Itraconazole-induced painful neuropathy in a man with type 1 diabetes.

OBSERVATIONS
Itraconazole-Induced
Painful Neuropathy
in a Man With
Type 1 Diabetes
The antifungal agent itraconazole isconsidered safe in patients with dia-betes (1,2) but has been associated
with acute painful neuropathy in patients
taking vincristine (3–5). We describe a
patient with diabetic neuropathy who de-
veloped acute painful neuropathy with
itraconazole therapy.
A 37-year-old man with poorly con-
trolled type 1 diabetes of 19 years’ dura-
tion presented with an excematous rash
affecting his lower abdomen. Skin scrap-
ings grew C. albicans. Itraconazole (200
mg/day) was prescribed, but on the 4th
day of treatment he developed acute bi-
lateral leg weakness with stabbing pain in
the hips and legs and difficulty walking.
He was taking no other medication apart
from insulin. He was known beforehand
to have peripheral neuropathy with ab-
sent ankle reflexes and mild postural
hypotension.
On examination, his BMI was 24 kg/
m2, his calf muscles were tender, and he
had moderate proximal leg weakness. His
muscle tone was normal and knee reflexes
preserved. There was no change in pos-
tural hypotension or sensory abnormali-
ties. Itraconazole was stopped after 7
days. Over the next 6 months his pain
improved, although he lost 6 kg in
weight. The power in his legs recovered,
but the knee reflexes were lost. Over the
next year his pain improved and he be-
came asymptomatic.
At the onset of his symptoms, blood
count, erythrocyte sedimentation rate, re-
nal function, and serum creatine kinase
were normal. On repeated follow-up,
these blood tests and his HbA1c remained
unchanged. The vibration perception
threshold at the medial malleolus rapidly
declined over the next year. Nerve con-
duction studies of the upper and proximal
lower limbs were normal, but conduction
velocities were slow in the distal lower
limbs. Over the next year, the distal
lower limb compound muscle action po-
tentia l ampli tudes reduced, sug-
gesting loss of these nerve fibers. Electro-
myography (EMG) was normal at presen-
tation, but after 6 months showed signs
of lower-limb muscle denervation. There
were no features to support lumbar plex-
opathy. As the abnormalities were most
marked in the distal muscles, this sug-
gests axonal degeneration, which tends to
affect the longest nerve fibers first.
There are a number of diagnostic
possibilities. Acute painful neuropathy
generally occurs within days of rapid im-
provements in glycemic control, which
was not the case in our subject. Diabetic
amyotrophy often presents with severe
anterior thigh muscle pain and weight
loss; however, EMG typically shows evi-
dence of lumbosacral plexopathy (6).
Acute diabetic radiculopathy can cause
acute pain and weakness, but the symp-
toms are usually in a single nerve root
distribution. Diabetic cachexia can be as-
sociated with profound weight loss, sym-
metrical peripheral neuropathy, and
autonomic dysfunction but is generally
associated with painful dysesthesias with-
out weakness (7).
In this man with preexisting diabetic
neuropathy, the sudden onset of neuro-
logical symptoms with itraconazole ther-
apy, axonal degeneration on electrophys-
iological testing, and stabilization after
stopping itraconazole suggest that a drug-
induced polyneuropathy is the most
likely diagnosis. Drug-induced neuropa-
thy is associated with axonal degenera-
tion, lasting weeks or months (8).
Discontinuing the administration of the
drug does not always lead to improve-
ment, as the damage to neuronal struc-
tures or function can be permanent.
RINKI SINGH, MBCHB1
TIM CUNDY, MD1,2
From the 1Auckland Diabetes Centre, Auckland,
New Zealand; and the 2Department of Medicine,
Faculty of Medical and Health Sciences, University
of Auckland, Auckland, New Zealand.
Address correspondence to Tim Cundy, MD, De-
partment of Medicine, Faculty of Medical and
Health Sciences, University of Auckland, Auckland,
New Zealand. E-mail: t.cundy@auckland.ac.nz.
© 2005 by the American Diabetes Association.
