Secondary Logo

Journal Logo

Intravenous Ketoprofen in Thyroid and Parathyroid Surgery

Basto, Emmanuel R. MD; Waintrop, Catherine MD; Mourey, François D. MD; Landru, Jérome P. MD; Eurin, Benoît G. MD; Jacob, Laurent P. MD

doi: 10.1097/00000539-200104000-00047
General Articles: Research Report

We compared the ketoprofen-propacetamol combination relative to propacetamol alone in thyroid and parathyroid surgery in terms of postoperative analgesic efficacy, bleeding, and incidence of nausea and vomiting to determine whether ketoprofen results in any benefit in this type of surgery. Patients were distributed in two parallel groups to be managed by anesthesiologists habitually prescribing (Ketoprofen group) or not prescribing (Control group) ketoprofen in this situation. The same anesthetic technique was used for all patients. Postoperative analgesia consisted of 2 g of propacetamol every 6 h and morphine boluses if the pain score measured by the numerical rating scale pain exceeded 40 (3 mg IV every 10 min in the recovery room, then 5 mg SC every 4 h in the ward). The Ketoprofen group received 100 mg of ketoprofen IV during surgery (starting on resection of specimen) and 8 h later. In the recovery room, patients received oxygen if the Spo2 while they were breathing room air was <95% on admission and at 1 and 2 h. Pain scores, opioid consumption, the volume of the cervical draining fluid, and the concentration and mass of hemoglobin in this fluid collected over 24 h were recorded. The 214 patients were distributed into two groups (n = 107 in each group) that were comparable in terms of age, weight, sex, duration of surgery, type of endocrinopathy, surgeon involvement, and the intraoperative dose of sufentanil (P > 0.2). The Ketoprofen group had lower numerical rating scale (P < 0.05), received less morphine during the first 24 h after surgery (7.4 ± 5 vs 11.7 ± 6 mg, P < 0.05), had fewer nausea and vomiting episodes (21 vs 38, P < 0.05), and were less likely to require oxygen breathing after 1 h in the recovery room (33 vs 59 patients, P < 0.05). The two groups had the same 24-h volume of cervical draining fluid (72.5 ± 43 vs 70 ± 42 mL, P > 0.2) and the same concentration (5.9 ± 3.4 vs 6.4 ± 2.8 g per 100 mL, P > 0.1) and mass of hemoglobin (3.9 ± 2.8 vs 4.2 ± 2.5 g, P > 0.2) in this collected fluid. Two cervical hematomas necessitating reintervention occurred in the Control group, compared with none in the Ketoprofen group. Ketoprofen reduces the pain score after thyroid and parathyroid surgery, as well as morphine requirements and related adverse effects, without increasing the risk of cervical bleeding.

Department of Anesthesiology and Intensive Care, Saint-Louis University Hospital, Paris, France

December 7, 2000.

Implications: In a prospective open study, ketoprofen reduced the pain score after thyroid and parathyroid surgery, as well as morphine requirements and related adverse effects, without increasing the risk of cervical bleeding.

Address correspondence and reprint requests to Emmanuel Basto, MD, Service d’anesthésie-réanimation, Hôpital Saint-Louis-1, Avenue Claude Vellefaux 75475, Paris, Cedex 10, France.

Thyroid and parathyroid surgery induces brief postoperative pain caused by several mechanisms (1). It is not relieved by paracetamol monotherapy in our practice. Opiates can increase postoperative discomfort such as nausea and vomiting (2) and may also increase the risk of respiratory depression (3). Nonsteroidal antiinflammatory drugs (NSAIDs), through a synergistic action with opiates (4–7), can reduce these adverse effects (8,9). The analgesic efficacy of NSAIDs combined with opiates has been well documented (9–16). In thyroid and parathyroid surgery, the efficacy of NSAIDs has not been assessed. We sought to determine in a prospective open study the efficacy of IV ketoprofen in terms of analgesic efficacy, the capacity to reduce the adverse effects of morphine, and postoperative bleeding in thyroid and parathyroid surgery.

