A comparison of the monofilament with other testing
We studied the number of testing sites and the proportion needed to be insensate for the optimal assessment of foot ulcer risk
sitivity and specificity of the 10 g monofilament with other
methodologies. Fifty-two individuals with either a current foot ulcer, a history of a foot ulcer or the presence of Charcot
Foot ulcers are the main cause of lower extremity
Diabetes Research and Clinical Pra
amputation in patients with diabetes [1,2]. It has been
estimated that 15% of patients with diabetes will
develop a foot ulcer in their lifetime and that 85% of
* Corresponding author. Tel.: +1 305 243 6146;
fax: +1 305 243 4484.
E-mail address: jsosenko@med.miami.edu (J.M. Sosenko).
0168-8227/$ – see front matter # 2005 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.diabres.2005.02.013
sites on each foot, the 128 Hz tuning fork at the halluces, the Biothesiometer at the halluces and the modified neuropathy
disability score. Sensitivities and specificities were calculated for the various modalities. The Biothesiometer and the neuropathy
disability score had the highest sensitivities (0.92 for both). The 128 Hz tuning fork tested only at the halluces (criterion: �1
insensate site) had the same sensitivity (0.86) as the 10 g monofilament tested at eight sites (criterion: �1 insensate site) with
similar specificities (0.56 and 0.58, respectively). The Biothesiometer and the modified neuropathy disability score tend to be
more sensitive than the 10 g monofilament for the assessment of individuals at risk for foot ulcers. The 128 Hz tuning fork tested
at only two sites is as sensitive as the 10 g monofilament tested at eight sites. These data suggest that the 10 g monofilament may
not be the optimum methodology for identifying individuals at risk of foot ulcers.
# 2005 Elsevier Ireland Ltd. All rights reserved.
Keywords: Diabetes mellitus; Foot ulcer; Neuropathy; Monofilament
1. Introduction
neuroarthopathy and 51 individuals with no history of any of
these conditions were assessed with the 10 g monofilament at four
with the 10 g monofilament. Also, we compared the sen
B. Miranda-Palma a, J.M. Sosenko a,*, J.H. Bowker b,
M.S. Mizel b, A.J.M. Boulton a
a Division of Endocrinology, Department of Internal Medicine, Diabetes Research Institute,
University of Miami School of Medicine, P.O. Box 016960 (D-110), Miami, FL 33101, USA
b Department of Orthopedic Surgery, Diabetes Research Institute, University of Miami School of Medicine,
P.O. Box 016960 (D-110), Miami, FL 33101, USA
Received 19 November 2004; accepted 18 February 2005
Available online 28 March 2005
Abstract
modalities for foo
t ulcer susceptibility
www.elsevier.com/locate/diabres
ctice 70 (2005) 8–12
Several instruments have been utilized to detect a
wer
a pr
amp
and
foot
path
2.2.
A
exa
com
Cha
of
earch and
lack of protective sensation [7]. The ideal instrument
for this should be readily available, easy to use, and
provide reproducible results with high sensitivity. In
recent years, the 10 g monofilament has gained
widespread use for assessing the loss of protective
sensation, since it is available and simple to use [8].
However, there is no consensus as to the specific sites
that should be tested and the minimum number of
insensate sites required for the optimal prediction of
foot ulceration [8].
A main objective of the present study was to
determine the number of testing sites and the
proportion needed to be insensate for the optimal
assessment of foot ulcer susceptibility with the 10 g
monofilament. Another objective was to compare the
sensitivity and specificity of the 10 g monofilament
with the vibration perception threshold (VPT) and the
modified neuropathy disability score (NDS), other
methods that have been shown to be useful predictors
of foot ulcers [9–11].
2. Materials and methods
The study was approved by the Institutional
Review Board of the University of Miami School of
Medicine, and all subjects signed an informed consent
prior to participation in the study.
2.1. Patients
One hundred and three patients were enrolled from
the General Medical Clinic and the Diabetic Foot
Clinic at Jackson Memorial Hospital. The diagnosis of
all non-traumatic amputations are preceded by a non-
healing foot ulcer [3]. The occurrence of a diabetic
foot ulcer is the result of multiple contributing factors,
but it is commonly accepted that peripheral neuro-
pathy with the associated loss of protective sensation
is most contributory [4]. It is therefore common
clinical practice to assess the presence and extent of
peripheral neuropathy in order to identify patients at
risk for developing foot ulcers [4]. This is based on the
rationale that individuals identified as being at high
risk for foot ulcers may benefit from more frequent
follow-up and from educational interventions [5,6].
