Quantity of Research Available
A total of 180 citations were identified in the literature search. Following screening of titles and abstracts, 154 citations were excluded and 26 potentially relevant reports from the electronic search were retrieved for full-text review. A grey literature search did not find any potentially relevant publications. Of the potentially relevant articles, 22 publications were excluded for various reasons, while four publications met the inclusion criteria and were included in this report. Appendix 1 describes the PRISMA flowchart of the study selection. The report of a previous CADTH review conducted in May 2015 on PFM crowns versus all-ceramic crowns7 is referenced in Appendix 5. Unlike the current review, the previous review included implant-supported restorations, bridges, dentures retained by special crowns, and composite resin restorations.
Summary of Study Characteristics
A summary of the characteristics of the included studies is presented in Appendix 2.
Study Design
One systematic review with meta-analysis,4 two prospective non-randomized studies8,9 and a retrospective non-randomized study10 were included in this review (Table A1). One study8 was published in 2016 while the other studies4,9,11 were published in 2015. The systematic review4 included 67 primary studies published between 1991 and 2013, five of which were randomized controlled studies (RCTs) while the remaining were prospective (n=37) and retrospective (n=25) non-randomized studies.
Country of Origin
The systematic review (Sailer et al.)4 was authored by investigators from Switzerland. One prospective study (Rinke et al.)8 was conducted in Germany, while the retrospective study (Valenti et al.)10 and other prospective study (Tartaglia et al.)9 were conducted in Italy.
Patient Population
A total of 6,095 patients ranging in age from 15 to 91 years were included in the primary studies of the systematic review.4 Fifty-nine patients were included in the study by Valenti et al.10 while Rinke et al.8 and Tartaglia et al.9 included 45 patients and 88 patients in their respective studies. The mean age of patients in the studies by Rinke et al.8 and Tartaglia et al.9 were 49.6 years and 57.0 years respectively. Although the patients included in the study by Valenti et al.10 were adults (described only as men and women) their ages were not specified.
Interventions and Comparators
The primary studies of the systematic review4 evaluated a total of 14,097 tooth-supported single-crowns (4,663 metal-ceramic and 9,434 all-ceramic crowns). Rinke et al.8 compared 50 metal-ceramic crowns with 55 zirconia crowns for up to six years. The study by Tartaglia et al.9 involved a non-comparative evaluation of 303 zirconia crowns (150 single crowns and 153 multiple units up to 6 elements) for a follow-up time of up to seven years. One hundred-and-seventy-nine (59%) of the restorations in this study9 were tooth-supported while the remaining were implant supported. A total of 110 lithium disilicate crowns with feather-edge finished line margins were evaluated by Valenti et al.10 over a period of nine years. The specifics of support for the crowns were not described.
Outcomes
The systematic review4 assessed the 5-year survival and complication rates of crowns using Poisson’s regression models. Survival was defined as the FDP remaining in situ with or without modification for the observation period.4 Failure and complication rates were calculated by dividing the number of events (failures or complications) by the total FDP exposure time.4 Complications were subdivided into technical and biological complications, with framework fracture, ceramic fracture, ceramic chipping, marginal discoloration, loss of retention and poor esthetics described as technical complication. Biologic complications referred to loss of abutment tooth vitality, abutment tooth fracture and secondary caries.
In the study by Rinke et al.,8 Kaplan–Meier plots were used to compute the 5-year estimated cumulative outcomes for survival and success of the restorations, and for veneering ceramic success. Survival was defined as the reconstruction remaining in situ at the follow-up examination visit without presenting an absolute failure that required the replacement of the entire restoration or extraction of the tooth.8 Success was defined as a reconstruction that remained unchanged and did not require any intervention to maintain function during the entire observational period.8 The success of the veneering porcelain was defined as event-free survival without any veneering ceramic failure (VCF).8 Assessed parameters were de-cementation, loss of vitality of the abutment teeth, secondary caries, tooth fracture, fracture of the framework and chipping of the ceramic veneer.8
The study by Valenti et al.,10 assessed the overall survival probability of crowns for up to nine years using the Kaplan-Meier method, with survival time defined as the period of time starting at baseline and ending when the clinician estimated that an irreparable failure of the crown had occurred. Irreparable failure referred to exposure of the tooth structure caused mainly by infiltration due to abutment decay, core fracture, or partial debonding.
The study by Tartaglia et al.,9 reported the 7-year Kaplan-Meier survival probability estimates for failures and complications, as well as the cumulative survival rate of crowns. Failure referred to situations where the prosthesis needed replacement or removal, while complications could be resolved without replacing the prosthesis. Survival rate was defined as surviving FDPs minus altered FDPs based on grades 2 and 3 of a 3-grade scale of chipping fractures, where surface chipping is graded 1 if the fractured surface is not extended into a functional area and polishing is possible. Further details of this scale were not provided.