Back to Top | Article Outline


The study protocol was approved by our institution’s ethical committee. Consecutive patients programmed for thyroid or parathyroid surgery between December 1998 and July 1999 were eligible for the study. Patients with conventional contraindications to antiinflammatory drugs, secondary or tertiary hyperparathyroidism with renal failure, or clotting disorders, and patients requiring sternotomy for intrathoracic adenomas or goiters, were ineligible. The four anesthesiologists working in the unit differed in their ketoprofen prescribing habits in this setting: two prescribed ketoprofen and two did not. Patients were assigned to two groups (Control or Ketoprofen) according to the order of admission and the anesthesiologists’ duty rosters. Two parallel groups were thus formed over 7 mo. One received propacetamol plus ketoprofen (Ketoprofen group), whereas the other received propacetamol alone (Control group).

During the preoperative anesthetic examination, the numerical rating scale pain (NRS, range from 0 to 100) was explained to all the patients. The anesthetic protocol was identical for the two groups, consisting of propofol induction (2.5 mg/kg) and isoflurane maintenance with 50% nitrous oxide in oxygen. Patients’ tracheae were extubated on the operating table at the end of the surgical procedure. Intraoperative analgesia consisted of sufentanil 0.2 to 0.25 μg/kg during the induction, then 5 μg every 30 min during surgery. Analgesia during the first postoperative 24 h consisted of the IV administration of propacetamol, a precursor of paracetamol, 2 g every 6 h, and morphine according to the NRS score (IV titration in the recovery room with boluses of 3 mg every 10 min until the NRS decreased to <40, and then from the 4th until the 24th postoperative hour, 5 mg SC every 4 h if the NRS was at least 40). After 24 h, the two groups received 1 g of paracetamol orally three times a day. The Ketoprofen group also received 100 mg of ketoprofen IV during surgery, starting on resection of the specimen, i.e., about 20 min before arriving in the recovery room, then again at the eighth postoperative hour.

Capillary hemoglobin oxygen saturation (Spo2) was measured with a Datex Cardiocap pulse oximeter (Datex, Helsinki, Finland) after patients breathed room air for 10 min (or less if Spo2 decreased to <95%). These measures were performed on arrival and at the 60th and the 120th minute in the recovery room. Patients were breathing room air if the Spo2 was at least 95%. They received oxygen via a nasal tube at a flow rate of 3 L/min if the Spo2 was <95%.

Vomiting and nausea, reported spontaneously by the patients, were routinely treated with 10 mg IV metoclopramide and with 4 mg IV ondansetron if they recurred.

The cervical drains, routinely used by the five surgeons participating in the study, were regularly emptied, and the volume of fluid collected by cervical drains was measured cumulatively until permanent removal at the 24th postoperative hour. At the end of the recovery room stay, the drains were emptied for the first time, and the hemoglobin concentration was measured with the HemoCue™ device (HemoCue AB [β-hemoglobin photometer]; M-Pact Corporation, ngelholm, Sweden). The hemoglobin level in the cervical draining fluid was measured at the bedside with the HemoCue device, for convenience, and had been validated, before the beginning of this study, in 40 patients, in whom measurement with the HemoCue device yielded a biased error of 0.6 ± 0.5 g/dL, i.e., inferior to 10% of the simultaneous hospital laboratory measurement (6.2 ± 2.4 g/dL). It yielded a correlation coefficient of 0.94 with the laboratory measurement. Patients were monitored for 2 h in the recovery room and were discharged if they met Aldrete’s criteria.

The following variables were measured: type of endocrinopathy; duration of surgery; respective involvement of the five surgeons; dose of sufentanil received during surgery; NRS every 10 min during the first hour in the recovery room, then on the ward every 4 h for 24 h and at the 36th and the 48th hours; morphine dose administered during the first 24 h after surgery; number of episodes of nausea and vomiting spontaneously reported and treated during the first 24 h; duration of postoperative oxygenation according to a 3-point score (0, no oxygen; 1, oxygenation <1 h; 2, oxygenation >1 h or continued on the ward); and total 24-h volume of cervical draining fluid and the concentration and mass of hemoglobin (volume × concentration).

The measures of all these variables were performed by recovery room nurses and ward nurses unaware of the study and according to usual hospital procedures for thyroid and parathyroid surgery. Data collection for statistical analyses was performed by an investigator after all patients were included.

