|Home | About | Journals | Submit | Contact Us | Français|
Supplemental intra-operative oxygen (80%) halves the incidence of nausea and vomiting after open and laparoscopic abdominal surgery, perhaps by ameliorating the subtle intestinal ischemia associated with abdominal surgery. It is unlikely that thyroid surgery compromises intestinal perfusion. We therefore tested the hypothesis that supplemental perioperative oxygen does not reduce the risk of postoperative nausea and vomiting (PONV) after thyroidectomy. As a positive control, we simultaneously evaluated the anti-nausea efficacy of droperidol. One hundred and fifty patients undergoing thyroidectomy were given sevoflurane anaesthesia. After induction, patients were randomly assigned to the following treatments: 1) 30% oxygen, balance nitrogen; 2) 80% oxygen, balance nitrogen; or 3) 30% oxygen with droperidol 0.625 mg. The overall incidence of nausea during the first 24 postoperative hours was 48% in the patients given 30% oxygen, 46% in those given 80% oxygen, and 22% in those given droperidol (P = 0.004). There were no significant differences between the 30% and 80% oxygen groups in incidence or severity of PONV, the need for rescue anti-emetics, or patient satisfaction. Droperidol significantly shortened the time to first meal. Supplemental oxygen was ineffective in preventing nausea and vomiting after thyroidectomy, but droperidol halved the incidence.
Postoperative nausea and vomiting (PONV) is among the five problems most feared by surgical patients.1 Uncontrolled PONV remains the leading cause of delayed discharge or unexpected admission after ambulatory surgery.2 The incidence of PONV varies widely and depends on numerous non-anaesthetic and anaesthesia-related factors.3–5 However, the risk is substantial in most populations.
The incidence of PONV after open colectomy is halved in patients receiving 80% perioperative oxygen compared to those receiving 30%.6 Supplemental oxygen has also been shown to halve the risk after laparoscopic gynaecologic surgery in one study7 whereas another found no benefit.8 The mechanism by which supplemental oxygen reduces PONV remains unknown, but may be related to hyperoxia ameliorating subtle intestinal ischemia secondary to bowel manipulation and/or compression. Any reduction in bowel ischemia would presumably decrease release of emetogenic substances, 7 9 and therefore, the risk of PONV.
The risk of PONV varies widely as a function of the surgical procedure. For example, PONV is common after thyroid surgery.3 10 11 The specific aetiology remains unknown, but is probably not related to perioperative intestinal ischemia. We therefore tested the hypothesis that supplemental perioperative oxygen does not reduce the risk of PONV after thyroidectomy. As a positive control, we simultaneously evaluated the anti-nausea efficacy of droperidol 0.625 mg.
With approval from the Institutional Ethics Committee at CHU de Liège and written informed consent, 150 patients scheduled for elective thyroid surgery were included in this prospective, double blind study. Enrolment was restricted to patients aged 19 to 70 years old who were American Society of Anesthesiology physical status I–III. Exclusion criteria included obesity (> 150% ideal body weight), gastro-oesophageal reflux, history of motion sickness, previous PONV, preoperative vomiting, antiemetic therapy, and any pulmonary disease.
Patients were informed preoperatively that pharmacologic treatment for nausea and vomiting would be available postoperatively upon request. Patients were given alprazolam 0.5 mg and hydroxyzine 25 mg orally 1 h before transfer to the operating theatre. Anaesthesia was induced with propofol 2 mg kg−1, sufentanil 0.2 μg kg−1, and rocuronium 0.6 mg kg−1. Anaesthesia was subsequently maintained with sevoflurane at a concentration that was adjusted to maintain arterial blood pressure within 20% of pre-induction values.
After induction of anaesthesia, patients were randomly assigned to the following three groups: 1) routine oxygen administration with 30% oxygen, balance nitrogen, (30% O2, n = 50); 2) supplemental oxygen administration with 80% oxygen, balance nitrogen (80% O2, (n = 50); and, 3) droperidol 0.625 mg immediately after induction of anaesthesia, combined with 30% oxygen, balance nitrogen (droperidol, n = 50). Patients and surgeons were blinded as to group assignment.
