Journal of Oral Rehabilitation 2009 36; 451–461

Coronal tooth structure in root-treated teeth prepared for

complete and partial coverage restorations
F. MURPHY, A. MCDONALD, A. PETRIE, G. PALMER & D. SETCHELL UCL Eastman Dental
Institute, Eastman Dental Hospital, 256 Gray’s Inn Road, London WC1X 8LD, UK

SUMMARY This in vivo study assessed the remaining three dies were of tooth structure prior to core
coronal tooth structure in teeth prepared for com- placement. All dies (n = 31) were scanned using a
plete and partial coverage restorations using 3D- laser profilometer and the volume of remaining
scanning and a Tooth Restorability Index (TRI). The tooth structure calculated. Four observers scored 31
cuspal coverage preparation designs selected by 10 dies using the TRI. The percentage loss of coronal
postgraduate dentists and 10 general dental practi- tooth volume following a complete instead of a
tioners were recorded in a questionnaire. Eighteen partial coverage preparation varied from 3Æ29% to
patients had molar root treatment completed at the 45Æ23% and the mean TRI fell from 10Æ7 to 7Æ5 units.
Eastman Dental Hospital and were prescribed a There was a strong correlation between mean TRI
coronal-radicular amalgam core and cast restora- and scanned volume of tooth structure (P = 0Æ013).
tion. Each tooth was prepared in vivo by one oper- Over 50% of the dentists altered their initial choice
ator for a cast restoration. Two clinical impressions of restoration design from complete to partial cov-
were made to produce two dies: one of remaining erage. Complete coverage preparations removed
coronal tooth structure before crown preparation more tooth structure than partial coverage.
and a second die of coronal tooth structure in vivo KEYWORDS: coronal, tooth structure, root-treated
after crown preparation. For teeth prepared for teeth, complete and partial coverage preparations,
partial coverage in vivo (n = 13), a third die was tooth restorability index
prepared in vitro representing remaining tooth
structure after complete coverage preparation. The Accepted for publication 8 March 2009

tooth structure. Recommendations have been made

Introduction
with regard to an acceptable minimum height of
Although not evidence-based, it is generally agreed that remaining coronal tooth structure (3–7), root dentine
the amount of residual tooth structure is important to thickness (8) and length of ferrules (9–11). However,
longevity of a tooth and its restoration (1). Root-treated the recommendations of these studies are based on
teeth frequently have a limited amount of remaining in vitro research, which is limited with respect to
tooth structure. It is recommended that root-treated simulating the elasticity of the periodontal ligament,
posterior teeth should be provided with cuspal coverage bone and tooth structure and the reproduction of the
restorations (2), although it is acknowledged that clinical environment including functional forces.
preparation of an already compromised tooth for this A partial coverage design of restoration is likely to be
type of restoration will further reduce the remaining more conservative of tooth structure than a complete
veneer restoration. Very few studies have been pub-
lished addressing the advantages and disadvantages of
Submitted in partial fulfilment of the degree of Master of Dental
Science in Conservative Dentistry, UCL Eastman Dental Institute, partial coverage restorations, other than examining the
University College London.

ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd doi: 10.1111/j.1365-2842.2009.01952.x
452 F . M U R P H Y et al.