The number of subjects required to show a lack of difference in bleeding between the two groups was estimated by studying the files of 550 patients operated on in 1997 and 1998 and who did not receive ketoprofen. The volume of cervical fluid collected in drains was 72.5 ± 36 mL, with a 95% confidence interval of 6.2 mL. Thus, to eliminate a real difference of 15 mL (i.e., 20% of mean) not being detected owing to a sampling bias, with a β risk of 10%, it was necessary to include at least 198 patients. Mean values were compared between the two groups by using a two-way analysis of variance with repeated measures with time for the pain scores and a Student’s t-test or a χ2 test for the other comparisons (postoperative morphine dose, nausea and vomiting, oxygen requirements, bleeding). The threshold of significance was set at 5%.

Back to Top | Article Outline


Two-hundred-fourteen ASA physical status I or II patients undergoing thyroid or parathyroid surgery were enrolled and were equally distributed between the Ketoprofen and Control groups. The two groups were similar in age, weight, sex ratio, operating time, type of endocrinopathy, and relative involvement of the five surgeons (Table 1). Pain intensity decreased significantly over time in both groups, but the Ketoprofen group had significantly lower pain scores than the Control group up to the 12th hour (P < 0.05) (Fig. 1). The two groups received the same intraoperative dose of sufentanil (27.1 ± 9 vs 26.3 ± 8 μg, P = 0.53) (Fig. 2). Eight (7.5%) patients in the Control group and 28 (26%) in the Ketoprofen group did not receive morphine (P < 0.05). The Ketoprofen group received less morphine in the recovery room (5.1 ± 4.2 vs 7.7 ± 3.8 mg, P = 0.008) and during the first 24 h (7.4 ± 5 vs 11.7 ± 6 mg, P = 0.009). In the Ketoprofen group, there were fewer episodes of nausea and vomiting after surgery than in the control group (21 [20%] and 38 [36%], respectively;P = 0.039) (Fig. 3). The number of patients not requiring oxygenation on arrival in the recovery room was similar in the two groups (33 vs 38, P = 0.47). The number of patients breathing oxygen for more than 1 h was smaller in the Ketoprofen group (33 vs 59, P = 0.045) (Fig. 3). As shown in Figure 4, 24-h draining fluid in the two groups was similar in volume (72.5 ± 43 vs 70 ± 42 mL, P = 0.23), the hemoglobin concentration (5.9 ± 3.4 vs 6.4 ± 2.8 g per 100 mL, P = 0.12), and the mass of hemoglobin (3.9 ± 2.8 vs 4.2 ± 2.5 g, P = 0.21). Two cervical hematomas necessitating surgical reintervention occurred in the Control group; one compressive cervical hematoma occurred early in the recovery room and required emergency surgical decompression, whereas the other occurred at the 14th hour on the ward and was not compressive. No cases of hematoma occurred in the Ketoprofen group.

Table 1

Table 1

Figure 1

Figure 1

Figure 2

Figure 2

Figure 3

Figure 3

Figure 4

Figure 4

Back to Top | Article Outline


During thyroid and parathyroid surgery, perioperative ketoprofen and propacetamol administration was associated with a reduction in pain scores, morphine requirements, nausea and vomiting, and the duration of postoperative oxygenation relative to propacetamol monotherapy. It was not associated with a increased incidence in local bleeding, assessed during the first 24 hours.

Thyroid and parathyroid surgery induces brief postoperative discomfort (1). The mechanism of this discomfort is complex, being linked to the cervicotomy itself, to the operative cervical hyperextension that causes postoperative muscular cervicalgia, and to the orotracheal intubation that causes postoperative irritation and laryngeal discomfort projecting into the operated region. These sources of discomfort are likely difficult to distinguish from the pain linked to the surgical tissue insult. Postoperative discomfort can also be increased by the cervical drains, which are kept in place for 24 hours (1), and by the very frequent nausea and vomiting in this at-risk population. Finally, this type of surgery is now well standardized and may be considered as a low-risk procedure. As a result, patients’ mean age is increasing and, with it, the risk of respiratory depression (17).

The postoperative pain observed in this study was brief. The pain scores fell markedly between the 24th and the 36th hours in both groups and were nil in most patients between the 36th and the 48th hours (Fig. 1). The pain severity is illustrated by the pain scores recorded in the Control group. Morphine was spared in only 7.5% of Control group patients and in 26% of the Ketoprofen group patients. Morphine was highly effective, because in the recovery room it decreased the pain scores from 54 ± 25 to 25 ± 10 in the Control group (Fig. 1), allowing all the patients to achieve a score <40 with a mean morphine dose of 7.7 ± 3.8 milligrams. The latter was below the mean dose necessary for postoperative pain reported in the literature, which ranges from 12.3 ± 5.1 milligrams in gynecological surgery (18) to 15.3 ± 6.8 milligrams in general surgery (19) and 17.4 ± 2 milligrams in orthopedic surgery (20). The relatively moderate morphine requirements in this type of surgery justified 10-minute intervals between two boluses, rather than the usual 5 minutes. This limited the adverse effects of the opiate, but perhaps at a cost of less rapid pain control (21).

