Journal of Dentistry (2005) 33, 805–815

www.intl.elsevierhealth.com/journals/jden

Outcome of direct restorations placed within

the general dental services in England and Wales
(Part 1): Variation by type of restoration
and re-intervention
P.S.K. Lucarottia, R.L. Holderb, F.J.T. Burkec,*
a
Dental Practice Board, Eastbourne, Sussex BN20 8AD, UK
b
Department of Maths and Stats, University of Birmingham, Birmingham B15 2TT, UK
c
Primary Dental Care Research Group, University of Birmingham School of Dentistry,
St.Chad’s Queensway, Birmingham B4 6NN, UK

Received 31 August 2004; accepted 18 March 2005

KEYWORDS Abstract Aim: It is the aim of this paper to consider the dental factors associated
Direct restorations; with the need for re-intervention on a restoration, such as the tooth position, size of
Outcome; cavity, and restoration material.
Restoration; Methods: Patients whose data were included in this study were those whose
Kaplan–Meier statisti- birthdays were included within a set of randomly selected dates, one of which was
cal methodology chosen in each possible year of birth. The restoration records consisted of all those
records containing directly placed restorations which related to courses of treatment
of patients 18 years or older with last date on the claim form after 31st December
1990, and with date of acceptance after September 1990 and before January 2002.
For each tooth treated with a direct restoration the subsequent history of
intervention on that tooth was consulted, and the next date of intervention, if any
could be found in the extended data set, was obtained. Thus a data set was created
of direct restorations with their dates of placement and their dates, if any, of re-
intervention.
Results: Data for over 80,000 different adult patients were analyzed, of whom 46%
were male and 54% female. A total of 503,965 tooth restoration occasions were
obtained from the data over a period of eleven years.
Single surface amalgam restorations were found to have the longest survival K58%
at 10 years, and glass ionomer the shortest 38% at 10 years. Factors which were found
to reduce restoration outcome included involvement of the incisal angle in
composite restorations—this resulted in a reduction in median survival of around
two years—and the placement of pins in a restoration. The presence of a root filling

* Corresponding author. Tel.: C44 121 237 2767; fax: C44 121 625 8815.
E-mail address: f.j.t.burke@bham.ac.uk (F.J.T. Burke).

0300-5712/$ - see front matter Q 2005 Published by Elsevier Ltd.

