Ken Tan A systematic review of the survival and

Bjarni E. Pjetursson
Niklaus P. Lang
complication rates of fixed partial
Edwin S. Y. Chan dentures (FPDs) after an observation
period of at least 5 years
III. Conventional FPDs

Authors’ affiliations: Key words: abutment fractures, biological, caries, fixed partial dentures, loss of retention,
Ken Tan, National Dental Center, Singapore,
loss of vitality, material fractures, periodontitis, success, survival, systematic review, tech-
Singapore
Bjarni E. Pjetursson, Niklaus P. Lang, School of nical complications
Dental Medicine, University of Berne, Berne,
Switzerland
Edwin S. Y. Chan, Clinical Trials and Epidemiology Abstract: The present study was done to determine the long-term success and survival of
Research Unit, Singapore, Singapore fixed partial dentures (FPDs) and to evaluate the risks for failures due to specific biological
and technical complications. A MEDLINE search (PubMed) from 1966 up to March 2004 was
Correspondence to:
Dr Ken Tan, conducted, as well as hand searching of bibliographies from relevant articles. Nineteen
Department of Restorative Dentistry studies from an initial yield of 3658 titles were finally selected and data were extracted
National Dental Center Singapore
5 Second Hospital Avenue independently by three reviewers. Prospective and retrospective cohort studies with a mean
Singapore 168938 follow-up time of at least 5 years in which patients had been examined clinically at the
Singapore
Tel.: (65)-63248947
follow-up visits were included in the meta-analysis. Publications only based on patients
Fax: (65)-63248900 records, questionnaires or interviews were excluded. Survival of the FPDs was analyzed
e-mail: tken@pacific.net.sg according to in situ and intact failure risks. Specific biological and technical complications
such as caries, loss of vitality and periodontal disease recurrence as well as loss of retention,
loss of vitality, tooth and material fractures were also analyzed. The 10-year probability of
survival for fixed partial dentures was 89.1% (95% confidence interval (CI): 81–93.8%) while
the probability of success was 71.1% (95% CI: 47.7–85.2%). The 10-year risk for caries and
periodontitis leading to FPD loss was 2.6% and 0.7%, respectively. The 10-year risk for loss
of retention was 6.4%, for abutment fracture 2.1% and for material fractures 3.2%.

Fixed partial denture (FPD) replacements retention and marginal defects (Swartz
for teeth have taken a variety of designs et al. 1996).
throughout the years. Many principles in- Biological and technical complications
volved in the preparation and construction have been reported in a variety of studies.
of fixed prostheses are still dominating, However, most of the reports deal with
although more compatible and resilient retrospective cohort studies after a defined
materials have been introduced in recent period of observation. Prospective con-
years. Owing to the variety of techniques trolled trials are almost non-existent as
Date:
Accepted 30 June 2004 employed and materials used, failing recon- there may be ethical problems with rando-
structions may be attributed to several mizing treatment procedures, while retro-
To cite this article:
Tan K, Pjetursson BE, Lang NP, Chan ESY. A causes. While some studies based only on spective studies or studies with historical
systematic review of the survival and complication rates
of fixed partial dentures (FPDs) after an observation
surveys, attribute over 50% of failures to controls have compared treatment out-
period of at least 5 years. III. Conventional FPDs. the dentists and materials used (Maryniuk comes over time. It has to be realized,
Clin. Oral Impl. Res. 15, 2004; 654–666
doi: 10.1111/j.1600-0501.2004.01119.x & Kaplan 1986), other self-reported failures however, that retrospective studies may
were attributed to biological and technical be biased towards the selection of the
Copyright r Blackwell Munksgaard 2004 failures including secondary caries, loss of treatment deemed to be optimal at that

654
 Tan et al . Systematic review of FPDs

time over other treatments. (Shugars et al. Obviously, results from studies of pa- Search Strategy

1998; Aquilino et al. 2001). tient cohorts should never be used to gen- First electronic search
3658 Titles
Another difficulty in comparing failure eralize the conditions encountered under
rates between studies arises from the fact circumstances not fulfilling the criteria Independently selected by 2 screeners
that the definition of failure used by the pursued in the given patient cohort. In 291 Titles

different authors may be highly variable. particular, patients treated in institutions Discussion
Kappa score 0.43 Discarded
While some authors define a failure when may differ from those from private practice 80 Titles
the entire fixed prostheses is no longer in situ as they may present with higher oral hy-
or requires immediate replacement (Leem- giene standards and be part of a strict Agreed by both
211 Titles
poel et al. 1985; Karlsson 1989) other authors maintenance care program (Nyman & Abstracts obtained

report on ‘functional’ survival (Izikowitz Ericsson 1982).

Discussion
1985) or the occurrence of biological and This present study is aimed at Agreed on 73 abstracts
Full text obtained
technical failures (Hämmerle et al. 2000).

The determination of the long-term
The lifespan of fixed partial dentures is Further handsearching
success/survival of FPDs. 8 Studies
usually in excess of 10 years. Longer ob-

The evaluation of the failure rates of
servation periods, however, have rarely Total full text articles
FPDs due to specific biological and 81
been reported (Palmqvist & Swartz 1993;
technical complications.
Valderhaug et al. 1997) owing to patient Final number of studies included
attrition. Some studies used telephone in- 19

terviews and questionnaires (Randow et al.

Fig. 1. Search strategy.
1986; Glantz et al. 1993) while others Material and methods
studies recalled patients for clinical exam-
Search strategy and study selection
ination. Consequently, these studies will
show a disparity in number of clinical cases Search strategy independently by three reviewers (K. T.,
actually examined. A MEDLINE search (PubMed) 1966 up to B. E. P. and E. S. Y. C.) using a data
Many studies on failures used databases April 2004 was conducted for articles pub- extraction form. Results of cantilever
from insurance claims (Rene et al. 1991; lished in Dental Journals in the English FPDs have been reported elsewhere (Pje-
Öwall et al. 1992) and others incorporated language searching for ‘fixed partial den- tursson et al. 2004a, 2004b).
patients who complained of problems at tures OR bridges’, and ‘partial edentulism’.
dental institutions (Schwartz et al. 1970; Manual searching of (i) bibliographies of Data extraction
Foster 1990a, 1990b). Due to the nature of previous reviews and (ii) bibliographies of The following parameters were analyzed.
self-reporting, the length of service of re- all publications complemented this review.
storations is usually approximated, and the Since following a screening literature Survival of FPDs
total number of a particular prostheses type search there were no randomized con- Survival was defined as an FPD that was in
is unknown. In such instances, failure rates trolled trials available on the subject, only situ at the examination visit irrespective of
are impossible to calculate. The progres- prospective or retrospective cohort studies its condition.
sion of periodontal disease, loss of retention with a mean follow-up time of at least 5
of one or several retainers without the years were considered. This was to elim- Success of FPD
entire bridge being detached (Ödman & inate obvious iatrogenic reasons for failure Success was defined as an FPD that
Karlsson 1988; Djemal et al. 1999), minor and faulty cementation techniques. Fur- remained unchanged and did not require
ceramic fractures and other events may go thermore, only publications in English any intervention over the observation
unnoticed by the patient. and in which patients had been examined period.
The longevity and complication rate of clinically at the follow-up visits were in-
FPDs will be influenced by the level of cluded. Publications based on patients re- Biological complications
skills and academic knowledge of the clin- cords, questionnaires or interviews were Caries.
ician. Reports from academic institutions excluded. Publications that combined find-
presented reports on FPDs incorporated by ings of both FPDs and single crowns were Loss of pulp vitality.
either undergraduate students under super- included only if at least two-thirds were
vision, graduate students, junior and senior FPDs. Periodontal disease progression.
faculty members, while patient cohorts Figure 1 describes the process of identify-
from private practice may also reflect a ing the 19 included studies from an initial Technical complications
variety of operator expertise, but conditions yield of 3,658 titles. This was indepen- Loss of retention.
clearly differ from the academic institu- dently done by two examiners (K. T. and
tions. Hence, failure and complication B. E. P.). The k statistics revealed a score of Abutment tooth fractures.
rates will always relate only to the patient 0.43 indicating modest agreement. Conse-
cohort treated and examined under the quently disagreements were resolved by Material fractures (fractures of veneer or
circumstances specified. discussion. Data were then extracted framework).