The analgesic synergy of NSAID-opiate combinations was first demonstrated by McQuay et al. (5) and Sunshine et al. (6). Ketoprofen reduces the respiratory depression linked to morphine (3), and the incidence of nausea and vomiting linked to morphine in urologic surgery (7) and pediatric ear-nose-throat surgery (9). The postoperative analgesic efficacy of NSAIDs has been confirmed in many settings, including orthopedic (10), ear-nose-throat (11), maxillofacial (12), dental (13), gynecologic (14), urologic (7), and thoracic surgery (15). The synergism is more controversial in abdominal surgery. Analyzing placebo-controlled trials published before 1991, Dahl and Kehlet (4) found only 7 out of 13 studies in which NSAIDs decreased both pain scores and opiate requirements in patients undergoing abdominal surgery. The efficiency of NSAIDs in thyroid and parathyroid surgery has yet not been assessed.

The pain scores, the number of patients who did not require morphine, and the dose of morphine administered in the recovery room and within the first 24 hours were significantly smaller in the Ketoprofen group, underlining the efficacy of ketoprofen in thyroid and parathyroid surgery. This brief dosing schedule (two injections, 8 hours apart) at the recommended maximal dose (equivalent to 300 mg/d) was adequate, because the pain scores were low after 16 hours and no longer justified opiate treatment in either group. Longer treatment with NSAIDs would have been inappropriate, being needless and carrying, in our study, an unacceptable risk of the adverse effects of NSAIDs.

Thyroid and parathyroid surgery carries a major risk of nausea and vomiting because it is a cervicofacial surgery (22) performed mainly in women (1,2). In a study of 118 patients, the frequency of nausea and vomiting during the 24 hours after thyroid surgery was 54%(2), which is significantly more frequent than the general postoperative incidence of 30%(23). In our series, there were 38 episodes (36%) of nausea and vomiting in the Control group, a relatively low rate compared with that in the series of Sonner et al. (2). This difference is probably linked to the fact that in the latter study, nausea was routinely screened for in a directive interview, whereas we recorded nausea only when spontaneously reported by the patients and treated. Some of our patients probably had mild nausea but thought that it was unnecessary to report it. The reduction in the incidence of postoperative nausea and vomiting in the Ketoprofen group may have been related to the reduction in morphine requirements during this period. Indeed, the two groups of patients did not differ regarding the duration of the operation or the dose of sufentanil administered perioperatively. Ketoprofen and NSAIDs have no known direct antiemetic action.

The number of patients needing to breathe oxygen on arrival in the recovery room was the same in the two groups. The respiratory depression resulting from the anesthesia, and especially from sufentanil, was therefore identical. In contrast, the number of patients needing to breathe oxygen for more than one hour was significantly smaller in the Ketoprofen group. The total dose of IV morphine injected according to the protocol, before the end of the first postoperative hour, was smaller in the Ketoprofen group, thereby reducing the risk of respiratory depression.

Ketoprofen administration may theoretically increase the postoperative bleeding risk (24) and thus the risk of compressive cervical hematoma. This is linked to the modifications of the coagulation system caused by treatments of even very short duration. After 48 hours of oral ketoprofen, there is a dose-dependent inhibition of aggregation platelet and thromboxane formation (16). The bleeding time and platelet aggregation are modified two hours after an injection of 1.4 mg/kg ketoprofen, normalizing 24 hours later (24). Although a retrospective study has shown an increase in perioperative bleeding and wall hematomas in patients treated for long periods with NSAIDs (25), prospective studies have not shown an increase in the risk of postoperative hemorrhage with treatments lasting less than five days and started during or after the operation. For example, a single injection of 75 milligrams of diclofenac or 50 milligrams of indomethacin did not lead to an increase in perioperative bleeding in adults undergoing tonsillectomy (26). Perioperative bleeding was not increased during resection of adenoid growths in children after an injection of ketoprofen, regardless of the dose (9). Finally, postoperative bleeding in drains did not increase after an injection of 200 milligrams of ketoprofen followed by a continuous infusion of 12.5 mg/h for 13 hours in hip and knee arthroplasty (10).

