The Up to 21-Year Clinical Outcome and Survival of

Feldspathic Porcelain Veneers: Accounting for Clustering

Danielle M. Layton, BDSc, MSc, MDSca/Terry R. Walton, BDS, MDSc, MS, FRACDSb

Purpose: This study aimed to investigate the clinical outcome and estimated cumulative
survival rate of feldspathic porcelain veneers in situ for up to 21 years while also
accounting for clustered outcomes. Materials and Methods: Porcelain veneers
(n = 499) placed in patients (n = 155) by a single prosthodontist between 1990 and
2010 were sequentially included, with 239 veneers (88 patients) placed before 2001 and
260 veneers (67 patients) placed thereafter. Nonvital teeth, molar teeth, or teeth with an
unfavorable periodontal prognosis were excluded. Preparations had chamfer margins,
incisal reduction, palatal overlap, and at least 80% enamel. Feldspathic veneers from
refractory dies were etched (hydrofluoric acid), silanated, and bonded. Many patients
received more than 1 veneer (mean: 5.8 4.3). Clustered outcomes were accounted for
by randomly selecting (random table) 1 veneer per patient for analysis. Clinical outcome
(success, survival, unknown, dead, repair, failure) and Kaplan-Meier estimated cumulative
survival were reported. Differences in survival were analyzed using the log-rank test.
Results: For the random sample of veneers (n = 155), the estimated cumulative survival
rates were 96% 2% (10 years) and 96% 2% (20 years). For the entire sample, the
survival rates were 96% 1% (10 years) and 91% 2% (20 years). Survival did not
statistically differ between these groups (P = .65). Seventeen veneers in 8 patients failed,
75 veneers in 30 patients were classified as unknown, and 407 veneers in 130 patients
survived. Multiple veneers in the same mouth experienced the same outcome, clustering
the results. Conclusions: Multiple dental prostheses in the same mouth are exposed
to the same local and systemic factors, resulting in clustered outcomes. Clustered
outcomes should be accounted for during analysis. When bonded to prepared enamel
substrate, feldspathic porcelain veneers have excellent long-term survival with a low
failure rate. The 21-year estimated cumulative survival for feldspathic porcelain veneers
bonded to prepared enamel was 96% 2%. Int J Prosthodont 2012;25:604612.

A porcelain veneer is a thin bonded ceramic resto-

ration used to restore the facial surface and part
of the proximal surfaces of teeth.1 Veneers allow for
veneers retained by the acid-etch technique. That
same year, Horn6 published the first report of the clin-
ical application of this method.
the conservative management of tooth misalignment Feldspathic porcelain is a predominantly glassy ce-
(instant orthodontics), unesthetic shape and form, ramic based on a naturally occurring glass: feldspar.
and discoloration. Feldspar primarily contains silica (silicon oxide) and
Indirect laminate veneers were first described by alumina (aluminum oxide) but also boric oxide, potash
Charles Pincus in 19372 as a temporary method to (potassium oxide), and soda (sodium oxide).
improve the tooth shapes of people in the film indus- Feldspathic porcelain has advantages and disad-
try. The introduction of acid etching by Buonocore3 vantages. The platinum foil and refractory die fabrica-
and composite resins by Bowen4 in the 1950s led to tion methods are technique sensitive, and the resulting
increased mainstream acceptance of the technique. porcelain is eggshell thin and requires careful handling
In 1983, Simonsen and Calamia5 published a labora- prior to bonding. The material is translucent, creating
tory study describing feldspathic porcelain laminate an extremely life-like restoration, but can be poor at
masking darkened tooth structure. The extremely thin
aPrivate
nature of the material allows for minimal tooth prepa-
Practice, Brisbane, Australia.
bPrivate ration, thus conserving enamel. Feldspathic porcelain
Practice, Sydney, Australia; Adjunct Professor, Faculty of
Dentistry, University of Sydney, Sydney, Australia. is also etchable, which is an essential prerequisite for
effective bonding to etched enamel.
Correspondence to: Dr Danielle M. Layton, 217 Wickham Terrace,
Brisbane, Queensland, Australia, 4000. Email: laytonpros@dlpros.
Many materials have been used for the fabrication
com.au of porcelain laminate veneers over the last 30 years.

