Hi, I'm Aaron Glick, and I'm excited to review the evidence of what works in screening and treating obstructive sleep apnea. I'm a diplomat of the American Academy of Dental Sleep Medicine, and I'm on faculty at the University of Texas Health Science Center at Houston School of Dentistry. Here are my conflicts of interest. I do healthcare innovations with the Texas Medical Center, and I work with a pulporegeneration startup Nangiotex. So in this hour, we'll review what it means to have an evidence-based practice, a general overview of what sleep apnea is, and from there, we'll highlight some screening and treatment options and the current evidence associated with these options. As a dental provider, having an evidence-based practice is made up of gathering the best evidence available, evaluating this evidence, and then incorporating this evidence as a component of your clinical decisions when you're treating patients. The evidence I'm talking about is rigorous scientific research, which forms the basis of clinical practice. The current available evidence should add to our clinical knowledge, and by having an evidence-based practice, it allows us dentists to improve our clinical decision-making to maximize patient outcomes. In conjunction with patient-centered practices, we can make a clinical decision that can best suit the patient with the most current technologies or methods available. So if we all care about our patients, why isn't everyone using evidence-based methods? I'll briefly review some of the difficulties in using evidence-based methods and why that might be the case. So why is it tough to keep up with evidence-based care? In general, there are three main reasons why it's tough to keep up with the evidence. Number one, dental seat medicine, like most health fields, is rapidly evolving, and the amount of new information and how long that new information stays current can make it difficult for us to keep up with the evidence. Two, given the amount of new information, some is not accurate and others can just sometimes be misleading. Three, like any busy clinician, you have multiple competing priorities for your time, and to be constantly reading articles takes effort to prioritize. We've tried to distill some of the current literature on the topics in the dental seat medicine field in this lecture. Just to further illustrate how time-consuming keeping current with just the literature that's published on obstructive sleep apnea, we see here a graph of only indexed articles that were published and found on PubMed on the topic of obstructive sleep apnea. You can see here that in the past couple years, there were roughly 2,000 articles, so that's about five articles per day that you would need to read just to keep up with the topic of obstructive sleep apnea, and that's not counting for any breaks on the weekend either. It's not surprising that we, as clinicians, don't have enough time to critically read this mass amount of articles and implement evidence-based changes in our office, even if we'd like to. As we further talk about the evidence, we'll be using this evidence pyramid for each subtopic. We can pictorially see the levels of evidence as a pyramid. The pyramid has multiple different types of articles within it. These types of articles are based on scientific rigor and the setup of the experiments. As you move upwards in the pyramid, the evidence is stronger and the quality of evidence increases. The number of articles also will decrease since systematic reviews and meta-analyses will combine multiple other single studies. So if you were to look at the evidence in a particular topic as a dentist for clinical decision-making, your best search strategy would be to find as many articles from the top of the pyramid, and if no evidence exists at the top of the pyramid, then work your way downward to find the highest level of evidence in the articles on the pyramid that you can. So the environment in which the study was conducted matters as well. For instance, there are studies that can be performed on humans or animals, and this would be called in vivo studies, versus ones conducted on a lab bench would be in vitro studies. When we look closer at in vivo studies, those in humans, these experiments could be done in highly controlled clinical lab, or they could be performed in a traditional dental office. Now there are different types of studies that will gather the qualitative data, like how subjects feel, or more quantitative data like pulse oximetry, readings, or AHI. We won't review all the different type of articles, but I do want to mention meta-analyses since we'll be reviewing a number of these types of studies. Since they're at the top of the pyramid and at the highest level of evidence, we'll be seeing these in frequency later. Now a meta-analysis is a method to statistically analyze multiple independent studies. Typically a systematic review will perform a meta-analysis, however not all systematic reviews will have a number of adequate studies or access to the quality results to perform the meta-analysis. So a systematic review is an article that has a standardized method of finding and critically evaluating independent single studies. Typically if the meta-analysis is appropriate and is performed, you'll likely see a graphic like you'll see on the right. This is called a forest plot. So we can use a forest plot to represent the heterogeneity of multiple independent studies. In a forest plot, each study is usually labeled with the first author's last name and date that the study was published. Here we see the studies labeled as 1, 2, and 3. The horizontal line to the right of the picture shows the confidence interval of the result for a particular outcome measure. So the longer the line represents the more variability of the results of the study. The box represents the point estimate of the standardized mean difference between the test and control group of that study. The size of the box represents the meaningfulness of the study, and this is how much the individual study was weighted in the final pooled data. So the larger the box, the more that that study contributed to the pooled data. The diamond at the bottom shows the pooled estimate of all the studies where the middle of the diamond represents the pooled point estimate and the horizontal edges of the diamond represent the pooled 95% confidence interval. So there will also be statistical tests for heterogeneity. If a treatment works ideally, we'd see a low amount of heterogeneity between studies, telling us the treatment is effective and the results are easily reproducible. The statistical test for heterogeneity is often reported as I squared. So in this example, we see that the diamond is to the right of the vertical line and this vertical line is the line of no effect. So anything on the line shows that there was no effect between the experimental group and the control group. Ultimately, the meta-analysis will tell us how precise the pooled results are and if the results between studies are heterogeneous. So how do