- Email: [email protected]
Pediatric adenoidectomy: What is the effect of obstructive symptoms on the likelihood of future surgery?
International Journal of Pediatric Otorhinolaryngology (2006) 70, 1467—1472
www.elsevier.com/locate/ijporl
Pediatric adenoidectomy: What is the effect of obstructive symptoms on the likelihood of future surgery?§,§§ Scott E. Brietzke a,*, Margaret Kenna b, Eliot S. Katz c, Elisabeth Mitchell a, David Roberson b a
Otolaryngology-Head and Neck Surgery Service, Walter Reed Army Medical Center, Washington, DC, United States b Department of Otolaryngology and Communication Disorders, Children’s Hospital Boston, Boston, MA, United States c Department of Pulmonology, Children’s Hospital Boston, Boston, MA, United States Received 9 December 2005; received in revised form 19 March 2006; accepted 21 March 2006
KEYWORDS Adenoidectomy; Tonsillectomy; Airway obstruction
Summary Objective: Determine if pediatric patients undergoing adenoidectomy for obstruction have an increased likelihood of undergoing eventual tonsillectomy and/or second adenoidectomy over extended follow-up. Methods: Retrospective cohort study within the ambulatory surgery component of a tertiary children’s hospital. Exclusion criteria consisted of the presence of significant co-morbid conditions (obesity, craniofacial syndrome, cerebral palsy, etc.) and less than 1 year of post-adenoidectomy follow-up. The study population included 100 children (mean age = 3.9 years, range 0.9—15 years) randomly selected from a financial billing database undergoing adenoidectomy either with pre-operative obstructive symptoms (OB n = 52) or without pre-operative obstructive symptoms (NOB n = 48) with a minimum of 1 year of retrospective follow-up. Electronic records were searched for subsequent surgery (tonsillectomy and/or revision adenoidectomy). Results: Overall, 29 of 100 children (29%) underwent subsequent surgery (tonsillectomy and/or revision adenoidectomy) over an average of 3.46 years (range 1.0—6.59 years) retrospective follow-up. Children in the OB group were three times more likely (age, sex adjusted odds ratio = 3.03, 95% confidence interval =1.18—7.78 p = 0.021)
§ The views herein are the private views of the authors and do not reflect the views of the Department of the Army or the Department of Defense. §§ Presented at the American Society of Pediatric Otolaryngology annual meeting, Las Vegas, NV, May 2005. * Corresponding author at: Walter Reed Army Medical Center, Department of Otolaryngology, 6900 Georgia Ave., Washington, DC 20307, United States. Tel.: +1 202 782 6638; fax: +1 202 782 4647. E-mail address: [email protected] (S.E. Brietzke).
0165-5876/$ — see front matter # 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2006.03.009
1468
S.E. Brietzke et al. than children in the NOB group to require tonsillectomy or second adenoidectomy. Age less than 2 years at time of initial adenoidectomy also suggested an increased likelihood of subsequent surgery. Conclusion: Children undergoing adenoidectomy with obstructive symptoms are more likely to require eventual tonsillectomy and/or second adenoidectomy than those undergoing adenoidectomy without obstruction. However, the low absolute likelihood of future surgery precludes a generalized recommendation for tonsillectomy in these children. # 2006 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Adenoid hypertrophy is one of the most common etiologies of upper airway obstruction in children [1]. Consequently, adenoidectomy is often considered as first line management by the Otolaryngologist evaluating a child with symptoms of upper airway obstruction, such as snoring and mouth-breathing. However, the potential contribution of tonsillar hypertrophy to the patient’s airway obstruction and the consideration of whether a tonsillectomy should also be performed in appropriate patients along with the adenoidectomy often complicates the decisionmaking. It is desirable to avoid tonsillectomy in younger patients and those with increased surgical risk because of the increased risk, pain and recovery time (for patient and parent) associated with tonsillectomy. Yet if the patient ultimately returns with recurrent or persistent symptoms and requires subsequent tonsillectomy, any advantage of performing adenoidectomy alone would potentially be lost. Knowing the likelihood of subsequent tonsillectomy and/or revision adenoidectomy in pediatric patients who undergo adenoidectomy alone for obstructive symptoms would be highly useful to the Otolaryngologist who is evaluating the child with upper airway obstruction. The purpose of this study is to evaluate the rate of subsequent tonsillectomy and/or revision adenoidectomy (for recurrent obstruction) in patients who undergo initial adenoidectomy alone. A retrospective cohort was assembled of patients that had undergone adenoidectomy for the indication of chronic otitis media, obstruction, or both. Rates of subsequent tonsillectomy and/or revision adenoidectomy were then determined for each group and compared.