604 The International Journal of Prosthodontics

2012 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
 Layton/Walton

A literature review that included studies in which prognosis were excluded. Teeth with large retained
feldspathic veneers were in situ for 5 years identified restorations, tooth loss of more than one-third the
six studies714 reporting Kaplan-Meier cumulative width of the incisal edge, or less than 80% of enamel
survival and two studies15,16 reporting straight per- remaining following preparation were not veneered.
centage outcomes. This literature review was com- Patients who showed extensive loss of tooth structure
pleted as part of a larger systematic review.17 through parafunction were excluded.
These feldspathic veneer studies included be-
tween 5014 and 1,1777 patients and between 8711,12 Clinical Procedure
and 3,25516 veneers. Patients received between 1 and
20 veneers each. Ten-year Kaplan-Meier failure rates Details regarding the clinical procedure were pre-
ranged from 5%9 to 47%,7 while 5-year Kaplan-Meier viously published10 and are outlined in Fig 1. Small
failure rates ranged from 2%9 to 42%.13 Clearly, these defective interproximal restorations were replaced,
outcomes are not in agreement, and the apparently and lesions were restored with composite resin.
contradictory outcomes likely relate to differences in Retraction cord was placed on the labial aspects of
methodology (Tables 1a and 1b). the teeth. The teeth were prepared with a labial re-
These studies differed in terms of setting, operator, duction of 0.5 to 0.7 mm, chamfer margins, and an
direction of inquiry (prospective versus retrospec- incisal reduction of 1 to 2 mm with a palatal over-
tive), and inclusion criteria. The studies also included lap. The palatal overlap was kept clear of the tooth
multiple veneers within the same patients. Veneers contact in maximum intercuspation; if this was not
are used to correct minor esthetic concerns; there- possible, the palatal overlap continued for at least
fore, restoration of multiple teeth simultaneously is 1 mm past the occlusal contact. Interproximal con-
common. Analyzing results per veneer rather than per tacts were reduced on the facial aspects only.
patient results in clustering. Minimum veneer thickness was determined on an
Unfortunately, in dental research, clustered out- individual basis.
comes are commonly ignored. Many researchers as- Impressions were taken with addition polyvinyl
sess the outcome of individual restorations in patients siloxane (President, Coltene). Feldspathic porcelain
mouths and report results at the restorative level, not (Durecem, Degudent; Mirage, Chameleon Dental
at the patient level. If the risk of restorative failure Products; Fortress, Chameleon Dental Products;
were isolated to an individual tooth, then there would Vita 900, Vita Zahnfabrik) was applied (usually in
be no need to account for clustering. However, if the three layers) to a refractory die (GG refractory die
risk of restorative failure is patient related, then spe- material, GC America). The restoration was etched
cific systemic conditions or patient habits may result (5% hydrofluoric acid, 10% sulfuric acid; Vita Ceramics
in a cluster of failures or a cluster of successes. In the Etch, Vita Zahnfabrik), steam cleaned, and delivered.
presence of clustering, analysis of individual restora- All laboratory procedures were completed by a single
tions may lead to biased results, with the outcomes commercial laboratory
artificially inflated or reduced. Clearly, the outcome The veneers were tried-in with either water or try-
of a single veneer in a mouth cannot be considered in paste. Following assessment, they were washed
independent of the outcome of another veneer in that with water and ethyl alcohol and then silanated.
same mouth. Retraction cords and rubber dam were not used. The
This study aimed to investigate the clinical outcome tooth substrate was cleaned with pumice and water
and estimated cumulative survival of feldspathic por- and etched (37% phosphoric acid). Each veneer was
celain veneers in situ for up to 21 years while also cemented individually with a dual-cure unfilled resin
accounting for clustered outcomes. cement (Vision 2, Mirage Dental Systems, Chameleon
Dental Products). The cementation and finishing pro-
Materials and Methods cesses were completed for each veneer before the
next was cemented. Occlusion was designed with
Inclusion/Exclusion Criteria anterior protrusive and canine laterotrusive guidance
when possible.
Porcelain veneers (n = 499) placed in patients (n = 155)
by a single prosthodontist in a private specialty prac- Clinical Follow-up
tice between 1990 and 2010 were sequentially includ-
ed in this prospective cohort study. Following veneer placement, patients were assigned
Nonvital teeth, molar teeth, and teeth subjec- an individualized maintenance schedule. The sched-
tively assessed to have an unfavorable periodontal ule was based on the patients dental and medical