Hematoma of the thyroid lodge requiring surgical investigation is relatively rare, with an incidence of 0.5% to 1.5%(27–29). There are multiple causes, including inadequate surgical hemostasis, drain obstruction or malposition, and a biological clotting disorder. The first two mechanisms could depend in part on the quality of surgical hemostasis (29), and therefore the surgeon’s experience, and also on the amount of cervical tissue dissected or endocrine tissue resected, i.e., the type of endocrinopathy. However, in our study the two groups were comparable regarding the type of endocrinopathy and the involvement of the different surgeons, eliminating an influence of these two equally distributed factors as confounding factors on the relation between the analgesic strategy (use of NSAIDs or not) and postoperative bleeding. Moreover, two studies showed no link between the surgeon’s experience (27) or the volume of endocrine cervical tissue resected (28) and the incidence of compressive cervical hematoma. Given the marked variance in the hemoglobin concentration in cervical draining fluid, we judged it necessary, in order to estimate the bleeding risk, to compare not only the volumes of drained fluid but also the concentration and mass of hemoglobin contained in the draining fluid. The fluid drained over 24 hours was similar in the two groups in terms of its volume and the concentration and mass of hemoglobin. There were two cervical hematomas requiring surgical reintervention (2 of 214, 0.9%), an incidence similar to that in others’ series (27–29). Both events occurred in the Control group, but the difference with the Ketoprofen group was not statistically significant. Ketoprofen at the dose regimen employed in our study did not increase postoperative bleeding (drainage fluid) or the incidence of compressive cervical hematoma.

Interpretation of the data in this study must consider several limitations related to the absence of double-blinded randomization. A double-blinded procedure was not chosen because two anesthesiologists did not wish to use ketoprofen in this setting before the absence of any hemorrhagic risk linked to this drug had been proven. Thus, from an ethical point of view, it was impossible to conduct a controlled double-blinded study. However, we can argue that the study protocol induced no change with respect to usual hospital procedures and that the data were recorded by nurses who were unaware of the study, who followed normal hospital procedures, and who were not surprised by intermittent ketoprofen prescription because it was a usual anesthesiologist-dependent variable.

Patient selection was randomized in a nonconventional way. All consecutive eligible patients were assigned to a study group according to the anesthesiologist’s duty roster. The group assignment was determined by the date of surgery (which was chosen by surgeon and patient and not by investigators). Because neither surgeon nor patient knew the alternative analgesic choice linked to the anesthesia, the assignment was random. We believe that the limitation of this study linked to this lack of conventional randomization was not significant because patients were not specifically chosen. The quality of this randomization was assessed by the same criteria as groups presented in Table 1.

In conclusion, in thyroid and parathyroid surgery, brief analgesia combining ketoprofen with paracetamol reduced postoperative pain scores, morphine requirements, and nausea and vomiting relative to paracetamol alone, without increasing the bleeding risk.