we as clinicians interpret and use this information? Well, in the example, the pooled point estimate of the diamond favors using oral appliance versus no treatment. In this case, the pooled confidence interval shown by the edges of the diamond are somewhat wide. We therefore can look at the boundary of the pooled confidence interval to think about what harm would be caused if crossing the line of no effect. Here would be that oral appliance does not work to treat sleep apnea. The additional harms would be whatever side effects the treatment of sleep apnea, sorry, the oral appliance caused. So you can imagine that in a surgical intervention, if you cross the line of no effect, this might cause a larger amount of harm and deciding not to use this intervention might be warranted. The risk of harm might be too high with little possibility of benefit for the patient to want to consider undergoing the surgery. As clinicians, we can then use this information in the context of the patient, treatment costs and other clinical factors to decide if continuing with the treatment is the best in the patient's case. I also want to introduce critically appraised topics. These are a type of secondary reporting that summarizes the evidence or recommendations that are clinically relevant. They typically summarize the best available research and are less rigorous than meta-analyses or systematic reviews. Some examples are here on the left of the screen. In this context, the lecture will be patterned from this concept of critically appraised topics where clinically relevant questions will be introduced, searched in the literature, and then recommendations will be formed. Now this is an example of our analysis that we'll dive into later. At the top, we'll see the clinical question. So as an introduction to our clinically directed question is, will this course be interesting? On the left, below the clinical question, we see the level of evidence. If you remember from the pyramid that we showed earlier, this is the level of randomized control trial. Typically, the level of evidence represented here will be the highest of the collection of articles that were found during our literature search. To the right, we see the evidence. In this case, there was only one randomized control trial that was found. The article is denoted by the last name of the first author and the year it was published. In parentheses, we have the type of article, the randomized control trial, and how many patients were investigated for this study. As you can see here, only one patient was tested, leading to a high risk of bias denoted below the answer on the bottom left. So the clinical question that we asked here was, will this course be interesting? And the answer at the bottom left says yes, although given the fact that there were no meta-analyses and only one patient tested, we might want to be skeptical of this answer. Here are some abbreviations that we might want to use during the rest of the lecture. So I've included this slide here just for your reference. The abbreviations are not used to confuse you, but to make it easier for everything to fit on one page, to summarize the evidence for each clinical question. Most abbreviations will be used throughout will be MA for meta-analysis, SR for systematic review, since it's the highest level of evidence and we're most interested in looking at these articles that summarize the vast amount of articles that have been done on the topics. If there were no meta-analyses or systematic reviews found using the preliminary search strategy, then feel free to refer back to the slide if you'd like. As a caution, these topics that were investigated, we didn't use a highly comprehensive or exhaustive search strategy, so it's possible that there is more evidence than was presented in this lecture. Additionally, in research articles for independent studies, if there's no statistically significant difference between a new treatment option, then publishing articles on that treatment, it's difficult to publish for researchers basically. If research is not then published, then it's likely that the information does not then enter common knowledge. A paucity of peer-reviewed research can suggest that the modality is either new, it's difficult to research, or it could suggest that the treatment has limited benefit. We see that the burden of proof would then lie with performing the treatment. Before we jump into the analysis, I want to briefly review a basic introduction to obstructive sleep apnea. Throughout this lecture, I'm assuming that you have limited or no knowledge of sleep apnea as a dentist. So let me describe how sleep apnea affects our patients. We can think of the airway as a compressible tube. When the patient falls asleep, their muscles relax, their jaw falls back posteriorly as the gravitational forces move on the jaw and impinge the airway. The tube, which we are calling the airway, then starts collapsing. So we start collapsing. This first collapse has an approximation of the tissues leading to snoring, which is then the vibration of the tissues as the patient breathes at night. Next as this tube further collapses, so that tube is going to now collapse all the way, we then have an apneic event. There's no air moving in or out of the respiratory system, so this leads to an increase of carbon dioxide in the brain, which is then detected by central chemoreceptors. The brain becomes alert and a cortical arousal will then occur, the brain wakes up, and this leads to overventilation where the patient breathes off all of their CO2. A patient will then be gasping for air, and in some patients, this is like a thermostat that overshoots and undershoots the temperature. These patients will then not breathe, and the cycle continues again and again. The patient falls back to sleep, and the airway collapses. Sadly, in some patients, this cycle can happen 100 times per hour in a very severe patient. So what causes sleep apnea? Well, obviously, there are many reasons that can cause sleep apnea. This is a busy slide, and it's not intended for you to know all of these etiologies. It's just presented so that way you know that sleep apnea is complex, and it's a multifactorial disease. Now, as the model of healthcare focuses on preventing the end-state diseases, I do want to introduce the concept of sleep disorder breathing. Now this is a continuum from primary snoring, upper airway resistance syndrome, and mild, moderate, and severe obstructive sleep apnea. When we look at the severities of sleep apnea, they're generally categories based on AHI, which is apnea hypopnea index. We'll be using the term AHI later, so it's good to know the approximate scale, and also that the scale is different depending on the lack of breathing in a pediatric or adult patient. AHI is a proxy for the severity of sleep apnea, and has become under question in past years. However, as we look at this evidence and ask ourselves what affects sleep apnea, this will be the sole measurement that we'll use to make our decisions a little bit easier, so