Volume 25, Number 6, 2012 605

2012 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Clinical Outcome and Survival of Feldspathic Porcelain Veneers

risk factors when the veneers were first issued and Table 1a Studies of Porcelain Veneers Reporting
then modified as those risk factors changed over Study Design/sample
time. In general, the regimen involved six monthly ap-
Burke and Lucarotti Retrospective cohort
pointments for 2 years and then yearly reviews until (2009)7 1,177 patients
year 5, at which point the frequency of recalls was 2,562 veneers (mean: 2.2 per patient*)
increased or decreased depending on the patients
requirements. The reviews involved examination, pro- Layton and Walton Prospective cohort
fessional prophylaxis, smoothing of minor porcelain (2007)10 100 patients
304 veneers (mean: 3.0 2.8 per patient)
chipping, and management of any complications.
Patients were also encouraged to return for review Dumfahrt (1999)8 Retrospective cohort
outside of these scheduled appointments if required. Dumfahrt and 72 patients
Regardless of the recall schedule, all patients in the Schaffer (2000)9 205 veneers (mean : 2.9 per patient*)
cohort were actively contacted and reviewed at regu- Peumans et al Prospective cohort
lar intervals over the study period. The most recent (1998)12 54 patients
Peumans et al 87 veneers (mean: 1.6 per patient*)
review occurred in 2010. (2004)11
Shaini et al (1997)13 Retrospective cohort
Clustering 102 patients
372 veneers (mean: 3.6 per patient)
Patients received a mean of 5.8 4.3 veneers each.
Therefore, the outcomes were clustered. There is no
reason to believe that the outcome of a single veneer Smales and Etemadi Retrospective cohort
(2004)14 50 patients
is independent of the outcome of another veneer in 110 veneers (mean: 2.2 per patient*)
the same mouth. Therefore, patients who received
more than one veneer may have a cluster of veneers
that failed, survived, or became lost to recall. To ex- NR = not reported.
*Mean not reported by authors and was estimated post hoc.
plore the effects of clustering, the results of this co- This simple average likely underestimates the true number.
Standard error or 95% confidence interval not available.
hort study were assessed via two methods and then
compared. First, the outcome was analyzed for all
499 veneers in all 155 patients without accounting for
clustering; second, the outcome was analyzed for 1 Table 1b Studies of Porcelain Veneers Reporting
randomly chosen veneer from each patient. Study Design/sample
A random number table was generated and used to Aristidis and Dimitra Prospective cohort
identify veneers for analysis.19 A single veneer in each (2002)15 61 patients
patient was included. Patients with multiple veneers 186 veneers (mean: 3.1 per patient*)
were listed alphabetically, and their veneers were Friedman (1998)16 Retrospective cohort
Unknown no. of patients
listed in order by tooth number (FDI system). The ran- 3,255 veneers
dom table was read in one direction from the top left
*Mean not reported by authors and was estimated post hoc.
corner toward the right, with the first number being This simple average likely underestimates the true number.
4. Therefore, the fourth veneer in the first patient
was included. If the patient had received fewer than
four veneers, the table would continue to be read until
a number equal to or less than the number of veneers
in situ was selected, at which point the corresponding
veneer was included.

Outcome Measure
to a biologically stable tooth, with an intact restorative
Waltons six-field classification20 was used to define margin and no requests for replacement made by the
veneer survival (Table 2). For Kaplan-Meier analysis, patient for any reason.
surviving prostheses were those defined as success- For surviving veneers, time in situ was defined
es, survivors, or repairs; censored prostheses were as the time between veneer placement and the last
those defined as deaths or unknowns; and failed follow-up appointment. For failed veneers, time in situ
prostheses were those defined as failures. Therefore, was defined as the time between veneer placement
a surviving veneer is one that remained in situ, bonded and the date the failure occurred.

