Postoperative pain is the principal cause of morbidity

The Effects of Glossopharyngeal Nerve Block on Postoperative Pain Relief After Tonsillectomy: The Importance of the Extent of Obtunded Gag Reflex as a Clinical Indicator Hee-Pyoung Park, MD* Jung-won Hwang, MD* Sang-Hyun Park, MD* Young-Tae Jeon, MD* Jae-Hyon Bahk, MD Yong-Seok Oh, MD* BACKGROUND: We evaluated the efficacy of glossopharyngeal nerve block (GNB) for the control of posttonsillectomy pain in adult patients, and correlated the extent of obtunded gag reflex as a clinical indicator of GNB with the extent of pain relief. METHODS: Seventy-five patients undergoing tonsillectomy received bilateral GNB with 0.75% ropivacaine with epinephrine (), 0.5% bupivacaine with epinephrine () at the end of the operation, or no intervention (). To evaluate the effects of GNB, we assessed throat pain (100 mm visual analog scale) and severity of gag reflex response 0.5, 8, and 24 h after surgery. RESULTS: In the immediate postoperative period, pain scores at rest and when swallowing in Groups R and B were significantly lower than those in (21 17 and 23 13 vs 42 16, 28 22 and 32 19 vs 62 14, P 0.001). The analgesic effect of GNB was strongly correlated with the extent of obtunded gag reflex (P 0.01). CONCLUSIONS: GNB is a useful method for the palliation of posttonsillectomy pain. An obtunded gag reflex response may be a clinical indicator for analgesia from GNB. (Anesth Analg 2007;105:267 71) Postoperative pain is the principal cause of morbidity after tonsillectomy. This pain can affect the patient s nutrition, ability to return to work or school, discharge from the hospital, and overall satisfaction with the procedure. Tonsillectomy produces severe pain on the first postoperative day (1). Several techniques have been described for the alleviation of this pain, including the use of opioids (2), steroids (3), and nonsteroidal antiinflammatory drugs (2,4), as well as local anesthetic sprays (5) and infiltration with local anesthetics around the tonsillar bed (6 10). There is some controversy regarding the efficacy of glossopharyngeal nerve block (GNB) for the control of immediate posttonsillectomy pain. Some authors have reported that GNB or modified peritonsillar infiltration with local anesthetic are effective for pain associated with tonsillectomy (11 15). Other reports indicated that From the *Department of Anesthesia and Pain Medicine, Seoul National University Bundang Hospital, Seongnam, Korea; and Department of Anesthesia and Pain Medicine, Seoul National University Hospital, College of Medicine, Seoul National University, Seoul, Korea. Accepted for publication March 12, 2007. Reprints will not be available from the author. Address correspondence to Jung-won Hwang, MD, Department of Anesthesia and Pain Medicine, Seoul National University Bundang Hospital, 300 Gumi-dong, Bundang-gu, Seongnam-si, Korea 463-707. Address e-mail to jungwon@snubh.org. Copyright 2007 International Anesthesia Research Society DOI: 10.1213/01.ane.0000266440.62424.f1 GNB with 0.5% bupivacaine or ropivacaine was ineffective for the reduction of early posttonsillectomy pain in adult and pediatric patients (16 19). However, no report provided a method to assess successful performance of GNB. We hypothesized that an obtunded gag reflex response is expected in patients receiving GNB, and that it might be a useful clinical indicator for evaluation of the extent of GNB. Our prospective, randomized study was designed to assess the efficacy of GNB for the control of pain experienced by adult patients in the immediate postoperative period after tonsillectomy, and to determine whether the extent of obtunded gag reflex as a clinical indicator of GNB correlates with the extent of postoperative pain relief. METHODS After approval of the Institutional Ethics Committee and obtaining written informed consent from each patient, 75 ASA 1 2 physical status adult patients scheduled for tonsillectomy under general anesthesia were recruited into this study (Table 1). Exclusion criteria included diabetes, cardiac conduction anomalies, liver or kidney disease, hypersensitivity to local anesthetics, chronic pain, regular analgesic use within 1 wk of surgery, peritonsillar abscess or swallowing disorder. During the preoperative visit, patients were instructed how to express pain on a 100-mm visual Vol. 105, No. 1, July 2007 267

