the tracheal intubation a dose of 0.4 mg/kg of rocuronium was administrated IV (11). The anesthesia was maintained with desflurane 2%–6% and 50% N2O in O2 titrated to keep the BIS index value between 50 and 60 towards the end of surgery. Additional fentanyl 0.05 mg IV was allowed when clinically indicated (SAP or heart rate more than 15% above the baseline values). All patients were mechanically ventilated to maintain an end-tidal CO2 concentration of 4.5–5.5kPa. A bolus dose of rocuronium (0.1 mg/kg IV) was administered if required (high peak inspiratory pressure values or coughing). A neuromuscular reversal drug was used if the TOF ratio was <0.8 (12) at the end of the operation. When the tourniquet was released, desflurane and N2O were discontinued. The patients were tracheally extubated when fully awake. The recovery status was assessed after extubation and every 20 min thereafter by using the new fast-track scoring system (9). A minimum score of 12 of 14 with no zero scores was required for the patient to be fast-tracked. Before transferring to ASU the patient had to be able to sit up, and those patients with vertigo or somnolence or those who wanted to lie down were kept in the PACU even if their fast-tracking score was ≥12.

The anesthesia preparation time (APT) was recorded as the time from ensuring the posture of the vertebral column until the end of the 10 min required for the patient to maintain the lateral decubitus position in the SSA group. In the GA group, the APT was considered as the time from the beginning of injection of the induction drugs until connection of the patient to a respirator. Immediate recovery time from anesthesia was recorded as the time from release of the tourniquet until tracheal extubation and orientation. Together, the APT and the immediate recovery time were considered as the anesthesia-related time. Walking, voiding, and home-readiness were measured from the release of the tourniquet. Home-discharge criteria consisted of absence of nausea, vomiting and bleeding, minimal or no pain, and ambulating. Voiding was not required before home-readiness but the time to void was recorded if the patient had not yet left the hospital (for example, if no escort arrived). Postoperative pain, PONV, headache, and notable tiredness were recorded. Four to seven days after operation, a telephone interview was conducted to inquire about possible headache, backache, pain, postdischarge nausea or vomiting, dysuria, transient neurological symptoms (TNS), or postdural puncture headache. TNS was described as transient pain/dysesthesia in the back radiating to buttocks or legs on areas not related to the surgery. The patient’s satisfaction with the anesthesia was also assessed.

The sample size of 28 per treatment group was calculated to detect a 20% difference in the home-discharge time (140 min (4) in the GA group versus 110 min (7,8) in the SSA group) after surgery with 40% sd using an α of 0.05 and a β of 0.2. Because of possible dropouts, we decided to randomize 32 patients into both groups. The outcomes between the groups were statistically compared by calculating the P value for the null hypothesis of no difference using the χ2 test, Fisher’s exact test, and the Mann-Whitney U-test when appropriate. A P value <0.05 was considered significant. The SPSS for Windows (version 11.0.1) statistical package (SPSS Inc., Chicago, IL) was used.


Sixty-four patients were enrolled (32 in both groups). The patients’ characteristics are shown in Table 1. Two spinal blocks failed (6%) and these patients received GA for the surgery. They were excluded from the analyses. In the GA group, three patients had to stay overnight in hospital. They were excluded from the analyses concerning voiding, ambulating, and home-readiness. One patient stayed because of surgical reasons and two because of pain. The surgical procedures of these two were partial excision of meniscus in one and excision of the plica and operation for osteochondritis in the other. Both procedures lasted longer (56 and 65 min) than the median duration of the surgery in this study.

Table 1
Table 1:
Demographic Characteristics

The APT and recovery times are given in Table 2. The overall anesthesia-related time in the operating room (OR) differed by 1.5 min between the SSA and GA groups (P = 0.033). No statistical differences were found in the fast-tracking possibilities or in the PACU time. In the SSA group, 12 patients were admitted to PACU instead of fast tracking because of partial motor block on the operative side, although the new fast-tracking criteria were fulfilled. Fourteen SSA patients stayed in PACU because of complete motor and/or sensory block above Th12 on the operative side. Home discharge criteria were fulfilled equally in both groups.

