General comment: This is interesting report in an important area of dental public health. My main concern is that the analysis is entirely qualitative and the authors do not seem to have considered extracting any data to look at the size of any effect on caries prevalence/incidence or other outcomes following withdrawal of water fluoridation. Both the two other major reviews in this area (McDonagh 2000 and Iheozor-Ejiofor 2015) have undertaken quantitative analyses, which I would find helpful when interpreting the findings of this review.
Other comments: The authors should consider structuring the methods using PICOS format. The outcomes they were interested in should be clearly outlined. This will obviously include measures of caries levels, such as DMFT and S, but also proportion of caries free in the population, as well as possibly enamel opacities and any other adverse or beneficial effects, although I appreciate that the authors might wish to concentrate on caries. I wonder if they should consider excluding the one study that looked at fluoride levels in plaque, as the authors state ‘The dental caries implications of that change are not known’, implying that this is not a very useful outcome to measure (which I agree with).
The specific databases searched can be found in the appendix; however I believe that it would be useful if a list of the databases (with the years searched) could be provided in the methods section, rather than the general statement that ‘diverse research databases’ were searched.
Did the investigators attempt to contact any authors to help identify any other studies? I would consider this routine when undertaking a thorough review. Did they undertake any hand searching? This would involve a lot of work, but they could target a few key publications.
The details about who undertook the screening of titles and abstracts, full articles and quality assessments are vague. The authors state that ‘The full list of titles and abstracts from the initial search was reviewed twice’. Does this mean by two reviewers, who undertook the assessments independently? Again this would be the usual approach when undertaking a systematic review. If the authors have not done this then they need to provide the justification. If the screening and quality assessments were undertaken by at least two reviewers independently how did they manage disagreements? Did they attempt to assess agreement?
A PRISMA flow chart should be provided.
I like the way that the details are summarised in tables.
The discussion is interesting. The authors overall conclusion is that ‘published research on CWF cessation and dental caries points more to an adverse effect of cessation on dental caries’ but I believe this would be strengthened by indicating the size of effect through a quantitative analysis.