David I.

Lamb, BDS, MDS, fDS

Senior Lecturer and Honorary Consultant
Department of Restorative Dentistry
Bryan Ellis, BSc. MSc
Senior Lecturer

Patient-Assessed Security
Changes When Replacing
Mandibular Complete Dentures

Department of Engineering Materials

University of Sheffield
Sheffield, England

Twenty edentulous patients had unsatisfactory dentures replaced in a student

ciinic, using conventional techniques, under consultant supervision. Three
visual analogue scale records of self-assessed mandibular denture security
were made while the dentures were fabricated, dnd two were made
following placement of the new dentures to measure any changes occurring.
The method appeared reliable, and significant improvements in security were
recorded. The percent increase in security and the coefficient of variation of
the preplacement visual anaiogue scale scores were each inversely
proportional to the preplacement visual analogue scale scores. This
relationship may be useful in the identification of patients with special
problems. Int i Prosthodont
1995;8:135-I41.

dentulous individuals suffer both disability and

handicap,' and their quality of life is markedly
affected by the amount of satisfaction they have
with their complete dentures.' Unfortunately,
patient satisfaction is difficult to measure. It is
reported that judgments of denture quality by clinicians correlate poorly with the patients' own judgments,'" possibly because the important qualities
determining patient acceptance, namely pain-free
function, security, and good appearance,' are
sometimes mutually incompatible, and clinicians
can find it difficult to make an appropriately
weighted treatment plan.
Patient satisfaction may be assessed on a graduated scale, sometimes with as few as two divisions
(ie, satisfactory and unsatisfactory),* but often with
as many as five." Such methods are simple to use
but do not discriminate well when there are few
divisions, and subjects tend to choose the central
section if scales with an odd number of divisions are
used. Alternatively, a visual analogue scale (VAS)
can be used. Such scales are often used to measure
feelings such as depression' and have been used in
dentistry to measure postoperative pain.''
Recently, in a preliminary study, VAS records

have been used to quantify denture-security

changes following relining.'" The purpose of this
study was to (1) use a visual analogue scale to
investigate changes in self-assessed mandibular
denture security that occurred when complete dentures were replaced under consultant supervision,
and 2) test the method's reliability within the ethical and other applicable constraints (Appendix]
and establish a measure that could be used as a
future basis for comparing subjective response and
clinical method.
Matertals and Methods
Selection of Subjects

Reprint requests: Or D.I. Lamb, Department ol Restorative

Dentistry, School of Clinical Dentistry. University of Sheffield,
Ctaremant CrefcenI, Shcllield SIO 2TA, England.

Number 2, J995

1 35

Twenty-two subjects were randomly selected

from patients attending the undergraduate clinic at
Sheffield School of Clinical Dentistry for replacement of their unsatisfactory complete dentures.
Complaints of pain were common and, whenever
possible, were relieved by the necessary adjustment prior to the initiation ofthe study.
Two subjects were eliminated during the study
when it became apparent that they did not understand the method of measuring security. The
remainder were experienced denture wearers who
had been provided with their first set of complete
dentures at least 5 years earlier. Sixteen subjects

The International Journal of Proslhodontics

^ When ReplacinR Miinrlihubr i3enturi

Denture firm
and secure
at all times

Denture loose
and insecure
at all times

Fig 1 Visual analogue scale as used in the study (not to scale). Ten-centimelre line represents spectrum of teeling between end-phrases. Degree of security measured as dislance in
mm of patient score marked from the left.

final study visit took place (V4). At the start of this

visit a record of mandibular denture security was
made (du) without the subjects seeing earlier
records, and at the end the subjects were asked to
rescore (c/*), having reviewed earlier records made
prior to denture placement.

had insecurity of their mandibular denture as one

of tbeir complaints (loose mandibular denture subgroup]. The remaining 4 had complaints thai did
not directly involve the mandibular denture, ie,
poor appearance, angular chelitis, chewing difficulty, and repeated fracture of the maxillary denture (non-loose mandibular denture subgroup).
Ten men and ten women (n=20), with an age
range of 55 to 72 years, completed the study.

