Abstract
PURPOSE:
Sleep related breathing disorders (SRBD) are seen at disproportionately higher rates in children with spina bifida compared with their same aged peers. SRBD such as obstructive sleep apnea (OSA) or central apnea are associated with developmental and cognitive consequences, and sudden death.
METHODS:
Participants aged 1 to 20 years with a diagnosis of spina bifida were recruited from a pediatric academic hospital spinal cord specialty clinic to evaluate the prevalence and impact of sleep disorders in the spina bifida population; 76 polysomnogram (PSG) reports spanning a 3-year period were reviewed in this retrospective cohort study.
RESULTS:
Of the PSGs reviewed, 37 (49%) indicated the presence of SRBD, and 28 (76%) of those children required an escalation of management (surgical intervention or additional respiratory support). These results are consistent with previous studies and further emphasize the clinical impact of SRBD on children with spina bifida by describing the interventions that followed an abnormal PSG.
CONCLUSION:
The high prevalence of SRBD in the spina bifida population supports the need for additional research to develop sleep questionnaires specific to spina bifida that can predict abnormal PSG clinically and to determine the standard of care following an abnormal PSG, chiefly in OSA and central apnea.
Introduction
Spina bifida is a birth defect that occurs when the neural tube develops abnormally in the first four weeks of pregnancy [1]. It is the third most common disability of childhood in the United States with over 166,000 persons affected [1]. Spina bifida is associated with comorbid hydrocephalus, Chiari II malformation and limb paralysis or impairment [3, 4]. Each comorbidity can potentially affect breathing responses while sleeping [3, 4]. Prior studies have reported a prevalence of sleep related breathing disorders (SRBD) between 2% and 11% in the general pediatric population, while the prevalence of SRBD in patients with spina bifida ranges from 41% in one study to as high as 81% in another [4–7]. In children older than one year, the most typically reported SRBD symptoms are apneas, shortness of breath, snoring, choking, and irritability [7]. Patients with SRBD can be asymptomatic. The two most common SRBD are obstructive sleep apnea and central apnea. Obstructive sleep apnea is a partial to complete airway obstruction causing sleep disruption and/or oxygenation and ventilation defects. Central apnea is a lack of respiratory effort, which is often due to compression of the brainstem from Chiari malformation and/or hydrocephalus in patients with spina bifida [7].
Typical clinical evaluations, including existing sleep questionnaires and routine exams, are often not accurate or sensitive enough to provide a SRBD diagnosis to individuals with spina bifida [6]. Therefore, polysomnography (PSG) is necessary to accurately diagnose SRBD and avoid morbidities and mortalities [4]. Others have previously observed that clinical presentation alone is not sufficient to exclude SRBD in the spina bifida population; Rocque and colleagues demonstrated that 42% of their spina bifida cohort had SRBD regardless of presence of specific symptomatology [4]. Prior studies used clinical evaluation and questionnaires to determine clinical concern about abnormal sleep in spina bifida through clinical evaluation and questionnaires alone [7]. The current study sought to investigate the prevalence of SRBD in patients with spina bifida from a multidisciplinary spinal cord disorders clinic to further define any additional potential risk factors for SRBD, identify specific PSG abnormalities to this group, and observe any emerging patterns in treatment.
Materials and methods
Retrospective chart review and data collection were completed for 203 patients with spina bifida currently receiving care at a multidisciplinary spinal cord clinic. To be considered for the study, patients must have been aged 1 to 20 years of age and must have had a PSG in the period between January 2019 and November 2021. Complete data were not available to indicate whether this was an initial or repeat PSG. Those over 20 were not served in the pediatric hospital that houses the spinal cord clinic (and thus data were not available). Patients under age one were excluded due to lack of reliable PSG standards for that age group. Patients who were ventilator-dependent were excluded, as this study was intended to evaluate the relationship of SRBD to spina bifida, not with respiratory failure from other causes (such as congenital malformations). There was no participant randomization since this study was a retrospective chart review. The pulmonologist was not blinded to conditions and was aware of patient medical history when reading sleep studies. Of the 203 pediatric patients who had a spina bifida diagnosis, 76 individuals met the inclusion criteria.