Table 2
Table 2:
Times of Preparation and Recovery, Bispectral Index Value and Fast-Tracking

Early and late side effects are listed in Table 3. In the GA group, the patients had significantly higher pain scores in the hospital and needed more postoperative opioids. In the GA group, the fast tracking failed in seven (22%) patients because of pain and five patients considered the pain to be severe (VAS ≥8). At home, the pain was mild in both groups. In ASU, somnolence was significantly more frequent in the GA group.

Table 3
Table 3:
Side Effects and Patient Satisfaction

After the PONV risk assessment, 13 (41%) patients received PONV prophylaxis in the GA group. Six patients (19%), compared with zero, suffered from PONV in the hospital in the GA and SA groups (P = 0.024), respectively. Of these six patients, three had received PONV prophylaxis, two had received reversal drugs, and four had received opioids for postoperative pain. At home the incidence of postdischarge nausea or vomiting did not differ between the groups (Table 3).

Bradycardia was treated in three versus eight patients (NS) and hypotension in zero versus seven patients (P = 0.011) after SSA and GA, respectively. Two of the patients who received medication for hypotension were elderly (ages 70 and 76 yr) and one belonged to ASA physical status III.

Ninety percent and 74% of the patients experienced the anesthesia to be superior to their expectations in the SSA and GA groups (NS), respectively. In the SSA group, one patient would choose GA instead of SSA in the future because of postdural puncture headache. In the GA group, six patients would like to have SSA instead of GA in the future. Four of them wished to be awake during the surgery and to be able to communicate with the surgeon and two of the patients would prefer SSA because of the side effects after GA.


Patients undergoing knee arthroscopy achieved equally fast home-readiness after both SSA with hyperbaric bupivacaine and GA maintained with desflurane. No significant difference was found in fast-tracking possibilities or time in PACU, but the sample size might have been too small to detect the difference between groups. Compared with GA, the use of SSA was associated with a significantly less side effects in the hospital.

Recovery from motor block is faster if intrathecal opioids are used in combination with small-dose local anesthetics (7). When we compared the SSA with a combination of 3 mg of bupivacaine + 10 μg of fentanyl with 4 mg of bupivacaine, we found a shorter PACU stay and a trend towards a better possibility of fast-tracking but equal time to home-readiness (171 versus 183 min, P = 0.172). As the reliability was equal (1 block versus 0 failed after B3F10 and B4) (7), we decided to use bupivacaine solely in the present study. In addition, even small doses of spinal opioids cause pruritus (7,13) and it seems that intrathecal opioids also cause dose-dependent PONV. In several studies, the 20–25 μg dose of intrathecal fentanyl has produced a 20%–30% incidence of PONV (14,15). Lennox et al. (3) found a 30% incidence of PONV in the SSA group with intrathecal lidocaine 10 mg mixed with sufentanil 10 μg compared with 0% in the GA group with desflurane. Neither in the present nor in our previous studies have any of the patients in the SSA group had PONV (7,8,13).

The time in PACU was relatively short in both groups. Although Jankowski et al. (2) fast tracked all patients after spinal anesthesia with 6 mg of bupivacaine and 15 μg fentanyl, we could fast-track only 12% of the patients in the SSA group. The different PACU bypass criteria between the studies might be an explanation. To fast track patients in the GA group, we estimated the White-Song fast-tracking criteria to be better for the GA patients as compared with the Aldrete scoring system (9). However, as the White-Song criteria were developed for patients undergoing ambulatory surgery with GA, we found them to be insufficient for patients undergoing SSA, and a complete recovery of motor block, the ability to sit up (without any symptoms), and sensory block not > Th12 was also required. Also, in the GA group, several patients with White-Song criteria ≥12 were admitted to the PACU instead of the ASU because, while they were sitting, they felt dizzy or somnolent. In the present study the PACU bypass criteria differed between the SSA and GA groups, a point that should be considered when interpreting the results of patients meeting fast-tracking criteria and time in PACU. In the future, when comparing GA and neuraxial anesthesia, PACU bypass criteria by Williams et al. (16) designed for both neuraxial and GA might be useful because they also include pain, PONV, shivering, and orthostasis.

