Description of the condition
Tooth staining/discoloration is traditionally classified as intrinsic or extrinsic stain. Intrinsic stain is found within the internal structure of the tooth and is caused by agents introduced during tooth formation that disrupt the normal tooth development. Causes include hematologic and developmental abnormalities, drugs such as tetracycline, high levels of fluoride and exposure to cationic metals, such as tin, iron and lead. Intrinsic staining is not responsive to the abrasive effects of home-use toothpastes. Extrinsic stain occurs after tooth development and is localized mainly to the salivary pellicle. The salivary pellicle is a thin layer of mucoproteins that absorb to the calcium phosphate in tooth enamel especially in areas of the crown that are not exposed to friction (Hannig 2006). The pellicle plays a role in the adherence of discolored materials to the tooth's surface, especially in areas of the teeth that are inaccessible to toothbrushing. Poor oral hygiene and the inability to remove stain-producing materials increase the degree of extrinsic staining. Certain tannin rich foods such as tea, coffee and red wine and tar from tobacco products are common promoters of the development of stain. Occupational or medicinal exposure to metallic salts such as iron, copper, silver nitrate and stannous fluoride can also be a source of extrinsic staining. Use of chlorhexidine, an antibacterial mouthwash, has been shown to commonly cause a superficial staining of the teeth and tongue (Alshara 2013). Since removal of the salivary pellicle often removes extrinsic stain/discoloration, many extrinsic stains are susceptible to the abrasive actions of tooth whitening toothpastes.
Many individuals are interested in the color of their teeth, with perceptions of white shades contributing to a greater acceptance of one's physical appearance, enhanced body image and self esteem (Carlsson 2008; Kershaw 2008; Van der Geld 2007). Studies suggest that the appearance of white teeth correlates positively with perceptions of physical and oral health, views of social competence and intellectual ability (Kershaw 2008; Tin-Oo 2011; Van der Geld 2007). In the context of socially favorable attitudes toward tooth color, recent studies have revealed up to 55% of individuals are dissatisfied with their tooth color (Akarslan 2009; Alkhatib 2005; Bauer 2012; Samorodnitzky-Naveh 2010). Dissatisfaction with tooth color is greatest among younger women who may value their physical appearance more, however, individuals across different age, socioeconomic, gender, and cultural groups have increasingly assessed their tooth color as unsatisfactory (Alkhatib 2005; Bauer 2012; Lajnert 2012; Samorodnitzky-Naveh 2007; Xiao 2007). The potential physical, psychological and social benefits associated with white teeth contribute to a consumer demand for tooth whitening products that has surpassed the desire for all other dental aesthetic treatments (Akarslan 2009). With the public interest in tooth whitening increasing, the production, marketing and use of over-the-counter, home-use whitening toothpastes have proliferated, highlighting a need for healthcare clinicians and consumers alike to be informed of the potential benefits and risks associated with these products.
Description of the intervention
The objective of teeth whitening is to remove tooth staining and improve tooth color. Whitening toothpastes designed for home-use primarily contain abrasive ingredients such as silicas and phosphates that mechanically remove and prevent the development of extrinsic stain without imposing significant damage or degradation of the enamel surface of the teeth. Although the abrasive agent is the primary ingredient, chemical agents such as hydrogen peroxide and calcium peroxide, pyrophosphates, or proteases may be added to intensify the effects of extrinsic stain removal. Additional substances added to the toothpastes include humectants for moisture, surfactants to enhance the foaming properties, fluoride for prevention of caries, and sweetening agents to enhance the flavor of the product (Joiner 2008). Over-the-counter whitening toothpastes and mouthrinses generally contain no more than 0.1% hydrogen peroxide, levels that do not pose a risk to the health of the consumer (ADA 2012; SCCP 2007).
