Difficult clinical decisions in sleep apnea surgery: identifying priorities for future clinical trials

Introduction

Obstructive sleep apnea (OSA) is a chronic condition causing marked morbidity and mortality, and is treated differently in children and adults (1-4). Pediatric OSA has a reported incidence of up to 5.7%, and may lead to behavioral and cognitive abnormalities, growth failure, systemic hypertension and cardiovascular effects such as cor-pulmonale, right and left ventricular dysfunction (5-8).

In Adults, partial or complete upper airway occlusion is temporized using conventional treatments such as: continuous positive airway pressure (CPAP); or dental appliance treatments including mandibular advancement splints (MAS) (9,10). However, in the pediatric population concerns of poor treatment adherence are magnified, raising the potential for targeted effective surgical treatments (11). The important considerations when initiating surgical intervention to treat pediatric OSA include integrating: age; symptom severity; underlying medical co-morbidities; with patient anatomy and polysomnographic findings.

The American Academy of Otolaryngology-Head and Neck Surgery (AAO-HNS) and American Academy of Pediatrics (AAP) recommend adenotonsillectomy as first-line treatment for OSA in healthy children over two years of age with moderate-severe OSA and adenotonsillar hypertrophy (12,13). Several randomized systematic reviews have concluded that adenotonsillotomy is associated with less postoperative morbidity than adenotonsillectomy, but with a similar effect on pediatric OSA symptoms (14-18). However, there is a lack of prospective long-term randomized controlled trials (RCTs) to determine the risk of regrowth and reoperation rates.

Drug-induced sleep endoscopy (DISE) is increasingly being used to evaluate anatomical sites of collapse prior to surgical intervention both in children with multilevel collapse or with persistent OSA following adenotonsillectomy (19-21). Multilevel upper airway surgical procedures are sometimes considered to treat persistent pediatric OSA following adenotonsillectomy after careful flexible fiberoptic nasolaryngoscopic evaluation (22,23).

Randomized clinical trials are the gold standard for testing treatment effectiveness. Fifteen years ago, both the funding and effectiveness of upper airway surgery for snoring and sleep apnea came into question due to a lack of high-quality RCTs (24,25). This incentivized the field into developing and conducting RCTs (26-29). Despite these well executed trials, additional high-quality studies in the medium to long term are required to justify funding and strengthen the indication for surgery in specified situations.

Clinical trials in surgery are difficult to successfully undertake and pose particular practical and methodological challenges that have been examined in the literature (30-32). A number of barriers accounting for a relative lack of surgical RCTs have been described including: selecting appropriate control group; blinding; standardizing surgical technique between individual surgeon and centers over time (33,34). However, the inherent value of a well-conducted RCT should not be overlooked. Collaboration between research and surgical communities is essential to enable the conduct of well-designed and appropriate trials (35).

Barriers to recruiting individual surgeons and participants for clinical trials include time constraints; risk to therapeutic doctor-patient relationship; and perceived importance of research questions (36). The dual-role of the surgeon-researcher is the main source of ethical conflict in surgical RCTs. Even if the surgical community is in equipoise about a clinical scenario, individual surgeons may have strong treatment preferences which may dissuade their enrollment of patients. To conduct a clinical trial a genuine state of clinical uncertainty or equipoise should exist within the surgical community when less than 70% of clinicians agree that a particular treatment option is superior (37).

A single surgeon may have “individual equipoise”, whereby they are genuinely uncertain of which treatment holds superior therapeutic merit. The surgical community can also demonstrate “community equipoise” by one of two outcomes: majority are uncertain; or equal proportions favoring each alternative treatment (37).

There is a need to identify appropriate and well-designed clinical trials in sleep surgery that are ethical, viable and broadly supported by the surgical community for addressing clinical questions in which community equipoise exists. Our previous study identified two potential RCT targets in adults with OSA. This study identified multiple potential RCT targets in paediatric OSA. We present five common, but untested, clinical scenarios for practicing otolaryngologists to gauge perceived clinical importance, community and individual equipoise and willingness to enroll patients in either a RCT or non-randomized follow up study. We present this article in accordance with the SURGE reporting checklist (available at https://www.theajo.com/article/view/10.21037/ajo-24-16/rc).