that way we can quantify the severity of the disease. So what is AHI? Well, AHI is an index of the number of times that a patient has restricted breathing throughout the night. It's agnostic to the amount of time slept during the night, since it's a total number of restrictions in breathing throughout the night, then you divide it by the total of hours slept, leading to an index per hour. Patients will typically have a sleep study conducted at their house. This would be called a home sleep apnea test, also known as a HSAT, or they could have it in a laboratory setting called a polysomnography, or PSG for short. The gold standard for detecting and scoring AHI is the PSG, so that's the in-lab sleep study. We prefer to see the PSGs conducted as we look at the research articles, since at this point it's considered the gold standard. So again, home sleep test, HSAT, is a sleep study performed at the home. The PSG is a sleep study performed at the lab, and the sleep studies are taken by the patient to diagnose if they have sleep apnea. Both of these sleep studies can be used to calculate the AHI, which stands for apnea hypopnea index, and the AHI measurement will then determine the severity of the sleep apnea disease. So we're going to break down the four main reasons why obstructive sleep apnea occurs. As we answer some of the clinical questions, having a brief understanding of the pathophysiology of sleep apnea will be helpful. All patients will have some sort of contributing anatomical issue that will predispose them to having sleep apnea, yet nearly all of these patients are not walking around during wakefulness not being able to breathe. So 70% of patients will have an anatomical impairment, plus one of these other reasons for sleep apnea. These non-anatomic reasons are ineffective dilator muscles, high loop gain, and low arousal threshold. First, though, we'll start with the major anatomical issues that can arise. Now, obesity plays a major factor in the impaired anatomy, but why is obesity not 100% correlated with sleep apnea? Well, obesity is an independent risk factor for sleep apnea, yet it doesn't explain all the cases of sleep apnea. Traditionally, we think of a large male that has sleep apnea, although this is not always the case. There are many petite females that also have the disease. Presented here is one theory as to why our anatomy is an important component of sleep apnea, and it's also good conceptually to learn about the anatomy. So we see here, this is the isotope box theory. We can think of our craniofacial skeleton as a bony enclosure, and we're going to line that with some soft tissue. So if you have a normal amount of bony enclosure, we have a normal soft tissue that we line it with, then we see here that we have a normal airway size on the top right. If now we have an obese patient, we have that similar size bony enclosure for the craniofacial skeleton, yet we line it with an increased amount of soft tissue. In this case, the airway size will be limited. If we take the same scenario with a patient that has a small mandible, we have a smaller bony enclosure, yet line it with the normal amount of soft tissue, you can imagine that you would also have, on the bottom right, a smaller airway size as well. Also, the distribution of fatty deposits is particularly important. Those with fat around the pharyngeal area, which is a more compressible area, is more likely also to cause displacement of the airway. So here's an fMRI image of an obese versus a normal BMI patient. As you can see, the fatty depositions on the neck, but also the tongue as well. Those with increased fatty depositions in the tongue are more likely to have sleep apnea, as you can see on the picture on the right. Ultimately, whenever the anatomy predisposes the airway to collapsibility, the more likely that sleep apnea will occur. This collapsibility can occur at different levels throughout the pharyngeal portion of the upper airway. It can also occur in different dimensions, directions, or it could happen with a complete or partial collapse. So we just talked about an impaired anatomy, of which 100% of patients will have some sort of component of the impaired anatomy. Of these 100% of patients, 70% will have another component that predisposes the patient to having sleep apnea. These can be ineffective dilator muscles, low arousal threshold, or high loop gain. So now we're going to go ahead and take a deeper dive into ineffective dilator muscles. So the upper airway is dramatically affected by sleep. The laryngeal, intrathoracic, and tracheal structures are supported by collagenous structures, and therefore are rigidly supported and support an open airway, as does the nose. However, the pharyngeal component, so this is the nasopharynx, velopharynx, oropharynx, and the hypopharynx, they don't have the same support. So mechanistically, we can think of the upper airway as a starling resistor. As the diaphragm contracts, this causes an increasingly negative pressure, or a suction, that allows us to breathe in. This is similar to being in a shower. So when you turn on a hot shower and you're located in that shower, when the shower curtain comes inwards, you're in an area of negative pressure. So you can think of yourself inside the airway, and the tissues have a propensity to move into the airway where there's a negative pressure. Because the pharyngeal tissues are not rigidly supported, this suction can cause a collapse, and we see this noted by the tissue pressure in this diagram. Concurrently, there is a, they're dilator muscles, right? So these act to open up the airway as muscular pressure, and that's denoted with the muscular pressure as you see here in the diagram. So these dilating muscle muscles, the force will try to overcome the suction of the tissue pressure. However, once they can't overcome these forces, the airway then collapses, and the pressure that the airway collapses would be called the critical pressure. In addition, Bernoulli's principle tells us that with increasing air velocity, the pressures along the wall of a tube will decrease, creating more negative intraluminal pressure. And changes to a more turbulent pattern of flow along the airway walls will further reduce the pressures leading to a collapsibility. So to prevent the negative tissue pressure, we have dilator muscles that increase activity during inspiration. Here are some examples of dilator muscles that have been identified as important during breathing. Some of these are the tensor valley palatini, levator valley palatini, and the patel pharyngeus. We also see one of the most well-studied muscles, the genioglossus and its relationship to the airway here. The hyoid is also an important structure that's suspended by muscles. The more inferior the hyoid position, the more likely a patient is to have sleep apnea. Similarly, the longer the neck of the patient, the more likely they are to have sleep apnea. So we can basically think about the muscles in three ways. First, the brain needs to be able to send a signal to the muscle to