Secondary Logo

Journal Logo

Pediatric Anesthesiology: Research Reports

An Anesthetic Management Protocol to Decrease Respiratory Complications After Adenotonsillectomy in Children with Severe Sleep Apnea

Raghavendran, Sreekrishna, MD*; Bagry, Hema, MD*; Detheux, Gregory, MD*; Zhang, Xun, PhD; Brouillette, Robert T., MD; Brown, Karen A., MD, FRCP(C)*

Section Editor(s): Davis, Peter J.

Author Information
doi: 10.1213/ANE.0b013e3181cfc435

Adenotonsillectomy is usually considered an uncomplicated surgery with a low risk for respiratory morbidity. However, postoperative respiratory complications occur more frequently when certain risk factors are present. These risk factors include age younger than 3 years, comorbidities such as airway anomalies, Down syndrome, and neuromuscular disease, and severe obstructive sleep apnea. A high apnea hypopnea index and repetitive episodes of profound hypoxemia during sleep increase the risk for postadenotonsillectomy respiratory morbidity at least 20-fold.14 Following reports that the obstructive sleep apnea syndrome (OSAS) may also negatively affect growth, neurocognitive function, and cardiovascular physiology,510 we implemented, in 2001, a clinical management protocol recommending expedited adenotonsillectomy and postoperative monitoring in the pediatric intensive care unit (PICU) for children with severe OSAS. However, children with the most severe OSAS, as indicated by repetitive decreases in hemoglobin saturation <80%, continued to experience a high incidence of major respiratory complications.11

We hypothesized that anesthetic technique was a potential factor contributing to this postoperative respiratory morbidity. Hence, our multidisciplinary group revised the perioperative guidelines to recommend a reduction in opioid use and administration of dexamethasone and atropine, specifically targeting children with severe OSAS. We compared data from 2 time periods: 2001 to 2002 (prerevision) and 2002 to 2006 (postrevision), focusing the analysis on children who demonstrated severe recurrent hypoxia on the perioperative sleep study.


Diagnostic Methods for OSAS

Evaluation of sleep-disordered breathing was coordinated through our Sleep Laboratory, and a detailed description of our home oximetry method has been reported in previous publications.1215 Before oximetry testing, the parents completed a questionnaire that included demographic data, information about their child's past and present medical conditions, previous surgery, including adenotonsillectomy, and questions regarding their child's sleep pattern. These questions included those about sleep, breathing during sleep, and the degree of parental concern about breathing during sleep. Particular care was directed to the responses given for 3 questions concerning (1) difficulty breathing during sleep, (2) obstructive apnea observed by parents, and (3) frequency of snoring during sleep. After the parents received 30 minutes of instruction on oximetry testing, they took the oximeter home, conducted the study that same night, and returned the oximeter the following morning. Parents also completed a sleep log of their child's nocturnal behavior.

We used a motion-resistant pulse oximeter set with a 2-second averaging time for hemoglobin saturation (Masimo Radical, Version, Irvine, CA). Saturation and pulse rate data were extracted and analyzed using software (Download 2001, Version 2.5.0, Stowood Scientific Instruments, Oxford, England). This program provided a number of metrics for oxygenation, including the nadir saturation. Periods of oximetry recording were excluded from analysis if the oximeter quality signal indicated low signal, low perfusion, unrecognized, defective, or no sensor, interference, and ambient light.

An oximetry test was considered positive for OSAS if it had 3 or more desaturation clusters <90%. An oximetry study was assigned a McGill Oximetry Score (MOS) of 2, 3, or 4, based on the presence of at least 3 desaturations less than 90%, 85%, and 80%, respectively.11 A MOS of 2 or higher was considered diagnostic of OSAS. An oximetry record not meeting the above criteria was inconclusive (MOS1), and in these children, the diagnosis of OSAS was established with polysomnography (Sandman, Pleasanton, CA). To establish a diagnosis of OSAS by polysomnography, we chose an objective measure, the polysomnographically determined mixed/obstructive apnea/hypopnea index ≥1 event per hour.14 Children with values <1 were considered to be normal or to have simple snoring.16

The Revised Guidelines for Adenotonsillectomy in 2002 to 2006

Both guidelines recommended the expedited scheduling of adenotonsillectomy for children assigned MOS4 and admission of these children to the PICU in the postoperative period. Revisions in the guidelines focused on the administration of opioids, dexamethasone, and atropine (see Discussion for rationale). For children assigned MOS4, the guidelines advised a reduction in the intraoperative opioid dosage, but because the choice of intraoperative opioid varied from IM codeine to remifentanil, the dose was not specified. In addition, the guidelines recommended a reduction in the increment of IV morphine administered in the postoperative period. The perioperative guidelines for adenotonsillectomy specify a postoperative oral codeine (dose 1 mg · kg−1) to be administered regularly. The revised guidelines, in 2002–2006, recommended that in the MOS4 group, postoperative oral codeine should be administered only for a complaint of pain and therefore should be prescribed pro re nata (PRN). The revised guidelines recommended the administration of IV dexamethasone (0.3 mg · kg−1, maximum 10 mg) and atropine after induction of anesthesia.