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
References
1. Albreski DA, Gross EG: The safety of itra-
conazole in the diabetic population. J Am
Podiat Med Assoc 89:339–345, 1999
2. Verspeelt J, Marynissen G, Gupta AK, De
Doncker P: Safety of itraconazole in dia-
betic patients. Dermatology 198:383–384,
1999
3. Jeng M, Feusner J: Severe vincristine tox-
icity potentiated by itraconazole therapy
in a child with acute lymphocytic leuke-
mia 21:329–330, 1999
4. Sathiapalan RK, El-Solh H: Enhanced vin-
cristine neurotoxicity from drug interac-
tions: case report and review of literature.
Pediatr Hematol Oncol 18:543–546, 2001
5. Bohme A, Ganser A, Hoelzer D: Aggrava-
tion of vincristine-induced neurotoxicity
by itraconazole in the treatment of adult
ALL. Ann Hematol 71:311–312, 1995
6. Subramony SH, Wilbourn AJ: Diabetic
proximal neuropathy: clinical and elec-
tromyographic studies. J Neurol Sci 53:
293–304, 1982
7. Blau RH: Diabetic neuropathic cachexia:
report of a woman with this syndrome
and review of the literature. Arch Intern
Med 143:2011–2012, 1983
8. Peltier AC, Russell JW: Recent advances in
drug-induced neuropathies. Curr Opin
Neurol 15:633–638, 2002
Evaluation of the
Portable HealthWear
Armband
A device to measure total daily
energy expenditure in free-living
type 2 diabetic individuals
L ifestyle modifications involving dietand exercise are effective in reducingthe incidence of type 2 diabetes (1–
3). In particular, lifestyle changes targeted
toward increasing daily energy expendi-
ture are a cornerstone treatment of type 2
diabetes (2). In the Diabetes Prevention
Research Group study (4), a goal of 150
min/week of physical activity was recom-
mended, and 74% of the patients in this
study group achieved this level by 24
weeks. This type of success in exercise
modification is made possible in research
protocols by intensive individual coun-
seling (5), which may not be easily repro-
duced in primary care settings. Because of
the critical importance of increasing daily
energy expenditure and its central role in
preventing and/or treating diabetes, one
can surmise that instrumentation provid-
ing accurate and simple feedback to type
2 patients may have clinical utility.
Therefore, we provide preliminary
L E T T E R S
DIABETES CARE, VOLUME 28, NUMBER 1, JANUARY 2005 225

data on the accuracy of the HealthWear
Armband (Roche Diagnostics, Indianapo-
lis, IN), an instrument designed to assess
total daily energy expenditure in free-
living adults, including those with diabe-
tes, versus the doubly labeled water
(DLW) technique. The development of
the DLW methodology provides a gold
standard from which the accuracy of
other devices to measure daily energy ex-
penditure can be determined (6). The
DLW method requires little subject coop-
eration, is unobtrusive, and accurately de-
termines daily energy expenditure
throughout a person’s daily routine. Un-
fortunately, the relatively high price of the
oxygen-18 water, the need for mass spec-
trometer instrumentation, and the high
level of technical expertise required has
limited its widespread application in clin-
ical research. The HealthWear Armband
evaluates daily energy expenditure based
on a proprietary algorithm. That is, it uses
a collection of sensors to gather informa-
tion such as movement, heat flow, skin
temperature, near-body temperature, and
galvanic skin response in conjunction
with body measurements such as sex, age,
height, and weight to calculate energy ex-
penditure. The heat flow sensor uses sen-
sitive thermocouple arrays and measures
the heat dissipated by the body. It is
placed between the skin and the side of
the Armband exposed to the environ-
ment. If such a practical tool was available
and validated, it would help the patient
monitor or increase their daily energy ex-
penditure levels to lose weight with the
ultimate goal of reducing the complica-
tions of diabetes.
As part of a larger study, we assessed
total daily energy expenditure in six dia-
betic patients treated with diet only
and/or oral hypoglycemic agents. We
tested two men and four women, aged
56.5 � 5.96 years with a BMI of 29.76 �
4.12 kg/m2, a fat-free mass of 50.08 �
12.17 kg, and a fat mass of 31.98 � 6.99
kg (determined by dual-energy X-ray ab-
sorptiometry). All subjects wore the
HealthWear Armband simultaneously
with the determination of DLW during a
10-day period to measure total daily en-
ergy expenditure. The armband was worn
around the right arm and was removed
only for showering and bathing purposes.
“On body time” for the armband was 99%
during the 10-day period for all subjects.