606 The International Journal of Prosthodontics

2012 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
 Layton/Walton

Kaplan-Meier Estimated Cumulative Survival

Characteristics Estimated cumulative survival
Setting: general dentists in the general dental services of England and Wales 5 y: NR
Inclusion: random selection by birth date of adults (age: > 18 y) who received at least 10 y: 53%
1 direct restoration and 1 veneer
Material: not stated, likely to include at least some feldspathic porcelain veneers
Setting: 1 operator, private specialist practice, Australia 5 y: 96% 1%
Exclusion: less than 80% enamel, loss of more than 1/3 incisal edge width, subjective assessment of 10 y: 93% 2%
high parafunctional risk 13 y: 91% 3%
Material: feldspathic porcelain (Mirage)
Setting: 2 operators, university, Austria 5 y: 95%
Exclusion: less than 50% enamel, compromised substrate 10 y: 91%
Material: feldspathic porcelain (Optec HSP)
Setting: 1 operator, location not reported (authors were from Belgium) 5 y: 92% 1%
Exclusion: poor oral hygiene, unfavorable occlusion, less than 50% enamel 10 y: 64% 6.5%
Material: feldspathic porcelain (GC Cosmotech)

Setting: students and staff, dental hospital, England 5 y: 58% 5.5%

Exclusion: poor oral hygiene, periodontal problems, severe tooth discoloration, extensive 6.5 y: 47% 7%
tooth structure loss, insufficient dental hospital records
Extra information: 90% of veneers were placed on unprepared teeth
Material: feldspathic porcelain (Vitadur N)
Setting: 2 operators, private specialist practice, Australia 5 y: 96% 5% (covered)
Exclusion: severe tooth discoloration, inadequate sound enamel, evidence of marked parafunction 5 y: 86% 5% (uncovered)
Extra information: 64 veneers prepared with an uncovered incisal edge and 46 with a covered incisal edge 7 y: 96% 5% (covered)
Material: feldspathic porcelain (Mirage) 7 y: 86% 5% (uncovered)
10 y: NR

Straight Percentage Outcomes

Characteristics Results
Setting: 1 operator, university, Greece 5 y: 98% judged to be clinically acceptable
Exclusion: signs of excessive occlusal forces
Material: feldspathic porcelain with 15% aluminum oxide (Ceramco)
Setting: 1 operator, private practice, United States 93% judged to be clinically acceptable
Inclusion: patients who returned for review; loss to follow-up not reported
Material: not stated, likely to include at least some feldspathic porcelain veneers

Statistical Analysis 5 years and 1 month would be in 5- to 10-year group.

All veneers were in situ for at least 1 year.
Demographics were reported as means and stan- Cumulative estimated survival was calculated us-
dard deviations. The veneers were grouped into ing the Kaplan-Meier method, and the standard er-
5-year intervals (1 to 5, 5 to 10, 10 to 15, 15 to 20, and ror was calculated with the Greenwood formula. The
> 20 years) based on treatment date. For example, a number at risk within each interval was considered
veneer that was in situ for 5 years exactly would be to be the number entering the interval minus half the
in the 1- to 5-year group, while a veneer in situ for number censored during the given interval. The 95%

Volume 25, Number 6, 2012 607

2012 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Clinical Outcome and Survival of Feldspathic Porcelain Veneers

a b

Figs 1a to 1c(a) Veneer preparations in the maxilla from canine to

canine. (b) Veneers in situ. (c) Schematic representation of the veneer
preparation. (Modified from Shillingburg et al18 with permission.)
Chamfer margin

Labial reduction
0.5 to 0.7 mm
Incisal reduction
1 to 2 mm

Table 2 Six-Field Classification System20

Field Definition
Successful Review of documentation or patient examination revealed no evidence of retreatment other than maintenance
procedures (eg, professional prophylaxis and smoothening of minor porcelain chipping). Smoothening was considered
minor when the veneer did not require further repair, the chip did not interfere with the marginal integrity, and the
result did not compromise the esthetics as determined by the patient.
Surviving Patient was not able to be examined by the author, but either the referring dentist or patient confirmed that there had
been no retreatment other than that previously described for a successful outcome.
Unknown Patient could not be located.
Dead Any patients who died during the survey period, regardless of whether they had experienced successful or surviving
treatment until their death. However, if previous documentation indicated some form of retreatment had been
undertaken before death, the relevant treatment episode was categorized as having a retreatment outcome.
Retreatment Patient underwent any form of retreatment other than maintenance procedures as previously described. Occlusal or
lingual perforation of a tooth for access to perform endodontic therapy was not considered retreatment. This category
was further subdivided to describe the result of the retreatment.
Repaired Original marginal integrity of the restorations and teeth was maintained.
Failed Part or all of the prosthesis was lost, the original marginal integrity of the restorations and teeth was modified,
or the restoration lost retention more than once.