Denture Construction
After making primary impressions using irreversible hydrocolloid (Alginoplast, Bayer Dental,
Leverkusen, Germany) in stock trays, working
impressions were made in special closely fitting
acrylic resin trays using zinc-oxide/eugenol
impression paste (SS White Manufacturing,
Gloucester, Fngland), Interarch relationship
records were made on registration bases using a
poly(vinyl siloxane) registration paste (Stat BR, Kerr
UK, Peterborough, England), Working casts were
poured in improved stone (Vel-Mix, Kerr UK), and
mounted on an average value articulator
(Condylator, Condylator Service, Z u r i c h ,
Switzerland), Anatomical, shallow-cusped teeth
(Cosmopolitan, De Trey Dentsply, Weybridge,
Surrey, England), set in balanced occlusion, were
used for the trial placement, and the dentures were
processed using a conventional heat-activated
acrylic resin (Trevalon, De Trey Dentsply) using a
long polymerization cycle (4 hours at 95C, and 2
hours at 95C).
All clinical stages of denture fabrication were
carried out by senior dental students under consultant supervision, and included a pre-centric occlusion check record and remount at the final stage to
refine the occlusion. This was recorded using a
poly(vinyl siloxane) registration paste (Stat BR, Kerr
UK), The dentures were again mounted on an average-value articulator, adjusted into balance, and
returned to the patient. To allow subjects to make
their visual analogue records more objectively,
explanations and assistance were provided by a
person not involved with the student supervision
and in the absence of the student involved.

Visual Analogue Scale

The visual analogue scale used (Fig 1) was a 10cm horizontal line representing a spectrum of feeling between extremes identified by end-phrases
that served to provide subjects with meaningful
limits to their feelings. The left side represented a
denture that was loose and insecure at all times,
and the righl side, a denture that was firm and
secure at all times. To improve comprehension, the
right side was further described to each subject as
being 'like having your natural teeth,' Subjects
rated the security of their mandibular denture by
placing a mark across the line at a point corresponding to their feelings at that time. The distance
(d) of the mark from the left-hand limit was measured; a low number represented an insecure denture and a high number a secure denture.
Technique
On the first study visit (VI), a written summary
of the purpose of the study was given to the subjects, supplemented by a verbal explanation.
Written consent was obtained and subjects were
asked to record the security of their present
mandibular denture on a visual analogue scale {d,).
On two later visits (V2 and V3), at weekly intervals
while the dentures were being made, further VA,S
records (d: and d-:) were taken to assess reproducibility, each time without the subjects seeing
their earlier records. When the new dentures were
comfortable, at least 3 and usually 4 weeks later, a

The International |ijrn.l of Proilhodontii

136

inj; M.inrifhular Dtrituri

Results

A bar chart showing the mean changes in

mandibular denture security for the 20 subjects is
given in Fig 2. Graphs of cumulative probability
against preplacement and postplacement VAS
score5 are given in Figs 3a and 3b. The cumulative
probability was calculated as:
P=

(1)

(n+1)

where ; is obtained by listing the VAS scores with

increasing values and numbering them in order (1
to 20), and the total number of observations was n
(20), Equation (1) is commonly used for the calculation of cumulative probability,' and has been
discussed in detail by Cumbel,'Figure 4 shows the percent increase in mean
VAS score plotted against mean preplacement VAS
score. One subject is seen to have a very small
negative change in VAS score. Figure 5 shows the
percent increase in mean VAS score plotted against
the reciprocal of the mean prepiacement VAS
score. In each instance the percent increase in VAS
score was calculated as equal to {{dp<,,-dp,^)/v,...] x
TOO, where dp,,,, = mean postplacement score (d,,,-id^,.)/2, and dp,^ = mean preplacement score
{d+d2+d,)/3; the large percent increases recorded
Fig 2 Bar chan of mean VAS scores in mm for 20 subjects
in some cases are indicative of a low initial satisover 5 visits. Using analysis ot variance mgan scores at V I ,
V2, V3 differ significanlly Irom V4a and V4b (P< 0001),
faction. Figure 6 graphs the coefficient of variation

O
O

0
O
O

0,8

0
0

0,8 -

o
o
o
o

O
0

0,6-

0,6

oo

p
0.4

0,2 -

O
O

O
O
O
0
0
0

O,rt

oo
o
0

0.2

O
O

o
o
o
0

o
20

60
40
VAS score (mm]

ao

100

40
60
VAS score (mm)

H
80

1
100

Fig 3b Cumulative probability (P) plotted against mean

posfplacement VAS score. P- i/{n+1).

