Background

In July 1998, the U.S. Food and Drug Administration (FDA) granted approval for the "NTI Clenching Suppression System" (now: "Nociceptive Trigeminal Inhibition Tension Suppression System": NTI-tss). According to the manufacturer, the NTI-tss device is indicated for the prevention and treatment of bruxism, temporomandibular disorders (TMDs), occlusal trauma, tension-type headaches and/or migraine [1].

The NTI-tss device is a small pre-fabricated anterior bite stop (Figure 1) which covers – in its most widely used form – the two maxillary (or mandibular) central incisors (Figure 2). The fit along the teeth is accomplished at the chair side by filling either an autopolymerizing acrylate or a thermoplastic material into the base of the device, which is subsequently adapted along the central incisors, thereby increasing the vertical dimension between the upper and lower jaw. Adjustments along the outer surface of the bite stop are made by the dentist to ensure that at jaw closure and during excursive movements tooth contacts are present only between the intraoral device and the incisal embrasures of the antagonistic teeth. This "miniature anterior bite appliance" [2] is typically worn during the night, although two variations of the bite stop are offered for daytime use [3].

Figure 1
figure 1

The NTI-tss device, standard type (left) and vertically reduced type (right).

Full size image
Figure 2
figure 2

Inserted NTI-tss device.

Full size image

In the lay press, most articles have reported positively about the NTI-tss splint in the therapy of long-lasting headache or facial pain [e.g., [4–6]]. Likewise, excited dental practitioners have published their personal impressions in local dental journals [e.g., [7]] or on the website of the international NTI-tss provider e-motion [3].

Conversely, some renowned clinical researchers [e.g., [8–12]] have tempered over-optimistic expectations by raising doubts on the claimed success and by pointing at the possibility of unwanted side effects, such as aspiration, ingestion, occlusal changes after prolonged unmonitored use, and mobility of anterior teeth. In 2003, Helkimo [8] delivered an expert statement on demand of the Swedish Dental Association and the Swedish National Board of Health and Welfare on the question whether the use of the NTI-tss device "is to be regarded as lege artis and according to science and empirical experience, both as to the treatment of stomatognathic problems as well as migraine." The author came to the conclusion that there is a "total lack of scientific documentation of its therapeutic effects and possible unwanted side-effects" [8]. As far as side effects are concerned, Jokstad et al [10] mentioned that one person in Norway [13] and three individuals in the United States were subjected to medical emergencies due to aspirated NTI-tss devices splints. For the three cases from the U.S., the author referred to the FDA's Manufacturer and User Facility Device Experience Database (MAUDE), which contains voluntary, user facility, distributor, and manufacturer reports of adverse events involving medical devices. Later, Wright and Jundt [12] repeated the contention of four aspirated NTI-tss devices by referring Jokstad et al's article [10].

Considering the controversy that exists within parts of the international dental community regarding the benefits and dangers of the NTI-tss device, it appears to be timely to

(a) systematically review the (dental) literature in order to summarize and appraise the currently available data on the efficacy and/or effectiveness of the NTI-tss device, and

(b) to summarize all identifiable documented cases, in which complications and/or side effects associated with this anterior bite stop have been reported in the dental literature.

Methods

Identification of publications about the efficacy and/or effectiveness of the NTI-tss device

To identify a maximum number of relevant publications (including dental congress abstracts), we performed a systematic search in the following electronic databases:

1. The Cochrane Library

2. PubMed

3. TRIP database

4. MEDPILOT.DE

5. BIREME

6. The database of the Deutscher Ärzte-Verlag, which comprises the four German-language dental journals "Deutsche Zahnärztliche Zeitschrift", „Zeitschrift für Zahnärztliche Implantologie“, „Oralprophylaxe & Kinderzahnheilkunde“, and „Zahnärztliche Mitteilungen“

7. The database of the Quintessenz Verlag, which considers the seven German-language dental journals "Die Quintessenz", „Kieferorthopädie“, „Parodontologie“, „Implantologie“, „Endodontologie“, „Quintessenz Team-Journal“ and „Quintessenz Zahntechnik“

8. Google Scholar

9. Web of Science (Cited reference search)

The key words and strategies for the searches in each of the nine databases are shown in Tables 1 to 9. Moreover, were searched NTI-tss-associated websites for relevant articles and references.