doi:10.1016/j.jdent.2005.03.008
806
was also found to reduce the survival of restorations in the crown of the root filled
tooth.
Conclusions: Small amalgam restorations have longer survival times before re-
intervention than large amalgam restorations such as MOD. Composite and glass
ionomer restorations perform less well than amalgam restorations. Pin placement
and root filling reduce the survival time of restorations.
Q 2005 Published by Elsevier Ltd.
Introduction
Cost-effective, predictable, pain-free dental treat-
ment is central to the achievement of patient
satisfaction and to the fulfilment of clinical
governance, let alone satisfying third party funders.
This paper describes the data and findings of a study
into the recorded intervals between placing direct
restorations and re-intervention on the same tooth
obtained from a large representative sample of
patients treated in the General Dental Services
(GDS) of England and Wales between 1991 and 2001.
The data consist of items gleaned from the payment
claims submitted by GDS dentists to the Dental
Practice Board (DPB) in Eastbourne, Sussex.
The interval between successive interventions is
a statistical proxy for the ‘life’ of a restoration, but
it must be recognized that there are many other
measures in use in the world of dental research.
These vary from a variety of administrative rules
(such as those developed by Robinson1) to assess-
ment by inspectors using, for example, the USPHS
criteria.2
The work described here relies on a comprehen-
sive, self-maintaining national longitudinal set of
data, rather than on a one-off cross-section or a
classical directly recruited longitudinal cohort. This
provides a greater assurance of results which can be
generalized and replicated. However, because the
data are taken from administrative records rather
than direct observation, conventions are necessary
to define every measurement in terms of the data
which are actually available.
The start of the life of a restoration is well
defined as a point of time, when the restoration is
actually placed in the tooth. This date is not
explicitly recorded in the administrative records
provided to the DPB. In this project, the date of
restoration placement was taken to be the last date
recorded in the payment claim in respect of the
course of treatment. In most cases this is the date
of completion, when the dentist discharged the
patient at the end of the course of treatment.
The end of the life of a filling is conceptually
more difficult, and it also strays into the issue of
P.S.K. Lucarotti et al.
censoring, which was discussed in the previous
article in this series.3
In this paper, and subsequent papers in this
collection of work, the definition of the end of the
life of a filling was taken to be the date of
acceptance for the next course of treatment in
which the tooth received an intervention other than
the maintenance activity of stoning and smoothing.
Re-intervention on a previously restored tooth is
statistically associated with the original restor-
ation, but it is generally possible that there is no
causal connection—the re-intervention may have
been required in response to a circumstance
unrelated to the origin restoration. For example,
a first molar tooth with an existing occlusal
amalgam restoration may receive a second occlusal
amalgam restoration, but it may be in a different
site.1
The above remarks have been made from the
point of view of the dentist providing treatment.
From the patient’s perspective, the need for
remedial treatment becomes apparent only if he
or she becomes conscious that a filling has fractured
or has fallen out or looks unsightly, or he or she
experiences pain or other discomfort. Depending on
the severity of the symptoms, the patient may
contact his or her dentist immediately, or wait until
the next routine check-up. If the need for repair is
asymptomatic, the patient relies on the pro-
fessional judgment of the dentist. However, it can
be argued that one of the expectations of patients
is that the number of unanticipated visits to the
dentist will be kept to a minimum—the regular
check-up should detect the need for intervention
before the patient becomes painfully aware of it.
The perspective of the person or organisation
that pays for the treatment is also relevant. In
terms of minimising the cost to the public purse,
the patient who does not come back to any part of
the NHS is a success. It could also be argued that a
patient does not return to his/her dentist because
the treatment provided has been successful. If the
ideal is that fillings should restore teeth to their
natural state, and that no natural tooth should be
lost, then every re-intervention on a restored tooth
is a failure, to the extent that it may reduce
Variation by type of restoration and re-intervention
the value, in years of useful tooth life, obtained
from the original filling. From a public health point
of view, the ideal would be that teeth never had to
be treated in the first place. However, once this
situation has been reached, the approach adopted
should be to preserve the teeth for life, possibly
through a series of restorative procedures, rather
than a one-off attempt to effect a permanent
repair, which, in many instances, is unrealistic,
particularly in younger patients.
This paper considers the dental factors associ-
ated with the restoration, such as the tooth
position, size of cavity, and restoration material.
Subsequent papers in this collection of work will
document the way in which the distribution of re-
intervention intervals varies with other factors.
Methods
The patients to be included in this study were
defined as those whose birthdays were included
within a set of randomly selected dates, one of
which was chosen in each year. The restoration
records consisted of all those records containing
directly placed restorations (including, initially,
root fillings without other direct restoration) which
related to courses of treatment with last date on
the claim after 31st December 1990, and with date
of acceptance after September 1990 and before
January 2002, and which were scheduled for
payment between January 1991 and March 2002
inclusive. The restoration records were further
restricted to those which related to courses of
treatment starting on or after the patient’s 18th
birthday. For all patients with at least one such
restoration record the data were extended to
include all treatment records on any of the payment
schedules from January 1991 to March 2002.
For each tooth treated with a direct restoration
the subsequent history of intervention on that tooth
was consulted, and the next date of intervention, if
any could be found in the extended data set, was
obtained. Thus a data set was created of direct
restorations with their dates of placement and their
dates, if any, of re-intervention.
A technique for analysing incomplete survival
data with individual dates of ‘life’ and ‘death’ was
developed by Kaplan and Meier in 1958 (described in
Collett, 19944). If it were known that all patients
with restored teeth which had not yet received re-
intervention were still attending their dentists at 1st
January 2002, then Kaplan–Meier analysis could be
applied directly, using 31st December 2002 as
censoring date.
807
However, it is unrealistic to suppose that all the
patients would re-attend, just as it is unrealistic to
assume, as Elderton5 did, that none of them would.
A technique has been devised by Lucarotti6 to
modify Kaplan–Meier analysis to allow for the
probability that a patient would eventually re-
attend, dependent on the time interval from the
last record of attendance by the patient to the end
of the observation period (31st December 2001),
together with the age of the patient.
This modified Kaplan–Meier analysis was used to
quantify the distribution of survival times, both
overall and according to the characteristics of the
restoration, the patient, the dentist, and date and
geographical location of the surgery. Except for the
analysis by type of restoration, records where the
only direct restoration was a root filling have been
excluded from these analyses. This paper considers
the type of restoration and the position of the tooth
in the mouth. Subsequent papers will consider other
factors.
The relationship between type of initial and type
of subsequent intervention, over time, was also
explored. Further analysis6 included the develop-
ment of a Cox-Regression model to quantify the
additional hazards associated with specific risk
factors, such as a root filling provided on the same
tooth as the restoration. Finally, the whole set of
analyses was repeated on a second and non-
overlapping random sample selected in the same
way.
Results
Characteristics of the sample population
The characteristics of the sample population have
already been set out in the first paper of this
series.3 The data are drawn from an age-stratified
sample of GDS patients, clustered by year of birth,
covering all claims scheduled between January
1991 and March 2002, inclusive. Just over eighty
thousand different adult patients were identified,
of whom 46% were male and 54% female. Between
them they accounted for 719,009 claims for
payment sent to the DPB. Each of these related to
a course of treatment involving all the care and
treatment necessary to secure and maintain the
oral health of a patient. There was a wide variation
in the number of claims associated with each
patient. Thirteen per cent of patients appeared in
the records only once, but they accounted for only
one per cent of the claims. On the other hand, those
one per cent of patients who appeared thirty times
808 P.S.K. Lucarotti et al.