In the six diabetic patients, we noted
no significant differences (� 78.3 � 158
kcal/day [mean � SD], 95% CI �87 to
245 kcal/day) in mean daily energy ex-
pendi ture between the armband
(2.237 � 568 kcal/day) and DLW
(2,315 � 625 kcal/day). Also, the corre-
lation between the armband and DLW
reached r� 0.9696 (P� 0.0014). In ad-
dition, the intraclass correlation coeffi-
cient (ICC) between the armband and
DLW reached ICC� 0.9625, with a tech-
nical error of measurement of 104 kcal/
day. Fourth, individual comparisons
between DLW and the HealthWear Arm-
band were examined using a Bland-
Altman plot (Fig. 1). The Bland-Altman
plot shows the difference in daily energy
expenditure between DLW and the
HealthWear Armband versus mean daily
energy expenditure determined between
DLW and the HealthWear Armband.
From these data, limits of agreement be-
tween DLW and the HealthWear Arm-
band were calculated (i .e. , mean
difference between DLW and the Health-
Wear Armband �1.96 SD of the differ-
ence). We hypothesized a narrow limit of
agreement (i.e.,� 100–300 kcal/day) be-
tween HealthWear Armband and DLW.
The measured range of under- and over-
estimation of the armband versus DLW
was �243 to 176 kcal/day, respectively,
suggesting an acceptable level of concor-
dance between the two methods.
Figure 1—Limits of agreement between DLW and the HealthWear Armband.
L E T T E R S
226 DIABETES CARE, VOLUME 28, NUMBER 1, JANUARY 2005

Although these results will need to be
confirmed in a larger sample size, prelim-
inary analyses suggest that the Health-
Wear Armband is an acceptable device to
accurately measure total daily energy ex-
penditure in type 2 diabetic patients over
a 10-day period. This information would
be useful in counseling patients regarding
appropriate levels of daily energy expen-
diture needed to prevent or offset diabetes
and its related complications.
DIANE MIGNAULT, BSC1
MAXIME ST.-ONGE, MSC1
ANTONY D. KARELIS, PHD,1
DAVID B. ALLISON, PHD2
REMI RABASA-LHORET, MD, PHD1
From the 1De´partement de Nutrition, Unite´
Me´tabolique, Universite´ de Montre´al, Montreal,
Quebec, Canada; and the 2Section on Statistical Ge-
netics and Clinical Nutrition Research Unit, Univer-
sity of Alabama at Birmingham, Birmingham,
Alabama.
Address correspondence to Diane Mignault, BSc,
Unite´ Me´tabolique De´partement de Nutrition Uni-
versite´ de Montre´al, 2405 Chemin Cote Ste-
Catherine, Pavillon Liliane de Stewart, Montreal,
Quebec, Canada, H3T 1A8. E-mail: diane.
mignault@umontreal.ca.
R.R.-L. has received honoraria and funds from
Roche Diagnostics.
© 2005 by the American Diabetes Association.
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
References
1. Tuomilehto J, Lindstrom J, Eriksson JG,
Valle TT, Hamalainen H, Ilanne-Parikka
P, Keinanen-Kiukaanniemi S, Laakso M,
Louheranta A, Rastas M, Salminen V,
Uusitupa M: Prevention of type 2 diabetes
mellitus by changes in lifestyle among
subjects with impaired glucose tolerance.
N Engl J Med 344:1343–1350, 2001
2. American Diabetes Association: Physical
Activity/Exercise and Diabetes (Position
Statement). Diabetes Care 27 (Suppl. 1):
S58–S62, 2004
3. American Diabetes Association: Stan-
dards of Medical Care in Diabetes (Posi-
tion Statement). Diabetes Care 27 (Suppl.
1):S15–S35, 2004
4. Knowler WC, the Diabetes Prevention
Program Research Group: Reduction in
the incidence of type 2 diabetes with life-
style intervention or metformin. N Engl
J Med 346:393–403, 2002
5. Lindstro¨m J, Louheranta A, Mannelin M,
Rastas M, Salminen V, Eriksson J, Uus-
itupa M, Tuomilehto J, the Finnish Diabe-
tes Prevention Study Group: The Finnish
Diabetes Prevention Study (DPS): lifestyle
intervention and 3-year results on diet
and physical activity. Diabetes Care
26:3230–3236, 2003
6. Schoeller DA, E van Santen: Measurement
of energy expenditure in humans by dou-
bly labeled water. J Appl Physiol 53:955–
959, 1982
Hypoglycemia
Following a
Nateglinide
Overdose in a
Suicide Attempt
V arious drugs are currently used forthe treatment of diabetes. Severalreports on overdose of these drugs,
especially on sulfonylurea and met-
formin, have been published (1,2); how-
ever, few such publications are available
on nateglinide, a rapid insulin secreat-
agogue. We report the first case of at-
tempted suicide by nateglinide overdose.