Table 1 The Cochrane Library.
Full size table
Table 2 PubMed.
Full size table
Table 3 TRIP database.
Full size table
Table 4 MedPilot.
Full size table
Table 5 BIREME.
Full size table
Table 6 Database of the Deutsche Ärzte Verlag.
Full size table
Table 7 Database of the Deutsche Quintessenz Verlag.
Full size table
Table 8 Google Scholar.
Full size table
Table 9 Web of Science: Cited reference search.
Full size table

The last update of all electronic searches was carried out on December 31, 2007.

In addition to the database search, textbooks related to the topics of TMDs, occlusion, and bruxism were considered. Furthermore, the Swiss provider (Karr Dental) and an international European provider of the NTI-tss device (e-motion) were requested to send us possible further material and publications. The abstracts, or, when available, the full-text papers were read in order to establish the acceptability of the publications to this review. Finally, the reference lists of the identified relevant articles were screened to find additional pertinent contributions (e.g., journal articles, textbooks, book chapters, master or doctoral theses, course material).

The strength of evidence related to the identified publications was evaluated using the classification suggested by Antes [14] (Table 10). Articles about randomized controlled trials (RCTs) were evaluated according to the quality score developed by Jadad et al [15]. The Jadad scale consists of five items, which focus on three dimensions of internal validity (randomization; double blinding; description of withdrawals and drop-outs) (Table 11). Since double-blinding (as required in the Jadad scale) is not possible when occlusal devices are used, single-blinding of the investigator(s) to the type of splint worn by the patient was used as criterion.

Table 10 Hierarchy of strength of evidence for therapeutic decisions [slightly modified after [14]]
Full size table
Table 11 Determination of the quality score proposed by Jadad et al [15]
Full size table

Uncertainties on data interpretation and discrepancies in scoring according to the classifications by Antes [14] as well as Jadad et al [15] were resolved by discussion between the two reviewers.

Identification of publications about complications and/or side effects of the NTI-tss device

For the identification of complications and/or side effects associated with the use of the NTI-tss device, pertinent reports found in the identified articles using the search strategies mentioned above were considered.

In addition, a specific search in the FDA's Manufacturer and User Facility Device Experience Database (MAUDE) was carried out with the following strategy: Go to Simple Search → Search term: NTI; Date Report Received by FDA: ALL YEARS

Results

Overall yield of the search

A total of 68 relevant publications of different levels of evidence were identified. Details about the search results in the nine electronic databases as well as in the NTI-tss-related websites are shown in Tables 1 to 9 and 12, respectively. Figure 3 (right column) reveals by which search strategy the publications were found: only 39 contributions were identified by the search in electronic databases and websites.

Table 12 NTI-tss-related websites. Results
Full size table
Figure 3
figure 3

Allocation of the identified 68 articles according to the hierarchy of strength of evidence. Red: hits in electronic databases (n = 20); green: publications identified by searching the websites (n = 19); orange: articles found in the reference lists of identified articles (n = 6); purple: contributions found in textbooks (n = 1); blue: publications mailed by providers of the NTI-tss device (n = 19); black: publications found in other sources (n = 3).

Full size image

Qualitative analysis

The overall qualitative distribution of the relevant publications according to their strength of evidence is shown in Figure 4. Figure 3 provides detailed information about the allocation of the identified 68 articles into the different levels of evidence.

Figure 4
figure 4

Qualitative distribution of the identified publications according to the strength of evidence (cf. Table 1).

Full size image

While no systematic review of RCTs (level I) could be identified, 13 contributions reported about the results of clinical studies. Nine of these publications [9, 10, 16–22] referred to 5 RCTs (level II). Seven of the nine publications about RCTs were published in dental journals; one was a congress abstract [19], and another one was a doctoral thesis [21]. Except the thesis, which was written in Turkish, and a Norwegian article [20], the papers were published in English. The allocation according to the Jadad scale is shown in Table 13.

Table 13 Assessment of the quality of the articles about randomized controlled trials (RCTs) according to the quality score proposed by Jadad et al [15]
Full size table

Among the four publications allotted to level III, there were three congress abstracts [23–25] and one article [26] about an uncontrolled clinical investigation. Twelve contributions referring to 11 case reports were categorized into level IV.

