Are alcohol containing IN BRIEF • Dentists need to be aware that there is a RESEARCH hypothetical risk for the development of oral cancer from repeated use of alcohol mouthwashes safe? containing mouthwashes. • The existing evidence to support or refute C .W. de A. Werner1 and R. A. Seymour2 such an association has been the subject of two recent confl icting reviews. • This paper critically evaluates the two reviews and further explores the data on the effi cacy of the addition of alcohol to mouthwashes. Alcohol (ethanol) is a constituent of many proprietary mouthwashes. Some studies have shown that regular use of such mouthwashes can increase the risk of developing oral cancer. Recently, the evidence has been reviewed by two separate authors. The conclusions from these reviews are confl icting. In this paper, we reconsider the epidemiological evidence linking alcohol containing mouthwashes with an increased risk of oral cancer. The evidence is considered in term of sample size, strength of association, confounding variables and data collection. In addition, clinical studies comparing alcohol versus non-alcohol mouthwashes are evaluated. The evidence suggests that the alcohol component of mouthwashes affords little additional benefi t to the other active ingredients in terms of plaque and gingivitis control. In view of this outcome and the hypothetical risk of oral cancer, it would seem prudent that members of the dental team advise their patients accordingly. INTRODUCTION reviews and the interpretation of the evi- problems: question formulation, verifi ca- Many proprietary mouthwashes contain dence and hopefully provide readers with tion of latest evidence, critical appraisal and alcohol (ethanol) and in some, the concen- an opinion on the safety or otherwise of value assessment of evidence and clinical tration of ethanol can be as high as 26%.1 alcohol containing mouthwashes. indication. Epidemiological studies have Two recent reviews2,3 have considered different weights on the evidence they pro- whether alcohol containing mouthwashes EPIDEMIOLOGICAL EVIDENCE vide. Additional information from in vivo increase the risk of the development of Epidemiological studies are complex and studies, especially those that investigate oral cancer. The conclusions from these aspects such as sample size, strength of the possible mechanism of alcohol toxic- reviews are confl icting, with one stat- association, confounding variables and ity to the oral mucosa, may complement ing ‘there is now sufficient evidence data collection bias can be rightfully chal- the epidemiological evidence and provide to accept the proposition that alcohol lenged. By contrast, arguments can also be possible benefi ts and risks for clinical use containing mouthwashes contribute to made for the exploitation of specifi c aspects of alcohol containing mouthwashes. All the increase of the development of oral of interest (reporting bias). This is obviously of this information is then evaluated to cancer’,2 while the other states that ‘criti- a controversial topic which can lead to dif- ascertain whether the benefi ts of alcohol cal review of the published data revealed ferent courses of clinical action. Dentistry containing mouthwashes outweigh the risk that a link between mouthwash use, spe- often deals with such issues and dentists of damage to the oral mucosa. cifi cally alcohol containing mouthwash have to make ethical decisions on clinical and oral cancers, is not supported by protocols and material based on less than QUESTION FORMULATION epidemiological evidence.’3 ideal types of scientifi c evidence. Do the clinical benefi ts of an alcohol The reviews were published in 2008 and Our aim is to guide the reader through containing mouthwash in the prevention 2009 and for the most part have quoted the a clear pathway which provides relevant of plaque related oral diseases outweigh same evidence to support their conclusion. information for refl ection and leads to the possible safety issues? The signifi cant In this paper we shall look critically at both an informed decision. Scientifi c evidence safety issue is developing oral squamous will be explored in light of the most recent cell carcinoma. This raises the further 1*Clinical Lecturer in Dental Public Health, 2Professor reviews so that we can empower all mem- question, is alcohol containing mouthwash of Restorative Dentistry, School of Dental Sciences, bers of the dental team with the best clini- a justifi able clinical treatment course to Newcastle University, Framlington Place, Newcastle upon Tyne, NE2 4BW cal advice on the use of mouthwashes. encourage patients to take? *Correspondence to: Dr Carlos Werner Email: [email protected] Centre for Evidence Based WEIGHTING EPIDEMIOLOGICAL Online article number E19 Dentistry (CEBD) EVIDENCE Refereed Paper - accepted 29 October 2009 DOI: 10.1038/sj.bdj.2009.1014 The CEBD suggests the following step- In an attempt to answer the questions set ©British Dental Journal 2009; 207: E19 wise approach for dealing with clinical out above via an evidence-based