The guidelines were discussed at departmental rounds and meetings, published in our anesthesia pocket manual, posted in the postoperative units, and implemented in October 2002. In addition, newly arriving anesthesia and surgical personnel, both resident and staff, were individually mentored regarding the new guidelines.

Method of Recruitment of Patients

Time Period 2002 to 2006 (Current Data)

The retrospective review received institutional approval. We retrieved consecutive cases from the hospital administrative database coded by surgical procedure for adenotonsillectomy with or without myringotomy and tube insertion. The hospital administrative database was also queried for both visits to the emergency room (ER Visit) and readmission to the hospital (Readmission) within 30 days of surgery.

Inclusion criteria were (1) OSAS and (2) documentation of a MOS. Exclusion criteria were (1) prior recruitment to research studies,11,17 (2) surgical procedures additional to adenotonsillectomy/myringotomy and tube insertion, (3) a major medical intervention (MMI) in the perioperative period, (4) delayed extubation of the trachea, and (5) discharge from hospital on the day of surgery.

The charts of 591 children who underwent adenotonsillectomy between October 11, 2002 and February 24, 2006 were reviewed. Two hundred ninety-eight children with OSAS had been assigned a MOS. Six children were excluded because they required an MMIRespiratory in the perioperative period leaving a study population of 292 children, of whom 97 had been assigned MOS4.

Time Period 2001 to 2002 (Historic Data)

Nixon et al.11 followed up prospectively 230 children referred to the sleep laboratory for evaluation between October 1, 2001 and September 30, 2002. The details of anesthetic management for those children who eventually underwent adenotonsillectomy had been recorded by K. Brown in a separate retrospective review that had also received institutional approval. In the relevant publication, Nixon et al.,11 publishing in a nonanesthesia journal, did not report the details of this anesthetic management. This database was, however, available for comparison. In 6 children, all assigned MOS1, the adenotonsillectomy was performed after October 2002 and therefore overlapped the study period of the new guidelines. These children were reported by Nixon et al. and therefore were allocated to the 2001–2002 database. One hundred twenty-six children from the 2001–2002 time period met the inclusion criteria, and 27 children had been assigned MOS4. The surgical technique for adenotonsillectomy was an electrocautery dissection technique.

Because the choice of intraoperative opioid varied, opioid doses were converted to morphine equivalents: codeine = 0.08, demerol = 0.1, fentanyl = 100, hydromorphone = 5, remifentanil = 0, and sufentanil = 1000.18 The intraoperative morphine equivalent (IOME) doses were expressed in mg · kg−1. Postoperative pain was assessed with the Children's Hospital of Eastern Ontario Pain Scale (CHEOPS).19 Pain was classified as mild, moderate, and severe based on the CHEOPS score recorded on admission to the postoperative unit, <6, 6 to 8, and >8, respectively. The cumulative postoperative dose of morphine, administered in the initial hour of recovery to achieve comfort, was recorded and expressed as mg · kg−1. If the postoperative oral codeine prescription was PRN, the codeine regimen was classified as PRN.


The primary outcome was an MMI for a respiratory event. MMIRespiratoy were identified from written comments in the anesthetic record and medical dossier. An MMIRespiratoy was defined as an intervention requiring a physician, such as instrumentation of the airway, bag/mask ventilation, and/or drug administration. (Neither antibiotic nor dexamethasone administration in the postoperative period was considered an MMI.)

Secondary outcomes were a minor medical intervention and prolonged hospitalization. Minor medical interventions, as might be performed by a nurse, included administration of oxygen and repositioning of the child's airway. It is our practice to administer blow-by-oxygen to all children recovering from anesthesia for the initial period in the postoperative unit. A notation of administration of oxygen implied that oxygen was required longer than usual to treat an episode of desaturation. The lowest saturation recorded in the postoperative period (SATpostop) documented in the medical dossier was recorded.

Delayed discharge was defined as discharge from the hospital after postoperative day (POD) 1.

Statistical Analysis

Categorical variables were summarized by proportions. Continuous data were presented as median (25th–75th percentile) or mean ± SD.