2. Methods This study was approved by the Children’s HospitalBoston institutional review board. Sample size calculations assuming an a = 0.05 and estimated that 100 randomly selected patients would provide 80% power for evaluation of the null hypothesis: ‘‘There
is no difference in the subsequent surgery rate for the obstructive symptom group versus the nonobstructive symptom group.’’ The retrospective cohort was assembled by randomly identifying potential subjects by search of the billing database of Children’s Hospital-Boston by CPT code 42830 (Adenoidectomy, age less than 12 years) from January 1998 to May 2004. Exclusion criteria consisted of (1) any significant co-morbidity that would preclude care at an ambulatory surgery center (e.g., craniofacial syndrome, cardiac disease, etc.), (2) previous history of adenoidectomy and/or tonsillectomy, or (3) post-operative follow-up of less than 1 year by electronic chart review. Once the subject was identified, the initial adenoidectomy operative report was reviewed, demographic data was recorded, and the surgical indication (obstruction, or no obstruction chronic) was categorized into either an ‘‘obstructive symptoms’’ (OB) or ‘‘no obstructive symptoms’’ (NOB) category based on the presence or absence of upper airway obstructive symptoms documented prior to surgery. The majority of patients undergoing adenoidectomy without obstructive symptoms underwent surgery for the indication of chronic otitis media with nearly all having concurrent myringotomy and tube placement. Tonsil and adenoid size were not uniformly recorded in the operative report and thus, could not formally be included in the analysis. The technique of adenoidectomy was also recorded (cautery only, curette only, or cautery and curette). The date of the initial surgery was used as time = zero and the patients were then followed forward in time retrospectively by reviewing the electronic database of all subsequent otolaryngology visits. Only patients with recorded otolaryngology clinic visits of at least 1 year after time = zero were included. There was no significant difference in the less than 1 year of follow-up rates among the OB and NOB groups. (Data not shown) Patients in the less than 1 year follow-up group were not significantly different to included subjects in respect to age, sex, or surgical technique. (Data not shown) The rate of subsequent tonsillectomy (for indication of obstruction only) and/ or revision adenoidectomy was then determined.
Pediatric adenoidectomy
1469
Statistical analysis was performed with the assistance of computer software (STATA version 8.2, College Station Texas). Test of significance for proportions were performed with Fisher’s exact test. Tests of significance for continuous variables were performed with one way analysis of variance (ANOVA). Odds ratios were calculated and adjusted for potential confounders using multivariate logistic regression. Differences were considered significant if the null hypothesis was rejected at p < 0.05.
3. Results One hundred patients met the inclusion criteria were included in the study. There was no significant difference in mean age, sex, average length of follow up, or choice of surgical technique amongst the two groups. (Table 1) Polysomnogram was rarely (less than 6% of the time) used to objectively evaluate obstruction. Curette only adenoidectomy was the predominant (75%) surgical technique. Tonsillectomy and revision adenoidectomy rates were significantly higher in the OB group compared to the NOB group. The age and sex adjusted odds ratio for subsequent surgery in the OB group versus the NOB group was 3.04 (95% confidence interval 1.18—7.80 p = 0.021). The overall rate of subsequent surgery was 38% (20/52) for the OB group and 19% (9/48) for the NOB group ( p = 0.023 Fisher’s exact test). (Table 2) The average time from initial adenoidectomy to tonsillectomy was 2.7 years (range 0.6—5.8 years) and to second adenoidectomy was 2.5 years (range 0.8—4.8 years). There was no significant difference in mean time to surgery amongst the two groups. The majority of the subjects were young (less than 6 years) at the time of initial adenoidectomy. Subjects were divided into three similar sized age categories (independent of the presence/absence of
Table 2 Number of patients requiring subsequent surgery (cumulative incidence)
Tonsillectomy (n%) Revision adenoidectomy Either procedure a
NOB (n = 48)
OB (n = 52)
7 (14%) 6 (12%) 9 (19%)
14 (27%) 14 (27%) 20 (38%)
p = 0.046 for the entire table (two-sided Fisher’s exact test). a Most second adenoidectomies were performed at the time of the tonsillectomy, therefore the ‘‘either’’ category is not necessarily a sum of the two individual categories.
obstructive symptoms) after examining the age histogram: less than 2 years, 2—4 years, greater than 4 years. There was no difference in presence or absence of upper airway obstruction (OB versus NOB) amongst the three age groups. A greater proportion of subjects in the youngest age group required subsequent surgery, although only rates for revision adenoidectomy reached statistical significance. (Table 3) This was not the result of increased follow-up time as the mean follow-up time for each age group was not significantly different. The sex adjusted odds ratio for subjects less than 2 requiring revision adenoidectomy was 3.33 (95% CI = 1.19—9.33, p = 0.022). There was no difference in mean time to tonsillectomy (2.7 years) or adenoidectomy (2.5 years) amongst the three age groups. There was no association between surgical technique (curette only adenoidectomy versus electrocautery and curette adenoidectomy) and subsequent surgery. Too few patients underwent electrocautery only adenoidectomy to allow for inclusion in the analysis.
4. Discussion This retrospective cohort study has demonstrated that patients who undergo adenoidectomy with
Table 1 Patient demographic and diagnostic data Mean age, years (range) Male (%) Female (%) Mean follow-up, years (range) Median follow-up, years Preoperative PSG (%) Surgical technique (%) Curette only Cautery only Both
NOB (n = 48)
OB (n = 52)
3.96 (1—11) 38 (79) 10 (21) 3.22 (1.00—6.33) 3.00 0
3.87 (0.9—15) 35 (67) 17 (33) 3.68 (1.08—6.59) 3.59 3 (6)
35 (73) 0 13 (27)
39 (75) 0 13 (25)
NOB: no upper airway obstruction symptoms prior to surgery; OB: upper airway obstructive symptoms as indication or co-indication for adenoidectomy; PSG: polysomnogram (sleep study).
