4 Evidence and interpretation
4.1.1 The clinical effectiveness of HealOzone treatment was compared with a control in a number of randomised controlled trials (RCTs), by comparing the number of dental caries lesions that reversed (for example, changing from leathery to hard) or progressed.
4.1.2 Ten RCTs of HealOzone treatment of primary dental caries were included in the systematic review conducted by the Assessment Group. Studies of less than 6 months duration were excluded from the review on the basis of clinical advice that follow-up periods of less than 6 months were inadequate to assess caries progression. Seven RCTs (two PhD theses, one pilot study [included in one of the theses] and four abstracts) evaluated the effect of HealOzone treatment on pit and fissure caries. Three RCTs (one published, one unpublished and one published in abstract only) evaluated the effect of HealOzone treatment on root caries. The studies were conducted in permanent teeth, with the exception of the study reported in one of the PhD theses. Two of the studies evaluated whether the addition of HealOzone treatment improved sealant retention: one in the treatment of pit and fissure caries, and another in the treatment of root caries.
4.1.3 The Assessment Group noted a number of factors in the studies that made it difficult to assess the effectiveness of HealOzone treatment alone and prevented a quantitative synthesis of the results. Several studies (particularly those published in abstract only) did not fully report the methodology used; for example, it was not always clear whether the assessors of treatment outcome were blinded to the treatment group. Also, the data analysis was not necessarily appropriate. For example, analysing the data by lesion without taking into account the fact that measurements derived from two or more lesions in the same patient are not independent could introduce bias into the results. Study participants received repeat treatments at different timepoints in different studies and in some studies no repeat treatment was given.
4.1.4 The Assessment Group reported the results of five RCTs (one PhD thesis and four abstracts) that evaluated the effect of HealOzone treatment in adults with non-cavitated pit and fissure caries. The PhD thesis (n = 90; 258 lesions) reported that at 12 months in the HealOzone treatment group (ozone plus reductant) 7% of lesions had reversed, 57% remained stable and 36% had progressed, compared with 6% reversed, 49% stable and 46% progressed in the control group (reductant only). The mean change from baseline in clinical severity score at 12 months was not statistically significant (p = 0.112). There was no statistically significant difference in sealant retention between treatment groups at 12 months (33% had partial loss in the sealant margins in the HealOzone group compared with 30% in the control group).
4.1.5 In three abstracts, the proportion of lesions reported as clinically reversed at 6 to 12 months follow-up ranged from 87% to 99% in the HealOzone treatment group. All studies reported that no significant clinical changes were observed in the control group, but no data were provided. The fourth abstract (n = 38; 76 lesions) reported that all lesions were hard at 3 months (that is, all caries had reversed) in the HealOzone treatment group (air abrasion, ozone, mineral wash plus glass ionomer sealant); reversal rates were not reported in the control group (conventional drilling and filling). The Assessment Group noted that the results of these studies should be interpreted with caution because the lack of detail meant that their methodology could not be easily assessed.
4.1.6 A small pilot study reported in the PhD thesis described in Section 4.1.4 (n = 8; 17 lesions per group) evaluated the effect of HealOzone treatment on cavitated pit and fissure caries. It reported statistically significant improvements in hardness and visual scores (p < 0.05) in the HealOzone group compared with the control group and no significant differences between groups in cavitation score, colour and perceived treatment need (p > 0.05).
4.1.7 The study in the second PhD thesis (n = 21; 74 lesions) evaluated the effectiveness of HealOzone treatment in non-cavitated pit and fissure caries in deciduous teeth (children aged 7 to 9 years). Data were presented graphically; at the 12-month follow-up, there was a small reduction in the severity of dental caries in the HealOzone treatment group (ozone plus reductant), and an increase in severity scores in the control group (reductant only). There was a statistically significant change in clinical severity scores with treatment over time (p < 0.01).
4.1.8 Three RCTs evaluated the effectiveness of HealOzone treatment for non-cavitated root caries; one published in full, one unpublished and one published only in abstract form. In the published study two lesions in each of 89 participants were randomised to treatment or control; 89 lesions per group. This study reported that 98% of lesions in the HealOzone treatment group (ozone application, reductant plus patient kit) became hard at 12 months compared with 1% in the control group (air treatment, reductant plus patient kit). The 21-month data from this study reported that 100% of lesions became hard with HealOzone treatment compared with 8% of lesions becoming hard, 80% remaining leathery and 12% becoming soft in the control group.
4.1.9 The abstract (n = 260 with two lesions each; 260 lesions per group) reported that 80% of soft lesions had reversed from clinical severity index 4 to 3 and that 94% of leathery lesions became hard and arrested in the HealOzone treatment group (ozone application) at 6 months. There were no statistically significant changes in lesion severity for either type of lesion in the control group (no treatment).