The overall anesthesia-related time was 1.5 min longer in the SSA than in the GA group (Table 2). Although the difference was statistically significant, such a short time has only minor clinical relevance. In the SSA group, the anesthesia-related time consisted only of APT (17 min) because all the patients were awake and orientated at the time of tourniquet release. Similar preparation time was also found by Fanelli et al. (17) after unilateral spinal anesthesia. In the SSA group, a 17-min reduction in the OR time could have been achieved if an induction room had been used. However, this imposes other costs and requirements in terms of both physical plant and personnel. On the other hand, had we used the laryngeal mask airway (LMA) and spontaneous ventilation in the GA group, a reduction in anesthesia related time in the OR might also have been achieved in the GA group. In the present study, the patients could be tracheally extubated and were orientated 7.5 min after the release of the tourniquet in the GA group, whereas the time to orientation with desflurane LMA anesthesia was 3.4–4 min in the studies of Dolk et al. and Tang et al. (5,6). In the present study, the use of tracheal intubation and muscle relaxants instead of the LMA can be criticized and might have caused a bias in favor of SSA versus GA.

In the present study, time to home-readiness was similar in both groups. Equal home discharge times between SSA and GA groups were also noted in earlier studies (2,4). Although the patients in the study of Lennox et al. (3) could ambulate 3 min after surgery in the SSA group, as compared with 60 min in the GA group, the overall recovery times were equal, 112 min versus 101 min, in the SSA and GA groups, respectively.

In the hospital, postoperative pain VAS scores were significantly higher in the GA group compared with those in the SSA group, resulting in a need for more postoperative opioids. No difference in pain scores was noted at home. Jankowski et al. (2) also reported higher VAS scores (although mild pain) in GA than in spinal anesthesia patients. In their work, the local anesthetic used by the surgeon on both the arthroscopic portal site and intraarticular (IA) could explain the lower VAS scores. We did not have this procedure in our protocol because in our earlier studies, the patients with SSA had no problem with postoperative pain, either in the hospital or at home after knee arthroscopy. The use of IA morphine 5 mg might, nevertheless, have been beneficial in the GA group in decreasing the need for postoperative opioids (18) and the risk of PONV. Although 13 (41%) of the patients received PONV prophylaxis, six (19%) patients developed PONV in the GA group. Half of them had received the prophylaxis. The triple prophylaxis with the combination of dexamethasone, 5-HT3 antagonists, and droperidol could have been more effective in preventing PONV (19), but we wanted to avoid the use of droperidol in outpatients because of the possible QT prolongation (20) and the sedative effect. In the present study, TNS occurred in both groups. Although the exact mechanism remains unclear, the TNS is most frequent after lidocaine spinal anesthesia (10%–37%) compared with other local anesthetics and in patients undergoing knee arthroscopy (18%–22%) or surgery in the lithotomy position (30%–36%) (21). Interestingly, TNS also occurred after GA (6% of the patients), which could support the suggestion of the musculoskeletal origin of these symptoms and that patients undergoing knee arthroscopy are in risk of developing TNS.

In conclusion, patients undergoing ambulatory knee arthroscopy had equal likelihood of meeting fast-track criteria and discharge times after SSA with 4 mg of hyperbaric bupivacaine and after endotracheal GA with desflurane. However, those patients receiving SSA had lower pain scores and need of postoperative opioids, less PONV, and somnolence in the hospital.

We are grateful for the help and support of Ms. Eija Ruoppa, RN, and the nurses at the Ambulatory Surgery Unit, to Ms. Marja-Liisa Putaansuu, RN, as well as to the personnel at the Anaesthesia Department of Lapland Central Hospital.