Differences between 2001–2002 and 2002–2006 in the MOS4 group were assessed with χ2 statistic or t test. The absolute risk reduction and number needed to treat for MMIRespiratory were determined. Multiple logistic regression was used to calculate the adjusted odds ratios (ORs) for known risk factors, age (<3 vs ≥3 years) and medical complexity,1,2,11,20 and the guidelines (SPSS Version 16.0.1, SPSS, Chicago, IL).21 In all statistical analyses, a P value <0.05 defined statistical significance.


Time Period 2002 to 2006 (Current Data)

Fourteen MMIRespiratory were required in 11 of the 97 children (11.3%) assigned MOS4: emergency succinylcholine (n = 2), reintubation (n = 4), insertion of oro- or nasopharyngeal airways (n = 3), administration of naloxone (n = 1), furosemide (n = 2), and nebulization of racemic epinephrine (n = 2). In 6 children, the MMIRespiratory occurred in the operating room (induction of anesthesia = 2). In the remaining 5 children, they occurred in the postoperative units: postanesthesia care unit (PACU) = 4 and PICU = 1 (Table 1).

Table 1
Table 1:
Summary of 19 Children Assigned MOS4 (Year 2002–2006 = 11) Requiring a MMIRespiratory

Time Period 2001 to 2002 (Historic Data)

Eight of the 27 children (29.6%) assigned MOS4 required an MMIRespiratory: emergency succinylcholine (n = 1), reintubation (n = 2), insertion of oropharyngeal airways (n = 3), bag/mask ventilation (n = 1), and nebulization of racemic epinephrine (n = 1). In 4 children, the MMIRespiratory occurred in the operating room (induction of anesthesia = 1). In the remaining 4, MMIRespiratory occurred in the PICU (Table 1).

The Concurrent Comparison Group: MOS1,2,3

Although the guidelines focused on the children with the most severe OSAS, we also examined children with less severe OSAS, the MOS1, 2, and 3 groups. No statistically significant differences in (1) demographic variables, (2) adherence to elements in the guidelines, or (3) outcome were found among these MOS groups. Therefore, for each of the 2 time periods, 2001–2002 and 2002–2006, these children were combined to form a concurrent comparison group, MOS1,2,3.

Demographic data, compliance with the guidelines, and outcome are presented in Table 2. In the MOS4 group, the incidence of MMIRespiratory decreased from 29.6% in 2001–2002 and to 11.3% in 2002–2006 (P = 0.032). The absolute risk reduction was 18.3%, and the number needed to treat was 6. Whereas in 2001–2002 the incidence of MMIRespiratory was 10-fold higher in the MOS4 group than MOS1,2,3, in 2002–2006, there was no statistical difference in the incidence of MMIRespiratory between MOS4 and its concurrent comparison group, MOS1,2,3.

Table 2
Table 2:
Summary of Demographic Data, Anesthetic Management, and Outcome Comparing the MOS4 and the Concurrent Comparison Groups MOS1,2,3 for Children Managed Under the 2001–2002 and 2002–2006 Guidelines

Secondary Outcome

Minor MIRespiratory

Overall, the lowest SATpostop was higher than the nadir saturation (nSATpreop), although in the majority of MOS4, the minimal postoperative saturation was <92% (Fig. 1).22 Forty children (41%) in the MOS4 group required a minor medical respiratory intervention, and in the majority (n = 37), this intervention was administration of oxygen. In 2002–2006, 34 children, assigned MOS4, were initially admitted to the PACU. An escalation in nursing care for episodes of severe desaturation required a transfer from PACU to PICU for 3 of these children. In 2001–2002, 7 children assigned MOS4 were initially admitted to the PACU and 1 required transfer to the PICU for an escalation in nursing care.

Figure 1
Figure 1:
The perioperative saturation nadir (nSATpreop) and the lowest recorded postoperative saturation (SATpostop) for McGill Oximetry Score (MOS)1,2,3 and MOS4 before and after adenotonsillectomy for obstructive sleep apnea syndrome (OSAS). The box plots display the median (line), 10th, 25th, 75th, and 90th percentiles and outliers. Outliers in MOS1,2,3 with nSATpreop >92% met polysomnographic diagnostic criteria for OSAS (see Methods). Some children in both groups had episodes of profound desaturation after adenotonsillectomy. The dotted line indicates 92%, the normative threshold saturation nadir during sleep in children (see Marcus et al.22).

The majority (55%) of children in the MOS4 group required a delayed discharge, a proportion not significantly different from the 2001–2002 time period. Reasons for delayed discharge in these 53 children were difficult to ascertain from the medical record but likely included respiratory concerns because desaturation on POD2 was documented in 65%.

