The Practice of Pediatric Sleep Medicine: Results of a Community Survey

DISCUSSION

The results of this survey of over 600 pediatric practitioners suggest that there still exist significant gaps both in basic knowledge about pediatric sleep and sleep disorders among pediatricians, and in the translation of that knowledge into clinical practice. The results also suggest that many pediatricians do not adequately screen for sleep problems in the clinical setting, and that this failure to screen is most likely to occur with older children and adolescents. Furthermore, despite the fact that the practitioners in our survey clearly acknowledged the importance of sleep disorders in children and adolescents, and the significant impact that sleep problems have on the health and well-being of their patients and families, the perceived level of confidence in their own ability to identify and successfully manage these disorders was low.

Despite the level of scientific progress that the field of pediatric sleep medicine has achieved in the last few years and the increased level of public awareness about pediatric sleep disorders, as evidenced by the number of books and articles published in the lay press and the level of media attention to these issues, the results of this survey are quite similar to those found in a comparable pediatric sleep knowledge survey of practicing pediatricians in the United States conducted 8 years before this study.¹¹ In comparing specific knowledge categories, pediatricians in both surveys clearly were most knowledgeable about developmental and behavioral aspects of children's sleep, and less knowledgeable about specific sleep disorders. In particular, the mean percentage of correct responses in our survey was <50% for the sleep disordered breathing, disorders of excessive daytime sleepiness, and sleep movement disorders category items. These findings also parallel those of a survey of Italian pediatricians,³⁷ in which knowledge scores for the categories of sleep apnea and narcolepsy were low compared with developmental sleep issues, as well as several surveys of adult practitioners³⁸ in which knowledge about disorders of excessive daytime sleepiness was also deficient. In no knowledge category did >50% of the pediatricians in our survey correctly respond to all the items, and the behavioral sleep knowledge category was the only one in which more than a quarter of the respondents answered all the items correctly. Interestingly, years of practice experience or proximity to training and academic versus community practice setting overall did not seem to significantly affect knowledge level; the exception was for the OSA subcategory, in which pediatricians who were more recently trained or practiced in academic settings appeared to be more knowledgeable.

Although a number of studies have demonstrated that pediatricians often fail to screen adequately for behavioral problems in general,^39–41 to our knowledge, ours is the only study which has specifically examined screening practices for sleep problems among a large sample of pediatricians. Although the majority of pediatricians in the survey did report routinely asking something about sleep issues in the context of the well-child examination, a significant percentage of those respondents asked only a single question for each age group, a practice that recent evidence suggests may be inadequate to identify all children with significant sleep problems.⁴² Furthermore, the combined percentage of pediatricians who did not screen or asked only a single question, ranging from 42% in infants to 52% in school-aged children to 74% in adolescents, is especially concerning in light of a number of recent studies documenting not only a high prevalence of sleep problems, particularly in adolescents,²³ but also resulting serious consequences on mood, behavior, and academic performance.^{21 ,43} Given the results of several studies suggesting that parents may underestimate the presence or magnitude of sleep problems in older children and adolescents,^{20 ,27} the impact of this age discrepancy is likely to be magnified by the fact that a low percentage of respondents also failed to ask the older child and adolescent directly about their own sleep habits.

Thus, as is the case with many behavioral concerns,⁴⁴sleep problems are likely to be underidentified by pediatric practitioners despite the prevalence and acknowledged importance of these disorders in children. A number of the reasons cited for underidentification of behavioral problems,⁴⁵ involving both practical considerations and issues related to practitioners' behavior and beliefs, may also apply to sleep problems as well. Contrary to surveys of pediatricians regarding other behavioral concerns such as ADHD,⁴⁶ in terms of more concrete concerns, our results indicated that lack of reimbursement did not play a major role in preventing pediatricians from asking screening questions about sleep. Similar to what has been reported in some surveys with adult primary care practitioners and sleep concerns,⁴⁷ however, time constraints were cited as an obstacle by a somewhat larger percentage (16%). Our results did not suggest that perceived low prevalence or lack of efficacy of available treatments for sleep problems are important deterrents to screening. Because we did not list the lack of available diagnostic and/or treatment services for pediatric sleep disorders as a reason for not screening, no specific conclusions can be drawn about this issue; however, the relatively high percentage of respondents in our survey who reported never or rarely referring patients to a sleep clinic or to a sleep lab for polysomnography suggests that relative lack of availability of such facilities might play a role.

Overall, our survey results suggest that a combination of practitioners' perceived low knowledge and comfort levels regarding sleep issues, and their beliefs, may have a relatively more important influence on screening practices. In particular, the assumption that parents will spontaneously raise concerns about sleep if they exist was cited as a reason for not screening for sleep problems. However, a number of studies which have examined this issue in regards to other types of behavioral problems have suggested that this assumption on the part of physicians about parents spontaneously raising concerns is often erroneous.^{48 ,49} Because studies of sleep screening practices of adult practitioners have also suggested that physicians tend to wait until the patient initiates discussion about sleep concerns,⁴⁷ it may be particularly important to directly address this potential miscommunication when educating both practitioners and parents about the identification of sleep problems in children.

Clearly one of the most important influences on physician knowledge levels, as well as beliefs and behavior in the practice setting, is previous exposure to both didactic instruction and role modeling during training. Previous studies have documented the scarcity of both didactic and alternative forms of instruction about sleep in general and pediatric sleep, in particular, during medical school and residency training. For example, a 1993 survey of medical school curriculae found that only 0.38 clinical teaching hours on average were devoted to pediatric sleep issues, the smallest time period of any surveyed topic.⁷ One study¹¹ that specifically investigated the level of pediatric residency education about sleep disorders in children and adolescents found that the mean number of hours of instruction on pediatric sleep was only 4.8 hours over 3 years. Fewer than one third of the programs offered didactic instruction on any of the basic pediatric sleep topics surveyed other than apnea and general sleep information.