Table 3 No. of Veneers Treated in Each 5-Year Period

confidence interval was 1.96 times the standard error.
Veneers Total Survival was expressed as a percentage standard
Time in situ n % n % error or as a percentage bounded by a 95% confi-
15 y (20102006) 145 29.1 499 100.0 dence interval. Results were expressed on Kaplan-
510 y (20052001) 115 23.0 354 70.9 Meier plots and delineated in life tables. Differences
in survival between groups were analyzed with the
1015 y (20001996) 157 31.5 239 47.9
log-rank test. Statistical significance was set at
1520 y (19951991) 77 15.4 82 16.4
P < .05. Data were analyzed using PASW Statistics
21 y (1990) 5 1.0 5 1.0 version 18.0 (IBM).

608 The International Journal of Prosthodontics

2012 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
 Layton/Walton

Maxillary teeth Maxillary teeth

15 14 13 12 11 21 22 23 24 25 15 14 13 12 11 21 22 23 24 25
100 50

80 40
No. of veneers

No. of veneers
60 30

40 20

20 10

0 0

20 10
45 44 43 42 41 31 32 33 34 35 45 44 43 42 41 31 32 33 34 35
Mandibular teeth Mandibular teeth

a b
Figs 2a and 2b Distribution of teeth (FDI tooth-numbering system) treated with veneers in the (a) entire sample (n = 499) and
(b) random sample (n = 155).

Results Table 4 Six-Field Outcome for the Entire Sample (n = 499)

and Random Sample (n = 155) of Porcelain Veneers
Demographics Entire sample Random sample
n % n %
In relation to treatment date, 145 veneers were in situ
Death 5 1.0 1 0.6
for 1 to 5 years, 115 veneers for 5 to 10 years, 157
veneers for 10 to 15 years, 77 veneers for 15 to 20 Failed 17 3.4 4 2.6
years, and 5 veneers for more than 20 years (Table 3). Repaired 3 0.6 2 1.3
Eighty-two percent of patients (n = 127) with 85% of Success 365 73.1 111 71.6
the total veneers (n = 424) were female. Eighteen per- Survival 39 7.8 10 6.5
cent of patients (n = 28) with 15% of veneers (n = 75) Unknown 70 14.0 27 17.4
were male. The age of patients at treatment ranged
from 15 to 73 years (mean: 41 14.1 years).
Patients received between 1 and 20 veneers each
(mean: 5.8 4.3 veneers per patient), with a distribu-
tion as follows: 1 veneer (n = 50, 32%), 2 to 6 veneers
(n = 87, 56%), 7 to 10 veneers (n = 15, 10%), 10 to Six-Field Outcome
15 veneers (n = 2, 1%), and > 15 veneers (n = 1, 0.5%).
Eighty-six percent (n = 426) of veneers were placed Table 4 shows the six-field outcome for both the en-
on maxillary teeth (incisors: 58%, canines: 18%, pre- tire sample and the randomly selected sample. Eleven
molars: 10%), while 14% (n = 73) were placed on patients with 56 veneers experienced more than one
mandibular teeth (incisors: 9%, canines: 3%, premo- outcome.
lars: 2%) (Fig 2a). All veneers were placed on vital Seventeen veneers failed in 8 patients, with half of
teeth. No veneers were placed on molar teeth. the patients experiencing multiple failures (13 failures
For the randomly selected sample of veneers, in 4 patients) and half experiencing a single failure
92% of veneers (n = 142) were placed on maxillary among other successful veneers (4 veneers in 4 pa-
teeth (incisors: 72%; canines: 10%; premolars: 10%), tients). The failures occurred between years 1 and 13.
while 8% (n = 13) were placed on mandibular teeth Seventy-five veneers in 30 patients were classified
(incisors: 6%; canines: 0.5%; premolars: 2%) (Fig 2b). as unknown and censored for Kaplan-Meier analysis.

Volume 25, Number 6, 2012 609

2012 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Clinical Outcome and Survival of Feldspathic Porcelain Veneers

100
90

Estimated cumulative survival (%)

80
70
60
50
40
30 All veneers (n = 499)
20 Randomly selected
veneers (n = 155)
10
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Time in situ (y)

Fig 3 Kaplan-Meier survival curves up to 21 years.

