Study	Reason for exclusion
Andersson 1999	Study design is neither parallel‐group nor split‐mouth. 69 prosthetic abutments have been randomised in 60 patients to 2 interventions. Some patients had 1 intervention, some others had both interventions
Andersson 2001	Study design is neither parallel‐group nor split‐mouth. 103 prosthetic abutments have been randomised in 32 patients to 2 interventions. Some patients had 1 intervention, some others had both interventions
Batson 2014	Follow‐up shorter than 1 year
Bindl 2005	Not randomised
Borg 2014	Study design is neither parallel‐group nor split‐mouth. 18 implant‐supported FDPs have been randomised in 16 patients to 2 interventions. Some patients had 1 intervention, some others had both interventions
Cehreli 2009	Study design is neither parallel‐group nor split‐mouth. 30 prosthetic crowns have been randomised in 20 patients to 2 interventions. Some patients had 1 intervention, some others had both interventions
Cehreli 2011	It is unclear whether randomisation was done according to patients or restorations. 33 consecutive patients included in the study were randomised to 2 interventions, feldspathic porcelain crowns (12 patients, 50 crowns) and glass‐infiltrated alumina all‐ceramic crowns (21 patients, 51 crowns). Authors were contacted via mail but did not answer
Chen 2008	Study design is neither parallel‐group nor split‐mouth. 35 implants in 23 patients have been randomised to 2 interventions. Some patients had 1 intervention, some others had both interventions
Christensen 2010	Unclear randomisation. Study is described as randomised with 10 different treatment modalities but 2 "were not available initially." 2 other groups were limited in size compared to the others because of investigators "concern about framework fractures shortly after placement"
Esquivel‐Upshaw 2012	Study design is neither parallel‐group nor split‐mouth. 36 posterior crowns in 31 patients have been randomised to 3 interventions. Some patients had 1 intervention, some others had more than 1 intervention
Esquivel‐Upshaw 2014	Study design is neither parallel‐group nor split‐mouth. 72 FDPs in 55 patients have been randomised to 2 interventions. Some patients had 1 intervention, some others had more than 1 intervention
Esquivel‐Upshaw 2014b	Study design is neither parallel‐group nor split‐mouth. 89 FDPs in 68 patients have been randomised to 2 interventions. Some patients had 1 intervention, some others had more than 1 intervention
Etman 2008	Study design is neither parallel‐group nor split‐mouth. 90 posterior crowns in 48 patients have been randomised to 3 interventions. Some patients had 1 intervention, some others had more than 1 intervention
Henriksson 2004	Not randomised
Jung 2008	Follow‐up shorter than 1 year
Li 2007	Not randomised
Ohlmann 2006	Study design is neither parallel‐group nor split‐mouth. 120 prosthetic crowns have been randomised in 66 patients to 3 interventions. Some patients had 1 intervention, some others had more than 1 intervention
Pelaez 2012	Study design is neither parallel‐group nor split‐mouth. 40 3 units FPDs have been randomised in 37 patients to 2 interventions. Some patients had 1 intervention, some others had both interventions
Sagirkaya 2012	It is unclear how randomisation was done. 59 subjects with excessive loss of tooth structure requiring full veneer crowns, crowns/FPDs needing replacement, and missing tooth or teeth requiring tooth‐ or implant‐supported crowns and/or 3 to 6 units FPDs were randomised to 4 interventions. According to materials and methods section patients were randomised, but the number of patient randomised to each group is not reported neither in materials and methods nor in results sections. Results report data according to units, and the 4 groups have large differences in size of reported units (Cercon, 24 units; ZirkonZahn, 118 units; Lava, 40 units; Katana, 85 units). Authors were contacted via mail but did not answer
Sailer 2009	Study design is neither parallel‐group nor split‐mouth. 40 implant sites have been randomised in 22 patients to 2 interventions. Some patients had 1 intervention, some others had both interventions
Sailer 2009b	Study design is neither parallel‐group nor split‐mouth. 76 FPDs have been randomised in 59 patients to 2 interventions. Some patients had 1 intervention, some others had both interventions
Vanoorbeek 2010	Study design is neither parallel‐group nor split‐mouth. 200 prosthetic crowns have been randomised in 130 patients to 2 interventions. Some patients had 1 intervention, some others had both interventions. Moreover randomisation was terminated after the first 120 restorations due to early complications in 1 group

FDPs = fixed dental prostheses.