Fig 3a Cumulative probability (P) plotted against moan preplacenien! VAS score, P = /7(n+1),

Number 2, 1995

20

137

The International lournal ol Proslhodontii

Security Changes When ieplcmg Mardibular Uentii

1400 -I

1200

1400 -[

O
t200

1000

1000

Q)
Increase
in VAS
Score 600

Increase
in VAS
score 600

o
0

400

o
0

400 -

200
0

o O

800

200 -

oo

1*^

20
40
60
30
^replacement VAS Score .mm)

0 -

100

0.15
0.2
0-1
Reciproca of VAS scores

0.05

0.25

Fig 4 Percent increase in VAS score plotted against mean

preplacement VAS score. Arrow indicates subject with negative change in score

Fig 5 Percent increase in VAS score plotted against reciprocal ot mean preplacement VAS score. Arrow indicates subject
with special problems

[v = (standard deviation/mean) x 100| against the

reciprocal of mean preplacement VAS score.
Where there are only three repeat measurements,
the standard deviation (s) can be expressed as the
approximation:"

scores should have been recorded wifb cross-over

and with a sufficient time interval between them
for the subjects to be uninfluenced by memory of
earlier records, but it appeared unjustified to
impose an extra visit on tbe patients when they
had shown such tolerance.
From the cumulative frequency graph (see Fig
3a) it is obvious that the preplacement distribution
is skewed and the plot is not the ogive shape
expected of a normal distribution. The initial
scores appear to be from a finite mixture distribution," but it is not proposed to analyze this in
detail here. The final scores (see Fig 3b] are also
skewed, and most subjects who felt that tbeir
mandibular denture was secure gave VAS scores in
the range 80 to 99. The distribution must therefore
have been non-normal, and the probability distribution function negatively skewed towards the
limit of 100. (Graphs plotted on normal probability
paper confirmed these findings and further illustrated the non-normality of the distribution). The
improvement in the security for all the patients
(except one, see below and Fig 7) is also graphically illustrated.
The percent improvement in security is shown
versus the mean preplacement VAS score iti Fig 4.
For patients with low VAS scores, there were very
large percent improvements that were a consequence of the low preplacement security values.
The inverse relationship is confirmed by the piot of

(2)
The VAS scores of a patient in the loose
mandibular denture subgroup are given in Fig 7.
Discussion
Results show that it is possible to be sure that a
true increase in mandibular denture security of
approximately 40 mm has taken place (see Fig 2|.
All three mean preplacement scores were similar,
with individual sets of measurements having a variability normally much less than the increases in
security achieved (for their coefficients of variation
see below and Fig 6). When using tests for normally distributed data (analysis of variance) the
mean preplacement scores were significantly.different from the postplacement scores (P < .0001),
which themselves were nearly identical in spite of
the second being recorded after subjects had consulted earlier preplacement scores. This confirms
that subjects can assess denture security reproducibly and denies the need for subjects to consult
earlier records. Ideally the two postplacement

The lolernaiioral lournal of Proilhodontics

138

Volume 8,

Security Cliann Wh

percent improvement in security against reciprocal

of mean preplacement VAS score (see Fig 5). This
also helps identify a patient with special problems;
the subject identified by the arrow had three dentures and several relines provided over the lasl 5
years. The graph of coefficient of variation of the
preplacement VAS scores against the reciprocal of
the mean of the same scores (see Fig 6) also shows
an inverse relationship between coefficient of variation and VAS score. This confirms a real problem
that, while repeating VAS scores does not greatly
affect the final mean scores, it is more difficult to
obtain a consistent measure of security at the low
end of the scale, and repeated recordings might
prove more useful for subjects having the lowest
VAS scores.
The method of visual analogue scales distinguished within the loose mandibular denture subgroup one subject whose preplacement and postplacement scores were both low (see Fig 7). This
subject was the one identified above and in Fig 5
as having special problems, and the method of
VAS scoring may thus allow identification of a
"dissatisfied denture patient."" Others with low
preplacement scores had the expected high postplacement scores of approximately 80 mm to 95
mm. In general, however, the postplacement
scores of the non-loose mandibular denture subgroup (mean d^ = 80 mm, mean d.,b = 83 mm,
n=16) were lower than those of the non-loose
mandibular denture subgroup (mean ct., and du, =
95 mm, n = 4), which might point to a tendency
for patients with specific complaints about their
loose mandibular dentures to be more difficult to
satisfy in this respect. Inspection of records after
analysis was of interest in that the two subjects
with the lowest preplacement and postplace men t
security scores had had a total of eight dentures
made within the last 3 years. Their progress will be
followed and further investigation of the potential
of this aspect of the VAS to monitor the dissatisfied
denture patient carried out. In contrast, the two
subjects in the loose mandibular denture subgroup
with the greatest differences between tbeir pre- and
postplacement scores had had their previous dentures for 8 years and 20 years, respectively.
The non-loose mandibular denture subgroup
which presumably comprised subjects who were
not sufficiently dissatisfied with the security of the
mandibular denture to warrant a complainthad
relatively high preplacement scores (mean d = 82
mm, mean d-, = 76 mm, mean dj = 79 mm) and it
would have been difficult to effect major Increases
in security when measured with a linear VAS scale.
One subject in this subgroup whose complaint was

Number 2, 1995

139

inu Mandihjirfr Dcntu

0
0.05
0.1
0.15
0.2
0.25
Reciprocal ol mean preplacemenl VAS scores
Fig 6 Coetticient ot variation (v) of preplacement VAS
scores plotted against reciprocal ot mean preplaoement VAS
score.

