Advising the avoidance of sleeping on the back and encouraging weight loss are the basic non-surgical therapies applied, when indicated, to patients with OSA. Nasal continuous positive airway pressure, or nCPAP—the principal treatment—is very effective, but often results in problems with patient tolerance.⁸ Surgical treatment of the soft palate is of low efficacy in patients with OSA,⁹ but has a definite role in patients with simple snoring.

	THE USE OF DEVICES IN MANAGING OSA AND SNORING

TOP ABSTRACT THE USE OF DEVICES... MATERIALS AND METHODS RESULTS DISCUSSION SUMMARY REFERENCES

 
Numerous dental devices with various designs are claimed to help manage snoring and OSA; suggested guidelines for their use have been published.¹⁰ Some of these AMP devices are what dentists describe as boil-and-bite appliances. These devices have a prefabricated shell containing a thermoplastic material and are heated and fit directly to teeth. Although such appliances may fit many patients, they do not fit all because of substantial variability of tooth position, arch form and jaw size.

Other AMPs include one-piece devices that are not adjustable and two-piece adjustable devices. The two-piece adjustable AMPs usually have some limited lateral movement capability and are adjustable in the sagittal plane.

There also are devices that attempt to capture the tongue and hold it forward. These devices can be custom-made or prefabricated.

How an AMP device works. Two basic theories explain how an AMP device works. One is that the device increases the airway caliber, thus making the airway resistant to collapse from the negative pressures produced during inspiration.¹¹ The typical magnitude of this anterior advancement is between 60 and 90 percent of the patient’s maximum possible protrusive jaw motion, which is usually 5 to 9 millimeters. Cephalometric studies have shown that changes will be induced in the airway with AMP insertion. These changes include an increase in the space behind the tongue and soft palate, and a change in the configuration of the uvula and tongue.^11,12

The second, less-endorsed theory is that the device causes stretch-induced activation of the pharyngeal motor system. This motor activation provides enough stiffness to the system to prevent collapse of the airway. This theory is supported by the observation that electromyographic activity in the tongue has been shown to increase with use of an AMP.¹³

Compliance with and complications of AMP use. Although dental devices have been widely used on patients with OSA for more than 10 years and are judged efficacious,¹⁰ there are only a few studies that systematically assessed relative compliance and complication rates.^14,15 Available case reports show good compliance (especially when compared with nCPAP) and infrequent morbidity with AMP treatment. Published reports suggest that temporomandibular joint pain and occlusal changes are relatively uncommon, but long-term risks of these complications have not been established.^14–19

AMP-use compliance, the regular nightly use of an AMP device across the entire sleep period, has not been widely or systematically studied. Existing studies include the following.

Reporting on 68 patients using a one-piece thermoplastic AMP device for OSA, Schmidt-Nowara and colleagues¹⁹ found that the rate of continued device use at a mean follow-up point of seven months was 75 percent. In a three-year prospective study, Clark and colleagues¹⁴ reported that 12 of 24 subjects were still using a two-piece adjustable AMP device. In another study on the two-piece adjustable AMP device, Eveloff and colleagues¹⁵ reported a compliance rate of 93 percent, with a mean follow-up of two years. In a five-year study, Ichioka and colleagues²⁰ reported 86 percent compliance with a one-piece mandibular advancement dental device used for OSA.

Ferguson and colleagues²¹ conducted a randomized crossover study of 27 patients with mild-to-moderate OSA (mean AHI = 24.5). In this study, patients received both nCPAP and a one-piece thermoplastic AMP. Outcomes were classified by treatment response and compliance. In this nine-month study, the subject used each treatment for four months. There was a two-week waiting period before and between each treatment. During this four-month period, the compliance rate was 62 percent for the nCPAP device and 76 percent for the AMP device. The comparison of the efficacy of these two treatments showed that the nCPAP device was more effective, but both groups experienced a substantial reduction in their apnea.

In a second crossover study in which patients with mild-to-moderate OSA were studied (mean AHI = 26.8), Ferguson and Ono¹⁶ compared a two-piece, adjustable AMP device with nCPAP. For the 20 patients who completed the study, nCPAP was helpful and used by 70 percent of the patients for the entire four-month treatment period. For the patients using the AMP device, sustained use occurred in 95 percent of the patients across the four-month treatment period. The patients stated that their reasons for discontinuing therapy were multiple and not always because the device was uncomfortable or ineffective.

While the literature supports the position that both one- and two-piece AMP devices are generally helpful for patients with mild-to-moderate OSA in the short term, it remains to be seen whether these devices will continue to be effective and therefore used over the long term. To date, no thorough crossover study has been conducted comparing one- vs. two-piece AMP devices. Because only limited data exist, we performed a study to assess the relative long-term (longer than one year) compliance levels for a two-piece, adjustable AMP device, and the relative rate and type of complications for patients who have worn the device for one year or more.