Summary of Critical Appraisal
A summary of the critical appraisal of the included studies is presented in Appendix 3.
All included studies4,8–10 clearly defined study objectives as well as the main outcomes measured. The interventions of interest and the main findings were also clearly described by all studies.4,8–10 Two reviewers independently selected relevant studies for inclusion in the systematic review4 based on well-defined criteria, following a comprehensive literature search of articles published from December 2006 up to and including December 2013. Two researchers independently performed data extraction, with disagreements resolved by consensus of three reviewers. The statistical analyses accounted for the risk of failure by restoration types in each study during follow-up, and made long-term projections for each of the different types of crowns. However, the analyses of failure of restorations were based on assumption of a constant risk over time, which may not reflect what happens in practice. Most of the primary studies (62/67) in the systematic review4 were non-randomized studies. Thus, the likelihood of uneven distribution of confounding factors between interventions that were compared create a greater potential for biases compared with randomized studies. In addition, without a comparator, it is difficult to accurately assess the incremental effect of an intervention under study. Furthermore, the methodological rigor and scientific quality of the included studies were not assessed.
While the studies by Rinke et al.8 and Tartaglia et al.9 provided inclusion and exclusion criteria and some demographic data, the study by Valenti et al.10 did not adequately describe inclusion and exclusion criteria and patients demographics were poorly reported. All the restorations in the study by Rinke et al.8 were placed by a single operator, thus reducing the potential for operator-related confounding risk. In the studies by Rinke et al.8 and Valenti et al.,10 a skilled dentist with appropriate training to assess outcomes, who was not involved with the restorative service and patient care, independently conducted follow-up clinical examinations to minimize evaluation bias. Except in one study9 where one out of four authors had received lecture fees from a company which develops and produces materials, instruments and devices for dental laboratories, the authors of all the included studies4,8–10 declared no conflict of interest.
None of the non-randomized studies8–10 performed a calculation for sample sizes to provide sufficient power to determine relevant differences between the compared interventions. Thus, it is unknown whether the patient-based sample size and the associated number of restorations used in the various studies were enough to rule out type II errors. In the study by Rinke et al.,8 eight patients (15.1%) were lost to follow-up, resulting in a 21.9% decrease in the number of metal-ceramic crowns and a 10.0% decrease in the number of zirconia crowns. The attrition rate and the resulting disproportionate reductions in types of crowns created an imbalance between the compared interventions and is a source of potential bias. There was a potential for a positive over-estimation of the cumulative survival and success outcomes as it is unclear whether patients who remained in the study were those who were satisfied with their treatment, while patients who dropped out were those who were not satisfied with their treatment.
In the study by Valenti et al.,10 the posterior crowns were manufactured as monolithic while the anterior crowns were veneered. The different production techniques conferred different abilities to the crowns to withstand masticatory pressure. Thus it is difficult to accurately assess the survival ability of the restorations at the various sites based on the material alone.
The study by Tartaglia et al.9 was an extension of an earlier published single center prospective study. However, analyses were based on only 88 of the original 138 patients without accounting for missing data of the 50 patients (36 %) with 142 crowns (32 %) from the original study who were lost to follow-up. Thus the potential for an error in the estimation of outcomes due to attrition cannot be ruled out.
The countries of origin of the primary studies of the systematic review4 were not declared, and all non-randomized studies in this review were conducted at single private centers in Germany8 or Italy.9,10 Therefore, it is unknown whether the reported findings are generalizable to the Canadian context.
Summary of Findings
A summary of the findings of the included studies is presented in Appendix 4.
1. What is the clinical evidence on the longevity of porcelain-fused-to-metal crowns?
From a meta-analysis of 17 primary studies, one systematic review4 reported that the estimated 5-year survival rate of metal-ceramic single crowns was 94.7% (95% confidence intervals [CI]: 94.1, 96.9). One prospective study8 reported that the 5-year estimated cumulative survival (ECSv) of the metallic-ceramic crown was 97.6%, (95% CI: 93, 100). The study8 also found that the 5-year estimated cumulative success (ECSc) of metal-ceramic crowns was 85.0% (95% CI: 77, 96), while the estimated cumulative veneering success (ECVCSc) was 95.0% (95% CI: 88; 100). The ECSs is a measure of event-free restorations, while the ECVCSc indicate the extent that ceramic veneers in meta-ceramic crowns remained intact without fractures.