Reasons for referral for PSG varied. Commonly reported symptoms in patients who were referred for PSG included snoring, grinding teeth, problems going to sleep, problems staying asleep, restless sleep, sleep talking, witnessed pauses in breathing (apnea), mouth breathing, daytime behavioral problems, headaches, attention issues, daytime sleepiness, restless legs, and restlessness. These symptoms were mentioned in either history of present illness or review of systems in clinical notes.
PSGs were performed at an academic pediatric hospital in monitored rooms with family available in adjacent rooms. The following variables were recorded on the PSG: electroencephalogram (EEG; C3/A2, C4/A1, O1/A2, O2/A1) right and left electrooculograms, submental electromyogram, tibial +/- upper extremity electromyograms, electrocardiogram (EKG), abdominal and chest respiratory effort (Sensormedics inductance plethysmography), end tidal CO2 (BCI capnograph), airflow (Dymedix thermistor), and SpO2 (Nellcor pulse oximetry). Expanded 8-channel EEG monitoring and nasal pressure monitoring (ProTech PTAF2) were also recorded in PSGs performed after publication of the latest American Academy of Sleep Medicine (AASM) criteria. Eighteen continuous time-linked sound/visual recordings were also obtained using audio and infrared technology. Events were scored using AASM criteria and were defined as central apneas, central hypopneas, obstructive apneas, obstructive hypopneas, oxygen desaturations, and periodic limb movement syndrome (PLMS) [10, 11]. Data were also evaluated for abnormalities in end tidal CO2 levels, abnormal EEG and abnormal EKG. Standardized sleep parameters were used to classify participants’ polysomnograms as normal or abnormal using AASM criteria [10, 11]; these sleep parameters included sleep efficiency, arousal index, respiratory rate, snoring, mouth breathing, apnea-hypopnea index, central apnea index, obstructive apnea-hypopnea index, oxygenation and ventilation.
All PSGs were reviewed by a pediatric pulmonologist with certification in sleep medicine. Each PSG was comprehensively reviewed considering the overall clinical presentation of the patient and sleep parameters described above. A PSG was defined as abnormal based upon AASM criteria as evaluated by the pediatric pulmonologist. For the purposes of this study, the severity of SRBD was not stratified, as the focus was on clinical outcomes as a result of PSG. PSGs were sorted into three groups: requiring escalation of management, minimal intervention (example: counseling or dietary supplementation), or no additional intervention.
The primary source of referrals for the sleep studies was a pediatric rehabilitation physician and advanced practice registered nurse (APRN) affiliated with the spinal cord disorders clinic, but other members of the transdisciplinary clinic including neurosurgeons were additional sources of referrals. Non-affiliated specialists, such as ear, nose, and throat (ENT) and pulmonology, comprised the remainder of referrals. Notably, none of the PSG referrals originated from primary care physicians. Limited sleep questionnaire data was collected but could not be included in the analysis due to the retrospective nature of the study and lack of complete data sets for questioning. However, each PSG required the referring provider to indicate the reason for referral, and this data was collected from the electronic medical record. Reason for referral was selected by the provider by choosing an indication from a list, discussed in more detail in the Results section. All data was checked for possible data entry errors or outliers using summary measures, boxplots, histograms, and bar charts. Numeric data was summarized using group means and standard deviations (SDs) or medians and other quantiles, such as the first and third quartiles. Much of the data analyzed was categorical in nature and was summarized using contingency tables. The primary outcome (whether a patient had an abnormal PSG for SRBD) was binary. When evaluating the association between this and other categorical variables, Chi-square and Fisher exact tests were employed, depending on the expected cell frequencies. For hypothesis tests involving numeric predictors, such as age and sleep efficiency score, ANOVA methods were utilized. When the necessary assumptions of ANOVA, such as homoscedasticity and normality, could not be satisfied, non-parametric tests were used. All data was summarized and analyzed using R software (v 4.1). Institutional Review Board approval was obtained, and informed consent was discussed and waived since this research presented no more than minimal risk to participants. Protected health information was deidentified.