Forty-three publications were classified to level V, representing narrative reviews, editorials, NTI-specific reviews, brief references to NTI-tss in articles, expert commentaries, and interviews.

Contents-based analysis of the articles on RCTs

The methods and results of the five RCTs are summarized in Tables 14 to 18. In four studies [9, 10, 21, 22], a complete-arch, hard acrylic resin occlusal stabilization (i.e. non-repositioning) appliance, which was worn at night, was used as therapeutic comparison (Tables 14, 15, 16, 17). In the other trial [18], a full-coverage occlusal splint similar to a bleaching tray was chosen; it was used at night and during stressful periods during wakefulness. Two RCTs focused on electromyographic (EMG) investigations of jaw-closing muscles: In one trial, participants were diagnosed with TMDs and concomitant bruxism [22] (Jadad score: 4); in the other study, the only inclusion criterion was bruxism (and associated symptoms) [21] (Jadad score: 2). In both investigations, the NTI-tss device, but not the occlusal stabilization appliance, showed a significant reduction of the EMG activity. A decrease in clinical symptoms and signs (i.e., pain; number of muscles with tenderness upon palpation; maximum unassisted jaw opening) could not be observed [22].

Table 14 Characteristics of the study of Baad-Hansen et al [22]
Full size table
Table 15 Characteristics of the study of Kavaklı [21]
Full size table
Table 16 Characteristics of the study of Jokstad et al [10, 19, 20]
Full size table
Table 17 Characteristics of the study of Magnusson et al [9]
Full size table
Table 18 Characteristics of the study of Shankland [16, 18]
Full size table

Two RCTs investigated clinical variables in TMD patients. One of these trials [10] (Jadad score: 5) found no difference between these two devices, while in the other study [9] (Jadad score: 3) a stabilization appliance led to a greater improvement than an NTI-tss device. However, only descriptive statistics were provided in the latter study report. In one RCT [18] (Jadad score: 1), patients with migraine or tension-type headache responded more favorably to the therapy with an NTI-tss device than with a bleaching tray-like splint.

Complications/side effects related to the NTI-tss-device

While no incidents occurred in the RCTs of Baad-Hansen et al [22] and Shankland [16–18], complications or side effects were observed in the other three RCTs [22] (Table 19). Two more incidents [13, 27] were identified in the reference lists of Magnusson et al [9] and Jokstad et al [10], respectively, while one report was found in a book chapter [2]. Finally, five cases were published in the FDA Medical Device reporting website (Table 20).

Table 19 Complications and side effects associated with the use of the NTI-tss device as described in the identified publications
Full size table
Table 20 Complications and/or side effects as described in the MAUDE database
Full size table

Different complications and side effects were documented; however, no case of an aspiration could be found.

Discussion

We were surprised to find that within the past decade as much as nine different trials related to the NTI-tss device were carried out, five of which were RCTs, while four were uncontrolled investigations [23–26] (Figure 1). As far as the estimation of the efficacy of a therapeutic measure is concerned, it is wise to consider primarily (or exclusively [28]) articles about RCTs, because they provide the strongest evidence on the efficacy of a therapy [29, 30]. Numerous examples have shown that observational studies, as compared to RCTs, are likely to generate over-optimistic (i.e., false-positive) conclusions [31–33]. Hence, it is not surprising that in clinical trials without control group the NTI-tss device was reported to be associated with a marked decrease in patients suffering from headache [25] and TMD-associated otologic symptoms (vertigo; otalgia; otic fullness; tinnitus; subjective hearing loss) [26].

In the five RCTs, which evaluated clinical and electromyographic effects of the NTI-tss device, different inclusion criteria and methods were used. Therefore, the results had to be analyzed separately. Each RCT has methodological weaknesses. Four studies had a limited number of participants (n = 10 to 38) (cf. Tables 14, 15, 16, 17, 18), which may have resulted in a statistical type II error (erroneous acceptance of the null hypothesis), as noted by Jokstad et al [10]. Shankland's investigation, in contrast, included 94 patients [18]; however, study and reporting quality were limited. The methodological flaws inherent to this study are not only reflected by the low Jadad score of 1, but also by the shortcoming to distinguish between the diagnoses of migraine and tension-type headache, which are completely different entities [34]. Shankland excluded patients with TMDs; however, there is a considerable symptom overlap between headache and TMD patients [35]. In fact, the same patient who suffers from pain in the temples may be diagnosed as having tension-type headache by a neurologist, whereas she may be diagnosed with myofascial pain in the temporal muscles by a dentist. Hence, an exact differential diagnosis between the two entities appears to be nearly impossible. Furthermore, there are limitations associated with the statistical analysis of the data gained in this study; for example, no information was provided about pre-treatment days of headache in that study [36].