Methods

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). This study was approved by the Illawarra Shoalhaven Local Health District Research Review Board, Low and Negligible Risk (ISLHD LNR/QA205, 26/07/23) Review. Informed consent was taken from all individual participants. All participants were attendees of the 5^th South Pacific ORL Forum Meeting in Fiji 2023 who were surveyed to quantify their attitudes toward five clinical scenarios. Surgeons were given a QR code link attachment to an online questionnaire. Most surgeons completed the questionnaire at the time of the meeting, however one month allowed to submit before the data collection phase was closed. Respondents who indicated they were not involved in sleep surgery were excluded from analysis.

We presented five clinical scenarios (A to D pediatrics, E adults) each with two treatment options. The questionnaire [see Supplementary file (Appendix 1)] methodology was developed by the Surgical Outcomes Research Centre in Sydney (REFS), modified from our previous study, published in Otolaryngology-Head & Neck Surgery, “Clinical Equipoise in Sleep Surgery” [2011] (32).

• Scenario A: intracapsular tonsillotomy compared to tonsillectomy in pediatric patients (aged 2–18 years) with sleep breathing disorder (no prior surgery or history of recurrent tonsillitis).
• Scenario B: is the addition of DISE to clinical examination better than clinical examination alone in pediatric patients (aged 2–18 years) post adenotonsillectomy with persistent symptomatic OSA with an apnea-hypopnea index (AHI) ≥2 on polysomnogram?
• Scenario C: multilevel stepwise upper airway reconstruction (any identifiable target can be addressed; revision (adeno)tonsillectomy, inferior turbinoplasty, lingual tonsil reduction, supraglottoplasty, other—performed at clinician discretion) compared to medical &/or device therapy (orthodontic procedures, MAS, CPAP) in pediatric patients (aged 2–≤18 years) post adenotonsillectomy with persistent symptomatic mild OSA (AHI ≤5) on polysomnogram.
• Scenario D: multilevel stepwise upper airway reconstruction (any identifiable target can be addressed; revision (adeno)tonsillectomy, inferior turbinoplasty, lingual tonsil reduction, supraglottoplasty, other—performed at clinician discretion) compared to medical &/or device therapy (orthodontic procedures, MAS, CPAP) in pediatric patients (aged 2–≤18 years) post adenotonsillectomy with persistent symptomatic moderate to severe OSA (AHI ≥5) on polysomnogram.
• Scenario E: is the addition of DISE to clinical examination better than clinical examination alone in adult patients post adenotonsillectomy with persistent symptomatic OSA (AHI ≥5 on polysomnogram)?

Participants were initially asked to quantify the importance of each clinical scenario (extremely important, very important, somewhat important, not at all important). We subsequently calculated an importance ratio (yes:no) by reducing these four options into two groups. The “Yes, it is important” option combined the responses of extremely important or very important. The “No, it is not important” option combined somewhat important or not at all important.

To quantify community equipoise, an 11-point bidirectional ordinal scale with a treatment option anchored at either end was used. The scale is centered on 0 to indicate “completely undecided” and numbered from 1 to 5 toward each end to represent increasing certainty of the preferred treatment option (refer to x-axes of Figures 1-5).

Figure 1 Scenario A: intracapsular tonsillotomy versus tonsillectomy in pediatric patients (aged 2–18 years) with sleep breathing disorder (no surgery or recurrent tonsillitis). Most surgeons willing to enroll in a RCT irrespective of treatment uncertainty. RCT, randomized controlled trial.

Figure 2 Scenario B: is addition of DISE to clinical examination better than clinical examination alone in pediatric patients (aged 2–18 years) post adenotonsillectomy with persistent symptomatic OSA and AHI ≥2 (polysomnogram)? RCT, randomized controlled trial; DISE, drug-induced sleep endoscopy; OSA, obstructive sleep apnea; AHI, apnea-hypopnea index.

Figure 3 Scenario C: multilevel stepwise upper airway reconstruction compared to medical +/− device therapy in pediatric patients (aged 2–≤18 years) post adenotonsillectomy with persistent symptomatic mild OSA (AHI ≤5, polysomnogram). RCT, randomized controlled trial; CPAP, continuous positive airway pressure; MAS, mandibular advancement splints; OSA, obstructive sleep apnea; AHI, apnea-hypopnea index.