Results
This retrospective chart review revealed that nearly half (49%) of the PSGs completed by patients followed by the spinal cord disorders team during this period were diagnosed with a SRBD. The prevalence of SRBD in this spina bifida population was comparable to the prevalence of SRBD in spina bifida populations in prior studies. A large majority of the children with spina bifida diagnosed with SRBD required an escalation of care, primarily referrals to ENT and pulmonology.
Of the 76 spina bifida pediatric patients who met study criteria, 50% were female and 50% were male, with ages ranging from 1 year to 19 years of age. Demographics of the patients are outlined in Table 1. Sixty-four (84%) patients who completed PSG had a Chiari malformation, and 53 (70%) had a ventriculoperitoneal shunt, which is reflective of the spina bifida population at large. Of the variables examined in Table 1, only biological male patients were significantly correlated with SRBD on PSG (p = 0.03).
Demographics and PSG results
Demographics and PSG results
1 n (%). 2 Wilcoxon rank sum test; Pearson’s Chi-squared test; Fisher’s exact test. PSG = polysomnography; VP = ventriculoperitoneal.
As presented in Table 2, the majority of SRBD diagnosed were obstructive sleep apnea [31]; seven were diagnosed with central apnea. Of the 76 patients, four were additionally noted to have PLMS, two of whom also had obstructive sleep apnea and two with no SRBD findings.
Diagnoses from PSG Results
1 n (%). 2One patient was diagnosed with both obstructive and central apnea. PSG = polysomnography; SRBD = sleep related breathing disorders.
Each referring provider indicated specific sleep related symptoms that prompted a referral for PSG. Table 3 lists these symptoms and whether they correlated with an abnormal PSG. Though several sleep symptoms were commonly reported (72% of patients had trouble sleeping), only those with reported apnea or breathing problems were more likely to have SRBD by PSG (p < 0.001). Of the patients without reported symptoms, two (5.4%) were diagnosed with a SRBD.
Comparing Sleep Symptoms with PSG Results
1 n (%). 2 Fisher’s exact test. PSG = polysomnography; SRBD = sleep related breathing disorders.
Table 4 describes subsequent management of patients following PSG. Of the 76 patients, 34 were referred for more specialized care regardless of PSG outcome: 21 (28%) were referred to ENT, 15 (20%) were referred to pulmonology or a sleep clinic, and three (4%) were referred to neurology. Of those diagnosed with SRBD, 57% were referred to ENT. None of these treatment or referral patterns were correlated with the presence or absence of SRBD on PSG.
Comparing Management Courses with PSG Results
1 n (%). 2 Fisher’s exact test. ENT = ear, nose, and throat; PSG = polysomnography; SRBD = sleep related breathing disorders.
This retrospective chart review revealed that 49% of the pediatric spina bifida patients in this sample had abnormal PSG results (37 with obstructive sleep apnea or central apnea). Most patients with abnormal PSG results were male and were diagnosed with obstructive sleep apnea. When referrals were necessary, most went to ENT or pulmonology. The prevalence of SRBD in these pediatric patients with spina bifida referred for PSG was 49%. This likely represents an underestimation of patients, as it does not include those who were referred for PSG but did not complete it, nor does it include those who were not referred for PSG or to the sleep clinic. Reasons for lack of referral for PSG vary and could include family buy-in, resource availability, prioritization of other concerns, and infrequent follow-up. Data did not exist that would allow for further characterization of patient selection in this retrospective study.
Referrals were largely from providers in a multi-disciplinary program serving children with spina bifida. It is worth noting that no referrals for PSG were from primary care providers and very few were from ENT. This suggests a need for further education in both communities; characterization of these referral patterns along with additional research would be necessary to better understand this relationship. It is worth noting that patients referred by primary care physicians to other sleep labs outside the study institution were not included in the study. There are no other pediatric sleep labs available in the state.