655 | Clin. Oral Impl. Res. 15, 2004 / 654–666

Tan et al . Systematic review of FPDs

Statistical analysis (Reichen-Graden & Lang 1989; Walton counting for 2094/3548 FPDs), 11.6%
The majority of the studies were character- 2003), information was obtained from per- were metal-ceramic while the others were
ized by an open cohort follow-up. Cohorts sonal communication with the authors to of gold–acrylic design. This distribution in
with open follow-up have varying follow- facilitate exposure time calculations. In sev- part reflects that there are few studies with
up times and censoring progressively de- eral studies, there was insufficient indivi- long follow-up of recent FPD design.
pletes the number of persons at risk. For dual or average follow-up information about The follow-up time was reported in var-
example, the loss of a FPD due to caries particular complications. Consequently, to- ious ways: in some studies a range was
would compete for the loss of the same tal exposure time was not calculable and indicated (Ericsson & Marken 1968; Ny-
FPD due to abutment fracture if the former hence, no failure rates were reported. man & Lindhe 1979; Hochman et al. 1992;
occurred first. As the number of events was often too Palmqvist & Swartz 1993; Yi et al. 1996;
In contrast, a closed cohort follow-up small for the normal approximation formu- Walton 2002; Hochman et al. 2003). In
of N-years is characterized by only the las to be valid, 95% confidence intervals others, the maximum observation times
event of interest (e.g. failure due to caries) for the rates were computed by exact (Gustavsen & Silness 1986; Karlsson
or survivors all with the same follow-up methods by assuming that the number of 1986; Valderhaug 1991; Jokstad & Mjör
time (N-years). events follow a Poisson distribution. Stu- 1996; Sundh & Ödman 1997; Glantz
For cohorts with open follow-up, only dies deemed similar enough by design were et al. 2002), averages (Libby et al. 1997;
complication rates (not risks) are directly pooled using negative binomial regression Nyman & Lindhe 1979; Bergenholtz &
calculable. Complication rates incorporate with robust standard errors. This is equiva- Nyman 1984; Fayyad & al-Rafee 1996a,
total exposure time (FPD time and in some lent to the random-effects Poisson regres- 1996b; Napankangas et al. 2002) or med-
instances abutment time) in the denomina- sion models as described elsewhere ians (Reichen-Graden & Lang 1989) were
tor. Total FPD exposure times were ex- (Pjetursson et al. 2004a, 2004b) with the reported.
tracted and calculated by different means: added conservatism from estimating robust Only six studies (Nyman & Lindhe
1. Summation of individual observation standard errors. Ten-year survival risks were 1979; Karlsson 1986; Sundh & Ödman
times for FPDs surviving and for FPDs calculated using the formula exp(-10x fail- 1997; Glantz et al. 2002; Napankangas
presenting with complications (Erics- ure rate) which assumes constant event et al. 2002; Hochman et al. 2003) reported
son & Marken 1968; Libby et al. 1997; rates. Ten-year failure risks were calculated on the sampling procedures. Four of these
Reichen-Graden & Lang 1989; Jokstad using 1-S(10). also reported the non-response rate (Karls-
& Mjör 1996; Hochman et al. 2003; son 1986; Yi et al. 1996; Sundh & Ödman
Walton 2003). 1997; Napankangas et al. 2002). The re-
2. From the average observation time of Results mainder did not provide adequate informa-
failed and surviving FPDs (Roberts tion to determine the quality of sampling.
1970a, 1970b; Fayyad & al-Rafee A total of 19 studies on conventional FPDs Non-responders included patients who
1996a, 1996b; Yi et al. 1996). met the inclusion criteria and were ana- died, changed their address, refused to be
3. Using acturial approximation and in- lyzed (Table 1). Seventeen were retrospec- re-examined, refused to fill in question-
formation given at fixed observation tive (Table 1a) and two were prospective naires or missed appointments due to
intervals (Valderhaug 1991; Sundh & cohort studies (Table 1b). All these studies chronic illness and work commitments.
Ödman 1997). incorporated 1764 patients with an age All studies represented open cohorts.
4. Using the mean of the minimum and range of 13–90 years. The sampling non- Therefore it was important to extract in-
maximum FPD follow-up times (Gus- response rate for the group ranged from 3% formation on the total FPD exposure time
tavsen & Silness 1986; Karlsson 1986). to 75% in the studies in which the in- in order for the event rates to be calculated.
5. From the standard exponential formula formation was available. Many studies This required mean follow-up times to
linking N-year risk and the assump- exceeded 10 years of follow-up and specific complications including censoring
tion of a constant complication rate. consequently, many patients were unable events (dropout, end of the study, com-
(Gustavsen & Silness 1986; Karlsson to return for follow-up examinations due to peting events). However, most studies
1986; Kelsey et al. 1996). physical and health limitations and death. reported complication risks as simple
In the 19 studies selected, a variety of proportions associated with the mean pa-
For the remaining studies, one study different types of FPD were represented. tient follow-up time instead of the mean
reported a mean patient follow-up time Despite this heterogeneity in FPD design, complication follow-up time.
instead of a mean follow-up time to the patient selection and clinical setups, 3548
first complication (Napankangas et al. FPDs met the inclusion criteria and formed Survival
2002), while another study did not report the basis for the analysis. However, canti- Thirteen studies provided data for FPD
the times to any specific complications at lever FPDs were not included and the survival (Table 3). A total of 2881 FPDs
all (Palmqvist & Swartz 1993). In two results of such reconstructions are reported were observed for 6–23 years, and 374 of
reports, the number of FPDs lost to fol- elsewhere (Pjetursson et al. 2004a, 2004b) . these had been lost. Nine studies permitted
low-up (Jokstad & Mjör 1996) and actual A total of 3548 FPDs were analyzed over the calculation of FPD exposure times
number of FPDs (Glantz et al. 2002) was a period 1–25 years (Table 2). Of the seven using methods outlined. In two studies,
not available. Finally for two studies studies that reported on FPD design (ac- the data set was made available by the

656 | Clin. Oral Impl. Res. 15, 2004 / 654–666

 Table 1. Study and patient characteristics
Study (year) Sampling frame and sampling Planned no. of Actual no. of Age range Mean age Operator Non-response (%)
method patients in study patients in study
(a) Retrospective studies
Hochman et al. (2003) Institutional patients, random Data not available 30 Dental students
selection, no periodontal involvement
Walton (1997), (2002), Private specialist practice Data not available 357 13–74 Single specialist
(2003)
Glantz et al. (2002) 43000 patients from a general dental 150 150 48.2 males and Private dentists
practitioners (GDP) requiring 4748.9 females
extensive restoration. Random
selection
Napankangas et al. (2002) 414 institutional patients random 159 132 39–82 56.8 Dental students 27 (17)
selection
Sundh & Ödman (1997) 265 institutional patients consecutive 101n 101 52.9 Dental students 30 (30)
selection
Libby et al. (1997) Institutional patients. Number not Data not available 50 21–90 63 Two specialists
stated.
Fayyad & al-Rafee (1996a, Institutional and private patients. Data not available 132 University and private
1996b) Number not stated
Yi et al. (1996) 4200 institutional patients 50 34 50–81 67 Specialist clinic 16 (32)
Palmqvist & Swartz (1993) Institutional patients from register. 122 66 Specialist clinic 56 (46)
Number not stated
Hochman et al. (1992) Institutional and private patients. 68 66 33–45 Private dentist 2 (3)
Number not stated
Reichen-Graden & Lang Institutional patients for board exams 88 58 26–72 Dental students 30 (34)
(1989)
Gustavsen & Silness (1986) Institutional patients 140w 108 50 Dental students 32 (23)
Karlsson (1986) Dental insurance plans, random 642 164 34–78 58 Private dentists 478 (75)
selection
Bergenholtz & Nyman Institutional periodontal patients. Data not available 52 21–68 47.1 Six specialists
(1983) Number not stated
Nyman & Lindhe (1979) Institutional periodontal patients 50 50 23–72 48.7 Specialist clinic
with 50% tissue support loss 251
patients, every consecutive fifth
patient
Roberts (1970a, 1970b) Institutional patients from register. Data not available Data not available 29 Specialist and postgraduates

657 |
Number not stated
Ericsson & Markén (1968) Private patients of one dentist. 98 72 One private dentist 27 (28)
Selection data unknown
(b) Prospective studies
Valderhaug (1991) 15 years Institutional patients. Selection data 102 102 25–69 48 Dental students z
unknown
Jokstad & Mjör (1996) Private patients by 3 dentists. 61 40 Private dentist 35
Selection data unknown
Total 1764
n
The 30 non-responders were replaced by re-sampling.
wRecalculated to be 140 instead of 143.
zStudy number indicative of studies with at least one follow up observation (prospective study).
Tan et al . Systematic review of FPDs