Crystalloid was given at a rate of 8 ml kg−1 h−1 throughout surgery. No patients required antagonism of neuromuscular relaxation. Inspired oxygen was increased to 100% for extubation. The patients were subsequently given oxygen at the designated concentration (30% or 80%, balance nitrogen) via facemask for 2 h at a rate of 9 L min−1. After two postoperative hours, the facemask was removed. We used a facemask with non-rebreathing valves and a 500 ml reservoir (Trident Medical International, Copenhagen, Denmark).
Postoperative analgesia was standardized and consisted of intravenous administration of propacetamol 2 g, which is endogenously converted into of paracetamol 1 g (acetaminophen in USA; Pro-Dafalgan®, UPSA Medica, Belgium) and tramadol 2 mg kg−1 30 min before the end of surgery. Upon arrival in the post-anesthetic care unit, piritramid, a synthetic opioid, (Dipidolor®, Janssen Pharmaceutica, Beerse, Belgium) was titrated intravenously by post-anaesthetic care unit nurses to provide satisfactory analgesia. These nurses were blinded to group assignment.
Thirty min after the end of surgery, provided there was no evidence of postoperative bleeding, ketorolac 30 mg was given IV followed by a continuous infusion of 30 mg over 6 h. Propacetamol (2 g, IV), paracetamol (1 g, orally), and tramadol (100 mg, IV or orally) were given as rescue medications.
Nausea and vomiting were treated, upon request, with alizapride 50 mg; treatment continued with tropisetron 2 mg, a selective 5-HT3 receptors antagonist, if PONV persisted more than 30 min.
Morphometric and demographic characteristics of the participating patients were recorded. Details of anaesthetic and surgical management were similarly recorded.
Postoperative pain scores were evaluated with 100-mm visual analogue scales 2, 6, and 24 h after surgery. Pain therapy was recorded during the first 24 h after the end of surgery. Patients were allowed to drink 4 h after surgery and, if tolerated, to eat after an additional hour. The postoperative times until the first drink and solid intake were recorded. Patient satisfaction was scored with 100-mm visual analogue scales, with 100 mm indicating complete satisfaction.
Blinded observers recorded the incidence and the severity of nausea and vomiting during the three following postoperative periods: 0–2 h, 2–6 h, and 6–24 h. Severity of nausea was assessed on a four-point scale as none, mild, moderate, or severe. Retching was recorded as vomiting.
Continuous variables were analyzed by one-way ANOVA or Kruskal-Wallis tests. Categorical variables were analyzed using Chi square analysis. Results are presented as means ± SDs, percentages, or medians and 10–90 percentile intervals; P < 0.05 was considered statistically significant.
Morphometric and demographic characteristics of the patients in all groups were similar except for age; the patients in the 30% O2 group were younger than those in the other groups were. Anaesthetic and surgical management (Table 1), postoperative pain (Table 2), and analgesic consumption were also similar in the groups. Six patients in the 30% O2 group and two in the droperidol group had O2 saturation below 95% after the operation. These patients were stimulated and encouraged to breath deeply. SpO2 increased spontaneously above 95% without giving oxygen supplement.
When the entire 24-hour observation period was considered, droperidol halved the incidence of nausea compared to the two other groups (Table 2, P = 0.004). The incidence of nausea and vomiting in the 30% and 80% O2 groups did not differ significantly (Fig. 1). This study achieved a 90% power to detect a 50% reduction in the rate of postoperative nausea and vomiting between the 30% oxygen group and the 80% oxygen group at an alpha level of 0.05.
In the droperidol group, time to the first meal was significantly shorter than in the 30% and 80% oxygen groups. Patient satisfaction was comparable in all groups (Table 3). Patient satisfaction was significantly greater in patients with no PONV as compared to patients with PONV: VAS scores of 91 [70–100] vs. 80 [44–98] (medians [10–90 percentile intervals], P < 0.006).
This study failed to observe any benefit of supplemental oxygen on PONV after thyroidectomy and confirmed the antiemetic efficacy of droperidol.
Gut manipulation during colon surgery and the associated splanchnic vessel dissection surely results in at least some intestinal ischemia.12 Indeed, the stress of major surgery per se stimulates the sympathetic nervous system and can substantially reduce splanchnic blood flow. Even slight intestinal ischemia triggers substantial release of 5-hydroxytryptamine and other emetogenic factors.13 Supplemental perioperative oxygen presumably reduces the amount of ischemic intestinal tissue and thus, the risk of nausea and vomiting.