B. Miranda-Palma et al. / Diabetes Res
diabetes had been made before enrollment. Patients
with major foot amputation proximal to the toe areas
as well as a hallux amputation were excluded. Other
exclusion criteria included evidence of peripheral
vascular disease assessed by absent foot pulses, a
history of claudication and/or peripheral bypass
surgery, a history of cerebro-vascular disease or any
other neurological disease that could be associated
with abnormal sensory findings, and a history of
alcohol abuse defined as >21 units per week.
2.3. Procedures
The following objective measures of peripheral
nerve function were assessed:
(1) Ten gram monofilament evaluation
All patients were assessed using the Bailey
10 g monofilament (Baileys Instruments Co.,
Chorlton, Manchester, UK), chosen because they
had previously been assessed to be accurate at
assessing pressure at 10 g when the filament
buckled [13]. Testing was performed at four
different sites in each foot: the plantar surface of
hallux, and the first (MTH1), third (MTH3), and
fifth (MTH5) metatarsal heads, according to the
recommendations of Mayfield and Sugarman [8].
With eyes closed, the patients were required to
provide a ‘‘yes/no’’ response to monofilament
pressure and in addition identify correctly the site
of contact. Each filament was placed against a
plantar surface of the four sites in a perpendicular
differ
defor
e included in the foot ulcer group if they either had
evious or an active foot ulcer, a history of minor
utation (limited to one toe other than the hallux)
/or Charcot neuroarthropathy. Patients without a
ulcer history, amputation or Charcot neuroarthro-
y were included in the control group.
Patient selection
ll patients had a history and a limited physical
mination. Foot ulceration was defined as a
plete break of the skin distal to the malleoli.
rcot neuroarthropathy was defined by at least two
the following: unilateral swelling, temperature
ential of �2 8C between the two feet, bony
mity and radiographic changes [12]. Patients
Clinical Practice 70 (2005) 8–12 9
fashion so that it bent with a constant force.
sensation using a 128 Hz tuning fork, dorsal
temperature sensation using warm and cold rods
(3)
2.4.
T
and
cha
thos
Sen
cuto
indi
3. R
O
he presence of a hallux amputation. Of those included
n the study, 52 patients (50%) had either a past or
resent ulcer, a history of a minor amputation (one
igit excluding hallux) or Charcot neuroarthropathy.
inety-three percent of the patients had type 2
iabetes mellitus (defined as treatment with diet or
n oral hypoglycaemic agent, or the institution of
nsulin treatment at least two years after diagnosis)
nd 56% were male. The vast majority of the patients
ere Hispanic (60%) or African–American (30%). In
he ulcer group there was a predominance of male
atients (73%).
Testing with the 10 g monofilament was assessed
ccording to different cutoff points for positivity
Table 1). We examined the sensitivities and
pecificities of the 10 g monofilament for detecting
oot ulceration with positivity cutoffs of �1 insensate
ites of 8, �2 insensate sites of 8 and �4 insensate
ites of 8. We found that the sensitivities of the
onofilament ranged from 65 to 86%, while the
earch nd Clinical Practice 70 (2005) 8–12
were
vasc
testing site of the hallux. Three consecutive
measures were then taken at variable speeds of
voltage increase and the median of the three was
taken as the final result for each foot.
Statistical methods
he t-test for the comparison of independent means
the chi-square test were used to compare
racteristics between the foot ulcer patients and
e without foot ulcers. P-values are two-sided.
sitivities and specificities were calculated after
ff values were determined according to the criteria
cated below.
esults
f 103 patients that were screened, 10 patients
excluded due to the presence of peripheral
d
tendon reflex was assessed with the patient supine
on a couch as previously described [11,14]. The
maximumscore for the twofeet is10,with a scoreof
�6 indicating moderate to severe neuropathy [14].
Vibration perception threshold (VPT)
This was measured using a Biothesiometer
(Biomedical Instrument Company, Newbury,
OH). This handheld device was balanced so that
the vibrating stylus rested on the apex of the
hallux. The normal sensation of vibration was
emonstrated on a proximal site and then over the
and Achilles tendon reflexes using a tendon
hammer. For one foot, each sensory test scored
zero for normal sensation or one for abnormal
sensation: ankle reflex scored zero if present, one if
present with reinforcement or two if absent. Pain
and temperature sensation were assessed on the
dorsal surface of the great toe after the stimuli were
demonstrated at a proximal, normal site. Vibration
perception was assessed using the 128 Hz tuning
fork over the apex of the hallux. The Achilles
(2) Modified neuropathy disability score (NDS)
Originally described by Young et al. [14], this
score (maximum 10) is derived from abnormalities
of pain sensation using a neurotip, vibration
B. Miranda-Palma et al. / Diabetes Res10
ular disease as determined by absent pulses and/or
specificities ranged from 58 to 71%.