Clin. Oral Impl. Res. 15, 2004 / 654–666

Tan et al . Systematic review of FPDs

Table 2. Information on FPDs in the included studies

Study (year) Original no. Total FPDs Ceramic Gold/resin Total Total Mean FPD Follow-up Mean
of FPDs abutments pontics length range follow-up
Hochman et al. (2003) 49 1–25 6.3
Walton (1997), (2002), (2003) 515 515 1209 885 1–15
n n
Glantz et al. (2002)
Napankangas et al. (2002) 540 204 2–15 7.6
Sundh & Ödman (1997) 163 138 21 117 4.3 18 18
Libby et al. (1997) 89 238 146 1–25 8.4
Yi et al. (1996) 43 256 224 11.5 14–15 14.5
Fayyad & al-Rafee (1996a, 1996b) 156 100 56 5.1
Jokstad & Mjör (1996) 81 59 88 1–10 10
Palmqvist & Swartz (1993) 103 103 34 69 487 5þ 18–23 18–23
Hochman et al. (1992) 138 569 4–17 4–17
Valderhaug (1991) 108 108 0 108 343 259 5.6 15 15
Reichen-Graden & Lang (1989) 73 182 150 3.5 4–8 6.4
Karlsson (1986) 238 60 178 944 662 6.8 10 10
Gustavsen & Silness (1986) 114 6 6
Bergenholtz & Nyman (1984) 82 4–13 8.7
Nyman & Lindhe (1979) 332 0 332 5–8
Roberts (1970a, 1970b) 1046 (1045) 1046 5–17
Ericsson & Markén (1968) 88 260 164 4.8 1–16 6.4
3548w
n
Total number of FPDs studied not stated.
w3548 does not include an indeterminate number of FPD from Glantz et al. (2002) and discrepancy of 1 FPD between Roberts (1970a, 1970b).
FPDs, fixed partial dentures.

Table 3. Survival of conventional FPDs

Study (year) Total no. Mean No. of FPDs No. of Total FPDs FPD failure
of FPDs follow-up time in situ (%) failures exposure time rate (95% CI)
Hochman et al. (2003) 49 6.3 43 (87.8) 6 324 0.019 (0.006, 0.04)
Walton (2002) 515 478 (92.8) 37 3363 0.011 (0.008, 0.015)
Napankangas et al. (2002)n 204 7.6 197 (96.6) 7 1478 0.005 (0.002, 0.01)
Sundh & Ödman (1997) 163 16–18 103 (76.8) 35 2532 0.014 (0.01, 0.019)
Yi et al. (1996) 43 14.7 43 (100) 0 632 0 (0, 0.006)
Palmqvist & Swartz (1993) 103 18–23 79 (77) 24 Data not available
Hochman et al. (1992) 138 4–17 129 (93.5) 9 Data not available
Valderhaug (1991) 108 15 59 (54.6) 26 1263 0.021 (0.013, 0.03)
Reichen-Graden & Lang (1989) 73 6.4 71 (97.3) 2 465 0.004 (0.001, 0.016)
Gustavsen & Silness (1986) 114 6 111 (97.4)w 3 676 0.004 (0.001, 0.013)
Karlsson (1986) 238 10 231 (97.1)z 7 2348 0.003 (0.001, 0.006)
Roberts (1970b) 1045 5.1 828 (79.2) 217 5132 0.042 (0.04, 0.05)
Ericsson & Markén (1968) 88y 6.4 87 (98.8) 1 560 0.002 (0.00004, 0.01)
Total 2881 374 18,773
Pooled rates Estimated FPD failure rate (95% CI) 0.0116 (0.0061, 0.022)
Estimated 10-year FPD survival (95% CI) 89.1 (81–93.8)
n
10-year FPD in situ survival risk is 84% (Fig. 1) (Napankangas et al. 2002).
wThe 6-year in situ failure risk for Gustavsen & Silness (1986) was 2.6%.
zKarlsson (1986) has different definition of FPD in function. He deemed those totally and partially removed as ‘no longer in function’.
yEricsson & Markén (1968) – one failure was added to the total of 87 observed FPDs as reported and given a follow-up time of 6 years (average follow-up
time).
FPDs, fixed partial dentures; CI, confidence interval.

authors (Reichen-Graden & Lang 1989; Figure 2 presents the Forrest plot of the 4.9–14%)/1000 FPD-years and a 10-year
Walton 2002), while for the other two, meta-analysis performed for all the studies risk of survival of 92%.
the data provided were inadequate (Hoch- with adequate information for FPD survi-
man et al. 1992; Palmqvist & Swartz val. It is evident that one study (Roberts FPD success
1993). 1970a, 1970b) represents an outlier with a Four studies provided information on FPDs
The pooled rate for FPD loss was 11.6 substantial higher rate of FPD loss than the that remained intact over the observa-
(95% CI: 6.1–22%)/1000 FPD-years. other studies. A sensitivity analysis ex- tion period (Fig. 3, Table 4). The pooled
Hence, the 10-year FPD survival was cluding this study (Roberts 1970a, 1970b) FPD complication rate was 34.1 (95% CI:
89.1% (95% CI: 81–93.8%). gives a pooled loss rate of 8.3 (95% CI: 16–74%)/1000 FPD years. Hence, the

658 | Clin. Oral Impl. Res. 15, 2004 / 654–666

 Tan et al . Systematic review of FPDs

Hochman 10-year risk of success of FPD was 71.1%

(95% CI: 47.7–85.2%). It is evident that
Walton
one study (Fayyad & al-Rafee 1996a,
Napankangas 1996b) represents an outlier with a sub-
Sundh stantial higher FPD complication rate than
the other studies. When excluded, the FPD
Yi loss rate was 21 (95% CI: 9.8–43)/loss
Valderhaug FPD-years and the 10-year risk of success
was 81.1%. The study with the lowest
Reichen-Graden
failure rate (Walton 2002) contributed
Gustavsen 63% of the total FPD exposure.
Karlsson
Biological complications
Roberts
Dental caries
Ericsson
Caries was reported on the surface, abut-
ment and the FPD levels:
11.6 (95% CI: 6.1 - 22.0) Only one study (Karlsson 1986) ad-
dressed dental decay on a surface level
0 10 20 30 40 50 60 70 80 and found 8.1% of all surfaces being de-
Failures per 1000 FPD-years cayed within 10 years.
Fig. 2. Failure rates (per 100 years) of conventional fixed partial dentures (FPDs). CI, confidence interval. Studies that reported the development of
caries of abutments were divided into those
reporting on caries of abutments
Walton
(1) not leading to the loss of the abutment
or FPD and
(2) those that resulted in the loss of the
Libby
abutment.

Six studies gave information on caries

occurring on abutments. In two studies
Fayyad
with a 10-year follow-up (Karlsson 1986;
Jokstad & Mjör 1996), high caries occur-
rence was reported (16–23%), while the
Reichen-Graden other 4 studies revealed a substantially
lower occurrence of caries ranging from
1% to 7% over the observation period.
Three studies reporting on caries resulting
in loss of abutments were excluded from
34 (95% CI: 16 - 74) Table 5a (Hochman et al. 1992; Palmqvist
& Swartz 1993; Yi et al. 1996) as minor
0 10 20 30 40 50 60 70 80 90 100 110
caries was not included in their reports.
Complications per 1000 FPD-years The pooled rate of caries on the abut-
Fig. 3. Complication rates (per 100 years) of conventional fixed partial dentures (FPDs). CI, confidence ment level was 9.9/1000 abutment years.
interval. Hence, the 10-year risk for caries on

Table 4. Success rate of FPDs in situ (without any complication)

Study (year) Total no. Mean Intact No. of Total FPDs FPD complication
of FPDs follow-up FPDs (%) complications exposure time rate (95% CI)
time
Walton (2002) 515 470 (91.3) 45 3363 0.013 (0.01, 0.018)
Libby et al. (1997) 89 8.4 76 (85.4) 13 759 0.017 (0.01, 0.03)
Fayyad & al-Rafee (1996a, 1996b) 156 5.1 100 (63.8) 56 792 0.071 (0.05, 0.09)
Reichen-Graden & Lang (1989) 73 6.4 56 (76.7) 17 465 0.037 (0.021, 0.059)
Total 833 131 5379
Pooled rates Estimated FPD complication rate (95% CI) 0.034 (0.016, 0.074)
Estimated 10-year FPD success (95% CI) 71.1 (47.7, 85.2)

FPDs, fixed partial dentures; CI, confidence interval.

