Snoring affects approximately 40%1 of the population, most commonly middle-aged males. Simple non-apnoeic snoring is characterised by loud snoring without the presence of any other abnormal respiratory breathing events, such as apnoea or hypopnoea.2 Apnoea3 is the complete cessation of breathing lasting at least 10 seconds, whilst hypopnoea is a less severe disturbance with a 50% reduction in ribcage and abdominal excursion for at least ten seconds, together with an associated reduction of 4% in oxygen saturation.4

Snorers can suffer from a poor quality of sleep and may complain of daytime sleepiness. Spouses and other family members can also be severely affected by this condition, so reducing the quality of life of not only the sufferer but also those around them. Management of these patients includes interventions such as weight loss, alcohol restriction, sleep position training, oral appliances, nasal appliances and pharyngeal surgery.5

These patients are also often seen by hospital-based orthodontists for management with mandibular advancement appliances (MAAs), due to their inter-disciplinary nature. The orthodontist may also need to liaise with other hospital disciplines, for example chest physicians and ENT consultants, to whom these patients may present.

Mandibular advancement appliances act to posture the mandible, drawing the tongue and soft palate forwards, therefore maintaining the pharyngeal airway during sleep. Studies have shown a good response to the reduction in snoring with such appliances.6,7,8

Various designs of appliance have been used in previous studies. These include vacuum formed monobloc bilaminate appliances and two piece interlocking appliances, for example the Herbst appliance.

Monobloc appliances are cheap, easy to make and fit and provide good retention. However, their main disadvantage is that they require a protrusive bite registration for construction and further adjustment to the amount of mandibular protrusion is not possible without re-making the appliance. Generally, these appliances are made with space between the upper and lower components, to facilitate breathing. In addition, vertical opening is kept to a minimum. Excessive vertical opening results in a downward and backward rotation of the mandible with resultant posterior movement of the tongue and soft palate. This can lead to further narrowing of the pharyngeal airway, so exacerbating the condition.9

Two piece interlocking appliances which allow incremental advancement provide more versatility than monobloc appliances. Recently, appliances which allow the patient to control the advancement themselves have been used.10,11

Previous research by Johnston6 and Bates8 has shown success rates of 76-84% and 70% respectively, using a monobloc design in non-apnoeic snorers. Recently, research in this area has taken into account the effects of the snoring on the patient's sleeping partner. Frequently, partners are severely affected, often more than the patient themselves. It has previously been shown that the use of these appliances provide a significant improvement on the partner's sleep pattern.6,8,12,13

Assessment of success utilising partners' views is a relatively new, but an important aspect in the provision of care. Bates and McDonald8 have previously assessed sleeping partners with a specifically designed questionnaire. They assessed 121 patients and their partners with a range of sleep-related breathing disorders, three months after provision of a MAA. Seventy percent of partners reported that their partner's snoring had improved. In addition, 64% of partners felt that their own sleep was improved. These findings are also reinforced by Johnston et al.,6 who found that 84% of partners reported a reduction in snoring loudness and 76% reported their partner snored on fewer nights per week, when assessed four to six weeks after treatment with a MAA. In addition, Johal et al.12 reported that 86% of partners reported a better quality of sleep following their partner receiving treatment with a MAA. However, this was assessed by the patient and not by their partner. In a later study, Johal et al.13 demonstrated an improvement of 72% in partner-recorded snoring and disturbance.

Treatment of patients with non-apnoeic snoring within a hospital environment has significant financial implications and may also involve a lengthy wait for treatment. Frequently, orthodontic consultants are unable to accept patients with simple non-apnoeic snoring for treatment with MAAs. This is often as a result of pressures to manage the orthodontic workload. In addition, access to this service may have restrictions placed by the primary care trust (PCT), for example, only offering treatment to patients with severe apnoea.

Recently, there has been increased interest in general dental practitioners (GDPs) treating these patients in their practices. Management of such patients in dental practice after appropriate screening by sleep clinics may offer a cost-effective alternative to treatment within a hospital setting. It is extremely important that the patient is adequately assessed by a sleep clinic, to ensure that the serious diagnosis of obstructive sleep apnoea is not missed, which if overlooked could have significant morbidity implications.

The effectiveness of GDPs in the provision of this treatment has never been assessed. Therefore this study aimed to assess whether GDPs can be successfully trained to treat non-apnoeic snoring using a monobloc design of mandibular advancement appliance, following a one day training course.