1470
S.E. Brietzke et al.
Table 3 Number of patients requiring subsequent surgery by age category ignoring initial adenoidectomy indication Tonsillectomy (%) Revision adenoidectomy * Either procedure *
<2 years (n = 31)
2—4 years (n = 36)
>4 years (n = 33)
7 (22%) 11 (35) 12 (39)
9 (25) 3 (08) 9 (25)
5 (15) 6 (18) 8 (24)
p = .023 (two-sided Fisher’s exact test).
obstructive symptoms require subsequent surgery in the form of tonsillectomy (and/or second adenoidectomy) more frequently relative to those who undergo adenoidectomy without any obstructive symptoms. Patients with initial obstructive symptoms were three times more likely (age, sex adjusted OR = 3.04, 95% CI = 1.18—7.80, p = 0.02) to require subsequent surgery for obstruction compared to those without initial obstructive symptoms. This may be a clinically useful observation. When considering the use of adenoidectomy alone for the management of upper airway obstruction, the likelihood of possible future tonsillectomy should be discussed with the patient/parents and be included in the informed consent process. However, exactly what level of risk to discuss with the patient/parents has been difficult to quantify. In this study, 27% (14 of 52) of patients with any obstructive symptoms at the time of initial adenoidectomy and 21% (21 of 100) of patients overall went on to require tonsillectomy a mean of 2.96 years (range 0.6—5.75 years) from the time of the initial adenoidectomy. These data may give some framework within which otolaryngologists can discuss the likelihood of future tonsillectomy with parents and patients during the informed consent process. However as this study represents only a retrospective cohort, a formal recommendation regarding performance of adenoidectomy alone versus adenotonsillectomy in patients with upper airway obstruction would be premature. The majority of patients (71%) did not require any future surgery for obstruction after adenoidectomy after an average of 3.5 years of retrospective follow-up. Why was adenoidectomy successful in the majority of patients? Before proceeding to discuss this point, it is important to note that in this study only subjective outcomes were considered. Only a limited number (6%) of patients underwent any objective measurement of airway obstruction with a polysomnogram (PSG). If available, the inclusion of objective PSG data could have changed the results significantly. Several plausible explanations exist for the high rate of subjective success of adenoidectomy alone in this cohort. Adenoidectomy (medical and surgical) is known to reduce pediatric snoring [2,3].
Consequently, the parent’s ability to monitor their child’s breathing during sleep also lessens and the parent’s ability to detect and report any persistent or recurrent obstruction to the physician is reduced, although obstruction may still exist. In the adult OSAS/snoring literature this concept has been termed ‘‘silent apnea’’ [4]. Another alternative explanation for the subjective success of adenoidectomy in this cohort is a type of surgical placebo effect. Parents who elected to address their child’s problem of airway obstruction by consenting to adenoidectomy may have considered this problem solved and did not monitor their children for any persistent symptoms (snoring, mouth breathing) afterwards which ultimately may have lead to a repeat evaluation and eventual tonsillectomy and/ or revision adenoidectomy. The final explanation to be considered is that adenoidectomy is highly successful in resolving the symptoms of upper airway obstruction in the majority of patients. This certainly is a plausible explanation as other scattered reports have subjectively and objectively validated the success of adenoidectomy (medical and surgical) alone in resolving/improving upper airway obstruction [3,5,6]. Yet, follow-up for the majority of these studies has generally been short term. One obvious weakness of this study is that adenoid and tonsil size were not recorded and analyzed. Clearly, assessment of tonsil size in a child presenting with upper airway obstruction is given great weight in the decision to pursue adenotonsillectomy. However, the issue is not as simple as the presence of large tonsils and adenoids equals a good candidate for surgical cure with adenotonsillectomy. There are numerous yet conflicting reports regarding the association or lack of association of tonsil and adenoid hypertrophy with objective and subjective measures of airway obstruction [9—13]. Moreover, there are numerous reports that adenotonsillectomy is less than 100% successful in resolving pediatric upper airway obstruction [14—16]. The age subgroup analysis was not conclusive in this study but did illustrate some interesting trends. As might be expected, the youngest patients had an increased rate of subsequent surgery compared to older children. However, this observation was not the result of increased follow-up time as follow-up
Pediatric adenoidectomy time was not significantly different between age groups. This observation was also independent of the presence of upper airway obstruction as the (co)indication of initial adenoidectomy. Instead, this observation may be the result of the fact that patients in the youngest age group were subsequently followed during the period (ages 2—6) in which they traversed what is considered to be the peak years of adenotonsillar hypertrophy [17]. During this time, many of the subjects regardless of initial adenoidectomy indication, may have gone on to independently develop symptomatic adenotonsillar hypertrophy to the point that subsequent surgery was performed. These age specific observations could be clinically useful in tailoring adenoidectomy preoperative counseling to the age of the patient. Only two other studies have directly addressed the same issue of subsequent tonsillectomy after initial adenoidectomy addressed in the current study. Kay, et al. recently published a large study reporting a 4.4% overall rate of subsequent tonsillectomy with mean follow-up of 2.4 years [7]. However, the study design did not ensure the patients were not lost follow-up during the study period and in the admitted words of the authors the reported subsequent surgery rate ‘‘likely underestimates the true rates in the general population’’. Young reported results in 1962 from a study very similar to the current study [8]. He found an overall subsequent tonsillectomy rate of 12% with follow-up of 6—9 years. However, in the subgroup of subjects less than 6 years (which would therefore be similar to the subjects in the current study), a strikingly similar tonsillectomy rate of 24% was reported. This study is a retrospective cohort study and along with the two previously mentioned similar studies, represents only the beginning of the consideration this clinical question. Dedicated prospective, controlled studies of the effectiveness of adenoidectomy with the use of objective measures of obstruction (PSG) and systematic scoring of symptoms with standardized questionnaires could lay any controversy to rest. The huge potential quality of life gains that could result from the selected use of adenoidectomy versus adenotonsillectomy should make this type of study a high priority within the public health scope of otolaryngology.