Back to Top | Article Outline


1. Daour R. Thyroidectomie sans drainage. Chirurgie 1997; 12: 408–10.
2. Sonner JM, Hynson JM, Clark O, Katz JA. Nausea and vomiting following thyroid and parathyroid surgery. J Clin Anesth 1997; 9: 398–402.
3. Moren J, Francois T, Blanloeil Y, Pinaud M. The effects of a nonsteroidal antiinflammatory drug (ketoprofen) on morphine respiratory depression: a double-blind, randomized study in volunteers. Anesth Analg 1997; 85: 400–5.
4. Dahl JB, Kehlet H. Non-steroidal anti-inflammatory drugs: rationale for use in severe postoperative pain. Br J Anaesth 1991; 66: 703–12.
5. McQuay HJ, Carrol D, Watts PG, et al. Codeine 20 mg increases pain relief from ibuprofen 400 mg after third molar surgery: a repeat-dosing comparison of ibuprofen and an ibuprofen-codeine combination. Pain 1989; 37: 7–13.
6. Sunshine A, Roure C, Olson N, et al. Analgesic efficacy of two ibuprofen-codeine combinations for the treatment of postepisiotomy and postoperative pain. Clin Pharmacol Ther 1987; 42: 374–80.
7. Tauzin-Fin P, Delort-Laval S, Guenard Y, et al. Etude comparative de la buprénorphine et de son association au ketoprofen et au paracétamol pour l’analgesie postopératoire en chirurgie urologique. Ann Fr Anesth Reanim 1996; 15: 41–6.
8. Beaver WT. Aspirin and acetaminophen as constituents of analgesic combinations. Arch Intern Med 1981; 141: 293–300.
9. Kokki H. I.V. intraoperative ketoprofen in small children during adenoidectomy: a dose-finding study. Br J Anaesth 1998; 81: 870–4.
10. Homeril JL. Ketoprofen for pain after hip and knee arthroplasty. Br J Anaesth 1994; 72: 383–7.
11. Ejnell H. Treatment of postoperative pain with diclofenac in uvulopalatopharyngoplasty. Br J Anaesth 1992; 68: 76–80.
12. Tarkkila P, Tuominem M, Rosenberg PH. Intravenous ketorolac vs diclofenac for analgesia after maxillofacial surgery. Br J Anaesth 1995; 43: 216–20.
13. Norholt SE, Sindet-Pedersen S, Larsen U, et al. Pain control after dental surgery: a double-blind, randomised trial of lornoxicam versus morphine. Pain 1996; 67: 335–43.
14. Rorarius MG. Diclofenac and ketoprofen for pain treatment after elective cesarean section. Br J Anaesth 1993; 70: 293–7.
15. Perttunen K. IV diclofenac in post-thoracotomy pain. Br J Anaesth 1992; 68: 474–80.
16. Stichtenoth DO, Tsikas D, Gutzki FM, Frölich JC. Effects of ketoprofen and ibuprofen on platelet aggregation and prostanoid formation in man. Eur J Clin Pharmacol 1996; 51: 231–4.
17. Sherperel P. Hasards de l’anesthésie pour chirurgie thyroïdienne. In: SFAR, Laxenaire MC, ed. Conférences d’actualisation, 36th congrès national. Paris: Masson, 1994;303–14.
18. Stamer UM, Maier C, Grond S, et al. Tramadol in the management of postoperative pain: a double-blind placebo and action drug controlled study. Eur J Anaesth 1997; 14: 646–54.
19. Butscher K, Mazoit JX, Sami K. Can immediate opioid requirement in the post-anaesthesia care unit be used to determine analgesic requirements on the ward ? Can J Anesth 1995; 42: 461–6.
20. Pico L, Zlotine N, Nègre I, et al. Avantages d’une titration de morphine au bloc operatoire [résumé]. Ann Fr Anesth Reanim 1998; 17: 864(R104).
21. Quaisser K, Mouschin M, Aubrun F, et al. Titration de la morphine IV postopératoire (résumé). Ann Fr Anesth Reanim 1997; 16: 626(R052).
22. Borgeat A. Nausées et vomissements postopératoires. In: SFAR, Conseiller C, ed. Conférences d’actualisation, 38th congrès national. Paris: Elsevier, 1998:33–42.
23. Dawson B, Reed WA. Anaesthesia for day-care surgery: a symposium. III. Anaesthesia for adult surgical out-patients. Can J Anaesth 1980; 27: 409–11.
24. Niemi TT, Taxell C, Rosenberg PH. Comparison of the effect of intravenous ketoprofen and diclofenac on platelet function in volunteers. Acta Anaesthesiol Scand 1997; 41: 1353–8.
25. Connelly CS, Panush RS. Should nonsteroidal anti-inflammatory drugs be stopped before elective surgery? Arch Intern Med 1991; 151: 1963–6.
26. Rorarius MG, Baer GA, Siirtola M, et al. Effect of intravenous diclofenac or indomethacin on the emergence from anaesthesia for tonsillectomy. Acta Anaesthesiol Scand 1993; 37: 616–21.
27. Shaha A, Jaffe BM. Complications of thyroid surgery performed by residents. Surgery 1988; 104: 1109–14.
28. Bergamaschi R, Becouarn G, Ronceray J, Arnaud JP. Morbidity of thyroid surgery. Am J Surg 1998; 176: 71–5.
29. Defechereux T, Hamoir E, Nguyen Dang D, Meurisse M. Le drainage en chirurgie thyroïdienne: est-ce toujours une necessité ? Ann Chir 1997; 51: 647–52.
© 2001 International Anesthesia Research Society