Fig 7 VAS score over 4 visits, for patrent no. 3 with complaint of loose manditiular denture.

The International lournal of Prosthcdonlics

Security Changes When Replacing Mandibul

design. General considerations necessarily applysuch as scientific validity and informed consent,
remembering that a bad experimental rlesign,
namely one incapable of yielding signilicant data,
is itself unethical and cannot be validated by
procuring patient consent."" Often to compensate
for the Hawthorne effect, whereby an improvement occurs in any observed group even when no
real change is introduced," a treatment is compared with a control and the responses are analyzed by suitable statistical methods. For patients
with pain and discomfort resulting from malfunction of their dentures, it would be unethical to
have a control group that did not receive the best
possible treatment, provided it was good clinical
practice. The placement of another loose or ill-fitting denture would be a suitable control, but
would justify a charge of malpractice.
The measurement of satisfaction is difficult.
Nevertheless, it would be very useful to have a
scale of measurement of patient satisfaction, and
an appreciation of the importance of patient satisfaction with treatment is an obvious attribute for
successful clinical practice. Yet, even in the physical sciences the establishment of scales of measurement has not been easy. For the present purposes, the aims were more modest and were to
establish that one aspect of patient satisfaction,
denture security, was increased by the placement
of a new denture, it is recognized that an absolute
scale of security has not been achieved and all
patients had a personal scale with their own
expectations. The measurements can only evaluate
the improvement in denture security which took
place, as measured by the difference between the
preplacement and postplacement VAS scores. The
improvement produced by conventional techniques was clearly demonstrated by the results
which showed that 19 out of 20 patients had
higher VAS scores after treatment. The one exception had such a high initial score that it made
improvement nearly impossible. Thus, the use of
relative differences in VAS score was shown to be
a valid means of assessing patient satisfaction as
represented by denture security.

repeated fracture of the maxillary denture showed

a slight deterioration in security, a finding that
emphasizes the need to identify complaints accurately and modify patients' expectations before
replacing dentures.
This study confirmed the VAS to be a simple
technique to apply to measurement of denture
security, and the results indicate that it is worthy of
further application in dental research. However,
final evaluation of the VAS method requires comparison with other assessment scales.
Conclusions
Twenty complete denture patients were evaluated
before and after receiving new prostheses, using a
visual analogue scale to rate patient-assessed
mandibular denture security. Within the limitations
of the study design imposed by ethical constraints,
the following conclusions may be made:
1. Visual analogue scores are a simple means of
measuring objectively a patient's perception of
denture security.
2. Visual analogue recording may be a useful
audit tool and help identify valuable clinical
techniques and patienls with special problems.
3. Visual analogue recordings show that patient
evaluation confirms clinical expectation and
that when replacing complete dentures using
conventional techniques, the expected
improvements in denture security are achieved.

Appendix
Design Problems in Patient Satisfaction Studies

Conceptually, the logistics of such studies are

simple. Select without bias a group of subjects satisfying the clinical criteria for the research, who
willingly give consent to participate and agree to
the extra investigations required. Many edentulous
patients are elderly or infirm, and visits to a dental
hospital present difficulties for themselves or their
caretakersfactors that may intioduce bias.
Additionally, studies must be designed to ensure
that as many subjects as possible (ideally all) complete the program and that minimum distress is
imposed. Consequently, the number of return visits
to hospitalespecially those following provision of
new denturesmust be limited.
Compared with other areas of medical research,
the ethical considerations for the reported study
are minor, but nevertheless complicate study

The International lournal of Proithodontii

Acknowledgments
The authors are graleful for the assistance provided by Mr.
Phillip Wragg, Consultant in Restorative Dentistry, Ctiarles
Cliiford Dental Hospital, for his help with studenl supervision
during ihe course of this study.