659 | Clin. Oral Impl. Res. 15, 2004 / 654–666

Tan et al . Systematic review of FPDs

Table 5a. Caries rate of abutment teeth (caries as a complication)

Study (year) Total no. Total no. of Mean No. of carious Total abutment Abutment
of FPDs abutments follow-up abutments (%) exposure caries rate
time time (95% CI)
Walton (2002) 515 1209 8 (0.7) 7894 0.001 (0.0004, 0.002)
Libby et al. (1997) 89 238 8.4n 15w (6.3) 2030 0.007 (0.004, 0.012)
Jokstad & Mjör (1996) 135 10 21 1078 0.019 (0.12, 0.03)
Reichen-Graden & Lang (1989) 73 z 6.4 median 5 901 0.006 (0.002, 0.013)
Karlsson (1986) 238 944 10 212 (22.5)y 9314 0.025 (0.02, 0.023)
Gustavsen & Silness (1986) 114 210 6 4 (1.9) 1260 0.003 (0.001, 0.008)
Total 1029 22,477
Pooled rates Estimated abutment complication rate (95% CI) 0.010 (0.0047, 0.021)
Estimated 10-year risk for caries at abutments (95% CI) 9.5 (4.6–18.9)
n
Follow-up time is calculated from survival curve (Fig. 1) for FPDs.
wFifteen abutments were computed from 238 retainers/89 FPDs approximated to three retainers/FPD. Multiplied by 5 FPDs affected by caries ¼ 15 retainers.
Total abutment time is upper limit.
zAbutments for conventional and cantilever FPDs not differentiated. Approximated that 182 abutments for 94 FPDs in total ¼ 1.94 abutments/FPD.
1.94  464.6 (total FPD time for conventional FPD) ¼ total exposure time for abutments carrying conventional FPD.
yDental caries on tooth surfaces as reported in Karlsson (1986) was omitted as the numbers were indeterminate.
FPDs, fixed partial dentures; CI, confidence interval.

Table 5b. FPDs lost due to caries

Study (year) Total no. Mean FPDs lost Total FPDs Failure rate
of FPDs follow-up due to exposure (95% CI)
caries (%) time
Hochman et al. (2003) 49 6.3 3 (6.1) 324 0.009 (0.002, 0.01)
Napankangas et al. (2002) 204 7.6 2 (1) 1478 0.001 (0.0002, 0.005)
Sundh & Ödman (1997) 138 18n 6 (4.3) 2532 0.002 (0.0009, 0.005)
Reichen-Graden & Lang (1989) 73 6.4 0 465 0 (0, 0.008)
Palmqvist & Swartz (1993) 103 18–23 1 (1) Data not available
Valderhaug (1991) 108 15n 5 (6.2) 1263 0.004 (0.001, 0.01)
Total 675 17 6062
Pooled rates Estimated FPD failure rate (95% CI) 0.0026 (0.0016, 0.0043)
Estimated 10-year risk of FPD loss due to caries (95% CI) 2.6 (1.6–4.2)
n
Follow-up time represented in these studies are the maximum follow-up time.
FPDs, fixed partial dentures; CI, confidence interval.

Table 6. Loss of abutment tooth vitality

Study (year) Total no. of No. of Mean No. of abutments Total abutment Complication rate
abutments vitality follow-up that lost exposure time (95% CI)
time vitality (%)
Jokstad & Mjör (1996) Data not available 86n 10 5 (5.8) 686 0.007 (0.002, 0.017)
Palmqvist & Swartz (1993) 487 327 18–23 49 (15) Data not available
Reichen-Graden & Lang (1989) – – 6.4 3 (3.7) 901 0.003 (0.001, 0.01)
Karlsson (1986) 944 650w 10z 73 (11.2) 6135 0.012 (0.009, 0.015)
Bergenholtz & Nyman (1984) Data not available 255 8.7 38 (14.9) 2219 0.017 (0.012, 0.024)
Total 1318 168 9941
Pooled rates Estimated abutment complication rate (95% CI) 0.011 (0.0059, 0.019)
Estimated 10-year risk for loss of abutment vitality (95% CI) 10.0 (5.7–17.3)
n
Not inclusive of dropouts.
wNine abutments with root fill through the retainer were presumed vital at time of cementation.
zFollow-up time represented in these studies are the maximum follow-up time.
CI, confidence interval.

the abutment level was 9.5% (95% CI: contain sufficient information for the cal- Loss of abutment vitality
4.6–18.9%). culation of a caries rate at the FPD level. Loss of abutment vitality was reported in
Caries reported on the FPD level leading The pooled rate for caries resulting five studies with a 6–23 years of follow-up
to the loss of the FPD was addressed in 6 in loss of FPD was 2.6/1000 FPD time (Table 6). One study (Bergenholtz &
studies (Table 5b). Seventeen out of 675 years. Hence, the 10-year risk for loss of Nyman 1984), specifically addressing loss
FPDs were lost as a result of caries. One FPD due to caries is 2.6% (95% CI: of vitality in patients reconstructed after
study (Palmqvist & Swartz 1993) did not 1.6–4.2%). successful therapy for advanced perio-

660 | Clin. Oral Impl. Res. 15, 2004 / 654–666

 Tan et al . Systematic review of FPDs

dontitis, reported the highest loss rate of 7). Of the 1080 bridges followed for periods tion affected 75 out of 1307 FPDs observed
17/1000 abutment years. Significantly ranging from 4 to 18 years, only seven were over a period of 5–23 years. Two studies
higher loss of vitality was observed in lost due to recurrent periodontitis. One (Gustavsen & Silness 1986; Karlsson 1986)
abutments when compared with non- study (Fayyad & al-Rafee 1996a, 1996b), reported high occurrences of loss of reten-
prepared control teeth. Several authors however, reported recurrent periodontitis tion. This correlated to the increased oc-
reported ‘endodontic complications’, with- to affect abutments of 12.8% of the FPDs currence of caries reported in one of them
out mentioning the baseline vitality status after only 5.1 years. Since the information (Karlsson 1986).
(Libby et al. 1997; Fayyad and al-Rafee provided did not differentiate between The pooled rate for loss of retention was
1996a, 1996b; Sundh & Ödman 1997; periodontitis reported as a complication or 6.6/1000 FPD-years. Hence, the 10-year
Napankangas et al. 2002). Consequently periodontitis leading to the loss of the FPD, risk for loss of retention is 6.4% (95% CI:
these studies were excluded. it was excluded from the analysis. Further- 3.9–10.4%).
The loss of vitality rates was not calcul- more, one study did not provide sufficient
able for one of the five studies (Palmqvist information for the calculation of total Fracture of the abutment teeth
& Swartz 1993), as total abutment time exposure time (Hochman et al. 1992). leading to FPD loss
was unavailable. The pooled rate for recurrent perio- The occurrence of fracture of abutments
The pooled rate for loss of abutment dontitis leading to FPD loss was 0.5/1000 leading to the loss of FPDs was reported
vitality was 11/1000 abutment years. FPD years. Hence, the 10-year recurrent in seven studies (Table 9). Over a period
Hence, the 10-year risk for loss of abut- periodontitis risk was 0.5% (95% CI: of 6–18 years, 16 of a total of 749 FPD
ment vitality is 10% (95% CI: 5.7–17.3%). 0.1–2.2%). were lost due to abutment fracture. In
one study, it was specifically stated that
Technical complications no FPDs had been lost as a result of
Recurrent periodontitis abutment fracture (Ericsson & Marken
Nine studies provided information on Loss of retention (fracture of the luting cement) 1968).
periodontal disease progression resulting Nine studies addressed loss of retention of The pooled rate of abutment fracture
in loss of abutment teeth or FPDs (Table the reconstruction (Table 8). Loss of reten- leading to FPD loss was 2.2/1000 FPD

Table 7. Loss of FPDs due to periodontal disease

Study (year) Total Mean FPDs lost due Total FPDs Failure rate (95% CI)
FPDs follow-up time to periodontitis (%) exposure time
Napankangas et al. (2002) 204 7.6 0 (0) 1478 0 (0, 0.003)
Sundh & Ödman (1997) 138 18n 4 (2.9) 2532 0.002 (0.0004, 0.004)
Libby et al. (1997) 89 8.4 0 (0) 759 0 (0, 0.005)
Yi et al. (1996) 43 14.7 0 (0) 632 0 (0, 0.006)
Hochman et al. (1992) 138 4–17 1 (0.7) Data not available
Valderhaug (1991) 108 15n 2 (2.5) 1263 0.002 (0.00002, 0.006)
Reichen-Graden & Lang (1989) 73 6.4 0 (0) 465 0 (0, 0.008)
Gustavsen & Silness (1986) 114 6n 0 (0) 676 0 (0, 0.005)
Nyman & Lindhe (1979) 173 6.2 0 (0) 1073 0 (0, 0.003)
Total 1080 7 8878
Pooled rates Estimated FPD failure rate (95% CI) 0.0005 (0.0001, 0.0022)
Estimated 10-year risk of FPD loss due to periodontitis (95% CI) 0.5 (0.1–2.2)
n
Follow-up time represented in these studies are the maximum follow-up time.
FPDs, fixed partial dentures; CI, confidence interval.