In addition to these educational gaps, there is a relative lack of emphasis on sleep issues, especially screening for sleep problems, in many pediatric textbooks and other resources. For example, although “Bright Futures”⁵⁰ recommends screening and includes trigger questions for sleep problems during well-child examinations in younger children, there is very little emphasis on both screening and anticipatory guidance regarding sleep issues, especially in comparison to other health issues such as nutrition and tobacco use, in older children and adolescents. We recently performed an informal survey of 8 of the most popular pediatric textbooks in current use which revealed that a range of only 0.3% to 2.0% of the total texts was devoted to pediatric sleep topics, and of that sleep information, on average, about 50% was devoted specifically to the topics of infantile sleep apnea and colic (compared to an average of 4% to OSA).

Finally, it should be noted that pediatricians' personal experience with and beliefs about sleep and sleep problems may have a potential impact on their level of understanding about their patients' sleep problems, as well as their ability to manage them. As might have been predicted, for example, pediatricians whose own children have had sleep problems had higher scores on the Behavioral Sleep items. The physicians in this survey reported personal sleeping habits, especially amount of sleep, that were very similar to what has been reported both in other primary care physicians⁴⁷ and in the general public,⁵¹ although there was somewhat of a discrepancy in this sample between the acknowledged need for and perceived importance of sleep in their own lives, and their actual sleep behavior. The combination of a high percentage of respondents who reported a personal history of sleep problems, and the almost one third of the sample who are potentially chronically sleep deprived (6 hours or less of sleep per night) suggests that there may be a significant degree of inadequate and/or disrupted sleep in this population.

The issue of the low level of screening for snoring found in this study, even in the age groups in which OSA is most prevalent, combined with the findings regarding knowledge levels, evaluation, and treatment practices for OSA, deserves additional comment. Numerous studies have not only documented a variety of clinical sequelae related to OSA in children, ranging from growth failure⁵¹ to a host of behavioral and academic problems,³⁰ but have also suggested that many of the neuropsychological consequences, in particular, are reversible with treatment.^{31 ,52} Thus, early identification, particularly in high risk groups such as children with Down syndrome, repaired cleft palate, and obese children, and prompt and appropriate treatment should be a priority. Furthermore, because clinical symptoms alone have been repeatedly demonstrated overall to have poor predictive validity for OSA in children,^{53 ,54} familiarity with and access to appropriate diagnostic tools (ie, overnight sleep studies) once symptoms have been identified are also key factors. Clearly, additional research is needed to understand the barriers that exist or are perceived to exist by practitioners in regards to using polysomnography diagnostically, given that half of our sample rarely or never did so. Finally, the treatment practices in regards to OSA reported in our study suggest that although pediatricians are knowledgeable about surgical and weight loss management options for children, they may not be as familiar with alternative treatments that are more commonly used in adults with OSA, such as continuous positive airway pressure.

In terms of other evaluation and treatment practices for sleep problems in children reported in this study, most of the pediatricians were aware of risk factors and seemed comfortable recommending behavioral management strategies for sleep problems in younger children with primarily behavioral issues. As a group, they were less appropriate in their evaluation and treatment of partial arousal parasomnias such as night terrors and sleepwalking. For example, they often failed to recognize potential diagnostic clues such as the timing of the nocturnal events,²⁴ and seemed to attribute considerable etiologic significance to psychological factors. A significant percentage also recommended treating night terrors with diphenhydramine, a medication with no demonstrated efficacy in this disorder.⁵⁵ Inappropriate use of or inadequate knowledge about sleep medications is also suggested by the 20% of respondents who had recommended melatonin, a hormone which has its' primary effect on circadian processes,⁵⁶ for adolescents with poor sleep hygiene. The majority of respondents, however, did recognize the value of basic sleep hygiene principles such as a regular sleep schedule, “stimulus control” (restricting in-bed activities), and sleep restriction in addressing adolescent insomnia.⁵⁷ Finally, as might have been predicted by the low level of knowledge about disorders of excessive daytime sleepiness, most respondents failed to recognize a significant risk for narcolepsy in the profoundly sleepy adolescent.⁵⁸

Because the survey was anonymous, we were not able to obtain information on the nonresponders. This, coupled with the limitation of a <50% response rate, although commensurate with other similar physician surveys,⁵⁹ may limit the generalizability of the results. It is possible that physicians who are more interested in and knowledgeable about sleep in children may have been more likely to complete the survey, skewing the results toward overestimating the knowledge base of pediatricians in the sample. This was also a survey of pediatricians in a single region of the country, and thus the findings may not be representative of pediatricians in the United States as a whole. Furthermore, all of the data collected was self-report and not accompanied by behavioral observations, and thus we could not validate the clinical practices reported by the respondents. However, because physicians are more likely to overestimate desirable practices such as preventive screening,⁶⁰ it is unlikely that the survey results represent a significant underestimation of what actually occurs in the clinical setting. Although every attempt was made to devise a survey instrument that would have adequate content and construct validity, formal psychometric assessment of the instrument itself was not done. Furthermore, the forced choice format we elected to use to enhance ease of completion, and thus increase response rate, may not accurately reflect actual clinical practice as regards evaluation and management of sleep problems.