Of the 37 children with SRBD and spina bifida, 28 required referrals and escalation of care, including surgical intervention (such as tonsillectomy) and increased respiratory support (such as continuous positive airway pressure [CPAP] or nighttime oxygen). Without PSG, some patients may not have sought further treatment. This study additionally affirmed what others have observed: the pediatric spina bifida population is at higher risk for SRBD with or without symptoms. In some cases, there was no sleep symptomatology indicated in the referral. While this may give some pause, the absence of symptomatology should not preclude a child with spina bifida from getting a PSG. Potentially harmful SRBD are not reliably accompanied by symptomatology in children with spina bifida.
The findings of SRBD in this pediatric spina bifida population were comparable to outcomes from other studies with the exception that others have noted more prominent central apnea than obstructive apnea. However, in this work, the ages were not directly comparable to other studies with more restricted age criteria [13]. Treatment strategies after PSG have not been fully elucidated in prior work but are presented here. A selection bias in this retrospective analysis is acknowledged, in that generally only spina bifida patients in which there was clinical suspicion of SRBD underwent PSG. Since this study was conducted retrospectively, many factors were not controlled, including referral to sleep clinic, standardized questionnaires or practices, and documentation for clinical reasoning. Standard sleep questionnaires were administered as a part of a distinct sleep clinic protocol; however, not all patients went to the sleep clinic prior to undergoing a PSG. The Pediatric Sleep Questionnaire (PSQ, sleep disordered breathing subset) and Children and Adolescent Sleep Checklist (CASC) were the standard sleep questionnaires used. Data were not available to determine any correlations these questionnaires had with demographics or PSG results. It should be noted that existing sleep questionnaires do not have sufficient specificity and sensitivity to be relied upon as a sole factor guiding sleep management [12]. Additionally, PSG prior to the study period could not be queried in detail due to lack of access to a remote and obsolete electronic medical record system.
The prevalence of SRBD in the children and adolescents with spina bifida at the study institution was 49%, far greater than typical peers. The results demonstrated that children and adolescents with spina bifida are far more likely to have SRBD and may have consequently higher risks for morbidity and mortality related to SRBD. Others have observed an alarming trend of sudden death in young adults with spina bifida who die during sleep; given the temporal connection with sleep and unexpected death, further investigation of SRBD in spina bifida is urgently needed.
Future research endeavors should include developing a sleep questionnaire specific to spina bifida, finding a pattern to predict which patients will have an abnormal PSG, and establishing standardized guidelines to manage steps following an abnormal PSG (for obstructive sleep apnea and central apnea particularly). Additionally, specific documentation of why referrals were not made may be useful in delineating barriers to PSG. The prevalence of SRBD in this pediatric spina bifida population was comparable to findings in previous studies, and this chart review confirmed that pediatric spina bifida patients are at higher risk of SRBD compared to their peers. This suggests that children and adolescents with spina bifida are under-diagnosed and under-referred for SRBD at the study institution.
Conclusion
Almost half of the patients with spina bifida from the multidisciplinary spinal cord clinic were diagnosed with SRBD, which falls within the range of previous studies’ findings. The results of this retrospective study illustrate the clinical importance of SRBD in the spina bifida population by demonstrating the prevalence of SRBD in patients with SB at this institution, the significant need for PSG screening regardless of symptomatic presentation, and the considerable number of referrals to escalate care following an abnormal PSG. Here the abnormal findings in PSGs as well as the potential treatment options in those cases were characterized. Further research is urgently needed to explore outcomes after interventions, creating sleep questionnaires specific to the spina bifida population, and barriers to accessing quality sleep care.
Footnotes
Acknowledgments
The authors gratefully acknowledge the Spinal Cord Disorders staff, Translational Research Institute at UAMS, and Arkansas Children’s Hospital Research Institution for their guidance and support. The project described was supported by the Translational Research Institute (TRI), grant UL1 TR003107 through the National Center for Advancing Translational Sciences of the National Institutes of Health (NIH). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
Conflict of interest
Research reported in this publication was supported by the National Center for Advancing Translational Sciences of the National Institutes of Health under Award Number UL1 TR003107. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The funding did not influence the study design, collection, management, analysis, interpretation of the data, or publication of this article.
Ethical considerations
The Institutional Review Board approval number is 239538 and the retrospective cohort study was performed from January 2019 to November 2021. Informed consent was waived since the study presented no more than minimal risk to participants.