Table 8. Loss of retention

Study (year) Total no. Mean FPDs that lost Total FPD Complication rates (95% CI)
of FPDs follow-up time retention (%) exposure time
Napankangas et al. (2002) 204 7.6 3 (1.5) 1478 0.002 (0, 0.006)
Sundh & Ödman (1997) 138 18 3 (2.2) 2532 0.001 (0.0002, 0.003)
Fayyad & al-Rafee (1996a, 1996b) 156 5.1 7 (4.5) 792 0.009 (0.004, 0.02)
Palmqvist & Swartz (1993) 103 18–23 8 (7.8) Data not available
Valderhaug (1991) 15 years 108 15 7 (3.7) 1263 0.006 (0.002, 0.01)
Reichen-Graden & Lang (1989) 73 6.4 3 (4.1) 465 0.007 (0.001, 0.018)
Karlsson (1986) 238 10 30 (12.6) 2348 0.013 (0.009, 0.018)
Gustavsen & Silness (1986) 114 6 8 (7) 676 0.012 (0.005, 0.023)
Nyman & Lindhe (1979) 173 6.2 6 (3.5) 1073 0.006 (0.002, 0.012)
Total 1307 75 10,627
Pooled rates Estimate FPD complication rate (95% CI) 0.0066 (0.004, 0.011)
Estimated 10-year risk for loss of retention (95% CI) 6.4 (3.9–10.4%)

FPDs, fixed partial dentures; CI, confidence interval.

661 | Clin. Oral Impl. Res. 15, 2004 / 654–666

Tan et al . Systematic review of FPDs

Table 9. Loss of FPDs due to abutment tooth fracture

Study (year) Total no. Total no. Mean FPDs lost due to Total FPD Failure rate (95% CI)
of FPDs of abutments follow-up time tooth fracture (%) exposure time
Hochman et al. (2003) 49 6.3 2 (4.1) 324 0.006 (0.0007. 0.01)
Napankangas et al. (2002) 204 7.6 5 (2.45) 1478 0.003 (0.001, 0.008)
Sundh & Ödman (1997) 138 18 4 (2.9) 2532 0.002 (0.0004, 0.004)
Libby et al. (1997) 89 238 8.4 1n (1.12) 759 0.001 (0.00003, 0.007)
Valderhaug (1991) 108 343 15 3 (3.7) 1263 0.002 (0.0005, 0.007)
Reichen-Graden & Lang (1989) 73 6.4 1 (1) 465 0.002 (0, 0.012)
Ericsson & Markén (1968) 88 260 0 (0) 560 0 (0, 0.007)
Total 749 16 7381
Pooled rates Estimated FPD failure rate (95% CI) 0.0022 (0.0015, 0.0032)
Estimated 10-year risk of FPD loss due to abutment fracture (95% CI) 2.1 (1.4–3.2)
n
Assumed fracture led to loss of FPD.
FPDs, fixed partial dentures; CI, confidence interval.

Table 10. Fractures of material: framework, post and core and porcelain veneers
Study (year) Total no. Mean Material Metal Ceramic Veneers Core Total FPD Complication
of FPDs follow-up fracture (%) exposure rate (95% CI)
time time
Hochman et al. (2003) 49 6.3 1 (2) 1 324 0.003 (0.00008, 0.01)
Walton (2002) 515 4 (0.8) 2 2 3363 0.001 (0.0003, 0.003)
Napankangas et al. (2002) 204 7.6 16 (6.9) 2 14 1478 0.011 (0.006, 0.018)
Sundh & Ödman (1997) 138 18 3 (2.2) 2 1 2532 0.001 (0.0002, 0.0003)
Libby et al. (1997) 89 8.4 2 (2.2) 1 1 759 0.003 (0.0003, 0.01)
Fayyad & al-Rafee 156 5.1 2 (1.3) 792 0.003 (0.003, 0.01)
(1996a, 1996b)
Palmqvist & Swartz (1993) 103 18–23 3 (2.9) Some data Data not
repeated available
Hochman et al. (1992) 138 4–17 4 (2.9) Data not
available
Valderhaug (1991) 108 15 1 (1.2) 1263 0.001 (0.0002, 0.004)
Reichen-Graden & Lang (1989) 73 6.4 4 (5.5) 3 1 465 0.008 (0, 0.012)
Karlsson (1986) 238 10 2 (0.8) 2 2348 0.001 (0, 0.003)
Nyman & Lindhe (1979) 173 6.2 3 (1.7) 3 1073 0.003 (0.001, 0.008)
Total 1984 45 14,397
Pooled rates Estimate FPD complication rate (95% CI) 0.0032 (0.0015, 0.0068)
Estimated 10-year risk of material fractures (95% CI) 3.2 (1.5–6.5)

FPDs, fixed partial dentures; CI, confidence interval.

Table 11. Comparison of previous meta-analyses and present review

Review Creugers et al. (1994) Scurria et al. (1998) Present review
Search dates 1970–1992 1966–1996 1966–2004
Language of paper Dutch, English, German English English
Number of studies 7 4 13
9 4
Minimum follow-up 5 3 5
Definition of failure In situ loss or in need of repair In situ loss Survival
In situ loss or in need of repair Success
Risk duration 5 and 10 years 10 and (15) years 10 years
Survival 90% 92% (75%) 89.1%
Success 87% (69%) 71.1%

years. Hence, the 10-year risk for abut- tion period between 4 and 23 years (Table FPD years. Hence, the 10-year risk for
ment fracture leading to FPD loss was 10). These included fractures of the frame- material complications was 3.2% (95%
2.1% (95% CI: 1.4–3.2%). work, the veneers or the core build-ups. CI: 1.5–6.5%).
Where information was available, the frac-
Material complications: framework, tures were recorded as separate categories. Discussion
veneer and core fractures Forty-five out of 1984 FPDs were affected
Twelve studies reported on the occurrence by material complications. The pooled rate Long term prospective cohort studies are
of material complications over an observa- of material complications was 3.2/1000 the gold standard for determining the sur-