contract. Second, the muscle needs to respond to that command of the brain. And lastly, the muscle needs to be effective in opening up the airway. So the muscle might contract, but if the muscle fibers aren't oriented properly, or they're not actually opening up the airway, then the muscle's really not being effective. So now let's go ahead and go back to our four reasons for sleep apnea. So we have impaired anatomy, ineffective dilator muscles, which we've already talked about. Next, we're gonna talk about low arousal threshold. Now, most of the times a patient has difficulty breathing throughout the night, the brain will not wake up. When the brain does wake up, this is known as a cortical arousal. Ideally, a patient would resolve their airway obstruction without a cortical arousal. However, 30 to 50% of sleep apnea patients will wake up with very small changes in negative pressure. So that was that critical pressure we talked about as well. These patients would be considered to have low arousal threshold. Constantly waking up throughout the night can lead to sleep fragmentation, which is a form of sleep deprivation. Again, let's go back to the reasons, the four reasons for sleep apnea. We already talked about impaired anatomy, so 100% of patients will have that. And 70% of patients will have one or more of the following, ineffective dilator muscles, low arousal threshold, and high loop gain. So now we'll focus on high loop gain. Now loop gain allows us to conceptualize the sensitivity of the respiratory control system. Now we can think of this as affecting the breathing stability of the patient. One third of sleep apnea patients will have high loop gain, and patients that have this breathing, basically it's unstable breathing, similar to a thermostat that over and undershoots the proper temperature. All right, so we've talked about now four reasons for sleep apnea. That's impaired anatomy, ineffective dilator muscles, high loop gain, and low arousal threshold. So let's now start the analysis. In this first section, we'll investigate some modalities for treating sleep apnea. Some of these include myofunctional therapy, frenectomy, maxillary mandibular advancement, rapid palatal expansion, and as a bonus, we'll dive into a little bit of sleep paroxysm. So what is myofunctional therapy? Well, myofunctional therapy is a repetitive exercise, particularly the isotonic and isometric exercises that are intended to build strength, proprioception, and coordination of muscles involved with the upper airway and mouth that would ultimately affect the collapsibility of the airway. So the theory of oropharyngeal myofunctional therapy is that patients can improve the effectiveness of their muscles and therefore reduce the collapsibility of the airway. This is based off the principle that some patients have impaired muscular dilator muscles during sleep, contributing to worsening sleep apnea, as we reviewed earlier. Therefore, exercising the oropharyngeal muscles could potentially improve functionality of the airway and improve the endurance of the muscle fibers. Exercises can include oral sounds, tongue movements, facial movements, swallowing, chewing, and other exercises with devices. So the question that we ask is, does myofunctional therapy reduce the severity of sleep apnea? The level of evidence was high as we had two meta-analyses and one systematic review. Compared with other meta-analyses, the number of subjects was lower, and in the DeFelicio article, since this was not a meta-analysis, we have denoted the number of studies that they investigated instead of the number of patients used for the analysis. The ultimate answer is that yes, myofunctional therapy does reduce the severity of sleep apnea, yet because there is some heterogeneity in the findings from these articles, we denoted that there was moderate risk for bias. Another point here is that we looked at both pediatrics and adult populations. Both of these populations are different, particularly when investigating sleep apnea. In children, sleep apnea is less prevalent, and given that children have a developing and growing craniofacial structure, sleep apnea can improve over time without intervention, whereas in adults, this is not the case. First, the first meta-analysis listed was solely focused on children, whereas the other two studies listed looked at both pediatric and adult populations. So we'll first look at the myofunctional therapy in adult population. The meta-analysis for the adult population had a total of 73 patients in the analysis, and the individual studies comprised of two randomized control trials and one retrospective case series. And all of those studies were based off the results of polysomnography, which is the in-lab sleep tests. The pooled results of this analysis show that the heterogeneity was significant, so the authors of this meta-analysis excluded two of the five studies. Both excluded studies were case series. In the studies where myofunctional therapy was continued for more than three months, they did find that, on average, there was a 50% reduction in the AHI. Additionally, the lowest oxygen saturation decreased in the majority of patients. However, overall, lowest oxygen saturation varied between studies. Individual studies suggest that snoring frequency is reduced as well as after myofunctional therapy. In the systematic review by DeFelicio, there was not an attempt to do meta-analysis. So as this publication came after the meta-analysis that we just reviewed, there are three articles of overlap. So the studies labeled four to seven were not covered in the last meta-analysis since they were published later. We see here that the final outcomes of AHI percentage decreased before and after myofunctional therapy in the adult population. In moving to the pediatric population, we see that in the next meta-analysis with six independent studies, making up of 160 patients total, that there's approximately a two AHI decrease for the children that underwent myofunctional therapy. Of the six studies, two used home sleep tests and four used the in-lab sleep studies. Additionally, two of the studies were randomized control trials and four were prospective case series. So from this table, three of the studies were excluded due to subgroup differences causing heterogeneity and two due to insufficient data to calculate. So in this analysis after they did that, there was no heterogeneity. So the actual process and procedures used for myofunctional therapy can vary widely. And this might account for a lot of the heterogeneity between the studies. Also in general, the patient compliance is another factor for this type of treatment. And in the meta-analysis, it wasn't reported how the investigators were tracking the compliance. We also don't know if this is a dose response, meaning that the more you treat a patient, the more the AHI decreases. Theoretically, we'd expect that like going to the gym, the more you work out, the stronger you'd be. Similarly for myofunctional therapy, we would expect that the benefits be a dose response