5. Conclusion Patients who undergo adenoidectomy with upper airway obstructive symptoms required subsequent tonsillectomy and/or revision adenoidectomy at an increased rate relative to those patients who under-
1471 went adenoidectomy for chronic otitis media alone. However, the fact that the majority of patients who undergo adenoidectomy alone do not require subsequent surgery over extended follow-up precludes any generalized recommendations for adenotonsillectomy over adenoidectomy.
Acknowledgements This work sponsored in part by the American Society of Pediatric Otolaryngology Research Award (2003 CORE Grant).
References [1] R.T. Brouillette, S.K. Fernbach, C.E. Hunt, Obstructive sleep apnea in infants and children, J. Pediatr. 100 (1982) 31—40. [2] C.Y. Chien, A.M. Chen, C.F. Hwang, et al., The clinical significance of adenoid-choanae area ratio in children with adenoid hypertrophy, Int. J. Pediatr. Otorhinolaryngol. 69 (2005) 235—239. [3] J.G. Demain, D.W. Goetz, Pediatric adenoidal hypertrophy and nasal airway obstruction: reduction with aqueous nasal beclomethasone, Pediatrics 95 (1995) 355—364. [4] B. Carlsson-Nordlander, H. Larsson, E. Svanborg, Warning against silent apneas after surgery for snoring, Lakartidningen 88 (1991) 1063—1065. [5] G. Criscuoli, S. D’Amora, G. Ripa, et al., Frequency of surgery among children who have adenotonsillar hypertrophy and improve after treatment with nasal beclomethasone, Pediatrics 111 (2003) e236—e238. [6] A. Shatz, Indications and outcomes of adenoidectomy in infancy, Ann. Otol. Rhinol. Laryngol. 113 (2004) 835— 838. [7] D.J. Kay, P.C. Bryson, M. Casselbrant, Rates and risk factors for subsequent tonsillectomy after prior adenoidectomy: a regression analysis, Arch. Otolaryngol. Head Neck Surg. 131 (2005) 252—255. [8] I.S. Young, Adenoidectomy: how often is tonsillectomy necessary later? J. Laryngol. Otol. 76 (1962) 791—796. [9] L.J. Brooks, B.M. Stephens, A.M. Bacevice, Adenoid size is related to severity but not the number of episodes of obstructive apnea in children, J. Pediatr. 132 (1998) 682— 686. [10] A.M. Li, E. Wong, J. Kew, et al., Use of tonsil size in the evaluation of obstructive sleep apnoea, Arch. Dis. Child. 87 (2002) 156—159. [11] J. Leach, J. Olson, J. Hermann, et al., Polysomnographic and clinical findings in children with obstructive sleep apnea, Arch. Otolaryngol. Head Neck Surg. 118 (1992) 741—744. [12] R.A. Franco Jr., R.M. Rosenfeld, M. Rao, First place—resident clinical science award 1999. Quality of life for children with obstructive sleep apnea, Otolaryngol. Head Neck Surg. 123 (2000) 9—16. [13] J. Ahlqvist-Rastad, E. Hultcrantz, H. Svanholm, Children with tonsillar obstruction: indications for and efficacy of tonsillectomy, Acta Paediatr. Scand. 77 (1988) 831—835.
1472 [14] P. Nieminen, U. Tolonen, H. Lopponen, Snoring and obstructive sleep apnea in children: a 6-month followup study, Arch. Otolaryngol. Head Neck Surg. 126 (2000) 481—486. [15] A.J. Lipton, D. Gozal, Treatment of obstructive sleep apnea in children: do we really know how? Sleep Med. Rev. 7 (2003) 61—80.