Focusing on the identified articles about RCTs, two main indications for the NTI-tss splint may be distinguished: bruxism and TMDs. As far as bruxism is concerned, the studies of Baad-Hansen et al [22] and Kavaklı [21] have shown that – when compared to baseline EMG recordings from the masseter muscle during sleep – insertion of the NTI-tss device leads to a significant reduction in EMG activity of jaw closing muscles during clenching or grinding. These findings are compatible with early results from Van Eijden et al [37] who found that maximal effort clenches on the incisal edges of the incisors resulted in a significant decline of EMG activity as compared to clenching in intercuspation.

The fact that Baad-Hansen et al [22] were unable to correlate their EMG findings with clinical symptoms (e.g., reduction of reported pain) may be due to the short duration of wearing the intraoral device (2 weeks). Data from other investigations suggests that a decrease of EMG activity may indeed be associated with a pain reduction in patients with masticatory muscle pain [38]. Moreover, a decrease of EMG activity is not a unique characteristic of the NTI-tss device, but may be achieved with any anterior bite stop [39], sometimes even with a full-coverage occlusal appliance [38]. Hence, we concur with Clark and Minakuchi [2] who suggest in their recent review on oral appliances that it is "reasonable to use a partial-coverage anterior bite appliance in those patients with a known tooth-clenching habit, because this habit cannot be controlled with a stabilization appliance." On the other hand, a hard acrylic resin stabilization splint may be helpful in patients with tooth grinding, because it helps preventing the unwanted consequence of sleep bruxism [40], such as tooth wear (attrition), tooth grinding sounds, and – often – associated pain [41].

Regarding the management of patients with (localized) TMDs (e.g., myofascial face pain and/or arthralgia of the TMJs), the use of stabilization appliances is sufficiently supported by evidence in the dental literature [42–45]. Türp et al [45], for instance, concluded after a systematic review of the literature that most patients with masticatory muscle pain are helped by the incorporation of a stabilization splint. Particularly patients with local pain of the myofascial muscles (as opposed to individuals with widespread pain) are likely to experience improvement from this therapy [46].

Helkimo's statement that there is "no peer-reviewed scientific publication that exhibits that the NTI-splint is superior to other well-established conventional splints for the treatment if [sic] functionally related facial pain or mandibular dysfunction" [8] still holds true: a superiority of this anterior bite stop has not yet been shown. Nonetheless, the currently best available evidence [10] suggests that for the management of TMDs the NTI-tss device is similarly efficacious as a stabilization appliance. This demonstrates, on the other hand, that the observed reduction in clinical symptoms is not a feature unique to the NTI-tss device. Instead, it may also be achieved with other splints, even with other forms of anterior bite stops: Nilner et al [47], for instance, observed in patients suffering from myofascial face pain that at 6- and 10-weeks follow-up appointments a prefabricated appliance covering the six upper front teeth may be as effective in pain improvement as a traditional stabilization appliance. In a patient group with the same diagnosis, Al Quran and Kamal [48] who inserted an NTI-tss-like device (AMPS: anterior midline point stop) reported similar clinical results after 1 month and 3 months.

The positive effect after the incorporation of any type of appliance has frequently been explained by patient-specific behavioral changes [2]. However, such an assumption is hard to justify given the fact that the splints are usually worn at night during sleep, as it was also the case in the RCTs reviewed here. Regarding localized myofascial pain of the masticatory musculature, an alternative explanation is based on the heterogeneous activation capability of these muscles [37, 49–51]: Experimental results in healthy volunteers indicate that a temporary positional alteration of the mandible (e.g., an increase of the vertical dimension after insertion of an occlusal appliance) changes the intramuscular recruitment pattern, which is often followed by a pain reduction [52–55]. Although this biomechanical hypothesis needs further studies to be validated [56], it provides a physiologically plausible explanation for the therapeutic success gained with oral splints.