A 30-year-old Japanese nondiabetic
woman was transferred to the emergency
department of our hospital 1 h after (6:30
A.M.) ingesting 3,420 mg (38 tablets, 90
mg each) of nateglinide, which was pre-
scribed to her diabetic partner. She had a
mild psychiatric history and used minor
tranquilizers occasionally. Upon arrival to
the hospital, she was able to walk unaided
though she seemed drowsy. Blood pres-
sure was 120/80 mmHg, pulse rate was 78
beats/min, and peripheral oxygen satura-
tion was 96% on room air breathing.
Blood glucose concentration measured at
arrival (1 h after ingesting nateglinide:
7:30 A.M.) was 2.0 mmol/l. A bolus dose of
40 ml of 50% glucose was injected intra-
venously (iv). At 8:30 A.M., blood glucose
was increased to 3.1 mmol/l. Another 40
ml of 50% glucose was iv injected again,
which resulted in normalization of blood
glucose concentration. However, at 10:30
A.M., blood glucose was 1.2 mmol/l. An-
other 40 ml of 50% glucose was iv in-
jected, and blood glucose concentration
returned to 4.2 mmol/l and immunoreac-
tive insulin was 11 �U/ml. At 11:30 A.M.,
blood glucose concentration was 2.2
mmol/l. At that stage, a bolus of 20 ml of
50% glucose was iv injected, and a 10%
glucose drip infusion was started (40 ml/
h). At 12:30 P.M., the blood glucose con-
centration was 3.0 mmol/l, and at 1:30
P.M., the blood glucose returned to normal
range under 10% glucose infusion (40 ml/
h), and no more hypoglycemic episodes
were observed. The next day, the blood
glucose was 5.5 mmol/l and immunore-
active insulin was 5.0�U/ml without glu-
cose infusion, and she was discharged
without complications. Thus, the pro-
longed hypoglycemic effect of nateglinide
continued for 6 h, and intravenous glu-
cose supplementation (totally 100 g) was
sufficient to maintain euglycemia.
Nateglinide is a novel, highly physio-
logic mealtime glucose regulator recent-
ly approved for the treatment of type 2
diabetes. Compared with sulfonylurea,
nateglinide causes rapid but short stimu-
lation of insulin secretion. Factitious sul-
fonylurea overdosages usually result in
prolonged hypoglycemia (usually �48 h
in a case report [3]), and intravenous glu-
cose supplementation or octreotide injec-
tion is necessary to maintain euglycemia.
To our knowledge, overdosage of nateg-
linide has not been previously reported.
Repaglinide-induced factitious hypogly-
cemia has been reported, but the dose
taken was equivalent to that used clini-
cally (4). In normal rats fasted for 17 h,
oral administration of nateglinide pro-
duced a prompt (within 1 h) and dose-
dependent reduction in blood glucose,
but significant reduction was no longer
noticeable after 3 h, even at the dose of
100 mg/kg (5). In comparison, gliben-
clamide showed a slower onset of its hy-
poglycemic action and caused a sustained
decrease in blood glucose levels for at
least 6 h at a dose of 1 mg/kg (5).
In conclusion, an overdose of oral
nateglinide elicited prompt but mild, rel-
atively short lived, and reversible hypo-
glycemia. These features of hypoglycemia
resembled the hypoglycemic effects of
nateglinide in regular therapeutic dosages.
SHIHO NAKAYAMA, MD
TAKAHISA HIROSE, MD
HIROTAKA WATADA, MD
YASUSHI TANAKA, MD
RYUZO KAWAMORI, MD
From the Department of Medicine, Metabolism and
Endocrinology, Juntendo University School of Med-
icine, Tokyo, Japan.
Address correspondence to Dr. Takahisa Hirose,
Department of Medicine, Juntendo University
School of Medicine, 2-1-1 Hongo, Tokyo 113-8421,
Japan. E-mail: hirosemd@med.juntendo.ac.jp.