662 | Clin. Oral Impl. Res. 15, 2004 / 654–666

 Tan et al . Systematic review of FPDs

vival experience of FPDs. Out of the origi- 1996. In one previous review, studies re- to 81%. The reason for the low risk of
nal yield of 3658 titles, only 19 studies ported in German and Dutch were in- success in this study was the unusually
qualified for the inclusion into the review cluded (Creugers et al. 1994) while the high occurrence of periodontal disease
and yet, for some aspects of analysis, the second review (Scurria et al. 1998) was (36.6%) and secondary caries (23.2%) lead-
number of studies with valid information limited, like the present, to studies re- ing to failure.
was reduced to a few studies only. Survival ported in English only.
and success rates of the FPDs could be Several papers that were incorporated in Biological complications
calculated for most of the studies included. the previous reviews had to be excluded There were a limited number of studies
However, the results of the analysis for the from the present analysis as they did not available that provided sufficient relevant
occurrence of biological and technical com- meet the inclusion criteria. In three of the information to calculate the occurrence
plications was drastically affected by the excluded studies, the mean follow-up time of biological complications such as caries,
incompleteness of reported data in the was less than 5 years (Reuter & Brose loss of abutment vitality and recurrent
majority of the studies. Furthermore, it is 1984; Cheung et al. 1990), while in an- periodontitis.
evident from the present analysis that most other two studies (Randow et al. 1986; Information on caries was divided into
of the studies on the longevity of FPDs date Leempoel et al. 1995) data on survival that which led to repair and that which led
back to the 1980s and 1990s, and there is a and complication of the FPDs were ob- to the loss of the FPD. The 10-year risk for
paucity of studies performed in the new tained from questionnaires without patient caries on abutments was 9.5%, but only
century. Consequently, caution must be contact. In addition, the previous analyses 2.6% of FPDs were lost as a result of caries.
exercised to the interpretation of technical required data to be reformulated from life Only one study (Karlsson 1986) classified
complications such as veneer fractures, table analyses that led to the calculation of marginal discrepancies and found a correla-
since most of the studies available for N-year risks. In the present review, how- tion between the worst marginal discre-
analysis would have reported on gold-ac- ever, the failure rate was directly estimated pancy (open margins) and the presence of
rylic FPDs. from the exposure times and the 10-year caries. Two studies reporting on a high
The focus of the present analysis was on survival from the relationship between occurrence of caries also reported high
a comprehensive evaluation of survival and event rate and survival function. The pre- occurrence of loss of retention of FPDs
success rates of FPDs. This, in turn, re- sent 10-year survival rate is affected by one (Karlsson 1986; Jokstad & Mjör 1996).
quired a calculation of the exposure times large study with over 1000 FPDs and a low However, no clear information on this
of FPDs. In the present review, all the 10-year survival (Roberts 1970a, 1970b). If association can be determined.
studies represented open cohorts that re- this particular study is excluded from the Prosthodontic treatment is known to
quired various methods to calculate FPD analysis, the 10-year survival rate of FPDs cause pulp trauma by mechanical prepara-
exposure times. rises to 92.1%. tion and the contact of various substances
A direct comparison on simple propor- From the Forrest plot of study-specific to opened dentinal tubules that may have
tions would seriously mislead due to the failure rates, it is evident that these vary precipitated an early response of pulpal
differing follow-up times. For the compli- widely among the various studies. This necrosis. (Langeland & Langeland 1968;
cations of loss of vitality of abutments and may be attributable to the patient cohort Bergenholtz & Nyman 1984; Hume &
loss of retention of FPDs the rate may be observed, the design and extent of the FPD, Massey 1990). However, from this meta-
underestimated due to a larger denomina- the maintenance care provided and the analysis on 5 studies, it was clear that loss
tor (loss of retention may have occurred experience and clinical setup of the clin- of vitality of abutment teeth occurred at a
earlier, but the FPD remained in situ and icians. Furthermore, in one study (Yi et al. later date than what could be attributed to
thus, accounted for a longer exposure 1996), some of the FPDs remained in the trauma from the preparation of the
time). To compare the risk of survival and function, but 30% had portions sectioned, teeth. This may either indicate a slow
success as well as the various complica- or abutments removed. Within the defini- progressive tissue degeneration induced by
tions risks, 10-year risks were estimated tions of survival, those prostheses re- the procedure or reflect the increased sus-
from the event rates. mained in function, thus leading to 100% ceptibility of pulpal infection by dentinal
survival despite the complications encoun- tubules in advanced periodontitis (Bergen-
Survival and success tered. Yet, the effect of the vigorous main- holtz & Nyman 1984). Pulpal necrosis was
As a result of the meta-analysis of the tenance regimen and regular follow-up for diagnosed primarily on the basis of the
present systematic review, the 10-year this group of patients cannot be overlooked presence of periapical radiolucency (Ber-
risk of survival of FPDs was 89.1%. This as being contributory to the high survival. genholtz & Nyman 1984; Gustavsen &
is similar to the 10-year survival risk of The 10-year success risk of FPDs was Silness 1986; Karlsson 1986; Reichen-Gra-
FPDs reported in previous meta-analyses 71.1%. This pooled rate was calculated den & Lang 1989; Palmqvist & Swartz
(90% and 92%, respectively) (Creugers from only 4 studies. One of these had 1993; Jokstad & Mjör 1996). One study
et al. 1994; Scurria et al. 1998). The latest yielded a very low success due to a high (Bergenholtz & Nyman 1984) compared
search date for the two analyses were 1992 rate of 71 complications/1000 FPD years 255 abutment teeth with 417 non-abut-
and 1996, respectively. In the present re- (Fayyad & al-Rafee 1996a, 1996b). If this ment teeth and found a higher incidence of
view, 10 additional studies have been in- particular study is excluded from the ana- pulpal necrosis in abutment teeth (15% vs.
cluded since 1992 and 6 studies since lysis, the 10-year success risk of FPDs rises 3%).

663 | Clin. Oral Impl. Res. 15, 2004 / 654–666

Tan et al . Systematic review of FPDs

The pooled 10-year risk for loss of abut- rate, since the number of admissible events dentures requiring intervention but leaving
ment vitality in the present review was is concomitantly reduced. Not all studies the FPD in situ were lower than that of
10%. In agreement with Bergenholtz & reported the cause of FPD failures. Due FPDs remaining totally intact during the
Nyman (1984), it may, therefore, be re- to the approximations used, the risks and observation period. Caries and loss of abut-
commended that patients treated with ex- rates should be cautiously interpreted. ment vitality were the most common
tensive FPDs should be closely monitored The highest 10-year risk was for loss of biological complications, while loss of re-
for the loss of vitality of abutments. retention amounting to 6.4%. In one long- tention of the FPD was the most common
The presence of cast post and dowels and term study (Ödman & Karlsson 1988) it technical complication. Fracture of abut-
non-vital abutments especially in distal was found that patients were often una- ment teeth and material complications
abutments has been shown to be associated ware of loose retainers or even abutment occurred less frequently.
with increased retention loss and fracture fractures. This questions the validity and In comparison with the two previous
of teeth and cores. This cautions against accuracy of survival figures from patient meta-analyses of 1994 (Creugers et al.
over dependence on non-vital teeth as stra- surveys and questionnaires. 1994) and 1998 (Scurria et al. 1998) (Table
tegic abutments. Far lower was the 10-year risk for the 11), the present systematic review by and
Based on nine studies, the 10-year risk of loss of FPD due to abutment tooth fracture. large confirmed the survival and success
loss of FPDs due to recurrent periodontitis Based on seven studies, the 10-year risk rates published. On the other hand, com-
was only 0.5%. One study (Nyman & was 2.1%. One study (Nyman & Lindhe plication rates for both biological and tech-
Lindhe 1979) found that patients placed 1979) reported that 8/332 FPDs experi- nical complications have hitherto not been
on a rigorous maintenance programme enced an abutment fracture. Five out of reported in a systematic way.
maintained gingival indices, probing depths these eight occurred in distal extension This analysis has revealed that within a
and alveolar bone heights for all 332 FPDs FPD abutments and six out of eight were 10-year time frame, FPD loss due to biolo-
throughout the entire observation period of on non-vital and root-treated abutments. gical complications amounted to 2.6% and
8–11 years. Another study (Reichen-Gra- Of particular interest, one study of 121 0.5% for caries and recurrence of perio-
den & Lang 1989) reported increased gin- bridges (Reuter & Brose 1984) found a dontitis, respectively, while technical com-
gival indices in abutment teeth compared higher incidence of abutment failure if plications within the same time frame
with control teeth with increased plaque, root canal treatment had been performed had much higher risks of FPD loss. The
bleeding on probing on abutment teeth, after bridge cementation compared with highest biological complication rate was
especially at sites where restoration mar- vital abutments and those that had been loss of abutment vitality, which later on
gins had been placed subgingivally com- root treated before construction of the may lead to technical complications. On
pared with control teeth and restorations bridge. the other hand, loss of retainer retention
with supragingival margins. One study Similar relatively low 10-year risks may result in unrestorable abutment caries.
(Ericsson & Marken 1968) found no sig- were obtained for material complications. Conflicts of interest: None declared.
nificant differences in probing pocket These included fractures of the framework,
depths between abutments and control veneers and/or cores and amounted to a
teeth. Overall, there seemed to be no ad- 10-year risk of 3.2%. Résumé
verse changes in FPDs incorporated into In one study (Libby et al. 1997) wear
periodontally well-maintained patients through the occlusal surface occurred in L’étude présente a été effectuée pour déterminer le
even if they presented with a history of two out of 89 FPDs while one out of 89 succès à long terme et la survie des prothèses fixées
et d’évaluer les risques d’échecs dü à des compli-
advanced periodontal disease. Where the FPDs showed a porcelain fracture. In a
cations biologiques et techniques. Une recherche
recall or maintenance was less stringent, study (Karlsson 1986) of both acrylic-ve- Medline de 1966 à mars 2004 a été effectuée ainsi
periodontal breakdown may occur, and neered FPDs and porcelain bonded to gold que manuelle pour les bibliographies des articles
may be more pronounced when margins FPDs, tootbrush wear on acrylic-veneered pouvant répondre à cette question. Dix-neuf études
were subgingivally located (Valderhaug & FPDs occurred in 78 out of 1207 units or ont été tirées à partir de 3 658 titres et les données
ont été extraites indépendamment par deux per-
Karlsen 1976). Secondary use of the bridge 6.4%, and porcelain fracture in 14 out of
sonnes. Les études prospectives et rétrospectives
for removable appliances also seemed to 331 units corresponding to 4.2% of porce- avec un temps de suivi de minimum cinq années
have a detrimental effect on the gingiva lain fused to gold units. A comparison of dans lesquelles les patients avaient été examinés
(Libby et al. 1997). the difference in survival between FPDs cliniquement lors du suivi ont été incluses dans
with acrylic facings and metal ceramic cette méta-analyse. Les publications uniquement
basées sur les dossiers des patients, les question-
Technical complications FPDs showed that over an 18-year period,
naires et les interviews ont été exclues. La survie des
The 10-year risk for technical complica- 38% of FPDs with acrylic facings and 4% prothèses a été analysée suivant les risques d’échec
tions such as loss of retention, loss of FPD with metal ceramic FPDs were replaced et de succès in situ. Des complications techniques et
due to abutment fracture and the occur- (Sundh & Ödman 1997). Reasons cited biologiques spécifiques telle que les caries, la perte
rence of material complications were cal- for the increase in failures were the greater de vitalité et la réapparition de la maladie parodon-
tale ainsi que la perte de rétention, la perte de
culated. For any given cause, all other incidence of discoloration and fracture after
vitalité, les fractures dentaires et de matériaux ont
causes of failure represent competing extensive wear of acrylics. également été analysées. La survie à 10 ans pour les
events that reduce the precision of the In conclusion, from the studies included, prothèses fixées était de 89,1% (intervalle de con-
estimate of that given cause-specific failure the pooled failure rates of fixed partial fiance de 95% : 81,0 à 93,8%) tandis que le succès