Table 1. Demographic Data Age (yr) 33 11 36 11 34 9 Gender (M/F) 17/8 15/10 16/9 Height (cm) 167 7 166 9 167 7 Weight (kg) 70 16 67 13 68 12 Operation time (min) 49 19 42 19 41 12 Anesthesia time (min) 75 19 71 22 71 17 Data are expressed as Mean SD. There were no statistically significant differences among groups. : patients received no glossopharyngeal nerve block, : patients received glossopharyngeal nerve block with epinephrine-mixed ropivacaine, : patients received glossopharyngeal nerve block with epinephrine-mixed bupivacaine. analog scale (VAS). All patients were empirically premedicated with 2.0 2.5 mg of midazolam in the reception room, and were noninvasively monitored during surgery. Anesthesia was induced with 2 mg/kg of propofol and 10 g/kg of alfentanil. Rocuronium (0.6 mg/kg) was used to facilitate neuromuscular blockade. After tracheal intubation, anesthesia was maintained with 50% nitrous oxide and 2% 3% sevoflurane. Tonsillectomies were conducted by the same two experienced surgeons, and there was no difference in surgical procedure or study group distribution between surgeons. Peritonsillar infiltration using 1% lidocaine with epinephrine (1.5 2.0 ml per tonsil) was performed by the surgeon before incision. Patients were randomly assigned to one of three groups using a sealed envelope method. At the end of surgery, bilateral GNB was done under direct vision using the McIvor gag with 0.75% ropivacaine with 1:200,000 epinephrine () or 0.5% bupivacaine with 1:200,000 epinephrine (), respectively. Group C did not receive any intervention. All GNBs were performed as described previously (11,12,17) by one anesthesiologist who was not involved in postoperative evaluation of the patients. In brief, a 25G spinal needle was angled to 45 at 1 cm from the tip of the spinal needle. The needle tip pierced the retropharyngeal mucosa at the middle point of the posterior tonsillar pillar (palatopharyngeal fold). The needle was directed behind the posterior tonsillar pillar as laterally as possible and was inserted through the pharyngeal wall about 0.5 1.0 cm in depth. After careful aspiration, 3 ml of study solution was injected slowly. The injection was repeated on the opposite side. After confirming no bleeding at the puncture site, all patients recovered from anesthesia and were transported to the postanesthesia recovery room. The status of each patient was evaluated by one assessor who was blinded as to group allocation. Thirty minutes after surgery, pain at rest and swallowing was assessed using the 100 mm VAS (0: no pain, 100: unbearable) in the recovery room. If pain scores were more than 60 mm at rest, 30 mg of ketorolac was administered IV. Gag reflex response was evaluated by lightly touching the posterior wall of the lower part of the oropharynx with a tongue depressor. The gag reflex response was objectively assessed according to our artificial scale (None: no response, Mild: grimace but tolerable, Moderate: facial flushing, Severe: facial flushing with cough or lacrimation or restlessness). The extent of difficulty in swallowing was also assessed on a 4-point artificial scale (None: normal or no difficulty in swallowing, Mild: mild difficulty in swallowing, Moderate: moderate difficulty in swallowing, Severe: no swallowing or swallowing only with maximal effort). Our evaluation of the degree of swallowing difficulty depended exclusively on the