© 2005 by the American Diabetes Association.
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
References
1. Klonoff DC, Barrett BJ, Nolte MS, Cohen
RM, Wyderski R: Hypoglycemia follow-
ing inadvertent and factitious sulfonyl-
Letters
DIABETES CARE, VOLUME 28, NUMBER 1, JANUARY 2005 227

urea overdosages. Diabetes Care 18:563–
567, 1995
2. Teale KFH, Devine A, Stewart H, Harper
JH: The management of metformin over-
dose. Anaesthesia 53:691–701, 1998
3. Klonoff DC: Association of hyperinsulin-
emia with chlorpropamide toxicity. Am J
Med 84:33–38, 1988
4. Hirshberg B, Skarulis MC, Pucino F,
Csako G, Brennan R, Gorden P: Repa-
glinide-induced factitious hypoglyce-
mia. J Clin Endocrinol Metab 86:475–477,
2001
5. Ikenoue T, Okazaki K, Fujitani S, Tsu-
chiya Y, Akiyoshi M, Maki T, Kondo N:
Effect of a new hypoglycemic agent,
A-4166 [(-)-N-(trans-4-isopropyl-cyclo-
hexanecarbonyl)-D phenylalanine], on
postprandial blood glucose excursion:
comparison with voglibose and gliben-
clamide. Biol Pharm Bull 20:354 –359,
1997
Increased
Prevalence of
Diabetes and
Obesity in Patients
With Salivary Gland
Tumors
I t has been recently revealed that type 2diabetes is a risk factor for tumorousproliferation at different sites (1).
Breast and colon cancers were the first tu-
mors for which an epidemiological corre-
lation was demonstrated between this
glucose metabolism disorder and tumor
incidence (2). The “western lifestyle,” first
of all obesity, increases the risk of not only
type 2 diabetes but also of overall cancer
promotion (3).
Swelling of the parotids and a de-
creased salivary flow rate caused by de-
generative alterations in the acinar cells
are common concomitants of diabetes
and dyslipidemias (4,5). However, no
data are available concerning the epide-
miological correlation between metabolic
disorders and salivary gland tumors.
In the present work, a retrospective
controlled study was performed to clarify
the possibility of an association between
type 2 diabetes, obesity, and salivary
gland tumors. Clinical data on 438 inpa-
tients were analyzed on the basis of their
case histories. In 224 patients, salivary
gland tumors had been surgically re-
moved and histologically diagnosed (SGT
group). The mean age of these patients
was 51.2 years (range 23–87). A total of
118 were men and 106 were women.
Their tumors were predominantly benign
(166 cases, 74.2%), and a majority of
them were located in the parotid. The rate
of malignancies was 25.8% (58 cases),
and the most frequent locations were the
accessory glands of the palate. Two hun-
dred and fourteen randomly selected pa-
tients undergoing dental surgery served
as the control group. They had no tumor-
ous lesions in the oral or maxillofacial re-
gion. Their mean age was 50.8 years
(range 24–80). A total of 104 were men
and 110 were women. Fasting blood glu-
cose levels were routinely measured re-
peatedly within 4 days before the surgical
intervention. The level was regarded as
pathological only if it was repeatedly
higher than normal (�6.9 mmol/l). Pa-
tients with known and treated disease or
with a newly diagnosed high fasting glu-
cose levels were included in the type 2
diabetic group. BMI was also registered in
the case histories. Patients with BMI �30
kg/m2 were regarded as obese. The �2 test
was used for the statistical analysis. A
probability level of 5% was taken as a
limit of statistical significance.
Type 2 diabetes was established in 51
patients (22.8%) of the SGT group and in
14 subjects (6.5%) in the tumor-free con-
trol group (P � 0.001). Among the dia-
betic and nondiabetic patients of the SGT
group, the rates of histologically malig-
nant tumors were 26.6 and 23.5%, re-
spectively (P � 0.05). Obesity was
registered in 102 patients (45.5%) in the
SGT group and in 38 subjects (17.7%) in
the control group (P � 0.001). The rates
of malignant tumors among the obese and
nonobese SGT patients were 27.8 and
23.5%, respectively (P � 0.05).
These findings permit a novel hy-
pothesis concerning the epidemiological
association between type 2 diabetes, obe-
sity, and salivary gland tumors. However,
there were no close correlations between
these metabolic disorders and the malig-
nancy rate of the tumors.