difference in resistance and retention form of various of Intermediate Restorative Material (IRM)* was
preparation designs (12, 13). Two studies compared the placed over the canal orifices and condensed into the
differences in the amount of tooth structure removal pulp chamber. The amalgam was left to set for 15 min
using various preparation designs on anterior Typodont prior to refinement of the preparation. All amalgam
resin teeth (14) and extracted human teeth (15). The cores and crown preparations were carried out under
extent to which tooth structure is preserved by using a rubber dam isolation.
partial coverage instead of a complete coverage design A second sectional impression (Impression 2) was
in posterior human teeth, however, remains unknown. taken of each tooth after crown preparation, without
The aims and objectives of this study were, first, to the core in place. These impressions were later used to
measure the remaining coronal tooth structure in root- assess the amount of remaining coronal tooth structure.
treated teeth prepared for both complete and partial A provisional restoration was made in Integrity)* by
coverage restorations; second, to compare the amount lining the internal aspect of a putty (President) matrix
of remaining coronal tooth structure when the same made from a diagnostic wax-up of the tooth. Retraction
teeth have been prepared for complete and partial cord (Ultrapak)‡ was used where necessary and a full
coverage restorations; and third, to examine the opin- arch impression (using President light and heavy body
ion of dentists with regard to the design of cuspal polyvinylsiloxane impression material) was taken in a
coverage restorations for root-treated teeth. Polytray stock tray.* The impression was poured in
Type IV Silky Rock Stone.§ The crowns (complete or
partial veneer gold crowns) were waxed-up, invested,
Materials and methods
cast and finished. On the final clinical visit, the crown
Eighteen human molar teeth were used in this study. was examined on the tooth preparation for proximal
The criteria for inclusion in this study were: (i) that the contours, occlusal contacts and marginal fit prior to
teeth should be molar teeth root- treated at the cementation with zinc phosphate cement.
Eastman Dental Hospital by endodontic post-graduates; The following method was used to produce a die and
(ii) that they should be suitable for an amalgam coronal allow the amount of remaining coronal dentine to be
radicular core; (iii) and that patients were happy to analysed. Impressions 1 and 2 were poured up in Type
accept a gold cuspal coverage restoration. IV Silky Rock Stone. Impression 1 produced a die
Once endodontic treatment had been completed and replicating the remaining tooth structure prior to core
before a core had been placed, a sectional impression placement and crown preparation. Impression 2 pro-
(Impression 1) was taken in a stock tray using duced a second die replicating remaining coronal tooth
President light and heavy body polyvinyl siloxane structure in vivo after crown preparation but prior to
material (Coltene, W Sussex, UK). This impression was core placement. For those teeth that were prepared for
poured to produce a die of remaining coronal tooth partial coverage restorations in vivo, a second die of
structure prior to tooth preparation. remaining tooth structure prior to crown preparation
Each tooth was prepared in vivo by one operator for a was poured in Type IV Silky Rock Stone. With the use
cast restoration. Thirteen were prepared for a partial of depth grooves, this die was prepared by one operator
coverage restoration and five for a complete coverage in the laboratory for a complete coverage restoration.
restoration. The preparations were completed without Thus 31 dies representing remaining coronal tooth
the core in place and with an occlusal reduction of structure were fabricated.
1Æ5 mm on the functional cusps and 1 mm elsewhere. Each die was carefully trimmed under a microscope
The measurement of the reduction was determined by at 20· magnification. The dies representing the com-
means of depth grooves. plete coverage restorations were trimmed to the crown
One of two types of amalgam core was placed, margin, which followed the gingival margin. Dies
depending on the depth of the pulp chamber. Where prepared for partial coverage restorations were trimmed
the pulp chamber was less than 4 mm in height, a to the level of the gingival margin.
conventional Nayyar core was placed by condensing
amalgam 2–4 mm into the root canals. Where the pulp *Dentsply, Surrey, UK.
chamber was greater than 4 mm in height, a modified ‡
UltraDent Products Inc., South Jordan, UT, USA.
§
Nayyar core was placed. For this type of core, 1–2 mm Whip Mix Corp., Kentucky, OH, USA.

ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd

 FULL ⁄ PARTIAL COVERAGE PREPARATIONS-REMAINING TOOTH 453

Fig. 1. Three-dimensional profile image

created by the laser profilometer of tooth
10 prepared for complete coverage
restoration.