to a limit. In the same line of thought, most of the adult studies were looking at improvements after three months. So it might be important that patients continue this therapy indefinitely to receive the continued benefits of this treatment. Lastly, there were a limited number of patients associated with these meta-analyses and the quality of the studies within the meta-analyses were not at the highest quality. Regardless, it does appear that generally the AHI will improve with myofunctional therapy up to 50% on average for the adult patients and approximately two AHI in the pediatric patients. So our next topic is ankyloglossia. Ankyloglossia or tongue tie is the excess connection between the tongue and the floor of the mouth. The thought here is that ankyloglossia creates a functional issue that leads to feeding, speech issues and sleep apnea. And for sleep apnea part of it, a limitation of tongue movement that reduces the ability of the tongue to be protruded out of the posterior area of the airway. As we know during the inspiration of the hypoglossal nerve stimulates the tongue to contract thus leading to a protrusion or contraction in the muscle of the tongue, which is the genioglossus, right? In most patients. So the theory of phrenectomy is to cut the lingual frenulum if it's functionally not allowing the tongue enough mobility thus we would relieve this connection. The tongue would then be able to move with more freedom potentially to protrude the tongue more during inspiration. Typically the lingual frenulum is cut using a laser scalpel or scissors and it's performed in both adults and children. There've been a number of articles that have reported the association between sleep apnea in adults or children with the presence of lingual frenulum. It seems like there might be a higher incidence of shortened lingual frenulum and those that have obstructive sleep apnea. Additional claims that we won't investigate and that are also more difficult to test is that is this limited mobility? Does it reduce the simulation between growth and development leading to less bone growth? There are also claims that we won't investigate that ankyloglossia has an association with mouth breathing and subsequent craniofacial changes. So our clinical question is, does phrenectomy reduce the severity of sleep apnea? There were a number of studies that were looked at for this topic that we found during the search strategy. Of the studies found they didn't have the objective data about the patient's sleep reported or the authors combined treatments like adenotonsillectomy. So only one randomized control trial was identified for this topic. Since this article did not report the AHIs specifically and didn't use any statistical tests to compare the experimental and control groups using the specific AHI measurements, unfortunately we couldn't come to a conclusion. Given the paucity of data, I don't think that we can conclude that phrenectomy does reduce the severity of sleep apnea. In this randomized control trial, 32 pediatric patients were randomly assigned to an experimental group where they had a PSG then laser phrenectomy, myofunctional therapy, speech therapy, and then were tested again with PSG three months later. The control group did not have the phrenectomy yet had the same myofunctional and speech therapy with the same PSG protocol. In their analysis, they did not report AHI and instead they categorized the severities of sleep apnea for both groups and then they performed a CHI-squared test. So, you know, which is a little bit unusual to report. Usually you would use the individual AHI data. Other concerns for this study were that the subjects were not matched between groups and the experimental group had 50% severe sleep apnea patients when the control group had only 18% severe sleep apnea subjects prior to the treatment. So ultimately there's little evidence to support that phrenectomy reduces the severity of sleep apnea. More literature focuses on the classification of what constitutes a shortened lingual frenulum and the increased presence of ankyloglossia in patients with sleep apnea. Now it's still unclear which classification or association of ankyloglossia is pathologic and if this leads to sleep apnea. So just the presence of ankyloglossia doesn't mean that there is a functional issue. However, ankyloglossia can cause other squalates such as speech and difficulty for infants to breastfeed. Both the American Association of Orthodontists and the American Academy of Otolaryngologists, I'm sorry, Otolaryngology do not support that phrenectomies be performed for the purposes of sleep apnea. Our next topic is maxillomandibular advancement. So what is this procedure? It's a surgical procedure where typically an oral surgeon will reset the maxillomandibular relationship in a more anterior location involving a Lefort I and a sagittal split osteotomy. As we reviewed earlier, retrusion of the craniofacial structures increase the potential for collapsibility of the airway. So this surgery repositions the craniofacial structures away from the airway. So that's the maxilla and the mandible. So our clinical question is, does maxillomandibular advancement surgery reduce the severity of sleep apnea? Now there are three meta-analyses found through the search strategy. All three meta-analyses had low heterogeneity and found that the surgical success rate of maxillomandibular advancement surgery was 85 to 86%. The success rates were defined as having an AHI of less than 20 with a 50% reduction from pre to post surgery. Now the percent of patients that had an AHI less than five after the surgery was found to be 39 to 46%. So due to the low heterogeneity between the meta-analyses and the findings of the meta-analyses themselves, the conclusion was that yes, maxillomandibular advancement does reduce the severity of sleep apnea and that this answer has a low risk of bias. When we look at some of the finer details of maxillomandibular advancement, we see that like we mentioned before, there's low heterogeneity as we see from the overall results of this meta-analysis by Zaghi. In this particular meta-analysis, they looked at 45 studies of which 36 studies had AHI data instead of RDI. 50% of patients had a benefit of 46 AHI after surgical correction. And some patient specific factors of patients that were more likely to achieve AHI less than five were those that were younger, lower BMI, lower preoperative AHI and higher preoperative SpO2 nadir. So basically these are patients that were younger with less severity of the disease before they started with the surgery. Similarly, the lower the severity of the disease preoperatively, the more likely that the patient was to have lower postoperative AHI. As you'd expect patients with the most severe sleep apnea to start had the greatest magnitude of reduction of AHI. So in the analysis, 518 patients all experienced some improvement in AHI or RDI except six patients. One last factor that was investigated was the amount of advancement of the mandible. So an advancement of 9.5 