664 | Clin. Oral Impl. Res. 15, 2004 / 654–666

 Tan et al . Systematic review of FPDs

était de 71,1% (52,2 à 83,6%). Le risque à dix ans Karies, Vitalitätsverlust und das Auftreten von Par- pérdida de vitalidad, fracturas dentales y del
pour des caries ou de la parodontite entraı̂nant la odontalproblemen sowie Retentionsverlust, Zahn- material.
perte de la prothèse était respectivement de 2,6 et und Materialfrakturen wurden ebenfalls analysiert. La supervivencia a los diez años para las dentaduras
0,7%. Le risque à dix années pour la perte de Die Ueberlebensrate von festsitzenden Brückenre- parciales fijas fue del 89.1% (95% intervalo de
rétention était de 6,4%, de fracture du pilier de konstruktionen über 10 Jahre betrug 89.1% (95% confianza (CI): 81–93.8%) mientras que el éxito
2,1% et de fracture de matériaux de 3,2%. Vertrauensintervall (CI): 81–93.8%), während die fue del 71.1% (95% CI: 52.2–83.6%). El riesgo de
Erfolgsrate 71.1% betrug (95% CI: 52.2–83.6%). caries a los 10 años y periodontitis que condujo a la
Das 10-Jahres Risiko für Karies und Parodontitis, pérdida del FPD fue del 2.6% y del 0.7% respecti-
Zusammenfassung welche zum Verlust der FPD führen, betrug 2.6% vamente. El riesgo de pérdida de retención a los 10
bzw. 0.7%. Das 10-Jahres Risiko für Retentionsver- años fue del 6.4%, para la fractura del pilar del 2.1%
Eine systematische Uebersicht der Ueberlebens- lust betrug 6.4%, für die Pfeilerfrakturen 2.1% und y para la fractura del material del 3.2%.
und Komplikationsraten bei festsitzenden Brücken- für Materialfrakturen 3.2%.
rekonstruktionen (FPDs) über eine Beobachtungs-
zeit von mindestens 5 Jahren III. Konventionelle
FPDs Resumen

Die vorliegende Studie wurde unternommen, um El presente estudio se llevó a cabo para determinar el
die Langzeiterfolgs- und Ueberlebensraten von fes- éxito y supervivencia a largo plazo de dentaduras
tsitzenden Brückenrekonstruktionen (FPDs) zu bes- fijas parciales (FPDs) y evaluar los riesgos de fracasos
timmen und die Risiken für Misserfolge aufgrund debido a complicaciones biológicas y técnicas espe-
spezifischer biologischer und technischer Kompli- cı́ficas.
kationen auszuwerten. Se llevó a cabo una búsqueda por Medline (PubMed)
Es wurde eine Medline Suche (PubMed) über einen desde 1996 hasta Marzo de 2004 al igual que una
Zeitraum von 1966 bis März 2004 durchgeführt und búsqueda a mano de bibliografı́as de artı́culos rele-
die Bibliographien von relevanten Artikeln wurden vantes. Finalmente se seleccionaron diecinueve ar-
manuell durchsucht. Von einer anfänglichen Aus- tı́culos de una cantidad inicial de 3,658 tı́tulos y se
wahl von 3658 Artikeln wurden schlussendlich 19 extrajeron los datos independientemente por tres
ausgewählt und die Daten wurden von drei Rezen- revisores. Se incluyeron en este meta-análisis estu-
senten unabhängig herausgelesen. Es wurden pros- dios cohorte prospectivos y retrospectivos con un
pektive und retrospektive Kohorten-Studien mit tiempo de seguimiento medio de 5 años en los que
einer mittleren Beobachtungszeit von mindestens 5 los pacientes se examinaron clı́nicamente en las
Jahren, in welchen Patienten klinisch nachunter- visitas de mantenimiento. Se excluyeron publica-
sucht worden waren, in die Meta-Analyse einbezo- ciones basadas solamente en las fichas de los pa-
gen. Publikationen, welche nur auf Einträgen in cientes, cuestionarios o entrevistas. Se analizó la
Krankengeschichten, Fragebogen oder Interviews supervivencia de los FPDs de acuerdo con los riesgos
basierten, wurden ausgeschlossen. Das Ueberleben de fracaso in situ e intacto. También se analizaron las
der FPDs wurde entsprechend des in situ und in- complicaciones biológicas especı́ficas tales como
takten Misserfolgsrisikos analysiert. Spezifische caries, pérdida de vitalidad y recurrencia de enferme-
biologische und technische Komplikationen wie dad periodontal al igual que la pérdida de retención,

References

Aquilino, S.A., Shugars, D.A., Bader, J.D. & White, Ericsson, S.G. & Marken, K.E. (1968) Effect of fixed Gustavsen, F. & Silness, J. (1986) Clinical and
B.A. (2001) Ten-year survival rates of teeth adja- partial dentures on surrounding tissues. Journal of radiographic observations after 6 years on bridge
cent to treated and untreated posterior bounded Prosthetic Dentistry 20: 517–525. abutment teeth carrying pinledge retainers. Jour-
edentulous spaces. Journal of Prosthetic Dentistry Fayyad, M.A. & al-Rafee, M.A. (1996a) Failure of nal of Oral Rehabilitation 13: 295–298.
85: 455–460. dental bridges. II. Prevalence of failure and its Hämmerle, C.H., Ungerer, M.C., Fantoni, P.C.,
Bergenholtz, G. & Nyman, S. (1984) Endodontic relation to place of construction. Journal of Oral Brägger, U., Bürgin, W. & Lang, N.P. (2000)
complications following periodontal and prosthe- Rehabilitation 23: 438–440. Long-term analysis of biologic and technical as-
tic treatment of patients with advanced perio- Fayyad, M.A. & al-Rafee, M.A. (1996b) Failure of pects of fixed partial dentures with cantilevers.
dontal disease. Journal of Periodontology 55: dental bridges: III – effect of some technical International Journal of Prosthodontics 13:
63–68. factors. Journal of Oral Rehabilitation 23: 409–415.
Cheung, G.S., Dimmer, A., Mellor, R. & Gale, M. 675–678. Hochman, N., Mitelman, L., Hadani, P.E. &
(1990) A clinical evaluation of conventional Foster, L.V. (1990) Failed conventional bridge work Zalkind, M. (2003) A clinical and radiographic
bridgework. Journal of Oral Rehabilitation 17: from general dental practice: clinical aspects and evaluation of fixed partial dentures (FPDs) pre-
131–136. treatment needs of 142 cases. British Dental pared by dental school students: a retrospective
Creugers, N.H., Kayser, A.F. & van’t Hof, M.A. Journal 168: 199–201. study. Journal of Oral Rehabilitation 30:
(1994) A meta-analysis of durability data on con- Glantz, P.O., Nilner, K., Jendresen, M.D. & Sund- 165–170.
ventional fixed bridges. Community Dentistry berg, H. (1993) Quality of fixed prosthodontics Hochman, N., Yaffe, A. & Ehrlich, J. (1992) Splint-
and Oral Epidemiology 22: 448–452. after 15 years. Acta Odontologica Scandinavica ing: a retrospective 17-year follow-up study. Jour-
Djemal, S., Setchell, D., King, P. & Wickens, J. 51: 247–252. nal of Prosthetic Dentistry 67: 600–602.
(1999) Long-term survival characteristics of 832 Glantz, P.O., Nilner, K., Jendresen, M.D. & Sund- Hume, W.R. & Massey, W.L. (1990) Keeping the
resin-retained bridges and splints provided in a berg, H. (2002) Quality of fixed prosthodontics pulp alive: the pharmacology and toxicology of
post-graduate teaching hospital between 1978 and after twenty-two years. Acta Odontologica Scan- agents applied to dentine. Australian Dental Jour-
1993. Journal of Oral Rehabilitation 26: 302–320. dinavica 60: 213–218. nal 35: 32–37.