Overall
1.0 Single surface amalgam
Tunnel amalgam
MO or DO amalgam
Two surface amalgam, not MO or DO
0.9 Root filling + indirect restoration
MOD amalgam
Proportion without re-intervention
Resin composite
0.8 Glass ionomer

0.7

0.6

0.5

0.4

0.3

0.2
0 500 1000 1500 2000 2500 3000 3500 4000
Time in days from placement to re-intervention

Figure 1 Eleven year survival of directly restored teeth until next intervention—by main type of restoration.

or more accounted for five per cent of the claims. then the case was classified according to the lowest
By these means, a total of 503,965 tooth restoration code, in the (ascending) order: single surface
occasions were obtained from the data.3 amalgam, occluso-buccal or occluso-lingual amal-
gam, MO or DO amalgam, MOD amalgam, tunnel
amalgam, composite resin, glass ionomer, root
Type of restoration filling. Cases classified as root fillings were there-
fore those which had no other direct restoration,
Figure 1 presents the survival curves to next and such root filling cases have been excluded from
intervention for all the types of direct restoration the remaining analyses in this and other papers in
distinguished by the codes used in the GDS. It should this series.
be noted that the regulations of the GDS preclude It should be noted from Figure 1 that, except
the use of composite resin or glass ionomer for the for these root fillings and tunnel restorations,
restoration of load-bearing surfaces of posterior which are a relatively rare type of restoration,
teeth, so the only cavities on which these materials there are clear non-overlapping distinctions to be
may be used on posterior teeth were class V lesions drawn between the different survival curves, with
on the gingival third of the tooth surface. single surface amalgam restorations consistently
Where two different restorations were placed on having the longest survival, and glass ionomer the
the same tooth in the same course of treatment shortest (overall log-rank P!0.0001). Table 1

Table 1 Restoration survival by treatment type.