millimeters compared with 7.9 millimeters was associated with better success. So the further the advancement, the better the surgical success rate. Ultimately the maxillomandibular surgical procedure has good surgical rates with almost all patients receiving some sort of benefit reduction of RDI or AHI. This is a surgical intervention. So there is morbidity and mortality risks. Major complications occurred in one to 3% of cases, major complications. So some of the examples of these would be the need to reoperate to remove a screw or a plate, maxillary non-union or a sudden difficulty in breathing. Our next topic is rapid maxillary expansion or rapid palatal expansion. This is the treatment to correct transverse deficiencies and deficiencies. The teeth are generally used as anchorage. However, bone can be used as well. This is generally performed in a pediatric population before the mid-palatal suture has ossified where the ossification occurs around 12 to 16 years of age. So we'll only be looking at the pediatric population for this clinical question. The expanding forces on the teeth or bone will split the mid-palatal suture with continued activation, will continue to increase the arch perimeter. The thought is that the rapid maxillary expansion will widen and flatten the maxilla where it's displaced inferiorly, leading to a change in the alignment of the mandible. After this inferior displacement of the maxilla, it would lead to a nasal cavity having an increased volume and this increase in nasal cavity volume possibly leads to a decrease in nasal resistance and improvement in nasal breathing. The increase in mandibular alignment would potentially improve the ability for the tongue to rest in a more protruded position. And this has also been described as muscular hydrostatic theory, basically where the tongue acts as a waterbed and if there's not enough room anteriorly, the tongue will then be pushed posteriorly into the airway. So our clinical question is, does rapid maxillary expansion in children reduce the severity of sleep apnea? To answer this question, we found four meta-analyses. So our level of evidence is high. However, there were limited randomized control trials reviewed in these meta-analyses and the heterogeneity between the studies were high. The studies did find that reductions in AHI for patients undergoing rapid maxillary expansion. However, we should view these results with some skepticism denoted by the medium risk of bias. Now this is a forest plot for one of the most recently published meta-analyses on the topic. Only three of these studies had control groups and only one randomized the selection of groups. In this meta-analysis, the authors found that the mean decrease of AHI as a result of the treatment was approximately six events per hour. So that's six AHI. This result was found to be statistically significant, yet did have high heterogeneity. In another of the more recent meta-analyses, these authors looked at 17 pediatric studies with 314 patients. These studies found a benefit in rapid maxillary expansion on average of five AHI. In the analysis, this was a decrease of approximately nine AHI to three AHI. The cure rate or rate of this treatment leading to an AHI of less than one was 15% to 38% in the individual studies. So it's also found that if the child had a small or no tonsils at all, that there was a larger reduction in the AHI with the rapid maxillary expansion alone. So ultimately, there was a reduction of AHI after the treatment of rapid maxillary expansion in the pediatric population. However, there was some limited amount of randomized controlled trials, and the results were somewhat varied. Further studies should be continued and completed to ensure that the results are not due to normal growth without any intervention. The next topic here is sleep bruxism. So this is the grinding of the teeth during sleep leading to tooth wear, fatigue of the musculatory muscles when awake, and potentially restorative failures. The theory is that the apneic event would lead to the patient thrusting their jaw forward to improve the airway stability and patency, resulting in subsequent bruxism. So our clinical question is, does sleep apnea cause bruxism? Now, there were no meta-analyses that were found. However, four systematic or scoping reviews were found. None of the systematic reviews concluded that sleep apnea caused bruxism or that there was a definitive link between the two. However, most did conclude that both sleep bruxism and sleep apnea have high prevalences in the same patient populations, and therefore share the same clinical features. So the answer to this question is no. There is no meta-analyses completed. And the individual studies investigated had methods that were heterogeneous. So there was also a medium risk of bias. Interestingly, some other findings were that after CPAP or mandibular advancement therapy, sleep bruxism does reduce. Additionally, higher frequencies of sleep bruxism are seen with times of increased apneic events. Ultimately, we do see an increased prevalence of bruxism in patients with sleep apnea, so approximately 40% in adults and 26% in children. However, there are just not enough studies to show any causal or specific relationship between the two conditions. There's also not one specific mechanism that can be identified either based on the current literature. So at this point, there's very limited evidence to support a relationship or causal relationship between bruxism and sleep apnea. In this next section, we'll look at some screening tools for obstructive sleep apnea. Some of the modalities for screening sleep apnea that we'll look at are questionnaires, home sleep tests, comium CT, and acoustic reflection. I just want to take a quick moment to review the basic differences between screening and diagnosing sleep apnea. So an initial population of patients would be screened to identify if these patients have an increased risk for having sleep apnea. If through the screening process, these patients are at a higher risk, then the subsection of the population would then be recommended to have additional testing to determine if they have the presence or absence of the disease. So that would be the diagnosis. In these patients, and even further subsection of the population, that test positive for the disease would then be recommended for treatment. Traditionally, we think about how well a diagnostic tool works by looking at the sensitivity and specificity. Well, we can do the same thing at looking at a screening tool. However, a screening tool has a different purpose, right? In screening, we're checking to see if a patient population should be recommended for further study, which, again, would be our diagnostic study. We will be looking at research studies that will investigate tools for the purpose of diagnosis and screening. So because our intent here is to look at just ones that have screening tools, as a screening tool, our aim would be to detect potential problems, right? So ideally, we would have high sensitivity and