Oral splints for patients with temporomandibular disorders or bruxism: a systematic review and economic evaluation
Philip Riley,Anne-Marie Glenny,Helen V Worthington,Elisabet Jacobsen,Clare Robertson,Justin Durham,Stephen Davies,Helen Petersen,Dwayne Boyers
DOI: https://doi.org/10.3310/hta24070
Abstract:Background: Splints are a non-invasive, reversible management option for temporomandibular disorders or bruxism. The clinical effectiveness and cost-effectiveness of splints remain uncertain. Objectives: The objectives were to evaluate the clinical effectiveness and cost-effectiveness of splints for patients with temporomandibular disorders or bruxism. This evidence synthesis compared (1) all types of splint versus no/minimal treatment/control splints and (2) prefabricated versus custom-made splints, for the primary outcomes, which were pain (temporomandibular disorders) and tooth wear (bruxism). Review methods: Four databases, including MEDLINE and EMBASE, were searched from inception until 1 October 2018 for randomised clinical trials. The searches were conducted on 1 October 2018. Cochrane review methods (including risk of bias) were used for the systematic review. Standardised mean differences were pooled for the primary outcome of pain, using random-effects models in temporomandibular disorder patients. A Markov cohort, state-transition model, populated using current pain and Characteristic Pain Intensity data, was used to estimate the incremental cost-effectiveness ratio for splints compared with no splint, from an NHS perspective over a lifetime horizon. A value-of-information analysis identified future research priorities. Results: Fifty-two trials were included in the systematic review. The evidence identified was of very low quality with unclear reporting by temporomandibular disorder subtype. When all subtypes were pooled into one global temporomandibular disorder group, there was no evidence that splints reduced pain [standardised mean difference (at up to 3 months) -0.18, 95% confidence interval -0.42 to 0.06; substantial heterogeneity] when compared with no splints or a minimal intervention. There was no evidence that other outcomes, including temporomandibular joint noises, decreased mouth-opening, and quality of life, improved when using splints. Adverse events were generally not reported, but seemed infrequent when reported. The most plausible base-case incremental cost-effectiveness ratio was uncertain and driven by the lack of clinical effectiveness evidence. The cost-effectiveness acceptability curve showed splints becoming more cost-effective at a willingness-to-pay threshold of ≈£6000, but the probability never exceeded 60% at higher levels of willingness to pay. Results were sensitive to longer-term extrapolation assumptions. A value-of-information analysis indicated that further research is required. There were no studies measuring tooth wear in patients with bruxism. One small study looked at pain and found a reduction in the splint group [mean difference (0-10 scale) -2.01, 95% CI -1.40 to -2.62; very low-quality evidence]. As there was no evidence of a difference between splints and no splints, the second objective became irrelevant. Limitations: There was a large variation in the diagnostic criteria, splint types and outcome measures used and reported. Sensitivity analyses based on these limitations did not indicate a reduction in pain. Conclusions: The very low-quality evidence identified did not demonstrate that splints reduced pain in temporomandibular disorders as a group of conditions. There is insufficient evidence to determine whether or not splints reduce tooth wear in patients with bruxism. There remains substantial uncertainty surrounding the most plausible incremental cost-effectiveness ratio. Future work: There is a need for well-conducted trials to determine the clinical effectiveness and cost-effectiveness of splints in patients with carefully diagnosed and subtyped temporomandibular disorders, and patients with bruxism, using agreed measures of pain and tooth wear. Study registration: This study is registered as PROSPERO CRD42017068512. Funding: This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 24, No. 7. See the NIHR Journals Library website for further project information.