665 | Clin. Oral Impl. Res. 15, 2004 / 654–666

Tan et al . Systematic review of FPDs

Izikowitz, L. (1985) A long-term prognosis for the 8 years following insertion. Journal of Oral Re- serviceability. Journal of the American Dental
free-end saddle-bridge. Journal of Oral Rehabili- habilitation 15: 55–63. Association 81: 1395–1401.
tation 12: 247–262. Öwall, B., Cronstrom, R. & Rene, N. (1992) Scurria, M.S., Bader, J.D. & Shugars, D.A. (1998)
Jokstad, A. & Mjör, I.A. (1996) Ten years’ clinical Prosthodontic claims in the Swedish Patient In- Meta-analysis of fixed partial denture survival:
evaluation of three luting cements. Journal of surance Scheme. Acta Odontologica Scandina- prostheses and abutments. Journal of Prosthetic
Dentistry 24: 309–315. vica 50: 365–374. Dentistry 79: 459–464.
Karlsson, S. (1986) A clinical evaluation of fixed Palmqvist, S. & Swartz, B. (1993) Artificial crowns Shugars, D.A., Bader, J.D., White, B.A., Scurria,
bridges, 10 years following insertion. Journal of and fixed partial dentures 18 to 23 years after M.S., Hayden, W.J. Jr & Garcia, R.I. (1998)
Oral Rehabilitation 13: 423–432. placement. International Journal of Prosthodon- Survival rates of teeth adjacent to treated and
Karlsson, S. (1989) Failures and length of service in tics 6: 279–285. untreated posterior bounded edentulous spaces.
fixed prosthodontics after long-term function. A Pjetursson, B.E., Tan, K., Lang, N.P., Brägger, U., Journal of the American Dental Association
longitudinal clinical study. Swedish Dental Jour- Egger, M. & Zwahlen, M. (2004a) A systematic 129: 1089–1095.
nal 13: 185–192. review of the survival and complication rates Sundh, B. & Ödman, P. (1997) A study of fixed
Kelsey, J. L., Whiltemore, A. S., Evans, A. S. & of fixed partial dentures (FPDs) after an observa- prosthodontics performed at a university clinic 18
Douglas Thompson, W. (1996) Methods in Ob- tion period of at least 5 years – I. Implant sup- years after insertion. International Journal of
servational Epidemiology, 2nd ed. NY: Oxford ported FPDs. Clinical Oral Implants Research Prosthodontics 10: 513–519.
University Press. doi:10.1111/j.1600-0501.2004.01117.x Swartz, B., Svenson, B. & Palmqvist, S. (1996)
Langeland, K. & Langeland, L.K. (1968) Cutting Pjetursson, B.E., Tan, K., Lang, N.P., Brägger, U., Long-term changes in marginal and periapical
procedures with minimized trauma. Journal of the Egger, M. & Zwahlen, M. (2004b) A systematic periodontal conditions in patients with fixed
American Dental Association 76: 991–1005. review of the survival and complication rates of prostheses: a radiographic study. Journal of Oral
Leempoel, P.J., Eschen, S., De Haan, A.F. & Van’t fixed partial dentures (FPDs) after an observation Rehabilitation 23: 101–107.
Hof, M.A. (1985) An evaluation of crowns and period of at least 5 years – IV. Cantilever or Valderhaug, J. (1991) A 15-year clinical evaluation of
bridges in a general dental practice. Journal of extensions FPDs. Clinical Oral Implants Re- fixed prosthodontics. Acta Odontologica Scandi-
Oral Rehabilitation 12: 515–528. search doi:10.1111/j.1600-0501.2004.01120.x navica 49: 35–40.
Leempoel, P.J., Kayser, A.F., Van Rossum, G.M. & Randow, K., Glantz, P.O. & Zoger, B. (1986) Tech- Valderhaug, J., Jokstad, A., Ambjornsen, E. & Nor-
De Haan, A.F. (1995) The survival rate of bridges. nical failures and some related clinical compli- heim, P.W. (1997) Assessment of the periapical
A study of 1674 bridges in 40 Dutch general cations in extensive fixed prosthodontics. An and clinical status of crowned teeth over 25 years.
practices. Journal of Oral Rehabilitation 22: epidemiological study of long-term clinical Journal of Dentistry 25: 97–105.
327–330. quality. Acta Odontologica Scandinavica 44: Valderhaug, J. & Karlsen, K. (1976) Frequency and
Libby, G., Arcuri, M.R., LaVelle, W.E. & Hebl, L. 241–255. location of artificial crowns and fixed partial
(1997) Longevity of fixed partial dentures. Journal Reichen-Graden, S. & Lang, N.P. (1989) dentures constructed at a dental school. Journal
of Prosthetic Dentistry 78: 127–131. Periodontal and pulpal conditions of abutment of Oral Rehabilitation 3: 75–81.
Maryniuk, G.A. & Kaplan, S.H. (1986) Longevity of teeth. Status after four to eight years following Walton, T.R. (1997) A 10-year longitudinal study of
restorations: survey results of dentists’ estimates the incorporation of fixed reconstructions. fixed prosthodontics: 1 Protocol and patient pro-
and attitudes. Journal of the American Dental Schweizer Monatsschrift für Zahnmedizin 99: file. International Journal of Prosthodontics 10:
Association 112: 39–45. 1381–1385. 325–331.
Napankangas, R., Salonen-Kemppi, M.A. & Raus- Rene, N., Owall, B. & Cronstrom, R. (1991) Dental Walton, T.R. (2002) An up to 15-year longitudinal
tia, A.M. (2002) Longevity of fixed metal ceramic claims in the Swedish Patient Insurance Scheme. study of 515 metal-ceramic FPDs: Part 1. Out-
bridge prostheses: a clinical follow-up study. Jour- International Dental Journal 41: 157–163. come. International Journal of Prosthodontics 15:
nal of Oral Rehabilitation 29: 140–145. Reuter, J.E. & Brose, M.O. (1984) Failures in full 439–445.
Nyman, S. & Ericsson, I. (1982) The capacity of crown retained dental bridges. British Dental Walton, T.R. (2003) An up to 15-year longitudinal
reduced periodontal tissues to support fixed Journal 157: 61–63. study of 515 metal–ceramic FPDs: Part 2. Modes
bridgework. Journal of Clinical Periodontology Roberts, D.H. (1970a) The failure of retainers in of failure and influence of various clinical char-
9: 409–414. bridge prostheses. An analysis of 2,000 retainers. acteristics. International Journal of Prostho-
Nyman, S. & Lindhe, J. (1979) A longitudinal study British Dental Journal 128: 117–124. dontics 16: 177–182.
of combined periodontal and prosthetic treatment Roberts, D.H. (1970b) The relationship between age Yi, S.W., Carlsson, G.E., Ericsson, I. & Wennstrom,
of patients with advanced periodontal disease. and the failure rate of bridge prostheses. British J.L. (1996) Long-term follow-up of cross-arch
Journal of Periodontology 50: 163–169. Dental Journal 128: 175–177. fixed partial dentures in patients with advanced
Ödman, P.A. & Karlsson, S. (1988) Follow-up study Schwartz, N.L, Whitsett, L.D., Berry, T.G. & Stew- periodontal destruction: evaluation of occlusion
of patients with bridge constructions performed by art, J.L. (1970) Unserviceable crowns and fixed and subjective function. Journal of Oral Rehabi-
private dental surgeons and at a university clinic, partial dentures: life-span and causes for loss of litation 23: 186–196.

666 | Clin. Oral Impl. Res. 15, 2004 / 654–666