patients answers. Most patients with severe swallowing difficulty were able to swallow with maximal effort. But, five patients (three in and two in ) with GNB reported no swallowing ability despite their maximal effort. In these patients, VAS scores during swallowing were replaced with VAS scores at rest. Any problems related to GNB, such as upper airway obstruction and foreign body sensation in the throat were recorded. On the ward, oral analgesic intake started 8 h after surgery and 600 mg of acetaminophen dissolved in water was administered every 8 h. If oral analgesia was not sufficient for pain control, 30 mg of ketorolac was administered IV as needed. The assessor reviewed all study patients on the ward 8 h after surgery, as well as on the following day, before discharge. Doses of parenteral analgesics administered until 8 h and 24 h after surgery were recorded, and VAS pain assessments were done 8 h and 24 h after surgery. The sample size was determined to have a power of 90% at a 5% significance level to detect a difference of 25 mm in VAS scores during swallowing between and the other groups. The comparison of pain at rest and on swallowing was conducted with a repeated measures ANOVA for time by treatment effect and was followed by ANOVA and Tukey test for comparison of VAS at each time point. ANOVA was used to determine the differences in demographic data. Other data were analyzed using 2 test, when appropriate. Kruskal Wallis test and Tukey test were used to determine a difference in the extent of gag reflex response among groups. Spearman s coefficients were used to determine significant correlations. Values of P 0.05 were considered statistically significant. RESULTS The time-by-gnb treatment interaction for postoperative pain was significant (P 0.002 pain at rest, P 0.001 pain during swallowing) and was different from or (P 0.01 at rest, P 0.001 during swallowing, Table 2). VAS pain scores at rest in were significantly higher 268 GNB for Posttonsillectomy Pain and Gag Reflex ANESTHESIA & ANALGESIA

Table 2. VAS Scores (mm) Measured at Rest and Swallowing at 30 min, 8 h, and 24 h After Surgery P* VAS scores at rest ( 0.01) Postoperative 30 min 42 16 21 17 23 13 0.001 Postoperative 8 h 31 15 20 16 22 16 0.01 Postoperative 24 h 18 8 19 17 20 15 0.676 VAS scores at swallowing* ( 0.001) Postoperative 30 min 62 14 28 22 32 19 0.001 Postoperative 8 63 15 51 19 52 16 0.034 Postoperative 24 h 50 18 51 20 50 16 0.976 No. of injection of ketorolac* Postoperative 30 min 9 1 0 0.001 Postoperative 8 h 6 4 2 0.574 Postoperative 24 h 3 4 3 0.891 No. of cases with VAS more than 50 at rest 30 min 9 2 0 0.001 after surgery No. of cases with VAS more than 50 during swallowing 30 min after surgery 19 5 4 0.001 Visual analog scale (VAS) was used as the pain measuring score. Data are expressed as Mean SD. * Comparison among groups at each time period. Time-by-glossopharyngeal nerve block (GNB) treatment interaction effect. P 0.05 compared with. : patients received no GNB, : patients received GNB with epinephrine-mixed ropivacaine, : patients received GNB with epinephrine-mixed bupivacaine. Table 3. Response of Gag Reflex in the Recovery Room Response of gag reflex * * None 1 13 10 Mild 4 6 7 Moderate 11 0 2 Severe 9 6 6 The response of gag reflex is different among groups (P 0.001 among groups, * P 0.01 compared with ). : patients received no glossopharyngeal nerve block, : patients received glossopharyngeal nerve block with epinephrine-mixed ropivacaine, : patients received glossopharyngeal nerve block with epinephrine-mixed bupivacaine. The response of gag reflex is expressed with four grades. None: no response, Mild: mild discomfort with grimace but tolerable, Moderate: moderate discomfort with facial flushing, Severe: severe discomfort with facial flushing and cough or lacrimation or restlessness. throughout the 24 h study period. There was no