We were unable to find evidence for the statement that the NTI-tss device was indicated for the prevention of bruxism, TMDs, chronic tension-type headaches, migraine, or occlusal trauma. Likewise, due to the poor quality of the publications by Shankland [16, 18, 17], its use for the therapy of migraine remains doubtful.

Contrary to claims made in the dental literature [10, 12], cases of aspirated NTI-tss devices could not be identified in the present review (c.f. Tables 19 and 20). None of the five reports published in the MAUDE database referred to such an event. Fleten and Gjerdet [13], who allegedly reported about an aspiration, presented a case where an NTI-tss splint was lodged at the esophageal entrance.

Nonetheless, possible side effects, particularly those related to the teeth and the occlusion, remain the greatest concern when wearing NTI-tss splints (cf. Table 20). Of course, adverse events may also occur when a stabilization appliance is used. Clark and Minakuchi [2] mention two concerns which, albeit rare, need to be considered in this context: (a) increase of clenching activity when wearing the bite splint; (b) unintended occlusal change after full-time use, due to alterations of the position of individual teeth or of the mandible. However, if a stabilization appliance is worn at night only – and, if needed (e.g., during stressful periods of life), additionally during one or two hours of the day – the development of an appliance-induced malocclusion is unlikely. Other side effects, such as increased or decreased salivation [57], are usually short-lasting, while tension in the teeth [57] can mostly be eliminated by careful adjustments along the labial and/or buccal surfaces that ensure the retention of the stabilization appliance. Since it is certain that published case reports represent only a fraction of all adverse events that have occurred while using any type of intraoral appliance, clinicians should be encouraged to frankly report any unintended negative outcome associated with splint therapy. In the hierarchy of scientific evidence, case reports represent a low level (cf. Table 9). Nonetheless, considering the absence of other data we believe that the publications summarized in Tables 19 and 20 deliver valuable information and convey an important message to the clinician (as well as to the patient), namely that the use of the device is not fee of risks. Unfortunately, no risk quantification is possible at this point in time. Meanwhile, it remains crucial to take care that a patient receiving an NTI-tss device remains compliant with follow-up appointments, especially when wearing it over an extended period of time.

It has been argued that an advantage of the NTI-tss splint as compared to a conventional occlusal bite splint is the reduced chair-side time. Therefore, some dentists are likely to be surprised when confronted with Magnusson et al's report [9] that on the average 27 minutes (range: 17–45 minutes) were needed to fit and adjust an NTI-tss device as opposed to 17 minutes (range: 11–26 minutes) for making interocclusal records, taking impressions, and adjusting the surface of a stabilization appliance. Indeed, this computation appears to be debatable. In a previous study it was shown that the average time required for chair-side adjustment of a stabilization appliances (fabricated without and with face-bow) was less than 11 minutes (range: 4–27 minutes) [58]. To determine centric relation by fabricating an interocclusal record and to make alginate impressions of the upper and lower jaw one may calculate additional 15 minutes. Hence, for making a stabilization appliance the dentist needs to be at the chair for about 25 to 30 minutes, which corresponds to the time frame reported by Magnusson et al [9] for preparing an NTI-tss device. Yet, it should be noted that an inexperienced clinician may need considerably more time for either device.

The occlusal stabilization splint [59] remains the "gold standard" for the therapy of patients with (particularly localized) temporomandibular pain (i.e., myofascial pain of the masticatory muscles; TMJ arthralgia) and/or bruxism because it "is an easily used, potentially long-term, and clinically effective treatment intervention with reasonable nightly patient compliance and good outcomes," thereby exhibiting "few potential complications [2]."

Conclusion

There is evidence from RCTs that the NTI-tss bite stop may be successfully used for the management of TMDs and bruxism. However, to avoid potential side-effects, it must be ensured that the patient is willing to return regularly to the dentist's office for control sessions and, if needed, re-adjustments. If this prerequisite is fulfilled, the NTI-tss splint may be particularly justified in the following clinical situations:

• A patient with acute and intense temporomandibular pain (possibly accompanied by a restricted jaw opening), who requires (as an emergency therapy) the rapid incorporation of an oral appliance to increase the vertical dimension of the jaws.

• A reduction of the EMG activity of jaw closing muscles during jaw clenching or tooth grinding is desired.