Sensitivities and specificities at individual sites of
the foot were also assessed with an abnormal site
defined as at least one foot being insensate at that site
(Table 2). In general, the individual sites had low
sensitivities (range: 65–77%). The highest sensitiv-
Table 1
Sensitivity and specificity of the 10 g monofilament according to the
number of insensate sites
Test Sensitivity Specificity
Monofilament �1/8 86 58
Monofilament �2/8 77 63
Monofilament �4/8 65 71
52 of 93 tested had either an ulcer, an amputation or Charcot
neuroarthropathy.
Table 2
Sensitivity and specificity of the l0 g monofilament according to the
location of testing (�1 of the two sites tested was considered
positive)
Testing site Sensitivity Specificity
Hallux 73 68
MTH1 71 63
MTH3 65 63
MTH5 77 68
52 of 93 tested had either an ulcer, an amputation or Charcot
t
i
p
d
N
d
a
i
a
w
t
p
a
(
s
f
s
s
m
a
neuroarthropathy.
earch
52 of 93 tested had either an ulcer, an amputation or Charcot neuro-
arthropathy.
ities were at the hallux and MTH5. When these two
sites were combined (�1 of 4 being positive), the
sensitivity was 81% with a specificity of 63%.
We compared the 10 g monofilament with two
other procedures for detecting decreased sensitivity,
the measurement of the VPT using a biothesiometer
and a modified NDS (Table 3). AVPT �25 [9,10] and
a modified NDS �6 [11] were used as cutoffs for
positivity. The VPT and the modified NDS both
demonstrated sensitivities of 92% for detecting
patients with foot ulceration with respective specifi-
cities of 39 and 53%. Both tests had higher sensitivies
than the 10 g monofilament test.
The 128 Hz tuning fork component of the NDS
(insensitivity of one of the halluces was considered a
positive test) was also compared with the monofila-
ment with a cutoff of �1 insensate site of 8 tested.
Testing with the tuning fork at the halluces alone had
the same sensitivity 86% and nearly as high a
a �1 of the two sites tested was considered positive.
B. Miranda-Palma et al. / Diabetes Res
Table 3
Sensitivity and specificity according to methodologies used to assess
insensitivity
Test Sensitivity Specificity
VPT � 25 92 39
NDS � 6 92 53
Monofilament
�1/8 86 58
128 Hz tuning forka 86 56
specificity (56% versus 58%) as that of the 10 g
monofilament.
4. Discussion
The 10 g monofilament is now commonly used
[5,7,8], yet there are few rigorous data available with
regard to how it is best utilized to detect foot
insensitivity and its resultant foot ulcer risk. Although
a number of studies have utilized monofilaments for
the assessment of neuropathy, there are no substantive
data that support any one standard method of
application of the 10 g monofilament [8]. It is
therefore not surprising that its optimal capacity to
detect insensitivity is not known. In examining
different methods of application of the 10 g mono-
filament and in comparing the performance of the
monofilament to other instruments, this study has
raised questions as to whether its widespread use is
warranted. The appropriate cutoff for the number of
insensate sites is dependent on screening objectives;
however, the data appear to indicate that requiring
more than one insensate site would lead to an
unacceptable number of false negatives. Even using
a cutoff of one or more insensate sites of eight tested
had only moderate sensitivity.
The 10 g monofilament did not compare favorably
with the other testing modalities. The sensitivities of
the VPT and the modified NDS were both higher than
that for the 10 g monofilament. The performance of
the 128 Hz tuning fork at two sites was even
comparable to the 10 g monofilament at eight sites.
Although the use of the VPT and the NDS would
apparently lead to higher sensitivities than the 10 g
monofilament, their specificities would be lower. One
cannot definitively say which performance character-
istic would be preferable from a public health pers-
pective; however, it would appear that there is more to
lose by a failure to initiate preventive interventions in
those who would benefit than the alternative.
A rationale for using the 10 g monofilament is that
it is available, inexpensive and easy to use [8].
However, if at least eight sites are required for its use,
it does require some time for testing. In fact, the use of
the 128 Hz tuning fork at two sites (the halluces)
would take less time than the 10 g monofilament tested
at eight sites and provide comparable accuracy. The
graduated tuning fork has also been shown to be a
useful, accurate and reproducible method of assessing
peripheral sensation [15].
There are other questions regarding the use of the
10 g monofilament apart from those raised by this
study. It has been shown that a number of monofila-
ments purported to be testing 10 g pressure actually
fail to do so [13]. Also, it is entirely possible that
monofilaments smaller in size would perform better.