2. What is the clinical evidence on the longevity of all-ceramic/porcelain crowns?
One systematic review4 reported that based on an analysis of 55 primary studies, the overall estimated 5-year survival rates were between 90.7% and 96.6%, depending on the construction material. Feldspathic or silica crowns had the lowest survival rate of 90.7% (95% CI: 87.5, 93.1) while crowns made of leucite/lithium-disilicate reinforced glass had the highest survival rate of 96.6% (95% CI: 94.9, 97.7) followed by densely sintered alumina crowns (96.0%; 95% CI: 93.8, 97.5); glass-infiltrated alumina crowns (94.6%; 95% CI: 92.7, 96) and densely sintered zirconia (91.2%; 95% CI: 82.8, 95.6).
One prospective study8 reported that the 5-year ECSv of all-ceramic (zirconia) crowns was 94.0%, (95% CI: 87, 100), with an event-free restoration rate (ECSc) of 74.3% (95% CI: 61, 87). During the observation period, 86.1% (95% CI: 75, 96) of ceramic veneers with zirconia crowns were intact without fractures.
One retrospective study10 reported that, the overall survival probability of lithium disilicate crowns for up to nine years was 96.1%, with the anterior placed crowns having a higher survival probability (100%) compared to posterior crowns (94.5%).
One prospective study9 reported that the 7-year cumulative survival rate in all-ceramic crowns was 94.7 %. All failures were observed in tooth-supported single crowns, with a significant difference (P < 0.001) between the kind of support (tooth versus implant), and number of crowns per restoration (single unit versus multiple unit). However, the position of the crown in the mouth (anterior versus posterior) did not influence the failure rates (P = 0.316).
3. What is the clinical evidence on the longevity of porcelain-fused-to-metal crowns compared with all-ceramic crowns?
One systematic review4 reported that the estimated 5-year survival rates of all-ceramic crowns made of leucite/lithium-disilicate reinforced glass, densely sintered alumina, glass-infiltrated alumina, or densely sintered zirconia were not significantly different (P > 0.05) from the 5-year survival rate of porcelain-fused-to-metal crowns. However, the survival rate of feldspathic or silica ceramic crowns was significantly lower (P < 0.05) compared with meta-ceramic crowns.
One prospective study8 found no significant differences between zirconia all-ceramic crowns and metal-ceramic crowns with respect to ECSv (P = 0.51), ECSc (P = 0.43) and ECVCSc (P = 0.36). Two (4.9%) metal-ceramic crowns and three (6.0%) zirconia crowns were replaced, and no fracture of a zirconia or a metallic framework was detected over the entire observational period.
4. What is the long-term (eight years and longer) cost-effectiveness of porcelain-fused-to-metal crowns compared with all-ceramic crowns?
The literature search for this review did not find any studies that evaluated the cost-effectiveness of porcelain-fused-to-metal crowns compared with all-ceramic crowns.
5. What are the contextual considerations for all-ceramic crowns or porcelain-fused-to-metal crowns that may affect their clinical or cost-effectiveness?
One systematic review4 stated that mechanically weaker ceramics like the feldspathic or silica glass-ceramics can only be recommended in anterior regions with low functional load. In addition, the study4 reported that densely sintered zirconia crowns exhibited significantly lower (P < 0.05) survival rates in the posterior region than the anterior region, with increased risk of chipping of the veneering ceramic and loss of retention. The authors stated that crowns made out of this material were therefore not recommended as a primary treatment option in the posterior location until further refinements to the production technology are made to overcome these limitations.4 However, one prospective comparative study8 did not find a statistically significant difference (P = 051) in the 5-year ECSv of metal-ceramic and zirconia-based molar crowns. In addition, one retrospective study9 found that the position of the crown in the mouth did not influence failure rates of zirconia crowns (P = 0.316).
Limitations
It is unknown whether the reported findings of the included studies are generalizable to the Canadian context since all the non-randomized studies8–10 were conducted at single private centers in countries other than Canada, and the countries of origin of the primary studies of the systematic review4 were not declared.
The main limitation of included systematic review4 is that its primary studies are predominantly non-randomized studies. Moreover, without an assessment of the methodological rigor and scientific quality of the included studies, the strength of evidence supporting its conclusions is indeterminate.
Besides the potential for bias inherent in their design, none of the non-randomized studies8–10 included this review performed sample size calculations to ensure adequate power and avoid a type II error. Furthermore, one study8 had a 15.1% drop-out rate resulting in disproportionate decreases from baseline in the numbers of the interventions it compared, and another study9 reported outcomes based on analyses that did not account for 36 % of patients and 32 % of crowns originally involved in the study. Moreover, one study10 constructed posterior crowns as monolithic while the anterior crowns were veneered, making it difficult to accurately assess the survival ability of the restorations at the various sites based on the material alone.