S.E. Brietzke et al. [16] J.S. Suen, J.E. Arnold, L.J. Brooks, Adenotonsillectomy for treatment of obstructive sleep apnea in children, Arch. Otolaryngol. Head Neck Surg. 121 (1995) 525—530. [17] C. Gaultier, Sleep-related breathing disorders. 6. Obstructive sleep apnoea syndrome in infants and children: established facts and unsettled issues, Thorax 50 (1995) 1204— 1210.
www.elsevier.com/locate/ijporl
Pediatric adenoidectomy: What is the effect of obstructive symptoms on the likelihood of future surgery?§,§§ Scott E. Brietzke a,*, Margaret Kenna b, Eliot S. Katz c, Elisabeth Mitchell a, David Roberson b a
Otolaryngology-Head and Neck Surgery Service, Walter Reed Army Medical Center, Washington, DC, United States b Department of Otolaryngology and Communication Disorders, Children’s Hospital Boston, Boston, MA, United States c Department of Pulmonology, Children’s Hospital Boston, Boston, MA, United States Received 9 December 2005; received in revised form 19 March 2006; accepted 21 March 2006
KEYWORDS Adenoidectomy; Tonsillectomy; Airway obstruction
Summary Objective: Determine if pediatric patients undergoing adenoidectomy for obstruction have an increased likelihood of undergoing eventual tonsillectomy and/or second adenoidectomy over extended follow-up. Methods: Retrospective cohort study within the ambulatory surgery component of a tertiary children’s hospital. Exclusion criteria consisted of the presence of significant co-morbid conditions (obesity, craniofacial syndrome, cerebral palsy, etc.) and less than 1 year of post-adenoidectomy follow-up. The study population included 100 children (mean age = 3.9 years, range 0.9—15 years) randomly selected from a financial billing database undergoing adenoidectomy either with pre-operative obstructive symptoms (OB n = 52) or without pre-operative obstructive symptoms (NOB n = 48) with a minimum of 1 year of retrospective follow-up. Electronic records were searched for subsequent surgery (tonsillectomy and/or revision adenoidectomy). Results: Overall, 29 of 100 children (29%) underwent subsequent surgery (tonsillectomy and/or revision adenoidectomy) over an average of 3.46 years (range 1.0—6.59 years) retrospective follow-up. Children in the OB group were three times more likely (age, sex adjusted odds ratio = 3.03, 95% confidence interval =1.18—7.78 p = 0.021)
§ The views herein are the private views of the authors and do not reflect the views of the Department of the Army or the Department of Defense. §§ Presented at the American Society of Pediatric Otolaryngology annual meeting, Las Vegas, NV, May 2005. * Corresponding author at: Walter Reed Army Medical Center, Department of Otolaryngology, 6900 Georgia Ave., Washington, DC 20307, United States. Tel.: +1 202 782 6638; fax: +1 202 782 4647. E-mail address: [email protected] (S.E. Brietzke).
0165-5876/$ — see front matter # 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2006.03.009
1468
S.E. Brietzke et al. than children in the NOB group to require tonsillectomy or second adenoidectomy. Age less than 2 years at time of initial adenoidectomy also suggested an increased likelihood of subsequent surgery. Conclusion: Children undergoing adenoidectomy with obstructive symptoms are more likely to require eventual tonsillectomy and/or second adenoidectomy than those undergoing adenoidectomy without obstruction. However, the low absolute likelihood of future surgery precludes a generalized recommendation for tonsillectomy in these children. # 2006 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Adenoid hypertrophy is one of the most common etiologies of upper airway obstruction in children [1]. Consequently, adenoidectomy is often considered as first line management by the Otolaryngologist evaluating a child with symptoms of upper airway obstruction, such as snoring and mouth-breathing. However, the potential contribution of tonsillar hypertrophy to the patient’s airway obstruction and the consideration of whether a tonsillectomy should also be performed in appropriate patients along with the adenoidectomy often complicates the decisionmaking. It is desirable to avoid tonsillectomy in younger patients and those with increased surgical risk because of the increased risk, pain and recovery time (for patient and parent) associated with tonsillectomy. Yet if the patient ultimately returns with recurrent or persistent symptoms and requires subsequent tonsillectomy, any advantage of performing adenoidectomy alone would potentially be lost. Knowing the likelihood of subsequent tonsillectomy and/or revision adenoidectomy in pediatric patients who undergo adenoidectomy alone for obstructive symptoms would be highly useful to the Otolaryngologist who is evaluating the child with upper airway obstruction. The purpose of this study is to evaluate the rate of subsequent tonsillectomy and/or revision adenoidectomy (for recurrent obstruction) in patients who undergo initial adenoidectomy alone. A retrospective cohort was assembled of patients that had undergone adenoidectomy for the indication of chronic otitis media, obstruction, or both. Rates of subsequent tonsillectomy and/or revision adenoidectomy were then determined for each group and compared.