Treatment type Percentiles (days) Survival rates n
Upper quartile Median 1 year 5 Year 10 Year
Single surface amalgam 1510 N/A 93% 72% 58% 76,418
Two surface amalgam, 1005 3461 89% 63% 49% 16,680
not MO or DO
MO or DO amalgam 1116 3488 90% 65% 49% 147,087
MOD amalgam 893 2759 88% 60% 43% 60,295
Tunnel amalgam 1443 N/A 93% 69% 54% 284
Resin composite 809 2645 87% 58% 43% 127,375
Glass ionomer 656 2138 84% 53% 38% 63,549
Root fillingCindirect 872 3166 86% 62% 46% 12,277
restoration
Variation by type of restoration and re-intervention 809

1.0 Overall
Tooth 1
Tooth 2
0.9 Tooth 3
Tooth 4
Proportion without re-intervention

0.8 Tooth 5
Tooth 6
Tooth 7
0.7 Tooth 8

0.6

0.5

0.4

0.3

0.2
0 500 1000 1500 2000 2500 3000 3500 4000
Time in days from placement to re-intervention

Figure 2 Eleven year survival of directly restored teeth until next intervention—by Tooth Position. Tooth position is
defined by counting outwards from the central incisor (tooth 1) to the third molar (tooth 8).

provides the key summary statistics for the factors affecting particular tooth treatment
distribution. occasions. In the analyses which follow, tunnel
restoration cases have been excluded, and the
Tooth position random method of adjustment3 has been used.
For composite resin restoration of incisors and
Tooth position is also clearly a relevant factor canine teeth there are supplementary codes on
associated with survival to next re-intervention payment claims which indicate whether the incisal
(Figure 2 and Table 2). Again overall log-rank P! angles or edge were involved in the restoration
0.0001. Third molar teeth have the longest survival (Figure 4). When the incisal angle is involved there
expectation, and incisors and canines the shortest. is an associated reduction in median survival of
Mouth quadrant does not have much influence on around two years, but survival is much better if only
survival expectation (Figure 3 and Table 3). The the incisal edge is restored.
better survival of restorations in mandibular teeth The overall log-rank statistic is significant (P!
is statistically significant (P!0.0001). 0.0001), as are all the pairwise log-rank statistics
except that between one incisal angle and complex
Additional risk factors angles (PZ0.3319). There were 95,805 cases of
composite resin restoration of incisors and canines,
Additional information, beyond tooth position and of which 15,756 involved one incisal angle, 3348 the
restoration material, is available about specific risk incisal edge only, and 1527 were complex.

Table 2 Restoration survival by tooth position.

Tooth position Percentiles (days) Survival rates n
Upper quartile Median 1 year 5 Year 10 Year
1 721 2340 85% 56% 40% 41,439
2 795 2548 87% 58% 42% 36,969
3 803 2473 87% 57% 42% 44,363
4 1010 3351 89% 63% 48% 59,311
5 1000 3278 89% 63% 48% 71,487
6 940 2951 89% 61% 45% 109,304
7 1113 3632 90% 65% 50% 97,634
8 1498 N/A 92% 72% 60% 31,181
810 P.S.K. Lucarotti et al.

1.0

0.9
Overall
Proportion without re-intervention Upper Right Quadrant
0.8 Upper Left
Lower Left
Lower Right
0.7

0.6

0.5

0.4

0.3

0.2
0 500 1000 1500 2000 2500 3000 3500 4000
Time in days from placement to re-intervention

Figure 3 Eleven year survival of directly restored teeth until next intervention—by mouth quadrant (from patient’s
perspective).