Subjects and methods

Study design

A prospective three-centred clinical study was undertaken for which ethical approval was obtained. The three centres involved were based at hospital units in Bradford, Leeds and Wakefield.

Selection of clinicians

A total of 258 GDPs from three regions of Yorkshire (Bradford, Leeds and Wakefield) were invited to partake in the study. No previous working knowledge of the management of sleep-related breathing disorders was required by the GDP. A response rate of 18% (47 GDPs) was achieved. Thirty-two percent of this sample (15 GDPs) was able to attend the training and clinical trial days. None of these GDPs had clinical experience of treating patients with MAAs. The one day training course involved lectures, followed by a two hour clinical session. During this, bite registration demonstrations were given to the GDPs. The GDPs then practised this skill clinically, using each other as subjects, as many times as they needed until they felt proficient.

Patient selection criteria

Suitable patients, who fulfilled the following inclusion criteria, were invited to participate in the study:

  1. 1

    Fit and healthy adults (18 years or over)

  2. 2

    Patients with a confirmed diagnosis of simple (non-apnoeic) snoring, as assessed by a sleep physician (assessment was carried out using overnight oximetry)

  3. 3

    Optimal dental and periodontal health

  4. 4

    A minimum of ten teeth present per dental arch

  5. 5

    Patients have a sleeping partner.

Exclusion criteria were as follows:

  1. 1

    Patients with obstructive sleep apnoea

  2. 2

    Inadequate dental or periodontal status to allow construction of appliance

  3. 3

    Patients with significant heart disease

  4. 4

    Patients with pulmonary disorders

  5. 5

    Patients with excessive daytime sleepiness (Epworth sleepiness scale (ESS) greater than or equal to 18).

Study protocol

Patients were seen by the GDPs on three occasions:

  • Alginate impressions and postured bite registration for MAA construction

  • Fitting of MAA and instructions for use

  • Review visit after three months.

All treatment was carried out within the orthodontic departments of the three centres, however, the GDPs worked independently.

Outcome measures

The primary outcome measure was a sleeping partner evaluation (SPE). This was a questionnaire which was completed before the start of treatment and repeated after three months of treatment. The initial questionnaire was administered before the start of treatment, at the time of obtaining consent for the study. The subsequent questionnaire was administered at the final review visit after three months of treatment.

The questionnaire has been previously used by Johal13 and questions the patient's sleeping partner with respect to both the severity of the patient's snoring and its level of disruption (Appendix 1). Scores range from a minimum of 4 (good quality of sleep for both the subject and their partner, with no snoring or disturbance of the partner's sleep), to a maximum of 16 (poor quality of sleep for both the subject and partner, with very loud snoring which always disturbed the sleeping partner). Improvement for the SPE questionnaire was defined as a reduction in the partner score to a value of 9 or less. This value represented an average quality of sleep for both the partner and subject, with mild snoring that sometimes disturbed the partner. Therefore treatment was considered to be effective if the score of the second questionnaire was 9 or less. Previous research13 has demonstrated an improvement of 72% is to be expected with this type of treatment. Therefore for GDPs to be effective, a success rate of 72% or greater needed to be achieved.

Secondary outcome measures were the Epworth sleepiness scale questionnaire and an outcome questionnaire.

The Epworth sleepiness scale (ESS) is designed to assess daytime sleepiness. These were administered prior to treatment and following three months of MAA use (Appendix 2). This is a validated questionnaire and asks the subject how likely they would be to doze in eight different situations. The final score is a sum of these eight scores, with a range from 0 to 24. Clinically normal scores range from 2 to 10, with a mode of 6.14

An outcome questionnaire previously described by Smith and Battagel7 (Appendix 3) was completed by the subject to assesses the short-term side-effects of the appliance at two time points: two to three days and one month. This questionnaire was completed by subjects at the three month review visit.

Sample size calculation

The sample size calculation was based on the primary outcome measure. Applying the improvement of 72% based on the findings of Johal et al.,13 it was decided that a clinically significant success rate would be achieved if this could be estimated to within 10%. Using 95% confidence intervals with an estimated proportion of 72% and a lower limit of 62% produced a sample size of 78. Clustering effects increased this sample size to 102. The estimated number of recruitable GDPs was 20, treating a minimum of five patients each.