2. Methods This study was approved by the Children’s HospitalBoston institutional review board. Sample size calculations assuming an a = 0.05 and estimated that 100 randomly selected patients would provide 80% power for evaluation of the null hypothesis: ‘‘There
is no difference in the subsequent surgery rate for the obstructive symptom group versus the nonobstructive symptom group.’’ The retrospective cohort was assembled by randomly identifying potential subjects by search of the billing database of Children’s Hospital-Boston by CPT code 42830 (Adenoidectomy, age less than 12 years) from January 1998 to May 2004. Exclusion criteria consisted of (1) any significant co-morbidity that would preclude care at an ambulatory surgery center (e.g., craniofacial syndrome, cardiac disease, etc.), (2) previous history of adenoidectomy and/or tonsillectomy, or (3) post-operative follow-up of less than 1 year by electronic chart review. Once the subject was identified, the initial adenoidectomy operative report was reviewed, demographic data was recorded, and the surgical indication (obstruction, or no obstruction chronic) was categorized into either an ‘‘obstructive symptoms’’ (OB) or ‘‘no obstructive symptoms’’ (NOB) category based on the presence or absence of upper airway obstructive symptoms documented prior to surgery. The majority of patients undergoing adenoidectomy without obstructive symptoms underwent surgery for the indication of chronic otitis media with nearly all having concurrent myringotomy and tube placement. Tonsil and adenoid size were not uniformly recorded in the operative report and thus, could not formally be included in the analysis. The technique of adenoidectomy was also recorded (cautery only, curette only, or cautery and curette). The date of the initial surgery was used as time = zero and the patients were then followed forward in time retrospectively by reviewing the electronic database of all subsequent otolaryngology visits. Only patients with recorded otolaryngology clinic visits of at least 1 year after time = zero were included. There was no significant difference in the less than 1 year of follow-up rates among the OB and NOB groups. (Data not shown) Patients in the less than 1 year follow-up group were not significantly different to included subjects in respect to age, sex, or surgical technique. (Data not shown) The rate of subsequent tonsillectomy (for indication of obstruction only) and/ or revision adenoidectomy was then determined.
Pediatric adenoidectomy
1469
Statistical analysis was performed with the assistance of computer software (STATA version 8.2, College Station Texas). Test of significance for proportions were performed with Fisher’s exact test. Tests of significance for continuous variables were performed with one way analysis of variance (ANOVA). Odds ratios were calculated and adjusted for potential confounders using multivariate logistic regression. Differences were considered significant if the null hypothesis was rejected at p < 0.05.
3. Results One hundred patients met the inclusion criteria were included in the study. There was no significant difference in mean age, sex, average length of follow up, or choice of surgical technique amongst the two groups. (Table 1) Polysomnogram was rarely (less than 6% of the time) used to objectively evaluate obstruction. Curette only adenoidectomy was the predominant (75%) surgical technique. Tonsillectomy and revision adenoidectomy rates were significantly higher in the OB group compared to the NOB group. The age and sex adjusted odds ratio for subsequent surgery in the OB group versus the NOB group was 3.04 (95% confidence interval 1.18—7.80 p = 0.021). The overall rate of subsequent surgery was 38% (20/52) for the OB group and 19% (9/48) for the NOB group ( p = 0.023 Fisher’s exact test). (Table 2) The average time from initial adenoidectomy to tonsillectomy was 2.7 years (range 0.6—5.8 years) and to second adenoidectomy was 2.5 years (range 0.8—4.8 years). There was no significant difference in mean time to surgery amongst the two groups. The majority of the subjects were young (less than 6 years) at the time of initial adenoidectomy. Subjects were divided into three similar sized age categories (independent of the presence/absence of
Table 2 Number of patients requiring subsequent surgery (cumulative incidence)
Tonsillectomy (n%) Revision adenoidectomy Either procedure a
NOB (n = 48)
OB (n = 52)
7 (14%) 6 (12%) 9 (19%)
14 (27%) 14 (27%) 20 (38%)
p = 0.046 for the entire table (two-sided Fisher’s exact test). a Most second adenoidectomies were performed at the time of the tonsillectomy, therefore the ‘‘either’’ category is not necessarily a sum of the two individual categories.
obstructive symptoms) after examining the age histogram: less than 2 years, 2—4 years, greater than 4 years. There was no difference in presence or absence of upper airway obstruction (OB versus NOB) amongst the three age groups. A greater proportion of subjects in the youngest age group required subsequent surgery, although only rates for revision adenoidectomy reached statistical significance. (Table 3) This was not the result of increased follow-up time as the mean follow-up time for each age group was not significantly different. The sex adjusted odds ratio for subjects less than 2 requiring revision adenoidectomy was 3.33 (95% CI = 1.19—9.33, p = 0.022). There was no difference in mean time to tonsillectomy (2.7 years) or adenoidectomy (2.5 years) amongst the three age groups. There was no association between surgical technique (curette only adenoidectomy versus electrocautery and curette adenoidectomy) and subsequent surgery. Too few patients underwent electrocautery only adenoidectomy to allow for inclusion in the analysis.
4. Discussion This retrospective cohort study has demonstrated that patients who undergo adenoidectomy with
Table 1 Patient demographic and diagnostic data Mean age, years (range) Male (%) Female (%) Mean follow-up, years (range) Median follow-up, years Preoperative PSG (%) Surgical technique (%) Curette only Cautery only Both
NOB (n = 48)
OB (n = 52)
3.96 (1—11) 38 (79) 10 (21) 3.22 (1.00—6.33) 3.00 0
3.87 (0.9—15) 35 (67) 17 (33) 3.68 (1.08—6.59) 3.59 3 (6)
35 (73) 0 13 (27)
39 (75) 0 13 (25)
NOB: no upper airway obstruction symptoms prior to surgery; OB: upper airway obstructive symptoms as indication or co-indication for adenoidectomy; PSG: polysomnogram (sleep study).