Another consideration is whether a root filling a reduction in median life expectancy of around
was placed in the tooth at the same time as the two years.
restoration in the crown of the tooth (Figure 5). As for root fillings, pair-wise log-rank tests for
Although restorations without root fillings on the presence of pin or screw retention are all highly
incisors and canines have shorter survival times significant (P!0.0001). The numbers of cases with
than those on molar and pre-molar teeth, the pin or screw retention were: 3218 out of 122,771
presence of a root filling on the same course of incisors and canines, 8810 out of 130,689 premolar
treatment reduces the time to next intervention of teeth, and 19,970 out of 237,944 molar teeth
all restorations of the crown of the tooth. restorations.
All log-rank tests for the effect of root filling
presence were significant (P!0.0001). The num- Type of re-intervention
bers of tooth restorations which also had root
fillings were as follows: incisors and canines had Of comparable interest to the time interval to the
5557 out of 122,771; premolar teeth 7065 out of next intervention is the type of re-intervention
130,689; and molar teeth 8912 out of 237,944. received, and the extent to which the type of re-
Dentine pins are placed with the aim of intervention varies over time.
enhancing the retention of a restoration, but the Type of re-intervention is closely associated with
presence of such additional treatment is also type of original restoration. The most common type
associated with shorter expected time to re- of initial restoration is an amalgam (Figure 7). It can
intervention (Figure 6). In the case of pin and be seen that the proportion of re-intervention with
screw retention, molar teeth appear to be the least amalgam increases during the first two years and
adversely affected, though even they suffer that the initial proportion of re-intervention with

Table 3 Restoration survival by mouth quadrant.

Quadrant Percentiles Survival rates n
Upper quartile Median 1 year 5 Year 10 Year
Upper right 960 3083 89% 62% 46% 139,168
Upper left 947 3025 89% 62% 46% 142,425
Lower left 973 3255 89% 63% 47% 106,228
Lower right 977 3332 89% 62% 48% 103,867
Variation by type of restoration and re-intervention 811

1.0

0.9
None
Proportion without re-intervention
One incisal angle
0.8
Incisal edge only
Complex angles
0.7

0.6

0.5

0.4

0.3

0.2
0 500 1000 1500 2000 2500 3000 3500 4000
Time in days from placement to re-intervention

Figure 4 Risk factors associated with composite resin restorations on incisors and canines. ‘Complex angles’ includes
cases where at least two angles or an angle and an incisal edge are involved.

indirect restorations or extractions is around one in the material of choice for re-intervention on teeth
three. previously restored using glass ionomer.
By contrast, there is much less variation with
time for composite resin restorations, with around
sixty per cent of the re-interventions also involving Discussion
composite resin.
For glass ionomer restorations the picture is The different life expectancies of the different
completely different, with less than half the re- types of amalgam restoration reflect the size and
interventions being in glass ionomer, and this complexity of the cavities being restored, with
proportion dropping steadily after the first year. single surface restorations surviving better K58% of
Increasingly over time, composite resin becomes these surviving ten years, compared with 43% for an

1.0

0.9 Incisors and Canines - No root filling

Premolar teeth - No root filling
Proportion without re-intervention

Molar teeth - No root filling

0.8
Incisors and Canines - Root Filling
Premolar teeth - Root Filling
0.7 Molar teeth - Root Filling

0.6

0.5

0.4

0.3

0.2
0 500 1000 1500 2000 2500 3000 3500 4000
Time in days from placement to re-intervention

Figure 5 Risk associated with root filling of directly restored teeth, by tooth position.
812 P.S.K. Lucarotti et al.

1.0

0.9 Molar teeth - No Pin or Screw

Premolar teeth - No Pin or Screw
Proportion without re-intervention
Incisors and Canines - No Pin or Screw
0.8
Molar teeth - Pin or Screw
Premolar teeth - Pin or Screw
0.7 Incisors and Canines - Pin or Screw

0.6

0.5

0.4

0.3

0.2
0 500 1000 1500 2000 2500 3000 3500 4000
Time in days from placement to re-intervention

Figure 6 Risk associated with dentine pins or screws in directly restored teeth, by tooth position.

MOD amalgam. The results of the present study the only type of tooth-coloured restoration per-
confirm the view put forward, in a systematic mitted within the GDS in loadbeaning situations in
review of restoration longevity,7 that amalgam posterior teeth. However, the majority of dentists
should be the material of choice for durability. In placed relatively few sealant restorations and the
this study, it was not possible to directly compare data on these restorations are dominated by the
the performance of amalgam and resin composite in work of a small number of dentists with unusual
load-bearing surfaces of posterior teeth because prescribing patterns, and therefore were not
the GDS Regulations (under which auspices the data considered suitable for comparison.
for this study were obtained) preclude the use of Resin composite restorations, generally placed in
resin composite in such situations, with the Class III, IV and V cavities, were found to have
following caveat. It would have been relevant to performance comparable only to the worst perform-
compare the performance of occlusal amalgam ing amalgam restoration, an MOD. Glass ionomer,
restorations with sealant restorations, which are which also may not be placed in load-bearing

100%

80%

60%

40%
Other
Dentures
Extractions
20% Crowns/Inlays
Glass Ionomer
Composite Resin
Amalgam
0%
0 4 8 12 16 20 24 28 32 36 40+
Interval to re-intervention in completed quarters

Figure 7 Type of re-intervention for teeth restored by amalgam restorations, by time since restoration.
Variation by type of restoration and re-intervention
situations under GDS Regulations, performed less
well—with only 38% of such restorations surviving
after ten years. It should, however, be remembered
that these different materials have not been
randomly assigned to different cavities—the
observed performance reflects the treatment plan-
ning of the dentist, as modified by the wishes of the
patient and the rules of the GDS, but not necessarily
what the relative performance of different
materials would have been had they been applied
to clinically similar cases. Resin composite and glass
ionomer restorations will predominantly be placed
in cavities such as Class V and Class III. With this in
mind, the comparison between the performance of
amalgam restorations (mainly placed in load-bear-
ing situations) and those of composite and glass
ionomer becomes even more salient. However,
despite the poorer performance of glass ionomer
and composite restorations, there could be a hope,
or an anticipation, that these restorations will be
placed in minimal intervention cavities which result
in little or no loss of tooth substance for retentive
features, using the adhesive properties of these
materials (when composite is used with a dentine
bonding agent). Additionally, glass ionomer restor-
ations may have been placed as transitional restor-
ations while the patient’s caries activity is brought
under control. Further investigation would appear to
be indicated to ascertain whether glass ionomer and
composite are indeed placed in non-retentive
cavities and whether there is a reduced incidence
of pulpal problems as a result.
When root fillings without other direct restor-
ations, and tunnel restorations (which are so rare
that the survival curve is ill-defined) are excluded,
it can be seen that the different types of restoration
fan out evenly over the eleven year observation
period, suggesting that a proportional hazards
model4 may be an appropriate way of describing
the relationship between them. This can be further
demonstrated using plots of the log cumulative
hazard against log time.6
There is clearly a close relationship between
tooth position and the type of restoration which is
used in a given tooth. Anterior teeth, for example,
are generally restored with composite resin, with
the aim of providing an aesthetic result which will
satisfy the patient. Third molars have a dispropor-
tionately high incidence of single surface (occlusal)
amalgam restorations, possibly reflecting the tech-
nical difficulty of placing more complex restor-
ations on such teeth, as well as the relative utility
of extraction or the fact that a proportion of these
teeth never erupt into position which is functional
enough to warrant placement of a large restoration.
This relationship is apparent from the chart of
813
survival by tooth position, with third molars (tooth
position 8) performing best, and incisors and canine
teeth worst. Again, the curves fan out evenly,
supporting the use of a proportional hazards model
to describe the relative hazards of different tooth
positions, as well as the material used. It also
appears that, although the incidence of direct
restorations on different jaws is very different,
with considerably more restorations on the upper
jaw than on the lower, this is not carried through,
except marginally, into the survival curves. This
may indicate the impact of indirect restorations on
the upper jaw, particularly for anterior teeth.
The analysis of additional factors quantifies some
unsurprising patterns. Although it may be expected
that restoration of incisal angles reduces the likely
interval to re-intervention when compared with
restorations not involving the incisal angle, there
have been no generally applicable estimates of
how great, on average, that reduction might be.
Figure 4 provides the necessary evidence. This may
not be considered surprising, since the incisal
angle restoration may not have any innate reten-
tion, although these types of cavity have been
considered to have low shrinkage-related stress
(low configuration factor) for resin composite
restorations.8 Conversely, the incisal edge compo-
site restoration will be likely to be subjected to
greater forces from opposing teeth than class III or
class V restorations. Further analysis of the data is
ongoing in order to determine whether incisal edge
involvement eventually predisposes to crown
placement.
Root fillings and pin or screw retention are likely
to be indications of a complex restoration, as well
as introducing risks of iatrogenic damage of their
own. The interest is not so much in the fact that
there is an effect, so much as the quantification of
the effect. In the case of a root filling, it appears
that the inherent risk to longevity from the root
filling is sufficient to swamp the risks associated
with the final surface filling, reducing median
survival to next re-intervention to under six years,
irrespective of tooth position. Pin or screw reten-
tion on anterior teeth (predominantly restored with
composite resin) predisposes to a reduction in life
expectancy of the restoration, both relatively and
absolutely, compared with such retention on a
molar tooth. This may simply reflect the mechan-
ical stress differential between the anatomies of
the two types of tooth. It may also indicate the ease
by which pulpal damage may occur in incisor teeth
in which the pulp is comparatively close to the
surface in comparison with molar teeth.
It could also be considered surprising that
sufficient data on pins in composite restorations
814 P.S.K. Lucarotti et al.