ZSUZSANNA SUBA, PHD
JO´ZSEF BARABA´S, PHD
GYO¨RGY SZABO´, DSC
DANIEL TAKA´CS
MA´RTA UJPA´L, PHD
From the Semmelweis University Faculty of Den-
tistry, Oral and Maxillofacial Surgery, Budapest,
Hungary.
Address correspondence to Dr. Ma´rta Ujpa´l, Sem-
melweis University, Department of Oral and Maxil-
lofacial Surgery, Ma´ria Str. 52., H-1085 Budapest,
Hungary. E-mail: suba@szajseb.sote.hu.
© 2005 by the American Diabetes Association.
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
References
1. Verlato G, Zoppini G, Bonora E, Muggeo
M: Mortality from site-specific malignan-
cies in type 2 diabetic patients from Ve-
rona. Diabetes Care 26:1047–1051, 2003
2. Bloomgarden ZT: Definitions of the insu-
lin resistance syndrome: the 1st World
Congress on the Insulin Resistance Syn-
drome. Diabetes Care 27:824–830, 2004
3. Pan SY, Johnson KC, Ugnat AM, Wen SW,
Mao Y, the Canadian Cancer Registries
Epidemiology Research Group: Associa-
tion of obesity and cancer risk in Canada.
Am J Epidemiol 159:259–268, 2004
4. Rusotto SB: Asymptomatic parotid gland
enlargement in diabetes mellitus. Oral
Surg Oral Med Oral Pathol 52:594–598,
1981
5. Coleman H, Altini M, Nayler S, Richards
A: Sialoadenosis: a presenting sign of bu-
limia. Head & Neck 20:758–762, 1998
COMMENTS AND
RESPONSES
Telecare for Patients
with Type 1 Diabetes
and Inadequate
Glycemic Control
Response to Montori et al.
The insightful study of Montori et al.(1) raises the question of whethertelecare can close the gap between
the necessary and frequent interventions
by health care professionals in order to
lower the HbA1c of patients. We would
like to comment on the control group and
meta-analysis in their study.
Telecare can be embedded into differ-
ent achievable scenarios, where the left-
most cornerstone considers a control
group and where conventional care
means no intervention (A). The right-
most cornerstone is to replace face-to-face
interventions by telecare with the same
L E T T E R S
228 DIABETES CARE, VOLUME 28, NUMBER 1, JANUARY 2005

frequency and quality as in the control
group (B). In reality, the typical group de-
fining “conventional care” would be situ-
ated between these two cornerstones. Due
to regional differences, the private situa-
tion, and, particularly, the distance to the
next specialist, the patients would not be
able to achieve the necessary frequency of
face-to-face visits that they would if the
doctor’s office were next door (C).
In a comparative study, medical out-
come (HbA1c) and costs would be differ-
ent depending on whether strategy A or B
or something in between was adopted. A
cost-effectiveness study is particularly
mandatory in strategy B, where the pri-
mary outcome parameter would be cost
and the secondary parameter HbA1c.
In a randomized cost-effectiveness
study (2), we adopted strategy B in order
to isolate the effect of telecare and not any
effect attributable to the frequency of in-
terventions. A similar study by Chase et.al
(3) is also situated closely to strategy B.
Both studies came to the conclusion that
telecare can save costs while maintaining
the quality of care (HbA1c).
In another recent study (4), we tried
to adopt an approach that better simulates
a patient’s “real-life” (C) by having the pa-
tients in the control group being educated
first in hospital and then returning to their
private practitioners. We believe that due
to the different goals of the studies and the
different control groups, a meta-analysis
would ultimately draw equivocal or false
conclusions.
In conclusion, we would suggest add-
ing the method of whether a cost-
effectiveness analysis or an outcome
analysis has been performed to a further
meta-analysis in telecare. However, the
use of excellent tools such as intention-to-
treat analysis and the high degree of trans-
parency makes the report of Montori et al.
groundbreaking and a critical starting
point for further studies in this field.
EBERHARD BIERMANN, MD
NORM BOGNER
JULIAN RIHL, MD
EBERHARD STANDL, MD
From the Institute for Diabetes Research, City Hos-
pital Munich-Schwabing, Muenchen, Bavaria, Ger-
many.