140

Liimb/Ellii

ritv Chindes When Replacing Mandibular Denlures

References

8, Ailken RCB, A growing edge of measurement uf feelings,

Proc Roy Soc Med 1969;62:989-993,
9, Seymour RA, Charlton |E, Phillips ME, An evaluation of dental pain using visual analogue scales and the McCill pain
questionnaire, | Oral Maxillofacial Surg 1983;41 :ti43-64e,
10. Lamb Dl, Ellis B, Kent C. Measurement of complete
mandibular denture security by visual analogue scales, Int
I Prosthodont 1994;7:30-34.
11. King IR, Probability Charts for Decision Making, New
York: Industrial Prssb, 1971:20-22,
12. Gumbel E|. StaLislics of Extremes, New York: Columbia
tJniversily Press, 1958:32-34.
13. Craddock JM, Statistics in ibe Computer Age, London:
English Universities Press, 1968:31,
14. EveriLt BS, Hand D|. Einlte Mixture Distributions. London:
Chapman and Hall, 19B1.
1,5 Beck CB, Bates |F, Basker RM, Gutteridgc DL, Harrison A,
A survey o the dissatisfied denture patient, Eur i
Prosthodont Rest Dent 1 993;2:73-7B,
16. Kelly Mi, Wadsworth |, What price inconclusive clinical
trials? Ann Roy Col Sutg Eng 1993,75:145-146,
17, Eason KD, New systems implementation. In: Wilson |R,
Coflett EN (edsl. Evaluation of Human Work, London:
Taylor and Francis, 1990:841-642,

1, Keni G, lohns RB, Psychological effects of permanently

implanted false teeth: A 2-year follow-up and comparison
with dentate patients, Psycliol and Health t993;B:2i3-222.
2, Reisine ST, Fertig i, Weber |, Leder S, Impact of denial
conditions on patients' quality ot life, Comriiunity Dent
Oral Epidemiol 1989:17:7-10,
3, t^eyink iW, Heezen iH, Schaub RMH. Dentist and paiieni
appraisal of complete dentures in a Dulch elderly popufatiori, Commiinity Dent Oral Epidemiol 19B6;14:3i3-326.
4, Mojon P, MacEntee Ml Discrepancy between need for
prosthcdontic treatment and complaints in an elderly
edenlulotJi populaiin. Community Dert Oral Epidenniol
1992;20:48-52.
5, Kalk W, de Baat C, Patients' complaints and satisfaction 5
years after complete denture treatment. Community Dent
Oral Epidemiol 199;18:Z7.-31,
6, Smith IP, Hughes D, A survey of referred patients experiencing problems witb complete deritures, | Prosthet Dert
19BB;60:583-586.
7, Vervoorn |M, Djinkerke ASH, LuLeijr P, van de Poel
ACM, Assessment of denture satisfacliiin. Community
Denl Oral Epidemiol 1 9HH:16:364-367,

Literature Absiracl -

Itnplants for Type II Diabetic Patetits: Interim Report

Diabetes is a serious illness affecting approximately 13 million Americans, half of whom liave
not been diagnosed. The number of diabetic patients is expected to increase and possibly
double every 15 years, HIstonoally, diabetes mellitus lias been considered a contralndlcalion
for dental implants. This prospective investigation was designed to determine Ihe success rate
of dental implants for a large number of type II (noninsulin-dependert ] diabetic patients. One
hundred seventy-eight implants were placed in 89 type II diabetic patients at 13 Department of
Veterans Affairs medioal oenters as part of a 5-year investigation. All patients were managed
by assigned physicians for a minimum of 14 days betre stage-one surgery. Two endosseous
implants were placed in the anterior mandible of each patient. The patients' medical status
was evaluated postoperailveiy at the end of the first, second, and fourth weeks, and then
monthly until placement of the prcstheses, Set:ond-stage surgery was accomplished 4 months
atter tirst-stage surgery and prostheses fabrication initiated. Four (2,2%] implants were mobile
at second-s1age surgery and classified as failures, tsline additional implant failures occurred
during tine tirst year, increasing the failure rate to 7,3%, Eight of the failures were of unknown
origin, tour mere attributed to uncontrolled diabetic episodes, and one was attributed to surgical complications. There was no initial Indication that the amount ot keratinized tissue or
smoking status influenced the implant failure rate The authors concluded that alttiough Ihe
initial success rale during the first year was favorable, ttie additional 4-year maintenance period will be important and will provide additional data regarding the prognosis
of diabetic patients receiving dental implants.
SfiernofI AF, Colwelt JA, Bingham SF. Implant Dent 1994:3(3):183-185. References: 27, Reprints: Dr
Alan F Shemoff, Dental Service (160, VA Medical Center, 1S70 Roanoke Boulevard, Salem, VA 24153,
Richard H, Seals. Jr. DDS. ME. MS. Depadmen! ol Prosltiodonl^cs, Unii/ersify ot Texas Healm Center at
San Antonio. San Amonio. Texas

141

The Intemational iournsl of Prosthodontic