1470
S.E. Brietzke et al.
Table 3 Number of patients requiring subsequent surgery by age category ignoring initial adenoidectomy indication Tonsillectomy (%) Revision adenoidectomy * Either procedure *
<2 years (n = 31)
2—4 years (n = 36)
>4 years (n = 33)
7 (22%) 11 (35) 12 (39)
9 (25) 3 (08) 9 (25)
5 (15) 6 (18) 8 (24)
p = .023 (two-sided Fisher’s exact test).
obstructive symptoms require subsequent surgery in the form of tonsillectomy (and/or second adenoidectomy) more frequently relative to those who undergo adenoidectomy without any obstructive symptoms. Patients with initial obstructive symptoms were three times more likely (age, sex adjusted OR = 3.04, 95% CI = 1.18—7.80, p = 0.02) to require subsequent surgery for obstruction compared to those without initial obstructive symptoms. This may be a clinically useful observation. When considering the use of adenoidectomy alone for the management of upper airway obstruction, the likelihood of possible future tonsillectomy should be discussed with the patient/parents and be included in the informed consent process. However, exactly what level of risk to discuss with the patient/parents has been difficult to quantify. In this study, 27% (14 of 52) of patients with any obstructive symptoms at the time of initial adenoidectomy and 21% (21 of 100) of patients overall went on to require tonsillectomy a mean of 2.96 years (range 0.6—5.75 years) from the time of the initial adenoidectomy. These data may give some framework within which otolaryngologists can discuss the likelihood of future tonsillectomy with parents and patients during the informed consent process. However as this study represents only a retrospective cohort, a formal recommendation regarding performance of adenoidectomy alone versus adenotonsillectomy in patients with upper airway obstruction would be premature. The majority of patients (71%) did not require any future surgery for obstruction after adenoidectomy after an average of 3.5 years of retrospective follow-up. Why was adenoidectomy successful in the majority of patients? Before proceeding to discuss this point, it is important to note that in this study only subjective outcomes were considered. Only a limited number (6%) of patients underwent any objective measurement of airway obstruction with a polysomnogram (PSG). If available, the inclusion of objective PSG data could have changed the results significantly. Several plausible explanations exist for the high rate of subjective success of adenoidectomy alone in this cohort. Adenoidectomy (medical and surgical) is known to reduce pediatric snoring [2,3].
Consequently, the parent’s ability to monitor their child’s breathing during sleep also lessens and the parent’s ability to detect and report any persistent or recurrent obstruction to the physician is reduced, although obstruction may still exist. In the adult OSAS/snoring literature this concept has been termed ‘‘silent apnea’’ [4]. Another alternative explanation for the subjective success of adenoidectomy in this cohort is a type of surgical placebo effect. Parents who elected to address their child’s problem of airway obstruction by consenting to adenoidectomy may have considered this problem solved and did not monitor their children for any persistent symptoms (snoring, mouth breathing) afterwards which ultimately may have lead to a repeat evaluation and eventual tonsillectomy and/ or revision adenoidectomy. The final explanation to be considered is that adenoidectomy is highly successful in resolving the symptoms of upper airway obstruction in the majority of patients. This certainly is a plausible explanation as other scattered reports have subjectively and objectively validated the success of adenoidectomy (medical and surgical) alone in resolving/improving upper airway obstruction [3,5,6]. Yet, follow-up for the majority of these studies has generally been short term. One obvious weakness of this study is that adenoid and tonsil size were not recorded and analyzed. Clearly, assessment of tonsil size in a child presenting with upper airway obstruction is given great weight in the decision to pursue adenotonsillectomy. However, the issue is not as simple as the presence of large tonsils and adenoids equals a good candidate for surgical cure with adenotonsillectomy. There are numerous yet conflicting reports regarding the association or lack of association of tonsil and adenoid hypertrophy with objective and subjective measures of airway obstruction [9—13]. Moreover, there are numerous reports that adenotonsillectomy is less than 100% successful in resolving pediatric upper airway obstruction [14—16]. The age subgroup analysis was not conclusive in this study but did illustrate some interesting trends. As might be expected, the youngest patients had an increased rate of subsequent surgery compared to older children. However, this observation was not the result of increased follow-up time as follow-up
Pediatric adenoidectomy time was not significantly different between age groups. This observation was also independent of the presence of upper airway obstruction as the (co)indication of initial adenoidectomy. Instead, this observation may be the result of the fact that patients in the youngest age group were subsequently followed during the period (ages 2—6) in which they traversed what is considered to be the peak years of adenotonsillar hypertrophy [17]. During this time, many of the subjects regardless of initial adenoidectomy indication, may have gone on to independently develop symptomatic adenotonsillar hypertrophy to the point that subsequent surgery was performed. These age specific observations could be clinically useful in tailoring adenoidectomy preoperative counseling to the age of the patient. Only two other studies have directly addressed the same issue of subsequent tonsillectomy after initial adenoidectomy addressed in the current study. Kay, et al. recently published a large study reporting a 4.4% overall rate of subsequent tonsillectomy with mean follow-up of 2.4 years [7]. However, the study design did not ensure the patients were not lost follow-up during the study period and in the admitted words of the authors the reported subsequent surgery rate ‘‘likely underestimates the true rates in the general population’’. Young reported results in 1962 from a study very similar to the current study [8]. He found an overall subsequent tonsillectomy rate of 12% with follow-up of 6—9 years. However, in the subgroup of subjects less than 6 years (which would therefore be similar to the subjects in the current study), a strikingly similar tonsillectomy rate of 24% was reported. This study is a retrospective cohort study and along with the two previously mentioned similar studies, represents only the beginning of the consideration this clinical question. Dedicated prospective, controlled studies of the effectiveness of adenoidectomy with the use of objective measures of obstruction (PSG) and systematic scoring of symptoms with standardized questionnaires could lay any controversy to rest. The huge potential quality of life gains that could result from the selected use of adenoidectomy versus adenotonsillectomy should make this type of study a high priority within the public health scope of otolaryngology.