1. Danelli G, Berti M, Casati A, et al. Spinal block or total intravenous anaesthesia with propofol and remifentanil for gynaecological outpatient procedures. Eur J Anaesthesiol 2002;19:594–9.
2. Jankowski CJ, Hebl JR, Stuart MJ, et al. A comparison of psoas compartment block and spinal and general anesthesia for outpatient knee arthroscopy. Anesth Analg 2003;97:1003–9.
3. Lennox PH, Vaghadia H, Henderson C, et al. Small-dose selective spinal anesthesia for short-duration outpatient laparoscopy: recovery characteristics compared with desflurane anesthesia. Anesth Analg 2002;94:346–50.
4. Song D, Joshi GP, White PF. Fast-track eligibility after ambulatory anesthesia: a comparison of desflurane, sevoflurane, and propofol. Anesth Analg 1998;86:267–73.
5. Tang J, White PF, Wender RH, et al. Fast-track office-based anesthesia: a comparison of propofol versus desflurane with antiemetic prophylaxis in spontaneously breathing patients. Anesth Analg 2001;92:95–9.
6. Dolk A, Cannerfelt R, Anderson RE, Jakobsson J. Inhalation anaesthesia is cost-effective for ambulatory surgery: a clinical comparison with propofol during elective knee arthroscopy. Eur J Anaesthesiol 2002;19:88–92.
7. Korhonen AM, Valanne JV, Jokela RM, et al. Intrathecal hyperbaric bupivacaine 3 mg + fentanyl 10 μg for outpatient knee arthroscopy with tourniquet. Acta Anaesthesiol Scand 2003;47:342–6.
8. Valanne JV, Korhonen AM, Jokela RM, et al. Selective spinal anesthesia: a comparison of hyperbaric bupivacaine 4 mg versus 6 mg for outpatient knee arthroscopy. Anesth Analg 2001;93:1377–9.
9. White PF, Song D. New criteria for fast-tracking after outpatient anesthesia: a comparison with the modified Aldrete’s scoring system. Anesth Analg 1999;88:1069–72.
10. Apfel CC, Laara E, Koivuranta M, et al. A simplified risk score for predicting postoperative nausea and vomiting: conclusions from cross-validations between two centers. Anesthesiology 1999;91:693–700.
11. Schlaich N, Mertzlufft F, Soltesz S, Fuchs-Buder T. Remifentanil and propofol without muscle relaxants or with different doses of rocuronium for tracheal intubation in outpatient anaesthesia. Acta Anaesthesiol Scand 2000;44:720–6.
12. Viby-Mogensen J, Englbaek J, Eriksson LI, et al. Good clinical research practice (GCRP) in pharmacodynamic studies of neuromuscular blocking agents. Acta Anaesthesiol Scand 1996;40:59–74.
13. Korhonen AM, Valanne JV, Jokela RM, et al. Ondansetron does not prevent pruritus induced by low-dose intrathecal fentanyl. Acta Anaesthesiol Scand 2003;47:1292–7.
14. Mulroy MF, Larkin KL, Hodgson PS, et al. A comparison of spinal, epidural, and general anesthesia for outpatient knee arthroscopy. Anesth Analg 2000;91:860–4.
15. Pollock JE, Mulroy MF, Bent E, Polissar NL. A comparison of two regional anesthetic techniques for outpatient knee arthroscopy. Anesth Analg 2003;97:397–401.
16. Williams BA, Kentor ML, Williams JP, et al. Process analysis in outpatient knee surgery: effects of regional and general anesthesia on anesthesia-controlled time. Anesthesiology 2000;93:529–38.
17. Fanelli G, Borghi B, Casati A, et al. Unilateral bupivacaine spinal anesthesia for outpatient knee arthroscopy. Italian Study Group on Unilateral Spinal Anesthesia. Can J Anaesth 2000;47:746–51.
18. Kalso E, Smith L, McQuay HJ, Andrew Moore R. No pain, no gain: clinical excellence and scientific rigour: lessons learned from IA morphine. Pain 2002;98:269–75.
19. Scuderi PE, James RL, Harris L, Mims GR 3rd. Multimodal antiemetic management prevents early postoperative vomiting after outpatient laparoscopy. Anesth Analg 2000;91:1408–14.
20. Dershwitz M. Droperidol: should the black box be light gray? J Clin Anesth 2002;14:598–603.
21. Pollock JE. Neurotoxicity of intrathecal local anaesthetics and transient neurological symptoms. Best Pract Res Clin Anaesthesiol 2003;17:471–84.
© 2004 International Anesthesia Research Society