approach, BRITISH DENTAL JOURNAL 1 © 2009 Macmillan Publishers Limited. All rights reserved. RESEARCH the epidemiological evidence needs to mouthwashes in the development of oral Table 1 Criteria for quantifying be quantifi ed. The criteria for quantify- cancer. This suggestion is based upon epidemiological evidence ing epidemiological evidence are shown case control studies which examined the Type I A least one good systematic review in Table 1. smoking-alcohol association in general, (including at least one randomised Neither of the review papers can be clas- with no stratifi cation of alcohol contain- clinical trial) sifi ed as systematic reviews as they have ing mouthwash use.4,5 The fi ndings shown Type II At least one good randomised not included a randomised clinical trial in a nation-based alcohol profi le and oral clinical trial (RCT) and do not follow strict reviewing cancer mortality correlation study failed Type III Well designed interventional studies without randomisation protocols. Although this is a criticism, it to indicate that the weight of this associa- should be identifi ed that a RCT to show tion is not as strong as the one from case Type IV Well designed observational studies (case control and cohort studies) an association between oral cancer and control studies.6 Type V Descriptive studies (correlational alcohol containing mouthwash use would The main conclusion of this review studies, cross sectional surveys), be diffi cult to carry out. Oral cancer is a is based upon the evidence provided by case reports, case series, infl uential chronic disease, can take many years to Guha et al.,7 which uses data obtained reports and expert opinion develop and an RCT will require a large from two multi-centred case control stud- follow-up population. ies. Nevertheless, a particularly relevant published in 2001 and a re-analysis of the The possible risk association between fi nding from this study7 curiously was not 1991 data by authors not involved in the alcohol containing mouthwashes and oral mentioned, weakening the review further. original study, followed in 2003.10 cancer is considered from fi ve directly The information related to self reported The 1991 study showed a statistically relevant case control studies (type IV use of more than twice a day mouthwash signifi cant increase in risk of oral cancer evidence). In such studies, subjects were indicated that it increased the chance of associated with regular mouthwash use selected on the basis of whether they do developing oral squamous cell carcinoma and suggested that the risk varied in pro- (cases) or do not have a particular disease by almost six fold (OR 5.86; 95% CI = 2.91, portion to dose, duration and frequency of (controls). Information is then determined 11.77) when compared to those patients mouthwash usage and alcohol concentra- based on the calculated odds ratios (OR). It who reported never having used mouth- tion. The authors concluded that regular is important to differentiate here the fact wash products. use of a mouthwash with high alcohol that risk in another type of study (cohort Lack of more in-depth explanation on content contributes to oral cancer risk. study) is derived from risk ratios (RR), existing evidence of association (or lack The re-analysis of the same data10 divided which are determined by a different math- of association) gives the impression that the cases of oropharyngeal cancers into ematical formulation. The terms OR and the most conclusive epidemiological evi- mucosal (true cases) and non-mucosal RR have been used interchangeably in the dence should be derived solely from the (pseudo cases). The re-analysis concluded literature evaluated and our assumption is most recent published study. The labora- that regular use of a mouthwash was asso- that the calculation appropriate for case tory studies quoted in this review, at best, ciated more strongly with the pseudo cases control studies was carried out, regardless explore the link between alcohol and oral than the true cases in women and there of how it has been reported. cancer as a facilitating factor rather than was a weak dose response relationship Some studies also refer to ‘excess risk’, a causal or risk factor. A more balanced between mouthwash use and oropharyn- which can introduce ambiguity in the ter- view on existing association between geal cancer in pseudo cases. In men, the minology. Excess risk should be defi ned alcohol containing mouthwashes from association between mouthwash use and as the difference between the proportion case control studies would have increased true disease was confi rmed, but considered of subjects in a population with oral can- the credibility of the conclusion from this weak evidence. The re-analysis also com- cer who were exposed to alcohol contain- paper. The review lacked depth of analysis mented that such weak associations might ing mouthwashes and the proportion of on key issues, which brings into question refl ect the use of mouthwashes by ‘smok- subjects with oral cancer who were not the validity of its conclusions.