In the MOS4 group, the readmission rate was 2%, and emergency room visits were 11%, values not significantly different from those in the concurrent comparison group, MOS1,2,3: 4% and 11%, respectively.

Did Anesthetic Management Differ in MOS4 During 2002 to 2006?

In the MOS4 group, one-third of children received an IV induction, and the usual drug was propofol. In both time periods, two-thirds of children in the MOS4 group received an inhaled induction with sevoflurane. Atropine usage decreased in 2002–2006 (Table 2). A subanalysis in 2001–2002 reported an MMIRespiratory in 28.6% of group MOS4 who received atropine and in 30.0% of those children who did not. In 2002–2006, an MMIRespiratory was recorded in 17.3% of group MOS4 who received atropine but only 4.4% of those children who did not.

Dexamethasone and opioid usage also differed between the 2 time periods. There was an increase in dexamethasone usage (P = 0.04), and the dose administered was 0.24 mg · kg−1 (0.17–0.34 mg · kg−1). Compared with 2001–2002, opioid usage in group MOS4 decreased as evidenced by a lower IOME (P = 0.03) and the PRN codeine regimen (P < 0.001, Table 2). Although in group MOS4 the median doses for postoperative morphine (POM) were not statistically different, the frequency histogram for POM in 2002–2006 was skewed leftward (Fig. 2). The POM dose required to achieve comfort was lower in the MOS4 group for all echelons of the CHEOPS (Fig. 3). In 2002–2006, compared with its comparison group MOS1,2,3, the MOS4 group received statistically lower doses of all 3 elements of opioid therapy (Table 2 and Fig. 4).

Figure 2
Figure 2:
Histograms showing the frequency of cumulative postoperative morphine (POM) doses administered to achieve comfort in the initial recovery period. Cumulative postoperative morphine dose category: 1 = 0 ≤ 0.02; 2 = >0.02 ≤ 0.04; 3 = >0.04 ≤ 0.06; 4 = >0.06 ≤ 0.08; 5 = >0.08 ≤ 0.1; and 6 = >0.1 mg · kg−1.
Figure 3
Figure 3:
The cumulative postoperative morphine (POM) doses required to achieve comfort for 3 levels of pain severity. Children in McGill Oximetry Score (MOS)4 (closed circles) were administered a lower POM dose than the concurrent comparison group, MOS1,2,3 (open circles) to achieve comfort, controlling for the admission Children's Hospital of Eastern Ontario Pain Scale (CHEOPS) score. Data are displayed as median (25th to 75th percentile). Numbers within the symbols indicate the number of patients per group.
Figure 4
Figure 4:
Trends in median (25th to 75th percentile) doses of intraoperative morphine equivalents (IOMEs) and the cumulative postoperative morphine (POM) doses administered between 2001 and 2006 in children with obstructive sleep apnea syndrome who were assigned a McGill Oximetry Score (MOS)4 (closed circles) and the concurrent comparison group, MOS1,2,3 (open circles). Note that in MOS4, the 25th percentile for POM doses was 0 for 2003 through 2006.

Was the Reduction in Opioid Dosage Associated with Evidence of Less Optimal Pain Control?

The reduced IOME in group MOS4, relative to its concurrent comparison group, MOS1,2,3, in 2002–2006 was not associated with a difference in the CHEOPS score recorded on admission to the postoperative unit (n = 218).

Overall, there was no statistical difference in the admission rate to PICU between the MOS4 groups. However, the trend over time was a decline in the proportion of children admitted to the PICU, decreasing from 100% in 2002 to 46% in 2005. By 2005, the majority of otherwise healthy children assigned MOS4 were not being admitted to the PICU (Table 3).

Table 3
Table 3:
Trends in Practice in the MOS4 Group Between 2001–2002 and 2006

Other Observations

In the 2002–2006 time period, the choice of intraoperative opioid in the majority of the MOS4 group was fentanyl or sufentanil. Combinations of intermediate and long-acting opioids were used in 14 children. The intraoperative opiates administered to the remainder were remifentanil = 3, IM codeine = 1, morphine = 17, and nil = 4.

The time interval between the sleep study and surgery in the MOS4 group was 3 days (2–6 days). There were no clinically important differences in the usage of other anesthetic medications in the MOS4 groups or the concurrent comparison group, MOS1,2,3. Operation time, that is, the time spent in the operating room, was longer in the MOS4 groups: 45 minutes (40–52 minutes) and 55 minutes (45–64 minutes) for time periods 2001–2002 and 2002–2006, respectively. Operation times for the concurrent comparison group, MOS1,2,3, were similar: 40 minutes (35–47 minutes) and 42 minutes (35–50 minutes) for time periods 2001–2002 and 2002–2006, respectively.

