Data relating to the likelihood of moving between health states (transition probabilities) were obtained by appropriate literature searches. Where probability data were unavailable, freedom of information requests were used, and where these were also not possible/available, appropriate estimates were determined in consultation with relevant experts and stakeholders to garner consensus.

Initial tooth states 

For the starting point of individuals in the model, data were obtained from Office for Health Improvement and Disparities Dental Epidemiology Programme results (including 3-year-olds (2020) and 5-year-olds (2022)) plus the decennial Child Dental Health Survey (2013) for ages 8-15, and Adult Dental Health Survey (2009) for ages 16+.

Caries progression

Likelihoods of progression from no decay to decayed (requiring restoration) were based on data from Ismail et al. (2015) but given this was a high risk child only cohort, were then adjusted based on work reported by Chaffee et al. (2015).

Treatment progression and reintervention

Probabilities of placing and replacing fillings (restorations), undertaking root canal treatments and extractions were calculated from previous analysis of dental claims data from pre-2006 (Burke and Lucarotti (2009) and Lucarotti and Burke (2018)). Unfortunately no more up to date data of sufficient depth and quality was available.   

Effectiveness of CWF

Data from the Cochrane systematic review of water fluoridation (Iheozor-Ejiofor, 2015) was selected for use given its relative recency and robust inclusion criteria. A sensitivity analysis was undertaken using more recent data from the CATFISH study (Goodwin et al. 2022) and either data source can be selected in the tool.