All the 31 dies were scanned by the Proscan 1000 Each point measured by the profilometer was repre-
laser scanning profilometer¶ to map their three-dimen- sented by a numerical value, which was used to create
sional profile. Mapping consisted of scanning each die a three-dimensional image of the die. This image was
at 0Æ05- or 0Æ07-mm intervals. The decision regarding transferred to Microsoft Excel spreadsheets (Microsoft
the choice of scanning interval was dependent upon the Corp., Redmond, WA, USA) where line graphs repre-
buccal-lingual width of the die along the finish line of senting approximately 20 (depending on the size of the
the preparation in a bucco-lingual and mesio-distal tooth) cross-sections were created. The opposing mar-
plane. The KL135A laser used in Proscan 1000 provided gin points on each line graph were joined with an
a vertical resolution of 1 lm. The maximum number of imaginary straight line, thus creating a virtual repre-
steps chosen for each profile was 250. A three-dimen- sentation of a cross section of the remaining coronal
sional image of a tooth prepared for a complete tooth structure at that point of the die. The areas
coverage restoration as scanned by the laser profilome- contained within the cross-section were mathemati-
ter is shown in Fig. 1. cally combined, based on the distance between the
As undercuts cannot be read by a laser profilometer, cross-sections, to determine the volume of the die
they were blocked out with a light body polyvinyl representing coronal tooth structure. Any impression
siloxane impression material (President). A definitive material used to block out undercuts was removed
finish line was needed to allow the profilometer to from the relevant die, weighed and its volume deter-
assess the start and end points of the measurement. The mined. This volume was then subtracted from the
finish line in partial coverage preparations did not volume of remaining coronal tooth structure computed
generally coincide with the gingival margin. A false from the cross-sections, and thus the actual volume of
finish line was created in teeth prepared for partial remaining coronal tooth structure was determined for
coverage restorations at the level of the gingival margin each die.
by blocking undercuts with President light body To ascertain whether coronal dentine had any struc-
impression material. This was used as a reference line tural value to resistance and retention, the Tooth
to which the profilometer could read, thus permitting Restorability Index (TRI) devised by McDonald and
the operator to compute the volume of remaining tooth Setchell (16) was used. This provided a numerical value
structure coronal to the gingival margin. as to whether a tooth or sections thereof were deemed
to be predictably restorable. Each tooth was notionally
¶
Scantron Industrial Products Limited, Taunton, UK. divided into sextants for this purpose.

ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd

454 F . M U R P H Y et al.

Fig. 2. Correlation between volume

of remaining tooth structure and
mean total Tooth Restorability Index.

Four clinicians were asked to analyse each die, using (ICC) was used to assess the total variation in the
their clinical judgement as they would in treating a overall analysis.
patient. A score of 0 to 3 was allocated to each sextant
according to the TRI in which 0 means no axial wall of
Results
dentine; 1 means inadequate dentine; 2 means ques-
tionable dentine; and 3 means adequate dentine. Thus, Table 1 shows the mean total TRI of each tooth as
a maximum score of 3 could be given to each sextant determined by four observers and the volume of
and a maximum score of 18 could be given to each remaining coronal tooth structure as determined from
tooth. The TRI of each die in this study comprised the the laser profilometer and subsequent calculations. A
mean score given by the four assessors. Spearman rank correlation of these results is shown in
A questionnaire was provided to 10 general dental Fig. 2. Table 2 shows the full set of TRI data in relation
practitioners and 10 dentists undergoing formal post- to one die (Tooth 10). The Spearman rank correlation
graduate training at the Eastman Dental Hospital.
Dentists were initially presented with dies of remaining Table 1. Mean total tooth restorability index scores and total
tooth structure prior to preparation (die 1). They were volume of tooth structure determined from laser profilometer
asked to record which of the following preparations
they considered as the most appropriate for each tooth:
complete veneer gold crown, complete veneer porce-
lain fused to metal crown, gold onlay, three-quarter
gold crown; or seven-eighths gold crown.
When this was completed, they were shown 10 dies
of teeth scheduled for complete coverage and subse-
quently for complete and partial coverage restorations.
In cases where a complete coverage design had been
indicated as being the most appropriate by the dentist,
they were asked to assess the dies prepared for both
complete and partial coverage and to indicate whether
they would reconsider their original decision in light of
the preparations they had examined. A Spearman rank
correlation was used to correlate the mean TRI as
determined by the four observers and the volume of
remaining coronal dentine as calculated from the laser
profilometer. Lin’s concordance correlation coefficient
(17) was used to compare the scores from each set of
two observers. The intra-class correlation coefficient

ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd

 FULL ⁄ PARTIAL COVERAGE PREPARATIONS-REMAINING TOOTH 455

Table 2. TRI Score for each sextant of tooth 10 (Complete and (a)
partial Coverage Preparations) as determined by four observers

(b)

coefficient of 0Æ67 (P = 0Æ013) indicated that there was a

significant correlation between the mean total TRI and
the volume of remaining coronal tooth structure.