The revised guidelines focused on the management of the MOS4 group, and no recommendations were made for the concurrent comparison group, MOS1,2,3. The only clinically important difference in anesthetic technique between the 2 time periods in this comparison group was an increased dexamethasone usage in 2002–2006 (45% vs 67%).

A subanalysis of the 2002–2006 MOS4 group showed that 54% (n = 38) of children who received dexamethasone were discharged on POD1, a significantly higher proportion than those who did not receive it (23%) (P = 0.004). This pattern was not seen in the concurrent comparison group, MOS1,2,3. Overall, however, management under the new guidelines did not achieve a statistically significant reduction in the duration of hospitalization, when compared with 2001–2002.

In 2002–2006, the incidence of reoperation for control of posttonsillectomy hemorrhage was 0.7%. One child (patient 2 in Table 1) required reoperation to control hemorrhage during the hospital admission, and a second child (MOS1,2,3) required readmission and reoperation. Two children in group MOS1,2,3 (dexamethasone = 1) experienced posttonsillectomy bleeding during the hospital admission but were managed conservatively. Eight additional children (dexamethasone = 7 and MOS4 = 4) were readmitted for posttonsillectomy bleeding and were managed conservatively.

Multiple Logistic Regression

The risk for MMIRespiratory was lower in the MOS4 group if managed with the 2002–2006 guidelines: OR 0.30 (95% CI: 0.11–0.86; P = 0.03). However, in 2002–2006, children in the MOS4 group were younger and had less comorbidity. Both young age and medical complexity are reported risk factors for respiratory morbidity after adenotonsillectomy.1,2,23 Multiple logistic regression for these risk factors and the guidelines revealed that the adjusted ORs for young age and medical complexity did not achieve statistical significance but the adjusted OR for MMIRespiratory under the 2002–2006 guidelines remained significant: 0.30 (95% CI: 0.10–0.85; P = 0.02).


After implementation of perioperative anesthetic management guidelines for children with severe obstructive sleep apnea undergoing adenotonsillectomy, the incidence of major respiratory complications decreased from 29.6% in 2001–2002 to 11.3% in 2002–2006. Furthermore, whereas in 2001–2002 the incidence of MMIRespiratory in the MOS4 group was 10-fold higher than its concurrent comparison group, MOS1,2,3, in 2002–2006, there was no statistical difference between these 2 groups. These findings remained significant after control for known risk factors for perioperative complications in this patient group, specifically young age and medical comorbidities. Indeed, neither young age nor medical complexity achieved statistical significance as risk factors for MMIRespiratory.

We defined a major respiratory complication as airway instrumentation or drug administration because these interventions were reliably identifiable in a retrospective review. However, 4 children from the MOS4 group, who had been admitted initially to the PACU, required transfer to a PICU for an escalation in nursing care. Reclassification of these 4 children in group MOS4 would have increased the incidence of MMIRespiratory in 2002–2006 to 14.4% and in 2001–2002 to 33.3%.

Elements common to both guidelines were expedited surgery and postoperative admission to the PICU. After reports that the OSAS may negatively affect neurocognitive and cardiovascular function,510 we have advocated the expedited scheduling of adenotonsillectomy for severe OSAS. Accordingly, the delay between the oximetry test and surgery was 4 days in 2001–200211 and 3 days in 2002–2006.

The postoperative care location for children undergoing adenotonsillectomy has been a controversial subject. Although Helfaer et al.24 reported an immediate improvement in sleep-related breathing after adenotonsillectomy, a high incidence of desaturation and respiratory complications occur on the first night after adenotonsillectomy in children with severe OSAS.1,2,12,20,25 Oxygen saturation improved in the majority of children in group MOS4, but some continued to have very low saturations after surgery (Fig. 1), and oxygen was administered to 40%. Both the 2002–2002 and 2002–2006 guidelines recommended admission to a PICU for the MOS4 group. However, the temporal trends suggest that clinicians became more comfortable with PACU recovery in otherwise healthy children assigned MOS4. No increase in the rate of MMIRespiratory was evident as the admission rate to the PICU decreased. However, fewer than half of the children in group MOS4 were discharged the day after surgery, and reasons for prolonged hospitalization likely included respiratory concerns given the high incidence of desaturation on the second POD.

Because OSAS arises from dysfunction of the upper airway muscles leading to obstructive apnea26 and because central muscarinic blockade is reported to enhance the function of genioglossus muscle,27 we hypothesized that atropine might be beneficial. We had, in fact, previously reported that atropine administration decreased respiratory morbidity in children with severe obstructive sleep apnea.20 These considerations notwithstanding, in 2002–2006 only 53% of children in the MOS4 group were given atropine, and its administration was associated with a 4-fold higher incidence of MMIRespiratory.

The critical elements in the guidelines allowing an improvement in respiratory outcome seem to relate to a reduction in opioid usage and dexamethasone administration. Waters et al.28 reported a heightened respiratory opiate sensitivity in children with severe OSAS. We reported an increased respiratory sensitivity to opioids after exposure to intermittent hypoxia in rat pups.29 We have also reported an increased analgesic sensitivity to opioids in children with OSAS who demonstrate profound recurrent hypoxia.17,30 These findings justified our recommendation to reduce opioid dosage in the revised guidelines. The lower doses of intraoperative opioids in group MOS4 were not associated with higher CHEOPS score on admission to the postoperative unit, despite a 10-minute longer duration of time spent in the operating room. Whereas one-third of group MOS1,2,3 required a cumulative POM dose in excess of 0.1 mg · kg−1 to achieve comfort, fewer than 10% of group MOS4 required this high dosage (Fig. 2). Furthermore, compared with the MOS1,2,3 group, the dose of morphine to achieve comfort for all pain severities was lower in group MOS4 (Fig. 3). These findings support our hypothesis that children with severe OSAS, who display profound recurrent hypoxia, have a heightened analgesic sensitivity to opioids.

There was a shift to lower dosage of morphine administered in the postoperative period in the MOS4 group. In 2002–2006, the median POM dose required to achieve comfort in these children was 0.02 mg · kg−1. Administration of morphine in dosing increments of 0.05 to 0.1 mg · kg−1 might have resulted in administration of excessive morphine to children in the MOS4 group, three-quarters of whom required a cumulative dose of POM <0.07 mg · kg−1. Indeed, in an earlier publication, we reported that a cumulative POM dose of 0.08 mg · kg−1, administered to children with a comparable severity of OSAS, was associated with a high (11%) reintubation rate after adenotonsillectomy.20

Termination of obstructive apnea in children with OSAS is achieved by the automatic recruitment of the upper airway dilating muscles to reopen the obstructed pharynx.3133 The recent report of a selective respiratory inhibition of the genioglossus muscle by fentanyl raises the possibility that this reflex mechanism may be impaired by opioids.34 Preemptive administration of codeine, with regularly prescribed doses, will require that the sleeping child is awakened for administration of codeine, regardless of the pain complaint. If the child then resumes sleep, the opioid might depress this reflex. The principal advantage of a PRN codeine regimen is avoidance of this scenario.

A major limitation of this study was the lack of a pain metric beyond the CHEOPS score recorded on admission to the postoperative unit. Indeed, Fortier et al.35 reported that postoperative pain in the majority of children undergoing ambulatory adenotonsillectomy is inadequately treated. Their findings led the authors to recommend the preemptive administration of codeine after adenotonsillectomy. Our data, in contrast, support the notion that a subgroup of children with OSAS and profound recurrent desaturation, the MOS4 group, displays a heightened sensitivity to opioids and are better managed when codeine is titrated to the severity of the pain complaint. There is an urgent need to determine the optimal codeine regimen in children undergoing adenotonsillectomy for OSAS.

Reported consensus opinion is that dexamethasone promotes an earlier return to oral intake after adenotonsillectomy.36 Its antiinflammatory and morphine-sparing properties37,38 coupled with a heightened analgesic sensitivity to opioids may have facilitated a reduction of opioid dosage in the MOS4 group, without adversely affecting pain control. There was good compliance with the recommendation for dexamethasone over time, and two-thirds of the children (n = 196) in 2002–2006 received dexamethasone. This suggests that dexamethasone administration was indeed a credible element in the guidelines. Although there has been recent concern that the use of dexamethasone might increase perioperative bleeding after tonsillectomy,39,40 we report only 2 cases of reoperation for control of posttonsillectomy hemorrhage, and only 1 of these 2 children received dexamethasone.

Limitations of the Study

The guidelines only focused on the MOS4 group, and one-third of patients in the concurrent comparison group, MOS1,2,3, had nSAT <80%. Although we have reported a strong inverse correlation between nSATpreop and analgesic sensitivity,17,30 the duration, frequency, and intensity of the stimulus by which intermittent hypoxia alters μ-opioid receptor functionality has yet to be demonstrated. Additional children with less profound degrees of recurrent hypoxia may also have benefited from a lower opioid dosage.

The study was retrospective. However, studies of postadenotonsillectomy respiratory compromise in children with severe OSAS are poorly suited to a prospective randomized clinical trial study design. Taken as a whole and spanning a decade of practice, management under the new guidelines reduced the incidence of major respiratory complications occurring outside of the operating room in children with OSAS and profound recurrent hypoxia from a high of 19.4% in 1999–200020 to 5.2%. Similarly, our reintubation rate in the postoperative units decreased from 9.8%20 to 1.3%.

These findings suggest that dexamethasone administration and a reduction in opioid dosage decrease the risk for respiratory complications in children with severe OSAS who undergo adenotonsillectomy with an electrocautery technique. In children who demonstrate profound recurrent hypoxia, we strongly recommend a reduction opioid dosage, specifically a decrease in the IV POM dosing increment to 0.02 mg · kg−1. Future studies should assess the implications of our findings to ambulatory adenotonsillectomy programs and, in the broader picture, to pain management in patients with obstructive sleep apnea.


1. Rosen GM, Muckle RP, Mahowald MW, Goding GS, Ullevig C. Postoperative respiratory compromise in children with obstructive sleep apnea syndrome: can it be anticipated? Pediatrics 1994;93:784–8
2. McColley SA, April MM, Carroll JL, Naclerio RM, Loughlin GM. Respiratory compromise after adenotonsillectomy in children with obstructive sleep apnea. Arch Otolaryngol Head Neck Surg 1992;118:940–3
3. Kendrick D, Gibbin K. An audit of the complications of paediatric tonsillectomy, adenoidectomy and adenotonsillectomy. Clin Otolaryngol Allied Sci 1993;18:115–7
4. Richmond KH, Wetmore RF, Baranak CC. Postoperative complications following tonsillectomy and adenoidectomy—who is at risk? Int J Pediatr Otorhinolaryngol 1987;13:117–24
5. Guilleminault C, Stoohs R. From apnea of infancy to obstructive sleep apnea syndrome in the young child. Chest 1992;102:1065–71
6. Guilleminault C, Pelayo R, Leger D, Clerk A, Bocian RCZ. Recognition of sleep-disordered breathing in children. Pediatrics 1996;98:871–82
7. McNamara F, Sullivan CE. Obstructive sleep apnea in infants: relation to family history of sudden infant death syndrome, apparent life-threatening events, and obstructive sleep apnea. J Pediatr 2000;136:318–23
8. Gozal D. Sleep disordered breathing and school performance in children. Pediatrics 1998;102:616–20
9. Tal A, Leiberman A, Margulis G, Sofer S. Ventricular dysfunction in children with obstructive sleep apnea: radionuclide assessment. Pediatr Pulmonol 1988;4:139–43
10. Amin RS, Kimball TR, Bean JA, Jeffries JL, Willging JP, Cotton RT, Witt SA, Glascock BJ, Daniels SR. Left ventricular hypertrophy and abnormal ventricular geometry in children and adolescents with obstructive sleep apnea. Am J Respir Crit Care Med 2002;165:1395–9
11. Nixon GM, Kermack AS, Davis GM, Manoukian JJ, Brown KA, Brouillette RT. Planning adenotonsillectomy in children with obstructive sleep apnea: the role of overnight oximetry. Pediatrics 2004;113:e19–25
12. Nixon GM, Kermack AS, McGregor CD, Davis GM, Manoukian JJ, Brown KA, Brouillette RT. Sleep and breathing on the first night after adenotonsillectomy for obstructive sleep apnea. Pediatr Pulmonol 2005;39:332–8
13. Constantin E, McGregor CD, Cote V, Brouillette RT. Pulse rate and pulse rate variability decrease after adenotonsillectomy for obstructive sleep apnea. Pediatr Pulmonol 2008;43:498–504
14. Brouillette RT, Morielli A, Leimanis A, Waters KA, Luciano R, Ducharme FM. Nocturnal pulse oximetry as an abbreviated testing modality for pediatric obstructive sleep apnea. Pediatrics 2000;105:405–12
15. Brouillette RT, Hanson D, David R, Klemka L, Szatkowski A, Fernbach S, Hunt CE. A diagnostic approach to suspected obstructive sleep apnea in children. J Pediatr 1984;105:10–4
16. Carroll JL, McColley SA, Marcus CL, Curtis S, Loughlin GM. Inability of clinical history to distinguish primary snoring from obstructive sleep apnea syndrome in children. Chest 1995;108:610–8
17. Brown KA, Laferrière A, Lakheeram I, Moss IR. Recurrent hypoxemia in children is associated with increased analgesic sensitivity to opiates. Anesthesiology 2006;105:665–9
18. Yaster M, Kost-Byerly S, Maxwell LG. Opioid agonists and antagonists. In: Schechter NL, Berde C B, Yaster M eds. Pain in Infants, Children, and Adolescents. Philadelphia: Lippincott Williams & Wilkins, 2003:181–224