	MATERIALS AND METHODS

TOP ABSTRACT THE USE OF DEVICES... MATERIALS AND METHODS RESULTS DISCUSSION SUMMARY REFERENCES

 
Subjects. We included in the study 65 patients who were referred by various sleep-medicine practitioners to the University of California-Los Angeles Sleep Orofacial Pain and Oral Medicine practice in Los Angeles from October 1990 through February 1997 for treatment of mild-to-moderate OSA with a dental device. The inclusion criteria were that they had undergone a successful NPSG study, and that it was of at least five hours in duration. All NPSG studies reviewed were judged to be acceptable (that is, they met the minimum standards for polysomnography). A second criterion was that at least one full year had passed since they began using an AMP device.

AMP device. The dental device being evaluated is custom-made to advance the mandible during sleep and has been described in two publications.^17,22 It consists of two separate, full-coverage, clear acrylic appliances that snap onto the dental arches. These appliances are connected with a rod and tube device (called a Herbst attachment), which allow for opening, and protrusive and some side-to-side movements, but no retrusive movements (Figure 1). Bilateral, interarch elastics keep the jaw closed, and interproximal ball clasps firmly retain each individual appliance. The appliance is carefully adjusted to fit only in an anterior mandibular position, is used only at night and usually advances the mandible by at least 75 percent of the patient’s maximum protrusive range of jaw movement. After insertion, one or more visits were scheduled to adjust the appliance to achieve best fit and comfort.

View larger version (74K): [in this window] [in a new window]   Figure 1. Custom-made anterior mandibular positioning device.

To assess the amount and type of complications, patients were asked to indicate, from a list of discomforts, how often they experienced each discomfort. The patients were asked to rate the frequency of complications by using the following scale: "never," "rarely," "sometimes," "frequently" and "always." The list of discomforts includes pain in jaw or facial muscles, pain in jaw joint area, pain in ear area, pain in teeth, pain in oral tissue, bite change, jaw joint noises, ill-fitting or loose appliance, excessive salivation and xerostomia. All patients were mailed a copy of the questionnaire, even if they were not contacted by telephone.

The primary reason for using a dual approach (telephone and mail follow-up) was to cross-validate one method against the other. It is difficult in a large city like Los Angeles to conduct a retrospective study and achieve a high yield of potential subjects. We elected to use both methods for surveying our subjects because some people are more likely to respond to one more than the other.

	RESULTS

TOP ABSTRACT THE USE OF DEVICES... MATERIALS AND METHODS RESULTS DISCUSSION SUMMARY REFERENCES

 
Of the 65 patients, 12 were not contacted because they had moved or their telephone was disconnected and no forwarding addresses or contact numbers were available. Only one of those contacted declined to participate. Of the 53 subjects contacted, 46 were men and seven, women. The mean age of the male subjects was 56.4 years, with a range of 28 to 80 years. The mean age of the women was 55.7 years, with a range of 31 to 68 years. Clinical examination showed that the majority had Angle’s Class 1 occlusion, with only seven subjects having an edge-to-edge (mild Angle’s Class 3) occlusion. None of the subjects were judged as having Angle’s Class 2 occlusion. The subjects in this study would be best characterized as having mild-to-moderate OSA severity. The mean AHI score, based on the available pretreatment NPSG data, was 20.9 ± 20.7 (mean ± standard deviation).

Of the 53 subjects contacted, 47 answered the questionnaire by telephone and five answered via the mail survey only. Mail surveys were returned by 19 of the 53 people contacted, since many subjects had already answered by telephone. Analysis of answers where both telephone and mail survey questions were available showed absolute agreement on 78 percent of the individual questions. For those questions where there was not complete agreement, 6 percent of subjects selected a response on the mail survey that was one category higher than that on the telephone survey and 4 percent selected a response on the mail survey that was one category lower. The remaining 12 percent of the responses were two or more category scores different.

Regarding dropouts or noncompliance, the 27 subjects with a mean use of 42.6 ± 17.8 months after AMP device insertion reported they were still using the appliance every night. The 26 dropouts had used their devices for a mean of 7.2 ± 8.3 months from the delivery date to cessation of appliance use. The reasons reported for discontinuing use are presented in Figure 2.

View larger version (48K): [in this window] [in a new window]   Figure 2. Reasons for terminating use of anterior mandibular positioning, or AMP, device.

View larger version (52K): [in this window] [in a new window]   Figure 3. Duration of compliance in using the AMP device.

View larger version (54K): [in this window] [in a new window]   Figure 4. A. Complications of use of anterior mandibular positioning, or AMP, device experienced by compliance group. B. Complications of AMP device use experienced by noncompliance group.