specificity. However, usually for screening tools, the sensitivity is maximized with a trade-off for lower specificity. In this case, the screening test would recommend more patients to get sleep studies as to not miss any patients that are falsely told that they don't have sleep apnea when, in fact, they do. Another point here is that when looking at this from a population health standpoint, the sensitivity and specificity rates are important to catch as many patients as possible with sleep apnea, yet not catch too many patients to be falsely recommending further diagnostic testing. No need to understand the details and formulas associated with the sensitivity and specificity that we see here. The main point here is just when we review a diagnostic or screening test that we will want to compare this to the gold standard. Now, in this case, the gold standard is going to be the in-lab sleep study or the PSG for the presence of or absence of sleep apnea. So the next topic here we'll focus on is questionnaires. As we mentioned before, we're specifically looking at methods for screening, of which our first would be questionnaires. Some of these screening questionnaires are StopBang, F4 Sleeping to Scale, Berlin Questionnaire. Typically, these screening questionnaires should be validated to the population you plan to use in your office. So for instance, if the screening questionnaire was developed and tested with a subject population between the ages of 2 to 18, so a pediatric population, you'd most likely not want to use this particular questionnaire for the adult population since you don't know if it would be valid for this new population that you would intend to use it for. So to phrase this as our clinical question, we're going to ask, are questionnaires accurate in the prediction of sleep apnea? There are a number of meta-analyses with a large amount of subjects. Since most of the articles listed here are meta-analyses, this would be considered at the highest level of evidence. Of course, our clinical question is vague, and there are a large number of questionnaires that can be used. Of the articles identified, they investigated questionnaires like StopBang most frequently. In the fourth article listed by Chu, they compared the diagnostic accuracy of Berlin Questionnaire, StopBang, Stop, and Epworth Sleeping to Scale. Their findings are consistent that StopBang is the most accurate. The StopBang has been used in a general population. It's been used in sleep clinics, surgical settings, and with commercial drivers to list some of the populations that it's been tested in. In these populations, it has been found that StopBang is a valid screening tool, particularly for those with moderate to severe sleep apnea. So for instance, one meta-analysis found that for mild sleep apnea, the sensitivity was 90% with a specificity of 46%. Moderate sleep apnea sensitivity was 94% with a specificity of 75%. And severe sleep apnea sensitivity of 96% with a specificity of 90%. So there are multiple questionnaires that can be used, and it's not within our scope today to review the differences and the vast amount of screening questionnaires that can be used in detail. However, some of the most common questionnaires that we mentioned are StopBang, Epworth Sleeping to Scale, and Berlin Questionnaire for the adult population. Since the presentation of symptoms are different for the adult and the pediatric versus the pediatric patient population, we only looked at the questionnaires in the adult population. However, there are other questionnaires that can be available for use in the pediatric population. Now, it's also possible that these questionnaires might miss some patients with sleep apnea or not provide a complete clinical picture of the patient. So combining clinical features can be of benefit to capture all patients that have sleep apnea. The next topic here is home sleep apnea testing. This is a portable device that records the objective measurements of sleep. Some of these measurements are airflow, oxygen saturation, and heart rate. Depending on the amount of the objective measurements, the home sleep apnea tests are classified into two groups. Although given new technologies, AI-enabled analysis, things like that, this has obscured the line between these two groups in some cases. Traditionally, these devices are used to diagnose sleep apnea, yet in practice, some of them are used as screening tools. So we'll be investigating this in a screening context. Our clinical question is, are home sleep apnea tests accurate in the prediction of sleep apnea in adults? We found meta-analyses that reviewed the diagnostic accuracy of home sleep apnea tests, showing the highest level of evidence. Again, since this question is very broad and there are two classifications of home sleep apnea testing devices, also many manufacturers within each category, but despite this limitation, the diagnostic accuracy is good in the prediction of sleep apnea, yet some sleep apnea tests, particularly those in the type four classification, don't have as strong diagnostic accuracy as compared with the in-lab sleep studies. Therefore, given the heterogeneity in devices found in the studies, we've concluded that there is medium heterogeneity. Ultimately, our answer is that yes, home sleep apnea tests are accurate in the prediction of sleep apnea in adults. So when we look at one of the meta-analyses that specifically identified a particular manufacturer's type four home sleep apnea test, they chose this one because there were more studies on this device, we can see an example of the sensitivity and specificity. For the AHI of greater than or equal to five, we see that a sensitivity of 88% and a specificity of 64%. For AHI, for the AHI of greater than or equal to 15, we see a specificity of 82% and a specificity of 88%. So a negative result with this home sleep apnea test means that a patient could still potentially have mild sleep apnea. And in which case the authors recommended a in-lab sleep study. So to remind us of our context, we're investigating how accurate our home sleep apnea tests for screening sleep apnea. Firstly, these devices are used diagnostically. So the bottom line is that these devices have high sensitivity and specificity for patients, particularly that have moderate to severe sleep apnea. Second, if our context is in the screening, not diagnosing, this is what we called harm or burden of care to have all patients sleep at home with some of these sometimes expensive devices might not be justified as a screening tool. Although there's no evidence to support this, the US Preventative Service Task Force published a detailed paper in JAMA asking these same questions because they're interested about the risks and harms of screening. As technology changes, these devices are getting better and better. Additionally, more phone apps and consumer sleep devices are coming to the market claiming to be able to predict sleep apnea. However, in one recent systematic review, they found that none of these smartphone apps were as accurate as