19. McGrath PJ, Johnson G, Goodman JT, Schillinger J, Dunn J, Chapman JA. CHEOPS: a behavioral scale for rating postoperative pain in children. In: Fields HL, Dubner R, Cervero F eds. Advances in Pain Research and Therapy. New York: Raven Press, 1985:395–402
20. Brown K, Morin I, Hickey C, Manoukian JJ, Nixon GM, Brouillette RT. Urgent adenotonsillectomy: an analysis of risk factors associated with postoperative respiratory morbidity. Anesthesiology 2003;99:586–95
21. Peat JK, Barton B. Medical Statistics: A Guide to Data Analysis and Critical Appraisal. 1st ed. Oxford, UK: Blackwell Publishing Ltd., 2005
22. Marcus CL, Omlin KJ, Basinski DJ, Bailey SL, Rachal AB, Von Pechmann WS, Keens TG, Davidson Ward SL. Normal polysomnographic values for children and adolescents. Am Rev Respir Dis 1992;146:1235–9
23. Wilson K, Lakheeram I, Morielli A, Brouillette RT, Brown K. Can assessment for obstructive sleep apnea help predict postadenotonsillectomy respiratory complications? Anesthesiology 2002;96:313–22
24. Helfaer MA, McColley SA, Pyzik PL, Tunkel DE, Nichols DG, Baroody FM, April MM, Maxwell LG, Loughlin GM. Polysomnography after adenotonsillectomy in mild pediatric obstructive sleep apnea. Crit Care Med 1996;24:1323–7
25. Schroeder JW, Anstead AS, Wong H. Complications in children who electively remain intubated after adenotonsillectomy for severe obstructive sleep apnea. Int J Pediatr Otorhinolaryngol 2009;73:1095–9
26. Remmers JE, DeGroot WJ, Sauerland EK, Anch AM. Pathogenesis of upper airway occlusion during sleep. J Appl Physiol 1978;44:931–8
27. Liu X, Sood S, Liu H, Horner RL. Opposing muscarinic and nicotinic modulation of hypoglossal motor output to genioglossus muscle in rats in vivo. J Physiol 2005;565:965–80
28. Waters KA, McBrien F, Stewart P, Hinder M, Wharton S. Effects of OSA, inhalational anesthesia, and fentanyl on the airway and ventilation of children. J Appl Physiol 2002; 92:1987–94
29. Moss IR, Brown KA, Laferrière A. Recurrent hypoxia in rats during development increases subsequent respiratory sensitivity to fentanyl. Anesthesiology 2006;105:715–8
30. Brown KA, Laferrière A, Moss IR. Recurrent hypoxemia in young children with obstructive sleep apnea is associated with reduced opioid requirements for analgesia. Anesthesiology 2004;100:806–10
31. Horner RL, Innes JA, Murphy K, Guz A. Evidence for reflex upper airway dilator muscle activation by sudden negative airway pressure in man. J Physiol 1991;436:15–29
32. McNamara F, Issa FG, Sullivan CE. Arousal pattern following central and obstructive breathing abnormalities in infants and children. J Appl Physiol 1996;81:2651–7
33. McNamara F, Sullivan CE. The genesis of adult sleep apnoea in childhood. Thorax 2000;55:964–9
34. Hajiha M, DuBord M, Liu H, Horner RL. Opioid receptor mechanisms at the hypoglossal motor pool and effects on tongue muscle activity in vivo. J Physiol 2009;587:2677–92
35. Fortier MA, MacLaren JE, Martin SR, Perret-Karimi D, Kain ZN. Pediatric pain after ambulatory surgery: where's the medication? Pediatrics 2009;124:e588–95
36. Steward DL, Welge JA, Myer CM. Steroids for improving recovery following tonsillectomy in children. Cochrane Database Syst Rev 2003;(1):CD003997
37. 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
38. Elhakim M, Ali NM, Rashed I, Riad MK, Refat M. Dexamethasone reduces postoperative vomiting and pain after pediatric tonsillectomy. Can J Anaesth 2003;50:392–7
39. Cardwell M, Siviter G, Smith A. Non-steroidal anti-inflammatory drugs and perioperatie bleeding in paediatric tonsillectomy. Review. Cochrane Database Syst Rev 2005;(2): CD003591
40. Czarnetzki C, Elia N, Lysakowski C, Dumont L, Landis BN, Giger R, Dulguerov P, Desmeules J, Tramèr MR. Dexamethasone and risk of nausea and vomiting and postoperative bleeding after tonsillectomy in children. JAMA 2008; 300:2621–30
© 2010 International Anesthesia Research Society