Mandibular advancement appliance

A monobloc MAA design was used in all three centres (Fig. 1). This was constructed from a dual laminate acrylic material with a soft fitting surface (Dual Laminate, Dental Resources, Inc., UK).

Figure 1
figure 1

Mandibular advancement appliance used in study

The occlusal registration was taken with a defined protrusion of 50% of maximum overall protrusion, using either softened pink wax or a Projet bite fork (Orthocare, Bradford, UK), dependent on personal preference of the treating GDP. This degree of protrusion has been shown to increase the cross-sectional area of the upper airway by 50%.15

All appliances were constructed by ten experienced technicians (two at Bradford, three at Wakefield and five at Leeds).

This appliance is simple to construct, however it does not allow for any adjustment to the amount of protrusion. Therefore, if any further adjustment is found to be necessary, the appliance needs to be re-made.


GDP recruitment

Problems were encountered with recruitment of GDPs. There was a low response rate (n = 47, 18%) to the invitation to be involved in the study. Of this sample, only 32% (15 GDPs) were able to attend the training and clinical trial days.

Patient recruitment

It was difficult to obtain complete data for all three questionnaires, due to lack of attendance of partners at all appointments. When partners were absent for appointments, questionnaires were sent home with the patient, to be returned at the next visit. When data was still incomplete at the end of the study, repeat questionnaires were posted to patients and their partners on two separate occasions. Despite these efforts, complete data were available for between 46-60 patients, dependent on the questionnaire.

Sleeping partner evaluation

Complete data were available for 48 (80%) of the sleeping partners. A success rate of 48% was achieved (95 percent CI 0.35, 0.61).

GDPs encountered problems with protrusive bite taking in 10% of patients, making construction of the appliance impossible. The main problems were as follows:

  • Bite recorded with a large mandibular asymmetry, when none was evident

  • Lack of definition, making location of casts impossible.

In these patients the bite registrations were repeated by the lead consultant orthodontist in each of the three centres to allow correct construction of the appliance. Although these appliances were subsequently successfully fitted, these patients were recorded as failures, as without the intervention of the orthodontist it would not have been possible to construct the appliance. Tables 1 and 2 summarise the success and failure data.

Table 1 Summary of success and failure data
Table 2 Summary of failures specifically related to bite taking problems

Epworth sleepiness scores (ESS)

Complete data were available for 83% of patients at the end of the study. The results are shown in Table 3 and demonstrate a small reduction between the start and end median scores. The median start score was 9, with the median post treatment score 7.5 (95% CI 0, 3).

Table 3 Epworth scores

Outcome questionnaire

Complete data were available for 83% of patients in relation to immediate (two to three days) short-term effects, and for 80% of patients with respect to the one month follow-up effects (Table 4).

Table 4 Outcome questionnaire results

Patients were also given the opportunity to report any other problems with the appliance. Problems reported included:

  • Patient removing the appliance inadvertently during the night

  • Appliance loose

  • Appliance cracked

  • Gagging on appliance.

However, the most frequently reported problem was pain on the teeth or gums, which was reported in 22% of patients.


Seventy-two percent of non-apnoeic patients have been previously successfully treated using this questionnaire as an outcome measure.13 In the original study, the questionnaire was used to assess the effect of the Herbst appliance in the management of a range of sleep-related breathing disorders. This included subjects with obstructive sleep apnoea and non-apnoeic snoring. Seventy-two percent of patients who had that treatment had a SPE score of 9 or less, indicating a successful treatment. Although this assessment is subjective and non-validated, it was at the time of designing this study the only available questionnaire in this area of research. Nevertheless, the previous success rate observed could be explained by the authors only selecting patients for MAA therapy on the basis of a favourable treatment prediction using sleep nasendoscopy. Sleep nasendoscopy was developed to identify those patients who would benefit from surgical management of their condition. During this procedure, patients are induced into a light phase of sleep and the upper airway is visualised using a flexible fibre-optic endoscope. The site (palatal, multi-level or tongue-based) and grade (partial or complete) of the obstruction is then assessed.16,17

Furthermore, the authors used an appliance which permitted sequential activation. Due to financial constraints, the current study was not able to use such an appliance, which permits self-adjustment. The latter offers greater potential for therapeutic benefit10,11 and patient adaptability. In addition, there is no need to record a postured mandibular position as they are self- adjustable. It could be proposed that the use of such appliances in this study may have resulted in greater success being observed.