One patient died, resulting in 5 veneers being clas- sample, the estimated cumulative survival rate was
sified as unknown after 13 years in situ. Sixty-eight 98% 1% at 5 years, 96% 2% at 10 years, 96% 2%
veneers in 27 patients were classified as unknown at 15 years, and 96% 2% at 20 years. The estimated
because the patients failed to return for follow-up cumulative survival rates of the entire sample and
visits, while 2 veneers in 2 patients were classified as random subsample were not significantly different
unknown because the successful veneers were re- (chi-square = 0.21, P = .65) (Fig 3).
placed with another prosthesis (eg, a tooth-supported
fixed dental prosthesis). Fifty-three of these unknown Discussion
veneers (76%) occurred in 12 patients (41%).
Four hundred seven veneers in 130 patients were Although feldspathic porcelain veneers have been
survivors. Three surviving veneers in 3 patients required commonly used for over 30 years, reports of their
repair, with all repairs occurring after 15 years of ser- survival rates appear contradictory. In this study, all
vice. Three hundred sixty-five surviving veneers were veneers completed over a 21-year period were in-
further classified as successful based on the six-field cluded sequentially, and the results highlight the ex-
criteria. Sixty-eight percent of patients experienced cellent clinical outcomes that can be obtained with
multiple surviving veneers (365 veneers in 88 patients). feldspathic porcelain veneers.
Of the 17 failed veneers, 11 were replaced with In comparison with the Kaplan-Meier survival rates
another veneer (and remain successful), 5 were re- reported by studies identified in the previously men-
placed with a metal-ceramic crown, and 1 required tioned literature review,17 these outcomes are comple-
complete removal of the tooth abutment and was re- mentary with three9,10,14 of the six studies. Each of
placed with a tooth-supported fixed dental prosthe- these studies reported strict assessment of remaining
sis. Reasons for failure included veneer shade (n = 2), prepared enamel and reported high survival rates. The
gingival recession adversely affecting esthetics (n = 8), 5-year survival rates were 96% 5%,14 95% (standard
porcelain fracture (n = 3), trauma (n = 1), tooth frac- error not available),9 and 96% 1%.10 A 7-year survival
ture (n = 1), loss of retention on more than one occa- rate of 96% 5% was reported by one study.14 Ten-
sion (n = 1), and extensive caries (n = 1). year survival rates of 93% 2%10 and 91% (standard
error not available)9 were also reported. One study
Survival and Clustering based on a similar patient cohort reported a 13-year
survival rate of 91% 3%.10 Further details regard-
For the entire sample of veneers, the estimated ing the methodology of these studies are available in
cumulative survival rate was 98% 1% at 5 years, Table 1.
96% 1% at 10 years, 91% 2% at 15 years, and Regarding earlier research reported by Layton and
91% 2% at 20 years. For the randomly selected Walton,10 that study included 48 veneers in 19 patients