5. Conclusion Patients who undergo adenoidectomy with upper airway obstructive symptoms required subsequent tonsillectomy and/or revision adenoidectomy at an increased rate relative to those patients who under-
1471 went adenoidectomy for chronic otitis media alone. However, the fact that the majority of patients who undergo adenoidectomy alone do not require subsequent surgery over extended follow-up precludes any generalized recommendations for adenotonsillectomy over adenoidectomy.
Acknowledgements This work sponsored in part by the American Society of Pediatric Otolaryngology Research Award (2003 CORE Grant).
References [1] R.T. Brouillette, S.K. Fernbach, C.E. Hunt, Obstructive sleep apnea in infants and children, J. Pediatr. 100 (1982) 31—40. [2] C.Y. Chien, A.M. Chen, C.F. Hwang, et al., The clinical significance of adenoid-choanae area ratio in children with adenoid hypertrophy, Int. J. Pediatr. Otorhinolaryngol. 69 (2005) 235—239. [3] J.G. Demain, D.W. Goetz, Pediatric adenoidal hypertrophy and nasal airway obstruction: reduction with aqueous nasal beclomethasone, Pediatrics 95 (1995) 355—364. [4] B. Carlsson-Nordlander, H. Larsson, E. Svanborg, Warning against silent apneas after surgery for snoring, Lakartidningen 88 (1991) 1063—1065. [5] G. Criscuoli, S. D’Amora, G. Ripa, et al., Frequency of surgery among children who have adenotonsillar hypertrophy and improve after treatment with nasal beclomethasone, Pediatrics 111 (2003) e236—e238. [6] A. Shatz, Indications and outcomes of adenoidectomy in infancy, Ann. Otol. Rhinol. Laryngol. 113 (2004) 835— 838. [7] D.J. Kay, P.C. Bryson, M. Casselbrant, Rates and risk factors for subsequent tonsillectomy after prior adenoidectomy: a regression analysis, Arch. Otolaryngol. Head Neck Surg. 131 (2005) 252—255. [8] I.S. Young, Adenoidectomy: how often is tonsillectomy necessary later? J. Laryngol. Otol. 76 (1962) 791—796. [9] L.J. Brooks, B.M. Stephens, A.M. Bacevice, Adenoid size is related to severity but not the number of episodes of obstructive apnea in children, J. Pediatr. 132 (1998) 682— 686. [10] A.M. Li, E. Wong, J. Kew, et al., Use of tonsil size in the evaluation of obstructive sleep apnoea, Arch. Dis. Child. 87 (2002) 156—159. [11] J. Leach, J. Olson, J. Hermann, et al., Polysomnographic and clinical findings in children with obstructive sleep apnea, Arch. Otolaryngol. Head Neck Surg. 118 (1992) 741—744. [12] R.A. Franco Jr., R.M. Rosenfeld, M. Rao, First place—resident clinical science award 1999. Quality of life for children with obstructive sleep apnea, Otolaryngol. Head Neck Surg. 123 (2000) 9—16. [13] J. Ahlqvist-Rastad, E. Hultcrantz, H. Svanholm, Children with tonsillar obstruction: indications for and efficacy of tonsillectomy, Acta Paediatr. Scand. 77 (1988) 831—835.
1472 [14] P. Nieminen, U. Tolonen, H. Lopponen, Snoring and obstructive sleep apnea in children: a 6-month followup study, Arch. Otolaryngol. Head Neck Surg. 126 (2000) 481—486. [15] A.J. Lipton, D. Gozal, Treatment of obstructive sleep apnea in children: do we really know how? Sleep Med. Rev. 7 (2003) 61—80.
S.E. Brietzke et al. [16] J.S. Suen, J.E. Arnold, L.J. Brooks, Adenotonsillectomy for treatment of obstructive sleep apnea in children, Arch. Otolaryngol. Head Neck Surg. 121 (1995) 525—530. [17] C. Gaultier, Sleep-related breathing disorders. 6. Obstructive sleep apnoea syndrome in infants and children: established facts and unsettled issues, Thorax 50 (1995) 1204— 1210.