	DISCUSSION

TOP ABSTRACT THE USE OF DEVICES... MATERIALS AND METHODS RESULTS DISCUSSION SUMMARY REFERENCES

It appears that the majority of those who quit using the appliance did so within the first three months. These compliance data are only generalizable to patients with OSA who have a similar severity (mild-to-moderate) and should not be applied to those with more severe problems. These data are similar to earlier research in a smaller case series report of a two-piece adjustable AMP device where 50 percent of the subjects were still using the device at the three-year point.¹⁴ In contrast, other studies^15,20 have reported a higher compliance rate for long-term AMP device use. To address these conflicting compliance rates, a larger and longer prospective study to further assess compliance is needed.

In this study, after initial fitting and adjustment of the appliance, patients were told to return to the office at one-year intervals or if they had a problem with the appliance. The appliance fitting and adjustment process typically involved one to two visits, conducted anywhere from one to three weeks after insertion. Of course, the nature and type of follow-up visits may have a substantial role in long-term use and, in hindsight, a more aggressive postinsertion recall program might have substantially increased the compliance rate.

Another confounder is that the majority of patients in this study elected not to follow our advice and participate in a second, postappliance sleep study. This situation is not uncommon in clinical practice; there is a general unwillingness by subjects with mild-to-moderate illness to undergo a second NPSG study, as it is often not covered by their insurance benefit plans.

The absence of this second NPSG study raises the question of whether noncompliance over time is potentially related to poor efficacy of the device. An answer to this question will require a comprehensive prospective study on AMP devices. Of course, the study of compliance within a research project, where closer-than-usual observation is done and possibly more rigorous entry criteria are used than might occur in clinical practice, is itself a strong confounder to a study.

Our data suggest that the most common self-reported reasons for discontinued use of the AMP device are that it was uncomfortable or painful to use, or it had no or little effect, or both. Unfortunately, it is not possible to determine which of these was the exact reason for appliance use cessation. Of the 26 patients who were categorized as AMP failure cases, one refused to participate in the posttreatment survey. Sixteen stopped using the AMP device within eight months, while the other nine used the device for more than one year.

Complications reported by the subjects in our study suggest that occlusal changes are more prevalent than previously thought.

While we recognize that neither a telephone nor a mailed survey is an accurate indicator of actual time of AMP use during the night, self-reported noncompliance can be determined. It may turn out that patients are removing the dental device during the night, but this will not be proved without a covert monitor of device use. We are confident that these data provide a reasonable estimate for noncompliance when using a two-piece adjustable AMP device for OSA.

It would be unfair to generalize these data to other AMP devices such as the one-piece "boil-and-bite" devices or the tongue-retaining devices. In general, we do not expect or predict that these other devices would have higher compliance rates.

Except for a study by Nakazawa and colleagues²³ in which 21 percent of the subjects had occlusal discomfort, actual induced malocclusion has not yet been reported as a sequela of AMP device use in the literature. Moreover, our data do not allow us to actually measure the magnitude of an occlusal change, as baseline occlusal records were not collected systematically. Nevertheless, complications reported by the subjects in our study suggest that occlusal changes are more prevalent than previously thought. These changes were reported more often in the compliance group (long-term users) and the occlusal changes do not appear to be transient, as 14 of 21 subjects who reported complications with their occlusion (or 26 percent of all subjects) stated that their occlusal changes were not transient but rather long-standing.

Overall, the data reported in this study lead us to predict that one in three patients will develop some minor complications of the jaw or mouth, or tooth pain from using an AMP device. Moreover, our findings suggest that a substantial number—14 of 27 long-term AMP device users in this study—will have a painless but possibly irreversible change in their occlusion. The magnitude and nature of this change will require a prospective study, but the data presented here indicate that annual follow-up office visits for early detection of these changes is advisable.

Finally, with regard to satisfaction, of the 27 subjects still using the device, 22 reported that they were very satisfied with it. The other five included four who were somewhat satisfied and one who was neither satisfied nor dissatisfied. The number of subjects in the compliance group shows an intent to continue using the AMP device.

	SUMMARY

TOP ABSTRACT THE USE OF DEVICES... MATERIALS AND METHODS RESULTS DISCUSSION SUMMARY REFERENCES

 
This study used a telephone and mail survey to assess consecutive patients with OSA who received an AMP device. Of the 65 subjects sought for follow-up, 53 were contacted (82 percent) and 27 were still using the device. Our compliance data (51 percent) are consistent with previous literature reports, but the prevalence and magnitude of the complications observed are greater than those that have been reported regarding dental devices for OSA. Based on the patient satisfaction survey data, it seems clear that approximately 50 percent of the patients will continue using the device short-term, with only minor complications and with high satisfaction. However, another 50 percent of the subjects who receive an AMP device stop using it, and approximately one in three patients had complications such as transient jaw muscle and tooth pain. Moreover, 26 percent developed changes—possibly irreversible—in their occlusion.