The American Dental Association (ADA) in the United States of America (USA) and the British Dental Health Foundation (BDHF) in the United Kingdom (UK), while not regulatory agencies, nonetheless grant 'Seal of acceptance' status to certain whitening toothpastes available on the market. For instance, the ADA grants the 'ADA seal of acceptance' to whitening toothpastes containing polishing or chemical agents to improve tooth appearance by removing surface stains through gentle polishing, chemical chelation, or other non-bleaching action (ADA 2012). Whitening toothpastes considered for the 'ADA seal of acceptance' are evaluated according to the Guidelines for Fluoride-Containing Dentifrices (which evaluate the safety of the dentifrice and its effectiveness against caries) as well as the ADA Program Guidelines for Home-Use Tooth Stain Removal Products (which evaluate the efficacy and safety of the whitening toothpaste) (ADA 2001; ADA 2012). Additionally, the abrasive potential of a dentifrice must be measured and conform to the requirements established by the International Organization for Standardardization (ISO) 11609:1995 (ADA 2012). The ADA has thus far granted the 'ADA seal of acceptance' to 16 whitening toothpastes (ADA 2013). The BDHF has thus far awarded the 'Seal of acceptance' to eight whitening toothpastes (BDHF 2013).
How the intervention might work
The main purpose of whitening toothpastes is the mechanical or chemical removal of stained plaque, either by insoluble abrasive substances or by enzymes that attach to proteins in the pellicle (Joiner 2010). Abrasives work in conjunction with toothbrush bristles to remove external stained plaque but have no effect on the intrinsic color of the teeth. The most commonly used abrasive substances are silica (hydrated silica, high cleaning silica, precipitated silica), oxides (aluminium oxide and titanium oxide), carbonates (calcium carbonate, sodium carbonate, sodium bicarbonate), phosphates (dicalcium phosphate or anhydrous dehydrate or both, calcium pyrophosphate, insoluble sodium metaphosphate, tricalcium phosphate), pumice and perlite (Davies 2010).
Chemical agents vary in their mode of action. Enzymes such as protease and papa ine work directly on degrading proteins in the pellicle that cause stain. Chelating/sequestering agents such as pyrophosphates have strong binding affinity with enamel, dentine and tartar and reduce the amount of adsorbed pellicle to which stains are bound. They also complex with calcium and magnesium in the saliva to prevent redeposition on the teeth. Surfactants are amphiphilic molecules that solubilize the organic matrix of the pellicle, facilitating the action of the abrasives (Joiner 2010). Additionally, agents such as peroxide compounds (hydrogen peroxide, carbamide peroxide) act by breaking down the stain molecules, bleaching the pellicle on the tooth surface as well as penetrating the tooth tissue to provide some minimal removal of intrinsic stains.
Moreover, increased tooth sensitivity may occur with the use of whitening toothpastes, especially for those that contain harsh abrasives. Other adverse effects, such as gingival irritation, are also of concern.
Why it is important to do this review
During the past decade, especially, home-use whitening toothpastes have become widely available for sale over-the-counter in both the USA and the UK. With the current rapid growth in demand for tooth whitening, it is imperative that the oral health community base its recommendations to patients and the public on sound scientific evaluations conducted in well-designed and independent randomized controlled trials (RCTs).
There are no published systematic reviews available to date that comprehensively analyze and cogently summarize the scientific evidence for RCTs on the effectiveness of home-use whitening toothpastes for the prevention and treatment of tooth staining in adults. A Cochrane review conducted by Hasson and colleagues (Hasson 2006) addresses the effectiveness of home-based chemically-induced whitening agents, such as gels applied in trays, paint-on films, and whitening strips, but not whitening toothpastes. The public should be made aware of any untoward effects due to the use of home-use whitening toothpastes. Hence, there is a clear and compelling need for a Cochrane review on the effectiveness and safety aspects of whitening toothpastes.
To assess the beneficial and adverse effects of home-use whitening toothpastes with an abrasive or chemical action or both, in the prevention and removal of extrinsic dental stains in adults.
Criteria for considering studies for this review
Types of studies
We will include examiner-blinded and participant-blinded, randomized controlled trials (RCTs) of whitening toothpastes compared with either placebo/non-whitening toothpastes or a different type, concentration or frequency of use of whitening toothpastes. We will distinguish between prevention studies (which will require a baseline prophylaxis) and treatment studies (which do not include a baseline prophylaxis). We will include both parallel group studies and cross-over studies (with a minimum washout period of one month). All relevant non-English-language papers will be translated, if possible.