There are certain limitations with this study. We
were testing the various modalities to detect foot ulcers
in individuals who already had previous or existing
ulcers, whereas the real issue is how the modalities are
able to predict the future development of foot ulcers.
and Clinical Practice 70 (2005) 8–12 11
The number of participants in this study was relatively
small and differences were not statistically significant.
Until a prospective study with sufficient power is
performed to compare these various testing modalities,
definitive conclusions cannot be drawn.
Despite the above limitations, this study does serve
to show that one must question the use of the 10 g
monofilament as a standard. Its sensitivity might be
insufficient, even with testing at multiple sites. The
tuning fork had been a staple in the clinic setting for
assessing insensitivity before the 10 g monofilament
became popularized. Until more definitive informa-
tion is obtained about the use of the 10 g monofila-
ment, alternatives should include the modified NDS or
even the 128 Hz tuning fork.
References
[6] G.D. Valk, D.M. Kriegsmann, W.J.J. Assendelft, Patient edu-
cation for preventing diabetic foot ulceration: a systematic
review, Endocrinol. Metab. Clin. N. Am. 31 (2002) 633–658.
[7] A.J.M. Boulton, R.A. Malik, J.C. Arezzo, J.M. Sosenko,
Diabetic somatic neuropathies: technical review, Diab. Care
27 (2004) 1458–1486.
[8] J.A. Mayfield, J. Sugarman, The use of the Semmes–Weinstein
monofilaments and other threshold tests for preventing foot
ulceration and amputations in people with diabetes, J. Fam.
Pract. 49 (Suppl.) (2000) 517–529.
[9] M.J. Young, J. Breddy, A. Veves, A.J.M. Boulton, The pre-
diction of diabetic neuropathic foot ulceration using vibration
perception thresholds: a prospective study, Diab. Care 17
(1994) 557–560.
[10] C.A. Abbott, L. Vileikyte, S.H. Williamson, A.L. Carrington,
A.J.M. Boulton, Multicentre study of the incidence and pre-
dictive factors for diabetic foot ulceration, Diab. Care 21
(1998) 1071–1075.
[11] C.A. Abbott, A.L. Carrington, H. Ashe, S. Bath, L.C. Every, J.
Griffiths, E.R. Van Ross, A.J.M. Boulton, North–west Diabetes
Foot Care Study: incidence of and risk factors for new diabetic
B. Miranda-Palma et al. / Diabetes Research and Clinical Practice 70 (2005) 8–1212
[1] E.B. Jude, A.J.M. Boulton, End-stage complications of dia-
betic neuropathy, Diab. Rev. 7 (1999) 395–410.
[2] W.J. Jeffcoate, K.G. Harding, Diabetic foot ulcers, Lancet 361
(2003) 1545–1551.
[3] A.J.M. Boulton, R.S. Kirsner, L. Vileikyte, Neuropathic Dia-
betic foot ulceraton, N. Engl. J. Med. 351 (2004) 48–55.
[4] G.E. Reiber, L. Vileikyte, E.J. Boyko, D.J. Smith, L.A. Lavery,
A.J.M. Boulton, Causal pathways for incident lower extremity
ulcers in patients with diabetic from two settings, Diab. Care
22 (1999) 157–162.
[5] J.A. Mayfield, G.E. Reiber, L.J. Sanders, D.J. Janisse, L.M.
Pogach, Preventive foot care in people with diabetes, Diab.
Care 21 (1998) 2161–2177.
foot ulceration in a community-based study, Diab. Med. 19
(2002) 377–384.
[12] J.E. Shaw, A.J.M. Boulton, Foot 5 (1995) 65–70.
[13] J. Booth, M.J. Young, Differences in the performance of
commercially available monofilaments, Diab. Care 23
(2000) 984–988.
[14] M.J. Young, A.J.M. Boulton, A.F. McLeod, D.R.R. Williams,
P.H. Sonksen, A multicentre study of the prevalence of diabetic
peripheral neuropathy in a United Kingdom hospital clinic
population, Diabetologia 36 (1993) 150–154.
[15] M.J. Hilz, F.B. Axelrod, K. Hermann, L.U. Haertl, M. Duetsch,
B. Neundorfer, Normative values of vibratory perception in
530 children, juveniles and adults age 3–79 years, J. Neurol.
Sci. 159 (1998) 219–225.
A comparison of the monofilament with other testing �modalities for foot ulcer susceptibility
Introduction
Materials and methods
Patients
Patient selection
Procedures
Statistical methods
Results
Discussion
References