significant difference in the VAS of compared to at any time (Table 2). The number of ketorolac injections for additional pain control during the immediate postoperative period was significantly higher in than in or (P 0.001, Table 2). Gag reflex obtundation was more intense in both the R and B groups than in group C (P 0.01, Table 3). In both R and B groups, VAS scores during swallowing in the recovery room were strongly correlated with the extent of obtunded gag reflex (P 0.01, Fig. 1). There were no significant differences in postoperative nausea and vomiting, difficulty in swallowing, foreign body sensation in the posterior pharynx, dyspnea, nasal obstruction, or dry mouth among groups (Table 4). An accidental intravascular injection of epinephrinemixed ropivacaine occurred in one female patient in. As soon as the ropivacaine was injected, the patient s arterial blood pressure and heart rate increased Figure 1. The figure shows that the extent of gag reflex response in (r 2 0.86) and (r 2 0.87) is strongly correlated with VAS scores during swallowing 30 min after surgery (*P 0.01) but, not in (P 0.11). : patients received glossopharyngeal nerve block (GNB) with epinephrine-mixed ropivacaine, : patients received GNB with epinephrine-mixed bupivacaine, : patients received no further intervention. The gag reflex response is expressed in four grades. None: no response, Mild: mild discomfort with grimace but tolerable, Moderate: moderate discomfort with facial flushing, Severe: severe discomfort with facial flushing and cough or lacrimation or restlessness. abruptly. However, no significant sequelae were observed in this patient. No patient developed secondary posttonsillectomy bleeding. Blood-tinged saliva was observed in 15 patients (six patients in, four in, five in ). DISCUSSION This study is differentiated from previous GNB studies by providing a relationship between the extent Vol. 105, No. 1, July 2007 2007 International Anesthesia Research Society 269

Table 4. Morbidity in the Recovery Room Nausea 4 5 5 Vomiting 2 2 2 The extent of swallowing difficulty Severe 7 14 10 Moderate 5 6 8 Mild 8 3 4 None 5 2 3 Foreign body sensation 14 16 18 in throat Dyspnea 0 4 3 Nasal obstruction 3 4 3 Dry mouth 7 11 12 Hoarseness 0 1 0 : patients received no glossopharyngeal nerve block, : patients received glossopharyngeal nerve block with epinephrine-mixed ropivacaine, : patients received glossopharyngeal nerve block with epinephrine-mixed bupivacaine. Swallowing difficulty response is expressed as four grades. None: normal or no difficulty in swallowing, Mild: mild difficulty in swallowing, Moderate: moderate difficulty in swallowing, Severe: no swallowing or swallowing with maximal effort. of diminished gag reflex and the degree of posttonsillectomy pain relief after GNB. To evaluate GNB s effectiveness for postoperative pain control, GNB itself should be performed successfully. We believed that the more successful GNBs resulted in less posttonsillectomy pain, because the glossopharyngeal nerve supplies sensory fibers to the tonsil and peritonsillar area. But whether the glossopharyngeal nerve was completely blocked, partially blocked, or not blocked, in all previous studies (11 19) was not fully investigated (Table 5). Ineffectiveness of GNB for pain control could be expected if local anesthetics did not actually reach the glossopharyngeal nerve terminals corresponding to the tonsillar area. Because the response to gag reflex decreases after successful GNB, the degree of obtunded gag reflex was used as a clinical indicator to assess how successfully the glossopharyngeal nerve was blocked. Our results indeed demonstrated a strong relationship between the extent of the obtunded gag reflex and the extent of posttonsillectomy pain relief (Fig. 1). Not all patients received satisfactory GNB in this study. Twelve patients (six in each and B) with severe gag