100%

90%

80%

70%

60%

50%

40%
Other
Dentures
30%
Extractions
Crowns/Inlays
20%
Glass Ionomer
Amalgam
10%
Composite Resin
0%
0 4 8 12 16 20 24 28 32 36 40+
Interval to re-intervention in completed quarters

Figure 8 Type of re-intervention for teeth restored by composite resin restorations, by time since restoration.

were available for analysis, since pins may exert a
weakening effect on a composite restoration,9,10
let alone the risk of dentinal cracking or crazing on
insertion of the pin11,12 and the potential for
increase in temperature in the pulp chamber when
drilling the pin hole.13 Additionally, pins may shine
through a tooth-coloured restoration producing a
poor aesthetic result and should be considered
redundant since the introduction of enamel etching
by Buonocore almost 50 years ago.14 Some re-
education of the dental profession in England
and Wales, in view of the poor performance of
pinned restorations in anterior teeth shown in this
study, would appear indicated, let alone the fact
that acid etching should produce a similar retentive
effect with less possibility of iatrogenic damage. In
this respect, however, it is reassuring to note that
the number of pins placed under the GDS Regu-
lations is dropping year on year.15–17 An alternative
viewpoint is that the Regulations should have been
criticized or challenged some time ago, insofar that
the use of pins and screws in resin composite

100%

80%

60%

40%

Other
Dentures
20% Extractions
Crowns/Inlays
Composite Resin
Amalgam
Glass Ionomer

0%
0 4 8 12 16 20 24 28 32 36 40+
Interval to re-intervention in completed quarters

Figure 9 Type of re-intervention for teeth restored by glass ionomer restorations, by time since restoration.
Variation by type of restoration and re-intervention
restorations in anterior teeth has been known to be
deleterious for over 30 years.9,10
A surprisingly large proportion of early re-
interventions, for all types of initial restoration,
consist of crowns and extractions, suggesting either
a revision of treatment plan or a planned transition
using an amalgam core for a crown. Regarding
amalgam restorations, the re-intervention is crown
or extraction in a third of cases. These findings
would appear to confirm the potentially weakening
effect of amalgam cavities on tooth structure, given
the need for retentive locks and keys, thereby
predisposing to a fracture that requires extraction
or crowning. In this respect, it is interesting to note
that Kelly and Smales have recently questioned the
cost-effectiveness of crowns as opposed to large
cusp-replacing amalgam restorations over a period
of 15 years.18 Further analysis of the DPB data,
which is ongoing, should confirm or otherwise if the
situation is the same in England and Wales. In the
case of glass ionomer, the proportion of ‘like-with-
like’ restoration never gets above fifty per cent,
suggesting that glass ionomer is used as a relatively
temporary and minimally invasive treatment.
From the point of view of the life of the tooth,
rather than just that of the restoration, the type of
re-intervention becomes important. In principle,
replacement of like with like could allow a tooth to
continue to survive indefinitely, even if the
intervals are short, whereas an extraction, even
after a long interval without re-intervention, is
terminal. The findings in Figures 7–9 could form the
basis of an economic evaluation of the relative
long-term prognosis for any initial intervention. In
this regard, when the fee currently payable for a
class I restoration is considered (£7.50) and when it
is considered that the survival of this restoration is
58% at ten years, the cost to the patient per year
must be considered excellent value. This may be
considered to compare favourably with the cost
effectiveness of other treatment systems for which
details are available.18
Conclusions
Small amalgam restorations, such as class I, have
longer survival times at ten years before re-
intervention (58%) than large amalgam restorations
815
such as MOD (43%). Composite and glass ionomer
restorations perform less well than amalgam
restorations. Pin placement and root filling reduce
the survival time of restorations.
References
1. Robinson AD. The life of a filling. British Dental Journal
1971;130:206–8.
2. Jokstad A, Bayne S, Blunck U, Tyas MJ, Wilson NHF. Quality
of dental restorations. International Dental Journal 2001;
51(3):117–58.
3. Lucarotti PSK, Holder RL, Burke FJT. Analysis of an
administrative database of half a million restorations over
11 years. Journal of Dentisty, in press; doi:
10.1016/j.dent.2005.06.011.
4. Collett D. Modelling survival data in medical research.
London: Chapman and Hall; 1994.
5. Elderton RJ. The prevalence of failure of restorations: a
literature review. Journal of Dentistry 1976;4(5):207–10.
6. Lucarotti PSK. The Life Expectancy of Dental Restorations
placed within the General Dental Services in England and
Wales. PhD Thesis. University of Birmingham. 2004.
7. Sheldon T, Treasure E. Dental restoration: what type of
filling? Effective Health Care 1999;5(2):1–11.
8. Feilzer AJ, De Gee AJ, Davidson CL. Setting stress in
composite resin in relation to configuration of the restor-
ation. Journal of Dental Research 1987;(66):1636–9.
9. Lugassy AA, Moffa JP, Hozumi Y. Influence of pins upon some
physical properties of composite resins. Journal of Prosthe-
tic Dentistry 1972;(28):613–8.
10. Dilts WE, Podshadley ASG, Neiman A. Effect of pins on some
physical characteristics of composite resins. Journal of the
American Dental Association 1973;(87):595–9.
11. Dilts WE, Welk DA, Laswell HR, George L. Crazing of tooth
structure associated with placement of pins for amalgam
restorations. Journal of the American Dental Association
1970;(81):387–90.
12. Pamejier CH, Stallard RE. Effect of self-threading pins.
Journal of the American Dental Association 1972;(85):895–8.
13. Anil N, Key F. Determining the temperature rise in the pulp
and at the root surface during pin hole drilling and placing of
dentine pins. European Journal of Prosthodontics and
Restorative Dentistry 1998;(6):111–4.
14. Buonocore MG. A simple method of increasing the adhesion
of acrylic filling materials to enamel surfaces. Journal of
Dental Research 1955;(34):849–54.
15. DPB Digest of statistics, Eastbourne: Dental Practice Board;
2002.
16. DPB Digest of statistics, Eastbourne: Dental Practice Board;
2001.
17. DPB Digest of statistics, Eastbourne: Dental Practice Board,
2000
18. Bogacki RE, Hunt RJ, del Aguila M, Smith WR. Survival
analysis of posterior restorations using an insurance claims
database. Operative Dentistry 2002;27:488–92.