Types of participants
Participants will be adults (aged 18 and above) with discoloration or extrinsic staining of teeth. Studies where participants have extrinsic staining or discoloration of teeth due to exposure to tetracyclines, or fluorides, or hereditary of systemic diseases affecting tooth enamel (e.g. amelogenesis imperfecta) will be excluded from this review. In studies that do not state whether or not staining was due to intrinsic staining, we will contact the authors for clarification. We will exclude participants with intrinsic stains only. Finally, we will also include studies where participants wish to have the 'whiteness/brightness' of their teeth increased, even though they do not have overt/specific patches of staining at baseline.
Types of interventions
The intervention of interest is home-use whitening toothpastes containing various whitening active ingredients that operate by chemical or abrasive action or both. The minimum duration of the intervention will be continuous use of the home-use whitening toothpastes for at least two weeks. The control groups will use placebo/non-whitening toothpastes that do not contain any whitening active ingredients. Studies in which whitening toothpastes are compared with strips, professional whitening procedures, and/or gels will be excluded.
Types of outcome measures
All outcomes will be grouped as short term (two weeks to four weeks of continuous use), moderate term (five weeks to 11 weeks of continuous use) and long term (12 or more weeks of continuous use) with a follow-up evaluation of up to three days after completion of the designated intervention period. Studies with an intervention period of less than two weeks or a follow-up evaluation period of more than three days will be excluded, as will studies that report only on plaque, gingivitis, tartar control, calculus, anticaries efficacy, or dentine hypersensitivity.
Measurement of tooth/color stain
Many methods are used clinically for the assessment of tooth color, stain measurement and analysis. Possible methods include:
- subjective measurements using various stain removal indices (e.g. Lobene Stain Index; modified Lobene Stain Index); and special visual color matching shade guides (e.g. Degudent Shadepilot (Classical Vitashade scale); Shaw and Murray Stain Index; Vita Easyshade (Vita 3D-Master scale); Vita Shade Guide); and
- objective measurements using instruments such as colorimeters (e.g. Minolta CR-221 Chroma Meter), and spectrophotometers which quantify the surface color of the teeth using the CIELAB color scale (Commission International De l'Eclairage (CIE) 1976 (L*, a* and b*)) where L* describes the lightness of a material, a* denotes the area on a red-green axis and b* relates to the site on a blue-yellow axis (Brook 2007; Chen 2010; White 2001).
The simultaneous use of instrumental and visual color matching methods are complementary and may lead to more precise and predictable esthetic outcomes (Chu 2010).
Subjective and objective measurements will be analyzed as continuous data. If a shorter ordinal scale is dichotomized, risk ratios will be analyzed.
Patient-reported improvement or reduction in tooth staining are obtained directly from patients through interviews or self completed questionnaires. Although patient-reported outcomes are of particular relevance in a topic of such esthetic importance, these are not commonly reported in the studies.
- Prevention and removal of tooth stains in adults as assessed by change in tooth stain area and intensity, and change in tooth color.
- Whiteness/brightness as assessed by change in tooth whiteness.
- Any adverse effects/reactions on the teeth or gums reported by the examiners or participants, which may include tooth sensitivity and gingival irritation.
Patient satisfaction with tooth whitening effects or stain removal effects or both as assessed by patient reports and clinical exams.
Search methods for identification of studies
We will attempt to identify all relevant RCTs involving human subjects on the effectiveness of home-use whitening toothpastes. No limitations will be placed on language or date of publication in the electronic searches of the databases.
We will search the following databases to identify all relevant RCTs:
- the Cochrane Oral Health Group's Trials Register (to date)
- the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, current issue)
- MEDLINE via OVID (1946 to date) (Appendix 1)
- EMBASE via OVID (1980 to date)
- LILACS via BIREME (1980 to date)
- Web of Science (limited to Conference Proceedings) (1990 to date)
- Proquest Dissertations and Theses (1806 to date).
We will search the following trials registries to identify ongoing studies:
- The US Institutes of Health Trials Register (http://clinicaltrials.gov)
- The metaRegister of Controlled Trials (www.controlled-trials.com/mrct)
- The World Health Organization Clinical Trials Registry Platform (www.who.int/ictrp/en/).
The subject search for MEDLINE (presented in Appendix 1) will be adapted for the search of other databases and will be revised appropriately to take into account differences in controlled vocabulary and syntax rules. We will link the MEDLINE search to the Cochrane Highly Sensitive Search Strategy for identifying RCTs, as referenced in Chapter 220.127.116.11 and detailed in box 6.4.c of theCochrane Handbook for Systematic Reviews of Interventions version 5.1.0 (updated March 2011) (Higgins 2011).
Searching other resources
We will handsearch the previous 12 months of the following journals to allow for the potential delay in indexing journals at the National Library of Medicine:
- Journal of Operative Dentistry
- Journal of Esthetic Dentistry
- Compendium of Continuing Dental Education
- Journal of Clinical Dentistry
- Journal of Prosthetic Dentistry
- Journal of Esthetic and Restorative Dentistry
- Journal of Dental Research.
We will identify additional study reports by searching the reference lists of relevant studies, systematic reviews and other articles found during the electronic searches.
Dental organizations, researchers, and experts known to be involved in the field will be contacted in an effort to trace unpublished or ongoing studies or both. In addition, the manufacturers of whitening toothpastes will be contacted.
Data collection and analysis
Selection of studies
We will import the downloaded set of records from each database to the bibiographic software package RefWorks and merge them into one core database to remove duplicate records and to facilitate the retrieval of relevant articles. All potentially relevant study reports identified when searching other sources (reference lists of relevant studies, reviews, articles, and text books) will also be obtained, and the records located from searching these ancillary sources will be entered manually into RefWorks.
We will screen the titles/abstracts/keywords of all records identified by the searches, independently and in duplicate, against the inclusion criteria of this review. We will obtain the full text reports of all eligible and potentially eligible studies (those with insufficient information in the title/abstract) to further establish whether the studies meet the inclusion criteria or not, using an inclusion criteria form, which will be previously prepared and pilot tested. Disagreements will be resolved by discussion. Where a resolution is not possible, another review author will be consulted.
We will attempt to contact authors of articles that cannot be classified in order to ascertain whether inclusion criteria are met. We will link multiple reports of the same study under one single study ID and will use the report with the most information/data as the primary study.
Data extraction and management
All studies meeting the inclusion criteria will then undergo data extraction and a risk of bias assessment. Studies rejected at this or subsequent stages will be recorded in the characteristics of excluded studies table, along with reasons for exclusion.
Two review authors will extract data from all included studies, independently and in duplicate, using a standardized data extraction form. A third review author will resolve disagreements until a consensus is reached. All study authors will be contacted as necessary regarding missing information or clarification of data or both. If a study is reported in a language unfamiliar to the review authors, appropriate translators will assist with data extraction. Duplicate publications will be reviewed as a single source.
The following trial characteristics will be extracted.
- Year of publication.
- Country of origin.
- Language of publication.
- Inclusion criteria.
- Exclusion criteria.
- Duration of the trial (weeks/months of follow-up).
- Number of participants at the start of the trial.
- Methods of stratifying/balancing groups pre-randomization.
- Methods of randomization.
- Number and type of teeth assessed.
- Comparability of baseline characteristics.
- Possible post-randomization adjustments for differences in groups.
- Reliability of primary outcome measurement.
Characteristics related to participants that will be extracted include.
- Age in years.
- Habits of use of tobacco, coffee, tea, and red wine.
- Prior use of chlorhexidine mouthrinse.
- Use of chlorhexidine mouthrinse during the trial.
- Prior dental prophylaxis.
- Year participation in trial began.
- Country where participation in trial occurred.
- Setting where recruitment into trial occurred.
Characteristics of the intervention (whitening toothpaste) that will be extracted include.
- Main whitening ingredient.
- Type of whitening ingredient (chemical or abrasive or both).
- Concentration of the whitening ingredient.
- Amount of toothpaste application.
- Duration of use.
- Frequency of use.
Characteristics of the outcomes that will be extracted include.
- Continuous outcomes such as various ways of assessing and reporting prevention and treatment of tooth staining (e.g. stain scores, shade guide scores, colorimeter values, and spectrophotometer values), where scores will be recorded at baseline and at specified follow-up times.
- Units of measurements used.
- Categorical outcomes (e.g. percentage and number of reduction in stain score, shade guide score, colorimeter values, and spectrophotometer values).
- Percentage and number of improvement in whiteness.
- Any adverse events/reactions assessed or reported.
- Contact information for the study authors, sponsoring institutions, and manufacturers.
- All other relevant outcomes assessed or reported.
Assessment of risk of bias in included studies
Two review authors will undertake the assessment of risk of bias of all the included studies, independently and in duplicate, as part of the data extraction process. A third review author will resolve any disagreements.
We will assess the following seven domains of potential bias.
- Random sequence generation (selection bias).
- Allocation concealment (selection bias).
- Blinding of participants and personnnel (performance bias).
- Blinding of outcome assessors (detection bias).
- Incomplete outcome data addressed (short-term outcomes (two to 12 weeks); long-term outcomes (> 12 weeks) - (attrition bias).
- Selective outcome reporting (reporting bias).
- Other potential sources of bias.
The assessment of the risk of bias will be carried out in two steps as described in Chapter 8 of the Cochrane Handbook for Systematic Review of Interventions (Higgins 2011). The first step will involve a description of the support for judgement of each domain (what was reported to have happened); and the second part will involve an assessment of the risk of bias (as low, high, or unclear).
The potential risk of bias for each study will then be summarized and the studies will be grouped into the following three categories.
- Low risk of bias: plausible bias not likely to seriously alter the results (if low risk of bias for all items).
- Unclear risk of bias: plausible bias that raises some doubt about the results (if unclear risk of bias for one or more key items).
- High risk of bias: plausible bias that seriously weakens confidence in the results (if high risk of bias for one or more key items).
A standardized risk of bias assessment form will be developed which will include the criteria for assessing the above domains of risk of bias and data will be entered in the risk of bias table in Review Manager (RevMan) (RevMan 2012).
The results of the risk of the bias assessment will be depicted graphically and a description of the results will be included in the narrative.
Measures of treatment effect
For continuous data, we will use mean difference (MD) scores with 95% confidence intervals (CIs) if the scales are similar.
If the scales are different, we will use the standardized mean difference (SMD) and 95% CIs will be calculated. Since it is expected that certain studies will be reporting mean change scores from baseline, while others will be reporting mean final scores, the SMD will be chosen as the measure of effect, as change scores can be combined with final scores when using SMD.
If binary data are identified, risk ratios (RRs) and 95% CIs will be calculated.
Studies with multiple treatment groups
In studies with more than one relevant intervention group and a common control group, such as those comparing whitening toothpastes with different chemical and/or abrasive active ingredients, to a placebo group, raw results (the numbers, mean scores and values, and standard deviations) from all relevant groups will be used. However, the control group numbers will be divided into approximately equally sized smaller groups to provide a pair-wise comparison for each modality. This will enable the inclusion of all relevant data in the primary meta-analyses.
Handling of reporting of results at different follow-up periods
In studies reporting results of more than one follow-up period, separate analyses will be performed for short term (two to four weeks), medium term (five to 11 weeks) and long term (12 or more weeks), and results will be categorized accordingly.
Unit of analysis issues
In this review, the unit of analysis will be the individual (summarizing information from teeth included in the study). In order to account for the two-period nature of the data in cross-over studies (with at least one month washout period), we will enter log effect estimates into RevMan using the generic inverse variance method (Elbourne 2002).
Dealing with missing data
Where standard deviations are not reported, they will be estimated from the reported standard errors or P values following the methods described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).
We will contact the authors of the studies to clarify incompletely reported data related to trial characteristics, methodology, and outcomes, whenever possible.
Assessment of heterogeneity
Heterogeneity in the results of the studies will be assessed by inspection of a graphical display of the estimated treatment effects from the studies. An I
- 0% to 40%: may not be important
- 30% to 50%: may represent moderate heterogeneity
- 50% to 90%: may represent substantial heterogeneity
- 75% to 100%: considerable heterogeneity.
Assessment of reporting biases
If more than ten studies are included in a meta-analysis, a test for asymmetry of a funnel plot will be undertaken to examine for possible publication bias (Egger 1997). Interpretation of the test is performed by visual inspection of the funnel plot. Since this test has low power, the absence of asymmetry does not exclude the existence of bias (Higgins 2011).
For continuous data, we will combine MDs (or SMDs if different scales are used) and 95% CIs. For dichotomous data, we will combine RRs and 95% CIs.
A random-effects model will be used if there is a minimum of four studies, otherwise, a fixed-effect model will be used.
RevMan will be used to conduct the meta-analysis (RevMan 2012).
Subgroup analysis and investigation of heterogeneity
Subgroup analysis will be performed on subgroups categorized according to the following factors, where adequate numbers of studies are available.
- Frequency and duration of toothbrushing with home-use whitening toothpastes.
- Types of ingredients in home-use whitening toothpastes (chemical versus abrasive).
- Concentrations of active ingredients.
We will assess the impact of excluding studies at unclear and high risk of bias from the analysis if sufficient numbers of studies are available.
Presentation of main results
We will produce summary of findings tables for the main outcomes of this review using GRADEPro software. We will assess the quality of the body of evidence by considering the overall risk of bias of the included studies, the directness of the evidence, the inconsistency of the results, the precision of the estimates, the risk of publication bias, the magnitude of the effect and whether or not there is evidence of a dose-response. We will categorize the quality of the body of evidence for each of the main outcomes as high, moderate, low or very low.
The authors wish to acknowledge the contributions of the former protocol authors, S Rajesh, VCC Marinho, R Croucher and Robert Norman, in the preparation of the updated protocol. Additional thanks to the staff and leadership of the Cochrane Collaboration Oral Health Group for their guidance in the conduct of this review.
We are grateful to Anne Littlewood, Trials Search Co-ordinator, Cochrane Oral Health Group, for refining the search strategy in MEDLINE via OVID and suggesting other databases and trial registries.
Appendix 1. Search strategy for MEDLINE (OVID)
1. exp Dentifrices/
2. (toothpaste$ or "tooth paste$" or tooth-paste$ or dentifrice$).tw.
3. ((tooth or teeth) adj5 gel$).tw.
5. Tooth bleaching/
6. Hydrogen peroxide/
7. (peroxide$ or carbamide$ or hexametaphosphate$).tw.
8. (bleach$ or white$ or bright$ or lighten$ or colour$ or color$).tw.
9. exp Urea/
11. Sodium bicarbonate/
12. ("sodium bicarbonate" or "baking powder" or "baking soda").tw.
14. 4 and 13
This subject search will be linked to the Cochrane Highly Sensitive Search Strategy (CHSSS) for identifying randomized trials in MEDLINE: sensitivity maximising version (2008 revision) as referenced in Chapter 18.104.22.168 and detailed in box 6.4.c of theCochrane Handbook for Systematic Reviews of Interventions version 5.1.0 (updated March 2011).
1. randomized controlled trial.pt.
2. controlled clinical trial.pt.
5. drug therapy.fs.
10. exp animals/ not humans.sh.
11. 9 not 10
Contributions of authors
Mary Brennan (MB) - Drafting of the protocol, searching for studies, selection of studies, writing to authors of papers for additional information, risk of bias assessment, data extraction, interpreting the analysis, drafting the final review, updating the review.
Donna Hallas (DH) - Drafting of the protocol, searching for studies, selection of studies, risk of bias assessment, data extraction, drafting the final review.
Susan Jacobs (SJ) - Developing a search strategy, searching studies, drafting the final review.
Miriam Robbins (MR) - Drafting of the protocol, selection of studies, interpreting the analysis.
Mary Northridge (MN) - Drafting of the protocol, data extraction, carrying out the analysis, interpreting the analysis, drafting the final review, updating the review.
Declarations of interest
There are no known potential conflicts of interest for the protocol authors.
Sources of support
- New York University College of Dentistry, New York, New York, USA.Salary support for authors.
- New York University College of Nursing, New York, New York, USA.Salary support for authors.
- New York University Division of Libraries, New York, New York, USA.Salary support for author.
- Cochrane Oral Health Group Global Alliance, UK.All reviews in the Cochrane Oral Health Group are supported by Global Alliance member organisations (British Association of Oral Surgeons, UK; British Association for the Study of Community Dentistry, UK; British Orthodontic Society, UK; British Society of Paediatric Dentistry, UK; British Society of Periodontology, UK; Canadian Dental Hygienists Association, Canada; Mayo Clinic, USA; National Center for Dental Hygiene Research & Practice, USA; New York University College of Dentistry, USA; and Royal College of Surgeons of Edinburgh, UK) providing funding for the editorial process (http://ohg.cochrane.org/).
- National Institute for Health Research (NIHR), UK.CRG funding acknowledgement:
The NIHR is the largest single funder of the Cochrane Oral Health Group.Disclaimer:
The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the NIHR, NHS or the Department of Health.