Encke 2009.

Methods	5‐year follow‐up, randomised, single‐blind, parallel‐group study
Participants	536 patients aged above 18 years recruited through newspaper advertisements Inclusion criteria: a vital or successfully, endodontically treated tooth in need of a crown in the posterior region (premolar or molar) no pathological signs on the X‐ray no clinical symptoms of inflammation periodontal treatable abutment tooth: after pre‐treatment PD < 4 mm, mobility < II, furcation involvement < 2 Exclusion criteria: addiction to alcohol or drugs. Psychologically unstable patients. Patients with sufficiently treated teeth patients with acute symptoms of functional disorders with the necessity of functional pre‐ treatment before prosthodontic therapy ASA IV or higher patients patients treated in the study who developed a severe systemic disease or a disease that would influence the treatment outcome, were not included in the statistical analysis. Patients who insisted on using a different crown system after explication 308 allocated to 2 groups (152 zirconia, 156 gold) with at least 1 tooth in the premolar or molar region in need of crowning. 224 patients received allocated intervention
Interventions	Single crowns on mandibular and maxillary premolars and molars Group 1: 123 patients with 123 all ceramic ZrSiO4 crowns (Everest HPC, KaVo Dental GmbH, Biberach/Riss, Germany) Group 2: 101 patients with 101 gold crowns (Degulor M, Degudent) All treatments were carried out in the University Dental Hospital Freiburg, Germany, by staff dentists Tooth preparation: for metal‐free crowns a deep chamfer margin was performed with 1.2 mm chamfer diamond burs, an occlusal reduction of 1.5 mm and rounded inner edges for gold crowns a chamfer type of margin was performed with 0.8 mm chamfer diamond burs, an occlusal reduction of 1.2 mm and rounded edges in both groups preparation margins were located at the gingival level or not more than 1 mm subgingival Provisionalization: autopolymerizing acrylic provisional crowns cemented with a eugenol‐free temporary cement Impressions and lab work: custom‐made trays and a silicone impression material master models were manufactured out of type IV dental stone models were mounted in a semi‐adjustable articulator using a face bow transfer and check bites for the metal‐free group, dies and wax‐ups of the crowns were scanned (KaVo Everest scan machine), then milled (KaVo Everest HPC blanks), subsequently sintered. Full crowns were fabricated without veneering for the gold group crowns were conventionally waxed up and casted using a gold alloy (Degulor M, Degudent) and fitted on the master model Try‐in and cementation: in both groups at the try‐in, proximal contacts and static and dynamic occlusions were checked and corrected with fine diamond burs with water cooling and silicone polishers a radiograph of the abutment and the crown was taken prior to cementation to evaluate the marginal fit and periapical region. If a periapical lesion was detected, root canal treatment was performed prior to insertion all crowns were cemented with a glass ionomer cement (Ketac Cem Maxicap)
Outcomes	Loss of vitality of abutment Surface roughness Fractures Marginal integrity Secondary caries at the crown margin Margin discolourations Marginal gap Crown loss
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Eligible patients were assigned a unique identification number, which was forwarded to a statistician (Institute of Medical Biometry and Medical Informatics, University of Freiburg). Randomization was carried out using the Bernoulli distribution for each patient identification number. Accordingly, the patients received either the ceramic (test) or gold crown (control). No balancing was carried out" Comment: of the 308 patients allocated to 2 groups (152 zirconia, 156 gold) only 224 patients received allocated intervention, resulting in 123 zirconia versus 101 gold crowns
Allocation concealment (selection bias)	Unclear risk	No information provided
Blinding of participants and personnel (performance bias) All outcomes	High risk	Due to the obvious differences between the 2 crown designs (gold versus ceramic), blinding was not possible
Blinding of outcome assessment (detection bias) All outcomes	High risk	Due to the obvious differences between the 2 crown designs (gold versus ceramic), blinding was not possible
Incomplete outcome data (attrition bias) All outcomes	High risk	In the ceramic crown group, 15 patients were not followed up at the 12‐month examination, 24 patients at the 24‐month follow‐up, 50 patients at the 36‐month recall and 46 patients at the 48‐ and 60‐month follow‐ups In the gold group, 7 patients were lost to follow‐up at the 12‐month recall, 16 patients at the 24‐month follow‐up, 27 patients at the 36‐month recall examination, 18 patients after 48 months and 19 patients at the 60‐month follow‐up 2 patients allocated to gold group were not included in the analysis, but reasons were provided
Selective reporting (reporting bias)	Low risk	None identified
Other bias	Low risk