4.1.10 The unpublished full-text study (n = 79; 220 lesions, the numbers in each group were not reported) investigated the effectiveness of HealOzone treatment (cleaning, ozone and reductant, with or without sealant) in cavitated and non-cavitated root caries at 12 months follow-up. The results for cavitated and non-cavitated lesions could not be disaggregated. Overall, HealOzone treatment was associated with a statistically significant reversal of dental caries compared with the control treatment (reductant, with or without sealant) (p < 0.001). In the treatment group 99% of lesions improved (47% lesions became hard, and 52% became less severe), compared with 12% (none became hard) in the control group. Sealant retention was also statistically significantly improved: 61% in the HealOzone group compared with 42% in the control group (p < 0.05).
4.1.11 Although the unpublished study described in Section 4.1.10 did not present separate data for cavitated and non-cavitated root caries, it was reported that the percentage of cavitated lesions in the HealOzone group that became hard decreased from 9% at 1 month to 1% at 9 months follow-up. No results were reported for the control group.
4.1.12 The Committee heard from experts that, although HealOzone treatment sterilises the surface of the tooth, microorganisms will immediately start to recolonise the area and will be well-established about 2 weeks after HealOzone treatment.
4.1.13 The experts expressed concerns about the clinical trials that reported strongly favourable results. Their concerns included the unexpectedly poor performance of the controls, problems in the accurate diagnosis of the severity of dental caries and the absence of objective outcome measures.
4.2.1 No published economic evaluations were identified on HealOzone treatment of dental caries. The manufacturer submitted an economic model. The Assessment Group developed a second model, but argued that, given the current state of the clinical effectiveness evidence, economic analysis is premature and the model should therefore be taken as illustrative only. The Assessment Group's model is therefore not described further here.
4.2.2 The submission from the manufacturer of the device assessed the cost effectiveness of adding HealOzone to conventional treatment that did not include preventive treatment. Effectiveness data for the addition of HealOzone treatment were based on average reversal rates of dental caries reported in the RCTs for non-cavitated (93.3%) and cavitated pit and fissure caries (79%), and for root caries (84.5%). The effectiveness of conventional treatment was based on the average annual progression rate of dental caries reported in clinical studies that were excluded from the Assessment Group's systematic review. The additional cost of HealOzone treatment per filling avoided was £9.58 in non-cavitated pit and fissure caries, £11.63 in cavitated pit and fissure caries and £5.18 in root caries.
4.3.1 The Committee reviewed the evidence available on the clinical and cost effectiveness of HealOzone treatment of tooth decay, having considered evidence on the nature of the condition and the value placed on the benefits of HealOzone by people with tooth decay, those who represent them, and clinical experts. It was also mindful of the need to ensure that its advice took account of the effective use of NHS resources.
4.3.2 The Committee considered the mode by whichdental cariesis reversed or arrested by HealOzone treatment. It accepted the evidence presented in the submission that ozone eliminated most microorganisms. However, given the testimony of clinical experts and lack of evidence to the contrary, the Committee concluded that it was logical that microorganisms would immediately start to recolonise the area and become well-established soon after HealOzone treatment. It considered the hypothesis that ozone could remove proteins in carious lesions by oxidation of amino acids, which could in turn enable remineralisation in the presence of the mineral reductant. However, the Committee noted a lack of evidence to support this theory.
4.3.3 The Committee considered the evidence in which patients in the intervention and control arm received fluoride treatment (mineral reductant and the patient kit). The Committee was aware that many of the measures used in the RCTs to monitor caries are not well validated and are unreliable. It discussed the validity of the evidence that reported little or no effect in the control group of the fluoride comparator treatment. The Committee noted that experts said they would have expected to see higher rates of caries reversal from fluoride treatment in the control arms. The Committee concluded that, in light of these concerns, the evidence could not be considered reliable.
4.3.4 The Committee accepted the Assessment Group's rationale for not examining the submitted trials of less than 6 months duration; the Committee agreed that shorter follow-up periods were inadequate to assess caries progress. Of the RCTs of HealOzone for non-cavitated pit and fissure caries that were of more than 6 months duration, the Committee took into consideration the fact that the non-significant results in the PhD thesis conflicted with the significant results of the abstracts. However, given the lower reliability of abstracts, the lack of information reported and the concerns over the robustness of the methodology used, it concluded that the benefits of HealOzone for non-cavitated pit and fissure caries had not been adequately demonstrated. Similarly, given the methodological concerns and the small sample size of the pilot studies in non-cavitated pit and fissure caries in deciduous teeth and cavitated pit and fissure caries in permanent teeth, the Committee concluded there is insufficient evidence to recommend this technology in these subgroups and further research is required.
4.3.5 For the reasons expressed in 4.3.3, the Committee concluded that the evidence from RCTs of cavitated root caries was unreliable, and that HealOzone should not be recommended for the treatment of cavitated root caries.
4.3.6 In light of the Committee's conclusion that the evidence from the RCTs was unreliable, the Committee did not discuss the findings of the cost-effectiveness evaluations.
4.3.7 In summary, the Committee concluded that there was insufficient evidence on the effectiveness of HealOzone treatment for this technology to be recommended, except as part of well-designed RCTs.