reflex after GNB had unsatisfactory pain relief. One of the advantages of better postoperative pain control with GNB is less use of parenteral or oral analgesics. There was a significant decrease in the use of ketorolac throughout the perioperative period. We experienced smoother emergence in the groups with GNB, however, this was not evaluated formally. Complications related to local anesthetics in the course of tonsillectomy have not yet been systematically studied. A serious problem related to GNB is upper airway obstruction due to blockade of the vagus nerve proximal to the origin of the recurrent laryngeal nerves by an excessive volume of local anesthetic (17). In this study, mild dyspnea was observed in some patients undergoing GNB and the patients were managed conservatively. This study had some limitations. First, we did not evaluate preoperative gag reflex for baseline data. Patients who have had no gag reflex after tonsillectomy might have no gag reflex even before surgery. It has been reported that the gag reflex is often absent in the normal population (20). Peritonsillar infiltration with epinephrine-mixed lidocaine was performed by surgeons to promote hemostasis during surgery. This Table 5. A Literature Comparison of the Effect of Glossopharyngeal Nerve Block (GNB) or Modified Peritonsillar Infiltration on Posttonsillectomy Pain Author Publication (year) No. of patients Patients population Analgesic comparison Time of initial measurement Result/Conclusion Spence (11)* 1996 100 Adults No comparison ND Satisfactory Bruin (12) 1994 ND Adults, children No comparison ND Satisfactory Naja (13) 2005 170 Adults, children No comparison 0 h Satisfactory Kaygusuz (14) 2003 80 Children Bupivacaine, dexamethasone infiltration versus lidocaine spray versus placebo Giannoni (15) 2001 64 Children Placebo versus ropivacaine versus ropivacaine clonidine Bell (16)* 1997 108 Adults GNB versus iv morphine 1 d Each group had better effect than placebo group Recovery room Each group had better effect than placebo group 10 min GNB was satisfactory only at immediate postoperative period Bean-Lijewski (17) 1997 566 Children GNB versus none ND GNB had more complications about airway obstruction El-Hakim (18) 2000 92 Adults GNV versus placebo versus none Park (19) 2004 130 Children Peritonsillar infiltration versus placebo GNB: glossopharyngeal nerve block. * Study with evaluation for GNB itself. This study described only obtundation of gag reflex, not the extent of obtunded gag reflex. The posttonsillectomy visual analog scale (VAS) scores in patients with unsatisfactory GNB are not described. ND: no description. 4 h Not different 15 min Not different 270 GNB for Posttonsillectomy Pain and Gag Reflex ANESTHESIA & ANALGESIA

may be a confounding factor in the interpretation of our results. Eeritonsillar infiltration could have affected the extent of gag reflex after tonsillectomy, especially in five patients in who had no gag reflex or diminished gag reflex but had high VAS scores. Second, the degree of gag reflex was measured only in the immediate preoperative period. Third, patients in did not receive any placebo, such as normal saline injection. Although one report showed that there was no difference in postoperative pain scores and analgesia consumption between GNB with saline and no injection (18), variations in injection technique may have affected our results. Finally, evaluation of the extent of swallowing difficulty was subjective and dependent on the patient s effort and report. In clinical and practical settings, however, it is very difficult to objectively evaluate the extent of pharyngeal muscle paralysis. In conclusion, this study demonstrated that effective GNB is a useful method for the palliation of postoperative pain after tonsillectomy in adult patients. The extent of an obtunded response to gag reflex correlated with the extent of postoperative pain relief. Therefore, an obtunded gag reflex response may be a clinical indicator for analgesia from GNB. REFERENCES 1. Toma AG, Blanshard J, Eynon-Lewis N, Bridger MW. Posttonsillectomy pain: the first ten days. J Laryngol Otol 1995;109:963 4. 