610 The International Journal of Prosthodontics

2012 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
 Layton/Walton

placed between 2001 and 2003, which overlapped The comparatively reduced survival rates reported
with the 260 veneers in 88 patients placed between in these studies could also be attributed to statistical
2001 and 2010 in the present study. Despite this over- methodology. First, this may relate to loss to follow-
lap, the similarities in the patient cohort are small and up. Differences in outcome could occur if a large pro-
do not preclude thoughtful comparison of the results portion of successful veneers were censored while a
and methodologies. large proportion of failed veneers returned for review.
The other three studies7,1113 all reported compara- Regarding outcomes reported by Peumans et al,11,12
tively reduced survival rates. Differences in survival high loss to follow-up did not occur, as nearly all ve-
can be attributed to differences in clinical and/or neers (81 of 87, 93%) returned for review at 10 years. In
statistical methodology. Clinically, differences could the study by Burke and Lucarotti,7 the loss to follow-
relate to environmental and patient-related factors; up was not reported. In the study by Shaini et al,13 42
statistically, differences could relate to loss to follow- of the original 372 veneers (11%) were available for
up and analysis of clustered outcomes. review at 6.5 years. It is unclear how many veneers
A prospective study by Peumans et al11,12 regarding were unavailable due to attrition or patient death and
a cohort of 87 veneers in 54 patients reported a high how many veneers were unavailable because they
survival rate of 92% 1% at 5 years, which dropped had been in situ for less than 6.5 years. Therefore, the
to 64% 6.5% at 10 years. The authors attributed the reduced survival may be partially related to bias in
10-year failure rate of 36% to reduced enamel under the results. Nevertheless, there is little evidence that
the preparations. Some of the veneers were placed the reduced reported survival rate was attributable to
on teeth with large interproximal restorations or a this issue.
high proportion of dentinal substrate exposed dur- Second, the present study accounted for clustered
ing preparation. Further, some veneers were not at- outcomes, while the six identified studies did not.
tached with adhesive bonding agents. Therefore, this Multiple patients in each of these studies received
reduced 10-year survival rate is likely attributable to more than one veneer, but the impact of these clus-
differences in clinical methodology and the increased tered outcomes on the results cannot be reviewed
prevalence of veneered teeth with reduced enamel retrospectively without access to individual patient
bonding substrate. data. Prior to accounting for clustering, patients in
Burke and Lucarotti7 reported an estimated cumu- the present study received between 1 and 20 veneers
lative 10-year survival rate of 53%. The authors retro- each (mean: 5.8 4.3 veneers per patient). Two-thirds
spectively evaluated the outcome of 2,563 veneers in (n = 105, 68%) of patients received more than 1 veneer.
1,177 patients. The material used for the veneers was Accounting for clustering was considered essential.
not specifically reported, but it is likely that multiple When a patient receives more than one veneer,
materials were used, including feldspathic porcelain. the individual characteristics of that patient may ad-
The tooth preparation and bonding substrate were versely or favorably affect the outcomes of all veneers
also not specifically reported. However, it is conceiv- placed, thus clustering the outcomes. Clustered out-
able that veneers placed within this environment may comes can be accounted for statistically, or the clus-
not have met the preparation criteria advocated by tered units can be separated prior to analysis. For this
prosthodontic specialists; likewise, it is conceivable research, the latter method was employed.
that the veneers may have been bonded to compro- From the entire veneer sample (veneers = 499,
mised tooth substrates. patients = 155), a nonclustered data sample of 155
A retrospective cohort study by Shaini et al13 of veneers was identified for survival analysis (ie, the
372 veneers in 102 patients reported 5- and 6.5-year random sample). In patients who received more than
survival rates of 58% 5.5% and 47% 7%, respec- one unit, a random number table was used to ran-
tively. The veneers were completed by students and domly identify a single veneer for analysis. In patients
staff at the Birmingham Dental Hospital in England. who received only one unit, each veneer was included
The authors reported that over 90% of the veneers for analysis. Survival of the randomly selected sample
were placed on unprepared teeth. The bond strength was analyzed and compared with the survival of the
to aprismatic enamel is lower than achievable with entire sample.
prepared enamel, and it is likely that this clinical tech- In the entire sample, almost three-quarters of ve-
nique resulted in the higher failure rate. This tech- neer failures (13 veneers) were clustered in half of the
nique does not adhere to traditional guidelines for patients with failures (4 patients). Reasons for failure
veneer preparation; therefore, while the results pro- varied between patients; however, within an individ-
vide useful clinical data, they are not comparable with ual patient, the failed veneers were attributable to a
the results of the present study. single reason. Qualitatively, failures were clustered.

Volume 25, Number 6, 2012 611

2012 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Clinical Outcome and Survival of Feldspathic Porcelain Veneers