Pneumoperitoneum, which is required for laparoscopic procedures, also induces significant hemodynamic changes and release of vasoactive hormones14 15 that can reduce intestinal perfusion.15–17 However, the extent to which laparoscopy impairs intestinal perfusion may depend on subtle details of surgical technique — perhaps explaining why supplemental oxygen was beneficial in one study of laparoscopic gynaecologic surgery7 but not in another.8
Thyroid surgery differs from abdominal surgery in that intestinal ischemia is unlikely because the surgical procedure does not directly involve the splanchnic circulation. Furthermore, the risk of significant bleeding is low. Hypotension, and consequent intestinal hypoperfusion, is thus rare during thyroid surgery. That supplemental oxygen did not reduce the risk of nausea and vomiting is consistent with the theory that oxygen prevents PONV by ameliorating subtle intestinal ischemia rather than by acting directly at the chemotactic trigger zone.
The difficulty with this theory is a recent study in which prophylactic supplemental oxygen was found to reduce the incidence of nausea and vomiting in victims of minor trauma who developed motion sickness during ambulance transport.18 It is unlikely that these patients had intestinal ischemia, suggesting that the benefits of supplemental oxygen were mediated by another mechanism. The most likely is that reduction of nausea by oxygen depends on dopamine release from the carotid bodies. The carotid bodies tonically release dopamine in amounts that are inversely related to arterial oxygen tension.19 The importance of this observation is that the chemotactic trigger zone is sensitive to dopamine as well as serotonin.20 Hyperoxia per se may thus reduce nausea and vomiting via a dopamine-dependent mechanism.
That then leaves the question of why oxygen proved ineffective in patients recovering from thyroid surgery. The most likely explanation is that nausea and vomiting in these patients was mediated by yet another mechanism. There are several possibilities. For example, intraoperative dissection of the recurrent laryngeal nerve(s) may contribute because this nerve is a branch of the vagal nerve that is highly involved in emetic reflexes. PONV after thyroidectomy may also result from nociceptive reflexes originating from the pharynx and larynx.3 (Pharyngeal pain, exacerbated by swallowing, is typical after thyroid surgery.) To the extent that PONV results from throat pain, it is unlikely to be affected by hyperoxia. Consistent with this theory, pain scores and analgesic consumption was similar in each of our three treatment groups. Postoperative nausea and vomiting has numerous aetiologies and many factors are known to influence the incidence of this complication. It is therefore unsurprising that a treatment effective for some aetiologies would fail in other circumstances.
We found that droperidol significantly reduced the incidence of PONV. This study thus confirms that droperidol prevents PONV.21–23 For example, low-dose droperidol (0.625 mg) compares favourably with ondansetron (4 mg) 22–23 — although droperidol is considerably less expensive than the 5-HT3 antagonists. The incidence of side effects are reportedly similar with each class of drugs.24–26 These data suggested that the least expensive was probably preferable for routine use. However, the Food and Drug Administration has recently added a “black box” warning to the label for droperidol because the drug has been associated with life-threatening QT prolongation. However, there is considerable question in the anaesthesia community about the justification for this warning.27
Patients were allowed to drink 4 h after surgery, and most did at that time or shortly thereafter. However, emetic episodes were triggered by drinking water in the 30% and 80% oxygen groups. This then delayed their first meal. As a result, droperidol not only reduced the risk of PONV, but also allowed patients to tolerate solid food significantly more quickly.
In summary, 80% perioperative oxygen was ineffective in preventing PONV after thyroidectomy. A difference in the aetiology of PONV after thyroidectomy and abdominal surgery may explain the discrepancy between our results and those reported after abdominal surgery. We also observed that droperidol 0.625 mg, a drug that is inexpensive but possibly associated with serious side effects, significantly reduced the incidence of PONV and shortened the time for first meal.
Presented in part at the annual meeting of the American Society of Anesthesiologists, New Orleans, October 13–17, 2001.
Supported in part by NIH Grants GM 58273 and GM 061655 (Bethesda, MD) and the Commonwealth of Kentucky Research Challenge Trust Fund (Louisville, KY). None of the authors has any personal financial interest related to this research