2. Ozalevli M, Unlugenc H, Tuncer U, Gunes Y, Ozcengiz D. Comparison of morphine and tramadol by patient-controlled analgesia for postoperative analgesia after tonsillectomy in children. Paediatr Anaesth 2005;15:979 84. 3. Hanasono MM, Lalakea ML, Mikulec AA, Shepard KG, Wellis V, Messner AH. Perioperative steroids in tonsillectomy using electrocautery and sharp dissection techniques. Arch Otolaryngol Head Neck Surg 2004;130:917 21. 4. Naesh O, Niles LA, Gilbert JG, Ammar MM, Phibbs PW, Phillips AM, Khrapov AV, Robert AJ, McClintock A. A randomized, placebo-controlled study of rofecoxib with paracetamol in early post-tonsillectomy pain in adults. Eur J Anaesthesiol 2005;22:768 73. 5. Valijan A. Pain relief after tonsillectomy. Effect of benzydamine hydrochloride spray on postoperative pain relief after tonsillectomy. Anaesthesia 1989;44:990 1. 6. Jebeles JA, Reilly JS, Gutierrez JF, Bradley EL Jr, Kissin I. Tonsillectomy and adenoidectomy pain reduction by local bupivacaine infiltration in children. Int J Pediatr Otorhinolaryngol 1993;25:149 54. 7. Stuart JC, MacGregor FB, Cairns CS, Chandrachud HR. Peritonsillar infiltration with bupivacaine for paediatric tonsillectomy. Anaesth Intensive Care 1994;22:679 82. 8. Ginstrom R, Silvola J, Saarnivaara L. Local bupivacaineepinephrine infiltration combined with general anesthesia for adult tonsillectomy. Acta Otolaryngol 2005;125:972 5. 9. Somdas MA, Senturk M, Ketenci I, Erkorkmaz U, Unlu Y. Efficacy of bupivacaine for post-tonsillectomy pain: a study with the intra-individual design. Int J Pediatr Otorhinolaryngol 2004;68:1391 5. 10. Apostolopoulos K, Labropoulou E, Samaan R, Bogris K. Ropivacaine compared to lidocaine for tonsillectomy under local anaesthesia. Eur Arch Otorhinolaryngol 2003;260:355 7. 11. Spence AG. Single injection field block of tonsillar fossa. Anaesth Intensive Care 1996;24:621 3. 12. Bruin G. Glossopharyngeal nerve block for tonsillectomy or uvulopalatopharyngoplasty. Can J Anaesth 1994;41:1236. 13. Naja MZ, El-Rajab M, Sidani H, Kabalan W, Ziade MF, Al-Tanni MA. Modified infiltration technique in tonsillectomy: expanded case report of 25 children. Int J Pediatr Otorhinolaryngol 2005;69:35 41. 14. Kaygusuz I, Susaman N. The effects of dexamethasone, bupivacaine and topical lidocaine spray on pain after tonsillectomy. Int J Pediatr Otorhinolaryngol 2003;67:737 42. 15. Giannoni C, White S, Enneking FK, Morey T. Ropivacaine with or without clonidine improves pediatric tonsillectomy pain. Arch Otolaryngol Head Neck Surg 2001;127:1265 70. 16. Bell KR, Cyna AM, Lawler KM, Sinclair C, Kelly PJ, Millar F, Flood LM. The effect of glossopharyngeal nerve block on pain after elective adult tonsillectomy and uvulopalatoplasty. Anaesthesia 1997;52:597 602. 17. Bean-Lijewski JD. Glossopharyngeal nerve block for pain relief after pediatric tonsillectomy: retrospective analysis and two cases of life-threatening upper airway obstruction from an interrupted trial. Anesth Analg 1997;84:1232 8. 18. El-Hakim H, Nunez DA, Saleh HA, MacLeod DM, Gardiner Q. A randomised controlled trial of the effect of regional nerve blocks on immediate post-tonsillectomy pain in adult patients. Clin Otolaryngol 2000;25:413 17. 19. Park AH, Pappas AL, Fluder E, Creech S, Lugo RA, Hotaling A. Effect of perioperative administration of ropivacaine with epinephrine on postoperative pediatric adenotonsillectomy recovery. Arch Otolaryngol Head Neck Surg 2004;130:459 64. 20. Davies AE, Kidd D, Stone SP, MacMahon J. Pharyngeal sensation and gag reflex in healthy subjects. Lancet 1995;345:487 8. Vol. 105, No. 1, July 2007 2007 International Anesthesia Research Society 271