Figure 4 Scenario D: multilevel stepwise upper airway reconstruction compared to medical +/− device therapy in pediatric patients (aged 2–≤18 years) post adenotonsillectomy with persistent symptomatic moderate to severe OSA (AHI ≥5, polysomnogram). RCT, randomized controlled trial; CPAP, continuous positive airway pressure; MAS, mandibular advancement splints; OSA, obstructive sleep apnea; AHI, apnea-hypopnea index.

Figure 5 Scenario E: is addition of DISE to clinical examination better than clinical examination alone in adult patients post adenotonsillectomy with persistent symptomatic OSA (AHI ≥5, polysomnogram)? RCT, randomized controlled trial; DISE, drug-induced sleep endoscopy; OSA, obstructive sleep apnea; AHI, apnea-hypopnea index.

Surgeon willingness to participate in each suggested trial was analysed by way of a willingness to enroll ratio (W:X & Y:Z) incorporating the following: would take part in a randomized clinical trial (W); would not take part in a randomized clinical trial (X); would take part in a non-randomized follow-up study (Y), and would not take part in a non-randomized follow-up study (Z).

Surgeons were subsequently asked a range clinical practice-related question including age, gender, location of practice, and type of appointment. This demographic information was obtained only for descriptive purposes.

Statistical analysis

Community equipoise was investigated by classifying the surgeon’s response using a bidirectional ordinal scale. Responses were plotted in a histogram and also classified into one of three groups to calculate the equipoise ratio: (A) preferred treatment 1, left-hand side; (B) completely undecided, and (C) preferred treatment 2, right-hand side). The equipoise ratios are calculated as percentages out of 100 and presented in proportions (A:B:C). For example, an equipoise ratio of 10:7:83 illustrates that 83% of surgeons selected preferred treatment 2. In this clinical scenario since more than 70% of surgeons perceive treatment 2 to be superior, no equipoise exists (37).

An example of community equipoise would be 50:5:45, where there is nearly equal preference to alternative treatments. An example of individual equipoise would be 10:75:15, where the majority are undecided not knowing which option is superior.

For each of the five clinical scenarios we have plotted the clinical equipoise in a histogram with a bidirectional ordinal scale ranging from +5 to −5 centered on zero (personal equipoise). Willingness to enroll their patients in a RCT is plotted with each bar of a clinical equipoise plot as a percentage of the total participants. Each Figures 1-5 demonstrates surgeons’ willingness to enroll patients in a RCT weighted to the clinical uncertainty of the respective scenarios A–E.

Results

The questionnaire was first presented in-person to 60 attendees at the 5^th South Pacific ORL forum meeting 2023. We then excluded the sample responses of those that indicated they were no longer practicing or had never undertaken surgery for the relief of snoring or sleep apnea because they were not part of our population description. This left 48 otolaryngologists. Demographic characteristics are shown in Table 1. The importance, equipoise ratios, and willingness to enroll patients for a RCT are presented in Table 2 together with our conclusions as to whether this is a viable clinical trial target. We have plotted the combined clinical equipoise data with willingness to enroll in each separate uncertainty score bar for all scenarios in Figures 1-5.

Scenario A demonstrated a moderate importance ratio, community equipoise and willingness to enroll patients in a RCT. Scenario B demonstrated a high importance ratio, no community equipoise and no willingness to enroll patients in an RCT. Scenario C demonstrated a high importance ratio, community and individual equipoise and a willingness to enroll in an RCT. Scenario D demonstrated a high importance ratio, community and individual equipoise and a willingness to enroll in an RCT. Scenario E demonstrated a moderate importance ratio, community equipoise and a willingness to enroll in an RCT. Scenarios A–E all demonstrated a willingness to enroll patients in a non-randomized follow up study.

Discussion

Feasible clinical trials in surgery must be perceived as important, unanswered (community equipoise), and supported (high willingness to enroll patients). Three of the five sleep surgery trial targets (C–E) we presented to surgeons were regarded as important, unanswered (equipoise—Figures 3-5) and supported (willingness to enroll their patients in a RCT). All trial targets demonstrated a willingness to enroll in a non-randomized follow-up study. Our survey suggests that three clinical trial targets (C, D and E, 2 pediatric and 1 adult) are feasible for a randomized trial.

Scenario A demonstrated low importance and no willingness to enroll patients in a RCT despite the existence of community equipoise. This discrepancy between community equipoise and low importance and an unwillingness to enroll patients in a RCT seems counterintuitive. However, this may indicate that surgeons perceive either option of similar efficacy with minimal long-term morbidity. Despite there being a lack of prospective long-term RCTs on intracapsular tonsillotomy to determine the risk of regrowth and reoperation rates, there remains an unwillingness to enroll pediatric patients in a clinical trial. Scenario A does not look like a strong candidate for a clinical trial based on these data.

Scenario B demonstrated high importance and willingness to enroll patients in an RCT, but low community equipoise. This may be explained by DISE in the pediatric population being relatively well understood with low morbidity. At this stage scenario B seems a difficult clinical trial target to us because once the trial starts otolaryngologists may not accept the reality of their patients being randomized into the clinical examination alone arm and the trial may suffer from a problem with cross-overs.

Scenarios C (mild OSA) and D (moderate-severe OSA) both demonstrated high importance ratios, similar community and individual equipoise ratios and moderate willingness to enroll pediatric patients in a RCT. These findings may indicate concerns surrounding pediatric OSA causing marked morbidity and mortality, and if left untreated may cause neurocognitive and cardiovascular abnormalities. Particularly in the pediatric population, poor adherence to non-surgical treatments severely limits disease alleviation, which may be better managed with surgical treatments.

Scenario E demonstrated low importance but reasonable community equipoise and moderate willingness to enroll in a RCT. These findings may suggest that some surgeons deem this to be a promising trial target, whilst others remain uncertain regarding the role of DISE in adults.

The limitations of the study include the relatively small sample size of 48 surgeons however this is partly explained by our methodology surveying participants of the 2023 South Pacific ORL Forum Meeting with a dominant interest in sleep surgery.

Our previous publication identified two suitable clinical trials in adult OSA comparing surgical vs. non-surgical interventions. This current study identified two potential trial targets in pediatric OSA. We also identified a trial target in adults with persistent symptomatic OSA post adenotonsillectomy.

Conclusions

We have identified three potential trial targets (scenarios C, D, E) and one less promising trial target (A). Multilevel stepwise upper-airway reconstruction versus medical therapy in pediatric patients post adenotonsillectomy with persistent symptomatic mild OSA (AHI <5), persistent symptomatic moderate-severe OSA (AHI ≥5), and DISE and clinical exam versus clinical exam alone to assess adult with persistent symptomatic OSA (AHI ≥5) post adenotonsillectomy. Significant equipoise exists surrounding the ideal treatment for these conditions.

Acknowledgments

Funding: None.

Footnote

Reporting Checklist: The authors have completed the SURGE reporting checklist. Available at https://www.theajo.com/article/view/10.21037/10.21037/ajo-24-16/rc

Data Sharing Statement: Available at https://www.theajo.com/article/view/10.21037/ajo-24-16/dss

Peer Review File: Available at https://www.theajo.com/article/view/10.21037/ajo-24-16/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://www.theajo.com/article/view/10.21037/ajo-24-16/coif). S.M. serves as an unpaid editorial board member of Australian Journal of Otolaryngology from January 2019 to December 2024. N.S.M. has received funding from the NHMRC and Bio Biosciences in the past 36 months for trials unrelated to the current investigation. He was Board Member and Finance Chair of Australasian Sleep Association (ASA) from 2018 to 2022. He is Deputy Chair Finance Committee of ASA from 2022 to the present. ASA is the peak professional body for sleep health professions in the region and is a non-profit. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). This study was approved by the Illawarra Shoalhaven Local Health District Research Review Board, Low and Negligible Risk (ISLHD LNR/QA205, 26/07/23) Review. Informed consent was taken from all individual participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Y Young T. Peppard PE, Gottlieb DJ. Epidemiology of obstructive sleep apnea: a population health perspective. Am J Respir Crit Care Med 2002;165:1217-39. [Crossref] [PubMed]
2. Marshall NS, Wong KK, Liu PY, et al. Sleep apnea as an independent risk factor for all-cause mortality: the Busselton Health Study. Sleep 2008;31:1079-85.
3. Punjabi NM, Caffo BS, Goodwin JL, et al. Sleep-disordered breathing and mortality: a prospective cohort study. PLoS Med 2009;6:e1000132. [Crossref] [PubMed]
4. Young T, Finn L, Peppard PE, et al. Sleep disordered breathing and mortality: eighteen-year follow-up of the Wisconsin sleep cohort. Sleep 2008;31:1071-8.
5. Kubba H, Downie LS. Trends in tonsillectomy surgery in children in Scotland 2000-2018. Clin Otolaryngol 2021;46:146-53. [Crossref] [PubMed]
6. Marcus CL, Moore RH, Rosen CL, et al. A randomized trial of adenotonsillectomy for childhood sleep apnea. N Engl J Med 2013;368:2366-76. [Crossref] [PubMed]
7. Ross RD, Daniels SR, Loggie JM, et al. Sleep apnea-associated hypertension and reversible left ventricular hypertrophy. J Pediatr 1987;111:253-5. [Crossref] [PubMed]
8. Redline S, Cook K, Chervin RD, et al. Adenotonsillectomy for Snoring and Mild Sleep Apnea in Children: A Randomized Clinical Trial. JAMA 2023;330:2084-95. [Crossref] [PubMed]
9. Gotsopoulos H, Chen C, Qian J, et al. Oral appliance therapy improves symptoms in obstructive sleep apnea: a randomized, controlled trial. Am J Respir Crit Care Med 2002;166:743-8. [Crossref] [PubMed]
10. Grunstein RR, Stewart DA, Lloyd H, et al. Acute withdrawal of nasal CPAP in obstructive sleep apnea does not cause a rise in stress hormones. Sleep 1996;19:774-82. [Crossref] [PubMed]
11. Weaver TE, Grunstein RR. Adherence to continuous positive airway pressure therapy: the challenge to effective treatment. Proc Am Thorac Soc 2008;5:173-8. [Crossref] [PubMed]
12. Marcus CL, Brooks LJ, Draper KA, et al. Diagnosis and management of childhood obstructive sleep apnea syndrome. Pediatrics 2012;130:576-84. [Crossref] [PubMed]
13. Mitchell RB, Archer SM, Ishman SL, et al. Clinical Practice Guideline: Tonsillectomy in Children (Update). Otolaryngol Head Neck Surg 2019;160:S1-S42. [Crossref] [PubMed]
14. Sjölander I, Borgström A, Nerfeldt P, et al. Adenotonsillotomy versus adenotonsillectomy in pediatric obstructive sleep apnea: A 5-year RCT. Sleep Med X 2022;4:100055. [Crossref] [PubMed]
15. Zhang LY, Zhong L, David M, et al. Tonsillectomy or tonsillotomy? A systematic review for paediatric sleep-disordered breathing. Int J Pediatr Otorhinolaryngol 2017;103:41-50. [Crossref] [PubMed]
16. Kim JS, Kwon SH, Lee EJ, et al. Can Intracapsular Tonsillectomy Be an Alternative to Classical Tonsillectomy? A Meta-analysis. Otolaryngol Head Neck Surg 2017;157:178-89. [Crossref] [PubMed]
17. Wang H, Fu Y, Feng Y, et al. Tonsillectomy versus tonsillotomy for sleep-disordered breathing in children: a meta analysis. PLoS One 2015;10:e0121500. [Crossref] [PubMed]
18. Acevedo JL, Shah RK, Brietzke SE. Systematic review of complications of tonsillotomy versus tonsillectomy. Otolaryngol Head Neck Surg 2012;146:871-9. [Crossref] [PubMed]
19. Truong MT, Woo VG, Koltai PJ. Sleep endoscopy as a diagnostic tool in pediatric obstructive sleep apnea. Int J Pediatr Otorhinolaryngol 2012;76:722-7. [Crossref] [PubMed]
20. Benedek P, Balakrishnan K, Cunningham MJ, et al. International Pediatric Otolaryngology group (IPOG) consensus on the diagnosis and management of pediatric obstructive sleep apnea (OSA). Int J Pediatr Otorhinolaryngol 2020;138:110276. [Crossref] [PubMed]
21. Parikh SR, Boudewyns A, Friedman NR, et al. International Pediatric Otolaryngology Group (IPOG) consensus on scoring of pediatric Drug Induced Sleep Endoscopy (DISE). Int J Pediatr Otorhinolaryngol 2023;171:111627. [Crossref] [PubMed]
22. He S, Peddireddy NS, Smith DF, et al. Outcomes of Drug-Induced Sleep Endoscopy-Directed Surgery for Pediatric Obstructive Sleep Apnea. Otolaryngol Head Neck Surg 2018;158:559-65. [Crossref] [PubMed]
23. Boudewyns A, Abel F, Alexopoulos E, et al. Adenotonsillectomy to treat obstructive sleep apnea: Is it enough? Pediatr Pulmonol 2017;52:699-709. [Crossref] [PubMed]
24. Elshaug AG, Moss JR, Hiller JE, et al. Upper airway surgery should not be first line treatment for obstructive sleep apnoea in adults. BMJ 2008;336:44-5. [Crossref] [PubMed]
25. Franklin KA, Anttila H, Axelsson S, et al. Effects and side-effects of surgery for snoring and obstructive sleep apnea--a systematic review. Sleep 2009;32:27-36.
26. MacKay S, Carney AS, Catcheside PG, et al. Effect of Multilevel Upper Airway Surgery vs Medical Management on the Apnea-Hypopnea Index and Patient-Reported Daytime Sleepiness Among Patients With Moderate or Severe Obstructive Sleep Apnea: The SAMS Randomized Clinical Trial. JAMA 2020;324:1168-79. [Crossref] [PubMed]
27. Fehrm J, Nerfeldt P, Browaldh N, et al. Effectiveness of adenotonsillectomy vs watchful waiting in young children with mild to moderate obstructive sleep apnea: a randomized clinical trial. JAMA Otolaryngol Head Neck Surg 2020;146:647-54. [Crossref] [PubMed]
28. Browaldh N, Bring J, Friberg D. SKUP(3) RCT; continuous study: Changes in sleepiness and quality of life after modified UPPP. Laryngoscope 2016;126:1484-91. [Crossref] [PubMed]
29. Browaldh N, Bring J, Friberg D. SKUP(3): 6 and 24 months follow-up of changes in respiration and sleepiness after modified UPPP. Laryngoscope 2018;128:1238-44. [Crossref] [PubMed]
30. MacKay SG, Chan L. Surgical Approaches to Obstructive Sleep Apnea. Sleep Med Clin 2016;11:331-41. [Crossref] [PubMed]
31. McCulloch P, Altman DG, Campbell WB, et al. No surgical innovation without evaluation: the IDEAL recommendations. Lancet 2009;374:1105-12. [Crossref] [PubMed]
32. Field CJ, Robinson S, Mackay S, et al. Clinical equipoise in sleep surgery: investigating clinical trial targets. Otolaryngol Head Neck Surg 2011;145:347-53. [Crossref] [PubMed]
33. Lilford R, Braunholtz D, Harris J, et al. Trials in surgery. Br J Surg 2004;91:6-16. [Crossref] [PubMed]
34. McCulloch P, Taylor I, Sasako M, et al. Randomised trials in surgery: problems and possible solutions. BMJ 2002;324:1448-51. [Crossref] [PubMed]
35. Cook JA. The challenges faced in the design, conduct and analysis of surgical randomised controlled trials. Trials 2009;10:9. [Crossref] [PubMed]
36. Ross S, Grant A, Counsell C, et al. Barriers to participation in randomised controlled trials: a systematic review. J Clin Epidemiol 1999;52:1143-56. [Crossref] [PubMed]
37. Rosen CL, Larkin EK, Kirchner HL, et al. Prevalence and risk factors for sleep-disordered breathing in 8- to 11-year-old children: association with race and prematurity. J Pediatr 2003;142:383-9. [Crossref] [PubMed]