traditional methods. In this article, they identified 300 apps and sorted through 44 articles on current technologies. The additional benefit though of some of these technologies is that it does offer an accessible and inexpensive way to allow for the patient to continuously monitor their condition. Now, our next topic is Comium CT for the use of screening for sleep apnea. So this would be a three-dimensional image to identify those at higher risk for sleep apnea. Our clinical question is in this case, are Comium CTs accurate in the prediction of sleep apnea in adults? There were three studies identified with two being systematic reviews and one meta-analysis representing the highest level of evidence. There were mixed conclusions between the studies. So at this point, the answer is no with high heterogeneity. Some major points to be considered that might lead to this heterogeneity is patient positioning, respiratory phase during the image capture, and the type of analysis that's used to predict the sleep apnea. In the meta-analysis done by Google, it was found that the authors did not find a difference between sleep apnea and the control groups when looking at the upper area. There were two force plots shown here. So the top force plot looked at the anterior-posterior linear dimension and the bottom force plot looked at the lateral linear dimension. In total, 216 sleep apnea subjects and 117 control subjects among nine studies were used in the meta-analysis. The anterior-posterior linear dimension was found to be significantly different in sleep apnea patients compared with controls. However, by just removing the subjects from one article, the results were no longer statistically significant. In looking at the bottom force plot, sleep apnea patients were found to have a smaller lateral linear dimension in supine. However, this effect was not significantly different when patients were upright. Ultimately, the cross-sectional area in the AP dimension seems to be the most reliable measure, yet there are multiple confounding factors for this. Because COMEM-CT is taken as a snapshot of time, the protocol of taking the COMEM-CT for airway purposes can potentially create many variances. Therefore, when taking a COMEM-CT, specific instructions to the patients can be important. Also, depending on the software used to automatically process the airway volume, this can lead to discrepancies. In most cases, the reliability of the software has been demonstrated to be acceptable, but just another thing to keep in mind. Additionally, when segmenting the airway volume in the different parts of the airway, standardized descriptions of the anatomical constraints could be beneficial as segmentation differences can lead to pretty large variances between providers. Ultimately, there is not overwhelming support that COMEM-CTs are accurate solely in the prediction of sleep apnea in adults. Our next topic is screening for sleep apnea using acoustic reflection. This is known as acoustic pharyngometry and rhinometry. This is a non-invasive measurement technique to assess the cross-sectional area of the upper airway. Our clinical question here is, is acoustic pharyngometry accurate in the prediction of sleep apnea? Now, based on our search strategy, no review studies were found, but we did find four cross-sectional studies. All four studies compared acoustic pharyngometry to the in-lab sleep study, PSG. The findings of the studies were varied. The studies with the least amount of bias mostly concluded that acoustic pharyngometry alone is not accurate in the prediction of sleep apnea. One of the articles, D. Young, 2013, they combined mild sleep apnea patients with those that were healthy. They also combined the minimum cross-sectional area of the acoustic pharyngometry with additional clinical factors. So in this article, they found that combining clinical predictors with acoustic pharyngometry, they could significantly differentiate between moderate to severe sleep apnea patients from the moderate to healthy patients. In looking at two of the other studies, they did report sensitivity and specificity for the upper airway cross-sectional area. So Kim, 2019, findings are from patients sitting, like in a sitting position, and the Kerzenska, 2016 article had patients sitting and supine. Now, the results from Kim, 2019, were that acoustic pharyngometry had a sensitivity of 45% and a specificity of 77%, while the Kerzenska article found that the sensitivity was 73% and a specificity of 46%. Both reported values represent fair discrimination of sleep apnea, yet they don't provide an advantage over clinical variables based on the sensitivity and specificity amounts. And they also don't provide any advantage over screening questionnaires as well for that same reason for the discernment of sleep apnea in patients. So ultimately, using acoustic reflection as a sole predictor for screening sleep apnea would not be sufficient. These techniques can identify anatomical areas of constriction, yet the passive physiology of sleep apnea is multifactorial, potentially leading to variances between patients and reduced discernment of sleep apnea in this case. And finally, there's been some reports that sleep apnea can be cured. Now, while there are some effective lifestyle changes and surgical interventions, there's no cure for sleep apnea. We won't go into detail on this topic, but as we saw earlier, maxillomandibular advancement has a 46% success rate when looking at a success rate of an AHI less than five. So this would be similar to what we call in literature a cure rate. Additionally, weight loss surgery has been shown to have a 25% success rate, achieving AHI less than five. In one man analysis. So there are many types of interventions that can be done to reduce the severity of sleep apnea without any additional long-term followup like surgeries. However, they're not completely curative. And despite a curative surgical procedure, it's always possible to then have the incidence of sleep apnea recur as time progresses and the patient become older. Since as the age increases, we know that the progression and likelihood of sleep apnea will increase. Lifestyle changes can also improve risk factors of BMI, obesity, and potentially reduce anatomical factors leading to a collapse of the upper airway. Lifestyle changes like surgeries might not necessarily require additional long-term intervention besides continuing these lifestyle changes. However, rarely are these lifestyle changes alone going to reduce the AHI to below five. So the bottom line here is that sleep apnea is similar to a chronic condition with a multifactorial etiology where there's no cure today. I would be highly skeptical of a claim that sleep apnea can be cured at this moment in time. And thank you very much. If you'd like further information about dental sleep medicine, please feel free to go to the AADSM website at aadsm.org.

evidence-based practices

screening

treatment

obstructive sleep apnea

pathophysiology

treatment modalities

screening tools

sleep apnea

research