Sleeping partners evaluation

The success rate of 48% was well below that of 72% achieved by previously by Johal.13 Problems were encountered in 10% of patients with the recording of their protrusive bites necessary for appliance construction. In each of these cases, it was necessary to record treatment as a failure. However, many of the GDPs involved commented that they felt that their skills improved in this area as the study progressed and their experience increased. Therefore, it would appear that this problem should be easy to rectify with further training and experience. Despite the low success rate achieved with the SPE evaluation of the MAA, 44% of patients where treatment was classed as a failure from this perspective, believed the treatment was beneficial, which is encouraging. These findings show the considerable impact these appliances can have on the lives of the patient and their sleeping partner.

Treatment with the MAA is only one aspect in the management of the snoring patient, and lifestyle factors such as weight loss and alcohol restriction may also play a part. GDPs can have a role in educating the patient as to the importance of these factors.

Daytime sleepiness (Epworth sleepiness score)

There was a small reduction between start and end median Epworth scores. This demonstrated that there had been no worsening of daytime sleepiness as a result of the appliance. Maximum start and end scores were higher than expected for patients with snoring rather than obstructive sleep apnoea, however the start median score was relatively normal. Whether the Epworth score can be utilised for distinguishing patients with obstructive sleep apnoea as opposed to simple snoring remains under debate. Johns14 previously reported that Epworth scores increased linearly with the severity of obstructive sleep apnoea. However, Osman18 failed to show this correlation.

Outcome questionnaire

Initial side-effects are to be expected with any oral appliance and the patient should be warned regarding these and encouraged to continue with the appliance. The most commonly reported side-effect found in the short-term was excessive salivation, affecting 67% of patients. This has also been reported as a major problem by Bates8 and Johnston,6 who both used similar bilaminate designs to the one used in this study. However, this side-effect was reported to be much lower by Smith and Battagel7 who used an adjustable Herbst-type appliance.

Discomfort of the muscles of the face and temporomandibular joint was also a commonly reported side-effect, affecting 58% of patients in the short-term. This symptom was much reduced in the longer term, when assessed again at three months, and is in agreement with previous research.7

Pain affecting the teeth or gums was reported by 22% of patients and is probably related to the design of the appliance used. This problem does not appear to have been reported with appliances that incorporate additional components for retention, for example Adams clasps, which are present in the removable Herbst appliance.7

Limitations of the study

A major limitation of this study is the low numbers of GDPs and consequently patients involved. There was a low response rate to the initial invitation to partake in the study from GDPs and subsequently this was further reduced by GDPs attending the training day but then being unable to attend the clinical sessions, despite the importance of both being stressed prior to the start of the study.

The problem encountered with lack of patient attendance at the review visit is a common occurrence with clinical studies. Repeat questionnaires were posted on two occasions to those who did not attend. However, of the questionnaires which were returned, not all questions had been answered, resulting in 17-20% of data being absent between the three questionnaires.

This study, although assessing the effectiveness of GDPs carrying out the treatment, was carried out in a hospital setting. This was felt necessary at the time to try and ensure clinical support was in place, if necessary. When problems occurred, such as the repeat taking of the protrusive bites being required, it allowed the patients' treatment to continue with minimum disruption. However, further work is obviously required to allow additional research to be carried out within a practice setting.

The problem encountered with the protrusive bite registration was interesting. This is a skill which orthodontists use on a daily basis. However, this was a new aspect for many of the GDPs involved. Unfortunately, the design of the MAA used meant that no further adjustment was possible without re-making the appliance. The study highlighted that further training in this area would be necessary, beyond that which we gave during the training day. Alternatively, the use of a different design of appliance which permitted incremental advancement and therefore avoided the need for a postured bite may have been more successful.

The outcome questionnaire was administered at the three month review visit, but questions the patient as to the side-effects of the appliance at two to three days and at one month. This may have led to errors associated with accurate recall of data. It may therefore have been better to ensure that the relevant sections of the questionnaire were completed accurately at the associated time points.

Patients were assessed in the sleep clinics by overnight oximetry. Ideally, full assessment using polysomnography would have given a more accurate diagnosis and more thoroughly excluded obstructive sleep apnoea.


GDPs were not effective in the management of non-apnoeic snoring using a monobloc appliance after a one day training course. Further training and/or selection of a different design of appliance should be considered for GDPs to become highly competent in this area.