Comparison of remaining coronal tooth structure in complete

and partial preparations

The dies of three teeth, prior to placement of the core

restoration and after preparation for either complete or
partial coverage restorations are represented in the
photographs in Figs 3a–c and 4a–c. The Sign test was (c)
applied and showed strong evidence that teeth prepared
for partial coverage restorations had, on average, a
greater volume of remaining coronal tooth structure
than the same teeth prepared for complete coverage
restorations (P < 0Æ001). The median of the differences
between complete and partial coverage was 51Æ4 with a
95% confidence interval of 31Æ6–74Æ9.
The second column of Table 3 represents the per-
centage of tooth volume loss that occurred by providing
a complete instead of a partial coverage restoration.
This percentage loss varied from 3Æ29% to 45Æ23%. The
third column of Table 3 represents the percentage
reduction of TRI, which occurred when preparations
for complete coverage restorations were completed on Fig. 3. (a) Image of tooth 10 before preparation for partial and
complete coverage castings. (b) Image of tooth 10 after prepara-
the same teeth which varied from )4% to 43%. The
tion for complete coverage casting. (c) Image of tooth 10 after
fourth column of Table 3 represents the percentage preparation for partial coverage casting.
reduction of sextants given a score of 3 by all observers,
which varied from )18% to 100%.

ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd

456 F . M U R P H Y et al.

(a) Table 3. Comparison of percentage reduction in tooth volume

and total Tooth Restorability Index (TRI) following preparation for
complete and partial coverage restoration

Percentage of Percentage of Percentage decrease

tooth structure reduction in of sextants with a
reduction from TRI from score of 3 in TRI
Tooth partial to partial to from partial to
number complete complete complete

1. 35 )4 )18
2. 3 7 50
3. 45 30 100
4. 9 34 60
10. 27 42 89
11. 13 8 22
12. 26 28 100
13. 18 28 75
(b) 14. 30 10 42
15. 26 25 50
16. 19 38 56
17. 23 18 41
18. 45 43 71
Mean 24Æ5 23Æ6 41

Table 4. Lin’s concordance correlation coefficient, British Stan-

dards Reproducibility coefficient (BSRC) and bias when compar-
ing the total Tooth Restorability Index (TRI) between each set of
two observers

Bias (mean
difference)
Comparison Lin (95% CI) BSRC and P-value

1 vs. 2 0Æ85 (0Æ76 to 0Æ95) 2Æ78 0Æ61 (P = 0Æ10)

1 vs. 3 0Æ92 (0Æ87 to 0Æ98) 2Æ34 )0Æ32 (P = 0Æ22)
1 vs. 4 0Æ92 (0Æ86 to 0Æ97) 2Æ35 )0Æ29 (P = 0Æ27)
(c) 2 vs. 3 0Æ84 (0Æ74 to 0Æ94) 2Æ76 )0Æ94 (P = 0Æ01)
2 vs. 4 0Æ87 (0Æ79 to 0Æ95) 2Æ54 )0Æ90 (P = 0Æ005)
3 vs. 4 0Æ92 (0Æ87 to 0Æ98) 2Æ32 0Æ03 (P = 0Æ90)

The value of Lin’s concordance correlation coefficient

assessing the agreement in total TRI between every pair
of observers ranged from 0Æ84 to 0Æ92 with the lowest
lower limit and the greatest upper limit of the 95%
confidence intervals being 0Æ74 and 0Æ98 respectively
(Table 4). This shows good concordance between observ-
ers. Each of the Bland and Altman diagrams (18) in
which the difference between each set of two observers’
Fig. 4. (a) Images of tooth 14 before preparation for partial and total TRI was plotted against their mean, showed a
complete coverage castings. (b) Images of tooth 14 after prepara- random distribution of points indicating that a single
tion for complete coverage casting. (c) Images of tooth 14 after measure of reproducibility is satisfactory [i.e. British
preparation for partial coverage casting. Standards Reproducibility coefficient (BSRC)] for each

ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd

 FULL ⁄ PARTIAL COVERAGE PREPARATIONS-REMAINING TOOTH 457

comparison. The BSRC represents the maximum likely Table 5. Number of times general dental practitioners and grad-
difference between the total TRI scores for any two uates initially chose complete coverage and changed from a
complete to a partial coverage design having been shown each
observers and this ranged from 2Æ32 to 2Æ38. When
tooth prepared for both designs
examining the bias, (i.e. whether there was a systematic
difference between two observers, assessed by determin- Initial choice Changed
ing whether the mean of their differences was signifi- of complete decision to
cantly different from zero), two P-values were significant coverage: teeth choose partial
and showed that Observer 2 tended to underestimate the 1–4, 10–18 design
total TRI score compared with Observers 3 and 4. General dental 44 22 (50Æ00%)
In the overall analysis assessing the agreement practitioner
between all four observers, ICC, a reliability coefficient Postgraduates 32 19 (59Æ38%)
comparable to Lin’s concordance correlation coefficient,
was 0Æ89. This implies that 89% of the total variation in
the total TRI scores was attributable to the differences undergoing re-root treatment may have been more
between teeth. In addition, the maximum likely differ- broken down than those undergoing root treatment for
ence between two measurements with different observ- the first time. The sample size reflected the number of
ers was 3Æ36 of a possible total score of 18. re-root treated teeth with sufficient tooth structure to
The total TRI score for each die was averaged over the allow for a partial coverage design.
four observers and this mean score was used as an It is widely agreed that root-treated molar teeth
estimate of the true score for that die. The difference should be provided with cuspal coverage to prevent
between a particular observer’s score and the ‘true’ fracture (2). All the 18 patients were provided with a
score for each die was calculated and the standard veneer gold crown as it was considered conservative of
deviation of these differences was plotted against their tooth structure when compared with porcelain fused to
mean. The resulting diagram indicated that Observer 2 metal or all-ceramic options. The crown preparation for
had slightly more variation in his ⁄ her scores than any each tooth was carried out under rubber dam isolation
of the other observers (suggesting increased error), and to minimize recontamination of the gutta percha.
Observer 2 also had scores, which tended to be further However, the rubber dam did not facilitate the use of
away from the ‘true’ mean than those of the other a reduction matrix and reduction grooves were used as
observers (suggesting bias). However, further analysis a tooth reduction guide.
showed that, relative to all the other observers, neither Two different types of core design were used in the
the increased error nor the bias were more than might prepared teeth. In more recent years, the literature has
be expected by chance. questioned the need to extend amalgam into the root
canals (19, 20) as described by Nayyar et al. (21). The
choice of core in this study was made by the
General dental practitioner and graduate assessment of dies
endodontic graduate and was dependent on the height
Table 5 summarizes the decisions made by 10 general of the pulp chamber. Thirty-nine per cent of the teeth
dental practitioners and 10 dentists undertaking formal were restored with a Nayyar core and the remaining
post-graduate training. Having been shown the dies 61% were restored with a modified Nayyar core
after preparation for both complete and partial cover- whereby the amalgam was not extended into the root
age, more than 50% of dentists who had initially canals and the coronal aspect of the root canals were
chosen a complete coverage restoration changed their sealed with IRM.* The coronal preparation was the
initial assessment to a partial coverage preparation. same for both the traditional and modified Nayyar
core. Therefore, the choice of core should not affect
the remaining coronal tooth structure. The type of
Discussion
core used may have an impact on the longevity of the
All the 18 patients included in this study were referred restoration. This can only be evaluated when moni-
to the Eastman Dental Hospital for specialist care. tored prospectively.
Seventy-two per cent of the sample involved molars Each tooth that was prepared for partial coverage
requiring non-surgical re-root canal treatment. Teeth in vivo had an additional die made of the unprepared

ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd

458 F . M U R P H Y et al.

tooth. This die allowed a complete coverage preparation The results showed that there was considerable
to be carried out in vitro. It would not have been variation in the volume of remaining tooth structure
appropriate to alter the partial coverage dies to com- present after preparation for coronal coverage. A tooth
plete coverage as, in some instances, grooves were of greater volume does not necessarily imply that it has
present in the partial coverage preparations, which a better long-term prognosis than a tooth with a lesser
would not be used in a complete coverage design. The volume. Teeth numbers 1, 14 and 18 showed the
method of obtaining the remaining tooth structure after greatest percentage loss of tooth structure when
preparation differed from our previous study [Bandlish prepared for complete coverage compared with partial
et al. (22)]. In that study, teeth were prepared with the coverage. This is likely to be related to the bulbous
core restoration in place and a series of interlocking buccal contour of these teeth. Considerable buccal
special trays and impressions were used to produce a tooth structure had to be removed to ensure an ideal
cast of the remaining coronal tooth structure after taper between buccal and lingual ⁄ palatal walls. Thus, a
preparation. In this study, the teeth were prepared partial coverage design option should be strongly
carefully without the core in place. The methodology considered in teeth with a bulbous contour.
was altered to simplify the process as interlocking trays In general, there was a good correlation between the
were then unnecessary. The preparation of teeth was volume of remaining coronal tooth structure and the
rendered slightly more difficult with this method as TRI. This correlation was weaker when the height-
thin sections of tooth structure were not supported by width ratio of tooth structure became unfavourable.
the core during preparation and making of impressions. The clinical significance of the volume calculations
The cores were placed prior to the final impression. and TRI is speculative at present. The real value of these
Bandlish et al. (22) described a method of blocking calculations in the restorability of a tooth and the
out undercuts with impression material. This method design of extra-coronal restorations cannot be evalu-
was adopted for external undercuts in partial coverage ated until long-term prospective data are available. The
preparations and was adapted to create a false finish TRI provides a structured classification to aid clinical
line at the level of the gingival margin. Although decision-making. A recent study (23) described classi-
coronal tooth structure ends 2–3 mm below the gingi- fication of the height and thickness of remaining walls
val margin, the gingival margin was chosen as a and size of the pulp chamber in root-treated teeth but
reference point in all partial coverage restorations to was relatively complicated to use in comparison with
which each die was trimmed. Coronal tooth structure, the TRI.
therefore, referred to supra-gingival tooth structure. It was interesting to note that the TRI scores varied
Dies of preparations for complete coverage restorations from observer to observer. Lin’s concordance correla-
were trimmed to their finish lines which also followed tion coefficient showed good agreement between pairs
the gingival margin. of observers. The BSRC indicated that the maximum
Scanning of the dies was performed at 0Æ05- or 0Æ07- likely difference between the TRI scores for each set of
mm intervals and was dependent on the bucco-lingual two observers ranged from 2Æ32 to 2Æ78 of a possible 18
width of the tooth. Where a 0Æ05-mm interval did not per tooth. This reflects the subjective clinical decision-
allow for a complete profile to be made of the tooth, an making, which as yet is not well documented. Observer
interval between steps of 0Æ07 mm was used. This 2 had slightly more variation in his ⁄ her scores and
resulted in a minor variation in the method of calcu- scores that tended to be further away from the overall
lating the amount of remaining tooth volume. mean than the other observers, but these were not
The volume calculation of residual coronal tooth significant effects.
structure does not give an indication of the strategic Inter-observer agreement of sextant scores occurred
value of the remaining tooth structure. The TRI was most often where the observers considered that there
devised (16) with the aim of scoring each sextant of a was no adequate tooth structure in two-thirds or
tooth as having inadequate, questionable or adequate more of the sextant (score of 0) or where they
remaining tooth structure. Although the mean TRI is considered there to be adequate coronal tooth struc-
certainly an indication of the restorability of a tooth, it ture (score of 3).
is likely to be less important than the number of Seventy-four per cent of the 31 dies had their two
sextants with a score of 3. lowest TRI sextant scores at the mesial and distal sites

ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd

 FULL ⁄ PARTIAL COVERAGE PREPARATIONS-REMAINING TOOTH 459

and agrees with previously published findings (22). This 1Æ5 to 2Æ5 mm (9, 10). However, these recommendations
reflects the high incidence of inter-proximal carious are based on in vitro models, generally on anterior teeth
lesions. and are often based on clinical opinions. None of these
With the exception of one tooth, all teeth had a studies attempted to evaluate either the strategic impor-
higher mean TRI score when prepared for partial tance of remaining tooth structure or the height to
coverage in comparison with complete coverage. When thickness ratio of remaining coronal tooth structure. In
prepared for partial coverage, tooth number one, had a addition, the height was assessed without regard to
mean TRI score, which was 0Æ5 units lower than when thickness of remaining coronal tooth structure. Whilst
prepared for complete coverage. Two of the observers height or thickness alone may be important factors in
gave a lower score for the mesio-buccal sextants in the longevity of a restoration, it is probably more useful to
partial coverage design. This was due to a groove that consider them in combination. The TRI and the volume
was placed at this site for the three-quarter crown calculations in this study attempt to assess both thickness
preparation. Two observers considered the tooth struc- to height ratio (by means of the TRI) and volume.
ture of this sextant as ‘questionable’ when a groove was One in vitro study considered volume (20). They
placed compared with ‘definitely adequate’ in the assessed the strength of core build-up materials in root-
complete coverage preparation. treated teeth including an assessment of the volume of
The findings of the questionnaire showed that there the pulp chamber of each tooth. The volume assessed in
was a slight trend for post-graduate dentists to choose this study was that of the pulp chamber rather than the
more partial coverage preparations than general dental remaining tooth structure. They found no significant
practitioners in the first instance. This choice may have correlation between pulp chamber volume and load on
been influenced by the student–teacher relationship failure of the core materials used. It is generally
and may have had a potential confounding effect. Fifty considered that it is the remaining coronal tooth
per cent of the general practitioners and 59% of the structure that confers strength and improved longevity
post-graduate dentists who originally chose complete to a tooth and its restoration.
coverage preparations changed their choice of design to Whether it is appropriate to place a plastic core or a cast
partial coverage when they were shown dies prepared post and core is a clinical decision commonly made
for both complete and partial coverage. This empha- without an evidence base. Nayyar et al. (21) in their first
sized the difficulty that dentists have when visualizing description of the amalgam coronal-radicular core did
the residual tooth structure after preparation. At the not give specific indications of when a Nayyar core should
time of formatting the questionnaire for dentists, it was be placed. Their only recommendations were that the
decided that the treatment options should be specified pulp chamber should be of adequate height and width
to reduce the number of variables. With multiple and that there should be adequate dentine thickness.
variables and a small sample size, it may have been Few studies have been carried out to assess the
difficult to obtain useful information. The design of the longevity of coronal-radicular amalgam cores. In their
questionnaire may have biased the results limiting the first report on the Nayyar core, Nayyar et al. (21)
range of coronal coverage restorations from which to reported a 100% success rate with 400 samples after
choose. It may have been useful to ask dentists to carry 4 years. A retrospective study (24) reported a median
out the preparations they considered to be the most survival rate of <10 years for the 39 coronal-radicular
appropriate on each die. This may have provided a amalgam restorations examined. Long-term monitoring
more realistic outline of the suggested preparation. of the cores in this study will provide further evidence
There is general agreement in the literature that of the longevity of these restorations.
conservation of tooth structure is one of the critical Very little literature has been published addressing
aspects to increasing the longevity of a tooth and its the advantages and disadvantages of partial coverage
restoration. It has been recommended (8) that a 2-mm restorations. Edelhoff and Sorensen (14) carried out an
thickness of dentine would improve resistance to the in vitro study to determine and compare the amount of
fracture of root-treated teeth, while other authors rec- tooth structure removed for laminate veneer and resin-
ommended approximately 2 mm coronal tooth structure bonded prostheses preparations in comparison with
as an ideal height (3–7). Lengths of ferrules have also conventional complete coverage restorations. They
been studied and their recommended lengths vary from found that the veneers and resin-bonded prostheses

ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd

460 F . M U R P H Y et al.

preparations required 25% to 50% less reduction than 4. Sorensen JA, Engleman MJ. Ferrule design and fracture
that required by complete coverage restorations. resistance of endodontically treated teeth. J Prosthet Dent.
1990;65:529–536.
Recording of the TRI and volume of remaining tooth
5. Assif D, Bitenski A, Pilo R, Oren E. Effect of post design on
structure will also allow future assessment of longevity resistance to fracture of endodontically treated teeth with
of each restoration in relation to these factors. The complete crowns. J Prosthet Dent. 1993;69:36–40.
clinical significance of these findings will not be 6. Ferrari M, Vichi A, Mannocci F, Masen PN. Retrospective
established until prospective data are available. study of the clinical performance of fibre posts. Am J Dent.
2000;13 (Spec No): 9B–13B.
7. Bolhuis HPB, De Gee AJ, Feilzer AJ, Davidson CL. Fracture
Conclusions strength of different core build-up designs. Am J Dent.
2001;415:286–290.
This study compared the remaining coronal tooth 8. Tjan AH, Whang SB. Resistance to root fracture of dowel
structure in root-treated teeth prepared for both com- channels with various thicknesses of buccal dentine walls. J
plete and partial coverage restorations by assessing the Prosthet Dent. 1985;53:496–500.
9. Libman WJ, Nicholls JI. Load fatigue of teeth restored with
relative volume of remaining tooth structure and the
cast post and core and complete crowns. Int J Prosthodont.
restorability of these teeth when prepared in vivo for 1995;8:155–161.
partial coverage and in vitro for complete coverage 10. Isidor F, Brondum K, Ravnholt G. The influence of post length
restorations. In addition, it examined the opinions of and crown ferrule length on the resistance to cyclic loading of
dentists with regard to the design of cuspal coverage bovine teeth with prefabricated titanium posts. Int J Prosth-
odont. 1999;12:78–82.
restorations for root-treated teeth.
11. Creugers NH, Mentink GM, Fokkinga A, Kreulen CM. 5-
Within the limitations of this clinical study, it can be
Year follow-up of a prospective clinical study on vari-
concluded that: ous types of core restorations. Int J Prosthodont. 2005;
1. A technique was devised for measuring the volume 18:34–39.
of teeth prepared for partial coverage restorations. The 12. Potts RG, Shillingburg HT, Duncansen MG. Retention and
volume of remaining coronal tooth structure after resistance of preparation for cast restorations. J Prosthet Dent.
1980;43:303.
complete coverage preparation varied from 112Æ23
13. Kishimoto M, Shillingburg HT Jr, Duncansen MG Jr. Influ-
mm3 to 296Æ60 mm3 compared with 157Æ54 mm3 to ence of preparation features on retention and resistance. Part I
386Æ34 mm3 for the same teeth prepared for partial MOD onlays. J Prosthet Dent. 1983;49:35.
coverage restorations. 14. Edelhoff D, Sorensen JA. Tooth structure removal associated
2. The percentage loss of volume of coronal tooth with various preparation designs for anterior teeth. J Prosthet
Dent. 2002;87:503–509.
structure that occurred by providing a complete instead
15. Hussain S, McDonald A, Moles D. An in vitro study
of a partial coverage restoration varied from 3Æ29% to
investigating the mass of tooth structure removed
45Æ23%. following restorative procedures. J Prosthet Dent. 2007;
3. The mean TRI for teeth prepared for complete coverage 98:260–269.
restorations was 7Æ5 compared with 10Æ7 for the same 16. McDonald AV, Setchell DJ. Developing a tooth restorability
teeth prepared for partial coverage restorations. index. Dental Update. 2005;32:343–348.
17. Lin L. A concordance correlation coefficient to evaluate
4. More than 50% of the 20 dentists who participated in
reproducibility. Biometrics. 1989;45:255–268.
the study altered their initial choice of complete 18. Bland JM, Altman DG. Statistical methods for assessing
coverage restoration to partial coverage when they agreement between two methods of clinical measurement.
were shown dies of the teeth prepared for both designs. Lancet. 1986;1:307–310.
19. Kane JJ, Burgess JO, Summitt JB. Fracture resistance of
amalgam coronal-radicular restorations. J Prosthet Dent.
References 1990;63:607–613.
20. Foley J, Saunders E, Saunders WP. Strength of core build up
1. Robbins JW. Restoration of the endodontically treated tooth.
materials in endodontically treated teeth. Am J Dent. 1997;
Dent Clin North Am. 2002;46:367–384.
10:166–172.
2. Sorensen JA, Martinoff JT. Intracoronal reinforcement and
21. Nayyar A, Walton R, Leonard LA. An amalgam coronal-
coronal coverage: a study of endodontically treated teeth.
radicular dowel and core technique for endodontically treated
J Prosthet Dent. 1984;51:780–784.
posterior teeth. J Prosthet Dent. 1980;43:511–514.
3. Hoag PE, Dwyer TG. A comparative evaluation of three post
and core techniques. J Prosthet Dent. 1982;47:177–181.

ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd

 FULL ⁄ PARTIAL COVERAGE PREPARATIONS-REMAINING TOOTH 461

22. Bandlish RB, McDonald VA, Setchell DJ. Assessment of the 24. Nanayakkara L. Retrospective analysis of factors influencing
amount of remaining coronal dentine in root-filled teeth. the longevity of post crowns. MSc Thesis, University of
J Dent. 2006;34:699–708. London; 1998.
23. Naumann M, Blankenstein F, Barthel CR. A new approach to
define defect extensions of endodontically treated teeth: Correspondence: Ailbhe McDonald, UCL Eastman Dental Institute,
inter- and intra-examiner reliability. J Oral Rehabil. 2006; 256 Gray’s Inn Road, London WC1X 8LD, UK.
33:52–58. E-mail:A.Mcdonald@eastman.ucl.ac.uk

ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd