Inspired Oxygen Fraction of 0.8 Does Not Attenuate Postoperative Nausea and Vomiting after Strabismus Surgery

With approval of our institutional review board (Ethik-Komission der Medizinischen Fakultät der Universität Duisburg-Essen, Essen, Germany) and written informed consent, we studied 210 patients (aged 5–79 yr, American Society of Anesthesiologists physical status I or II) scheduled to undergo strabismus surgery. Patients with a predisposition for malignant hyperthermia, previous antiemetic treatment, psychiatric disorders, or pregnancy were excluded from the study.

On the day of surgery, patients were premedicated with midazolam orally (patients with a weight of less than 30 kg received 0.5 mg/kg liquid midazolam; patients with a weight above 30 kg received 7.5-mg tablets). Anesthesia was standardized. Atropine (0.01 mg/kg), alfentanil (0.02–0.04 mg/kg), etomidate (0.15–0.3 mg/kg), and mivacurium (0.1 mg/kg) were administered for induction of general anesthesia. Alternatively, inhalational induction with sevoflurane was performed if the anesthesiologist in charge of the case considered it appropriate. After placement of an intravenous catheter, these latter patients then received the same doses of intravenous anesthetics.

The trachea was intubated, a gastric tube was placed, and, regardless of age, all patients received paracetamol (20 μg/kg) rectally for analgesia. Anesthesia was maintained using sevoflurane (1–1.5 minimal alveolar concentration). Additional alfentanil was administered at the decision of the anesthesiologist in charge of the case. Mechanical ventilation was adjusted to maintain end-tidal carbon dioxide partial pressure at 35–40 mmHg. Inspired oxygen fraction was chosen according to randomization. Before extubation, all patients received 50% oxygen for approximately 2 min, and the gastric tube was removed under suction. Patients were hydrated using Ringer’s lactate (2–4 ml·kg⁻¹·h ⁻¹intravenously) during surgery.

Appropriate morphometric and demographic characteristics, and potential confounding factors likely to influence PONV, such as current smoking, history of motion sickness, or PONV after previous operations, were recorded. Intraoperative variables (heart rate, blood pressure, oxygen saturation, end-tidal sevoflurane, inspired oxygen fraction) were recorded 10 min after induction of general anesthesia, as were episodes of hypotension or bradycardia.

Patients were interviewed preoperatively and 6 and 24 h postoperatively by an investigator unaware of treatment assignment. Vomiting was categorized as none, one, two or three, or more than three episodes, and severity of nausea and postoperative pain were each rated on a four-point scale as none, mild, moderate, or severe.5Patient satisfaction was measured using a four-point scale (1 = very satisfied; 4 = not satisfied). Use of rescue antiemetic medications and postoperative pain treatment were also recorded.

Before induction of general anesthesia, patients were randomly assigned to one of three treatments: (1) 30% oxygen in air plus intravenous administration of saline, (2) 80% oxygen in air plus intravenous administration of saline, or (3) 30% oxygen in air plus intravenous administration of ondansetron (75 μg/kg) during induction. Computer-generated randomization codes were kept in sealed, opaque envelopes.

Regardless of group assignment, patients with PONV received dimenhydrinate (1–3 mg/kg rectally) on request. If nausea and vomiting persisted, droperidol was administered intravenously (25 μg/kg). All patients were allowed to drink as soon as they were awake. For postoperative analgesia, patients received additional paracetamol on request. If pain persisted, the opioid piritramide (0.1 mg/kg intravenously) was administered.

Our primary, prospectively defined outcome variable was the incidence of nausea and/or vomiting over the initial 24 postoperative hours. For analysis, any score for nausea or vomiting exceeding “none” was considered positive. The incidence of nausea and vomiting and the potential impact of confounding factors were evaluated by univariate and multivariate statistics. The primary null hypothesis was that inspired oxygen fractions of 0.3 versus  0.8 do not result in statistically significant differences in the incidence of PONV. One-way analysis of variance, Kruskal-Wallis test, and chi-square tests were used for univariate analyses, and multivariate regression was performed to evaluate potential contributions of confounding factors on the combined incidence of nausea and vomiting.

Results are presented as number of patients, percentage, mean ± SD, median and range, or odds ratio and 95% confidence interval. An α-error P  of less than 0.05 was considered statistically significant. Prestudy power analysis had determined that a minimum of 53 patients/group would be necessary to have a 90% chance of detecting a 50% relative reduction in PONV from a presumed prevailing incidence of 60%. To allow some margin, we decided to study 210 patients.

During a 1-yr period, 373 patients were scheduled to undergo strabismus surgery, 106 refused to participate in the study, and 55 did not meet inclusion criteria. Two patients consented but were not included for medical reasons. Therefore, data sets from 210 patients were analyzed (table 1).

Nausea and vomiting was worst in the first six postoperative hours, and its incidence subsequently decreased in all groups. The overall incidence of nausea and vomiting during the postoperative 24-h observation period was 41% in the 30% oxygen plus saline group, 38% in the 80% oxygen group, and 28% in the 30% oxygen plus ondansetron group (P = 0.279). Therefore, there was no statistically significant difference in PONV incidence among groups (table 2).

Compared with placebo, ondansetron tended to decrease the incidence of both nausea and vomiting, with a greater impact on vomiting. However, these differences also were not statistically significant (table 2).

The incidence of oculocardiac reflex was not associated with an increased incidence of PONV. There were no differences in postoperative pain scores, use of postoperative analgetic drugs, or patient satisfaction. Few patients requested antiemetic rescue medication, and the use of antiemetic rescue medications did not differ significantly among groups (table 2).

Multivariate logistic regression confirmed that treatment did not significantly influence PONV. In contrast, age, PONV after previous operations, and current smoking had a highly significant influence on the incidence of PONV during the initial 24 h after strabismus surgery (table 3). In our population, younger patients, patients with previous PONV, and nonsmokers were more likely to experience PONV.

There were no statistically significant differences among groups with regard to demographic, morphometric, or intraoperative parameters. In particular, the dosage of alfentanil also did not differ significantly among the groups. Furthermore, the incidence of hypotension (systolic blood pressure < 80 mmHg) and oculocardiac reflex as well as other potential confounding factors, such as current smoking, motion sickness, or previous PONV, were comparable among groups (table 1).

An inspired oxygen fraction of 0.8 during alfentanil–sevoflurane anesthesia did not decrease the incidence of PONV in a large cohort of patients undergoing strabismus repair.

Our study was designed to include patients regardless of age, representing the whole population undergoing strabismus repair in our institution, because ocular surgery evokes emesis in both adults and children. We used sevoflurane for maintenance of anesthesia because of its fast kinetics and to be able to provide inhalational induction if necessary. In the case of inhalational induction, patients received the same intravenous induction agents after placement of an intravenous catheter. Therefore, our study design provided a highly standardized anesthesia.

The incidence of PONV after sevoflurane anesthesia is comparable to other volatile anesthetics, such as halothane, which is frequently used in studies on pediatric strabismus repair.2,7However, halothane is mostly combined with nitrous oxide, the latter increasing the incidence of PONV.8Still, the 38% incidence of PONV in the 30% oxygen plus placebo group was less than expected, and therefore, statistical power was slightly decreased. However, a presumed oxygen effect, if present, still would have been detectable.

The incidence of PONV after colon resection5or gynecologic laparoscopy6has been halved by administration of 80% oxygen, decreasing nausea more than vomiting. Furthermore, an increased inspired oxygen fraction has been shown to decrease symptoms of motion sickness when applied during ambulance transport of duration similar to anesthesia duration in our study.9In our study, however, incidences of PONV or nausea did not differ between the 80% and the 30% oxygen plus placebo groups (PONV: 41% vs.  38%; nausea: 38% vs.  37%, respectively).

The impact of an increased inspired oxygen fraction on PONV is controversial. During abdominal surgery, oxygen presumably counteracts the consequences of potential intestinal hypoperfusion and ischemia with subsequent release of vasoactive hormones, i.e. , 5-hydroxytryptamine, or other emetogenic substances.5,6Furthermore, hyperoxia per se  has been suggested to affect the dopamine-sensitive chemotactic trigger zone by decreasing the tonic release of dopamine mediated via  carotid body afferent nerve activity.9However, currently, there are no data available to strengthen either assumption.

The extent to which these mechanisms contribute to PONV also remains questionable because another study found no benefit of 80% oxygen in patients undergoing ambulatory laparoscopy.10Furthermore, in patients undergoing thyroid resection or breast surgery, an increased inspired oxygen fraction also did not diminish PONV.11,12Certainly, splanchnic hypoperfusion or ischemia are unlikely to occur during thyroid, breast, or strabismus surgery.

During strabismus surgery, any episodes of hypotension due to the oculocardiac reflex and bradycardia are short lasting, and no data exists as to whether oculocardiac reflex results in mediator release. The incidence of oculocardiac reflex was slightly higher in the 80% oxygen group (50%) as compared with the 30% oxygen group (42%) and ondansetron group (33%). However, these differences were not statistically significant. Most studies13–15found no evidence of an association between oculocardiac reflex and PONV. Therefore, because retrobulbar neural blockade does not diminish the incidence of PONV,16other mechanisms are relevant for evoking PONV after strabismus surgery, such as type of anesthetics, surgical techniques, or patient characteristics.14,17Accordingly, in our study, age, history of PONV, and smoking significantly influenced the risk of PONV after strabismus surgery.

Mechanisms involving the 5-hydroxytryptamine receptor system seem of importance for PONV after strabismus surgery, because prophylactic ondansetron, a 5-hydroxytryptamine antagonist, decreases PONV.1Several studies on strabismus surgery have reported a significant decrease of PONV incidence after prophylactic administration of ondansetron.18–21Although not statistically significant, history of previous PONV was slightly higher (26%) in the 80% oxygen group as compared with 21% in the 30% oxygen group and 14% in the ondansetron group. Therefore, according to the Apfel score,22approximately 10% of patients in the ondansetron group had a slightly reduced risk for PONV. In our study setting, ondansetron tended to decrease the incidence of PONV compared with placebo (28% vs.  41%), mostly by decreasing vomiting (24% vs.  13%), which is a common finding for this drug.23–25However, these differences were not statistically significant. Nearly all studies on ondansetron have used nitrous oxide,1,19–21which we omitted. Furthermore, we used alfentanil, which has less proemetic effect on PONV incidence than other opioids,26and administered paracetamol rectally to all our patients, which might have caused less need for postoperative opioids and reduced PONV incidence.27Therefore, ondansetron was less effective in our study settings. Nevertheless, we used ondansetron to compare its effects to the magnitude of effects of increased inspired oxygen, if present. In women after laparoscopy ondansetron was less effective than oxygen in preventing PONV.6In our patients after strabismus surgery, although ondansetron tended to decrease the incidence of PONV compared with placebo (28% vs.  41%), 80% oxygen had no effect (38%). Therefore, an inspired oxygen fraction of 0.8 was not superior to ondansetron in decreasing PONV incidence.

In conclusion, an inspired oxygen fraction of 0.8 during alfentanil–sevoflurane anesthesia for strabismus surgery does not decrease PONV. Possibly, increased inspired oxygen is not as efficient during superficial tissue surgery as during colonic surgery.