Glass hybrid (glass ionomer) versus composite for non-carious cervical lesions: Survival, restoration quality and costs in randomized controlled trial after 3 years.

OBJECTIVE: This study compared survival, restoration quality and costs of glass hybrid (GH; EQUIA Forte Fil/EQUIA Forte Coat) and resin composite restorations (RC; OptiBond FL/Filtek Supreme XTE) of sclerotic non-carious cervical lesions. METHODS: This is a cluster-randomized trial (ClinicalTrials.gov: NCT02631161). 88 patients (50-70 years) with 175 sNCCLs were randomized to receive GH or RC. Restorations were placed without mechanical cavity preparation and followed for a mean 36 (min/max: 31/55) months (variable follow-up due to COVID-19 lockdown). Restoration quality was re-evaluated at 1-, 18- and 36-months using FDI-criteria. Survival was assessed using multi-level Cox-regression analysis. Costs were estimated from a payer's perspective in Germany. Initial costs were determined based on micro-costing using time recordings and hourly costs, and follow-up costs based on statutory insurance fee-item-catalogues. RESULTS: 88 patients (175 restorations) were treated; 43 received GH (83 restorations), 45 RC (92 restorations). 17  GH and 19 RC showed total retention loss, 5  GH were partially lost (p = 0.396/Cox). FDI ratings were not sufficiently different for any domain except surface luster, where RC showed higher score (p < 0.001). Costs were initially lower for GH (32.57; SD 16.36 €) than RC (44.25; SD 21.40 €), while re-treatment costs were similar (GH: 9.15; SD 15.70 €; RC: 7.35; SD 14.51 €), resulting in significantly lower costs for GH (GH: 41.72; SD 25.08 €) than RC (51.60; 26.17 €) (p < 0.001/GLM). CONCLUSIONS: While survival was not significantly different, GH was significantly less costly both initially and long-term than RC for restoring non-carious cervical lesions. CLINICAL SIGNIFICANCE: Within this trial, survival was not significantly different between GH and RC to restore sclerotic NCCLs. As GH was significantly less costly both initially and long-term than RC, using RC was only cost-effective for payers willing to invest high additional expenses per minimal survival gains.