In total, 75 veneers in 30 patients were classified References

as unknown. One hundred percent of unknowns due
to death were clustered in 1 patient, and nearly 75% 1. The glossary of prosthodontic terms. J Prosthet Dent 2005;
94:1092.
of losses to follow-up were clustered in half of the
2. Pincus C. Building mouth personality. Presented at the
patients in this category (53 unknowns occurring in California State Dental Association, San Jose, California, 1937.
12 patients). Qualitatively, unknown outcomes were 3. Buonocore M. A simple method of increasing the adhesion of
clustered. acrylic filling materials to enamel surfaces. J Dent Res 1955;
Eighty-four percent of patients (130 patients) ex- 34:849853.
4. Bowen R. Development of a Silica-Resin Direct Filling Material.
perienced at least 1 surviving veneer (407 veneers).
Report 6333. Washington: National Bureau of Standards, 1958.
Approximately 90% of these occurred in 70% of pa- 5. Simonsen R, Calamia J. Tensile bond strength of etched porce-
tients (365 veneers in 88 patients). Qualitatively, sur- lain [abstract 1154]. J Dent Res 1983;62:297.
viving veneers were clustered. 6. Horn HR. Porcelain laminate veneers bonded to etched enam-
The Kaplan-Meier method analyzes failures, un- el. Dent Clin North Am 1983;27:671684.
7. Burke F, Lucarotti P. Ten-year outcome of porcelain laminate
knowns, and survivors to estimate the cumulative
veneers placed within the general dental service in England
survival. Clustering of results will affect the calculat- and Wales. J Dent 2009;37:3138.
ed outcome. Accounting for clustering of outcomes 8. Dumfahrt H. Porcelain laminate veneers. A retrospective eval-
at the study level improves the validity of the results. uation after 1 to 10 years of service: Part IClinical procedure.
For the randomly selected sample, the estimated Int J Prosthodont 1999;12:505513.
cumulative survival rate was 96% 2% at 10 years 9. Dumfahrt H, Schaffer H. Porcelain laminate veneers. A ret-
rospective evaluation after 1 to 10 years of service: Part II
and 96% 2% at 20 years. For the entire sample, the Clinical results. Int J Prosthodont 2000;13:918.
estimated cumulative survival rate was 96% 1% at 10. Layton D, Walton TR. An up-to 16-year prospective study of
10-years and 91% 2% at 20 years. The differenc- 304 porcelain veneers. Int J Prosthodont 2007:20:389-396.
es between groups were not statistically significant. 11. Peumans M, De Munck J, Fieuws S, Lambrechts P, Vanherle G,
Quantitatively, the distribution of failures and sur- Van Meerbeek B. A prospective ten-year clinical trial of porce-
lain veneers. J Adhes Dent 2004;6:6576.
vivals within patients did not significantly affect the 12. Peumans M, Van Meerbeek B, Lambrechts P, Vuylsteke-
estimated cumulative survival. Nonetheless, analysis Wauters M, Vanherle G. Five-year clinical performance of por-
of outcomes without accounting for clustering could celain veneers. Quintessence Int 1998;29:211221.
prove misleading. 13. Shaini FJ, Shortall AC, Marquis PM. Clinical performance of
The method chosen to account for clustering in porcelain laminate veneers. A retrospective evaluation over a
period of 6.5 years. J Oral Rehabil 1997;24:553559.
this study required no additional software and was 14. Smales R, Etemadi S. Long-term survival of porcelain laminate
simple to apply, accurate, and time efficient. Although veneers using two preparation designs: A retrospective study.
accounting for clustering decreases the number of Int J Prosthodont 2004;17:323326.
prostheses in the analysis and thus decreases the 15. Aristidis G, Dimitra B. Five-year clinical performance of porce-
lain liminate veneers. Quintessence Int 2002;33:185189.
power to detect differences between study variables,
16. Friedman MJ. A 15-year review of porcelain veneer failure
failure to account for clustering may result in errone- A clinicians observations. Compend Contin Educ Dent 1998;
ous statistical findings and incorrect identification of 19:625628.
prognostic survival factors. 17. Layton D, Walton T. A systematic review and meta-analysis
of the survival of porcelain veneers over 5 and 10 years. Int J
Prosthodont 2012;25:590603.
Conclusions
18. Shillingburg HT Jr, Hobo S, Whitsett LD, Jacobi R, Brackett
SE. Metal-ceramic restorations. In: Shillingburg HT Jr, Hobo
When bonded to prepared enamel substrate, feld- S, Whitsett LD, Jacobi R, Brackett SE (eds). Fundamentals of
spathic porcelain veneers have an excellent long- Fixed Prosthodontics, ed 3. Chicago: Quintessence, 1997:458.
term survival rate and low failure rate. The 21-year 19. StatTrek website. http://stattrek.com/Tables/Random.aspx.
Accessed 20 January 2011.
estimated cumulative survival rate was 96% 2%.
20. Walton TR. The outcome of implant-supported fixed prosthe-
Multiple dental prostheses in the same mouth are ex- ses from the prosthodontic perspective: Proposal for a clas-
posed to the same local and systemic factors, result- sification protocol. Int J Prosthodont 1998;11:595601.
ing in clustered outcomes. Efforts should be made by
future researchers to account for clustering in their
analyses.

612 The International Journal of Prosthodontics

2012 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.