Address correspondence to Eberhard Biermann,
MD, Institute for Diabetes Research, City Hospital
Munich-Schwabing, Koelnerplatz 1, Muenchen,
80804, Germany. E-mail: eberhard.biermann@
lrz.uni-muenchen.de.
© 2005 by the American Diabetes Association.
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
References
1. Montori VM, Helgemoe PK, Guyatt GH,
Dean DS, Leung TW, Smith SA, Kudva
YC: Telecare for patients with type 1 dia-
betes and inadequate glycemic control: a
randomized controlled trial and meta-
analysis. Diabetes Care 27:1088–1094,
2004
2. Biermann E, Dietrich W, Rihl J, Standl E:
Are there time and cost savings by using
telemangement for patients on intensified
insulin therapy? A randomised, con-
trolled trial. Comput Methods Programs
Biomed 69:137–146, 2002
3. Chase HP, Pearson JA, Wightman C, Rob-
erts MD, Oderberg AD, Garg SK: Modem
transmission of glucose values reduces
the cost and need for clinic visits. Diabetes
Care 26:1475–1479, 2003
4. Biermann E, Rihl J, Standl E: Outcome
and efficiency of restoration of metabolic
control of insulin dependent diabetics by
telemonitoring (Abstract). Diabetologia 47
(Suppl. 1):A348, 2004
Telecare for Patients
With Type 1
Diabetes and
Inadequate Glycemic
Control
Response to Biermann et al.
W e thank Biermann et al. (1) fortheir kind remarks about theclarity and importance of our re-
search and for proposing a novel taxon-
omy of telecare interventions by purpose:
as a supplement to usual care (A in their
taxonomy) or as a replacement of face-to-
face visits (sometimes called nonvisit en-
counters; B or C in their taxonomy).
Two of the trials included in our
meta-analysis used, to some extent, tele-
care as replacement. In their 6-month
trial, Chase et al. (2) added telecare en-
counters to usual care and replaced the
face-to-face visit at 3 months with a tele-
care nonvisit encounter. Biermann et al.
(3) replaced all ad hoc visits with telecare
encounters but kept face-to-face visits at
2, 4, 6, and 8 months for patients allo-
cated to both telecare and control. These
two replacement trials found no signifi-
cant difference between telecare and con-
trol; however, this is not enough to claim
that telecare visits are similar or superior
(i.e., not inferior) to face-to-face visits.
To determine noninferiority, one
needs to evaluate the magnitude of the
smallest reduction in HbA1c that is still
consistent with the trial data, i.e., the
lower limit of the confidence interval (4).
Data from the two trials of telecare as re-
placement were consistent with differ-
ences in HbA1c as large as 0.7% in favor of
usual care. Thus, evidence of noninferior-
ity from each of these trials alone is incon-
clusive.
Following the suggestions of Bier-
mann et al. (1), we evaluated a subgroup
effect across trials in our meta-analysis by
telecare purpose (replacement versus
supplement). There was no difference
(test for interaction, P � 0.44) in pooled
treatment effects between trials of telecare
as replacement (pooled random-effects
estimate 0.5% [95% CI 0.0–1.0%]) and
as supplement (0.2% [�0.4 – 0.8%]).
Therefore, the results of our meta-analysis
(0.4% [0.0–0.8%]) apply to both types of
trials, are not misleading, and suggest that
telecare may be noninferior to usual care
when it supplements usual care or when it
replaces some visits.
We agree that a systematic assessment
of costs with attention to the opportunity
costs of implementing telecare and to the
resulting freed-up resources available to
patients who are most likely to benefit
from face-to-face visits rather than tele-
care is critical when considering new
technologies. Unfortunately, to our
knowledge, a formal cost-effectiveness
analysis of telecare as a supplement or re-
placement to usual care is not available.
VICTOR M. MONTORI, MD, MSC1
GORDON H. GUYATT, MD, MSC2
STEVEN A. SMITH, MD1
YOGISH C. KUDVA, MD1
From the 1Divisions of Diabetes and Internal Medi-
cine, Mayo Clinic College of Medicine, Rochester,
Minnesota; and the 2Department of Clinical Epide-
miology and Biostatistics and Medicine, McMaster
University, Hamilton, Ontario, Canada.
Address correspondence to Yogish C. Kudva,
MD, MSc, W18A, Mayo Clinic, 200 First St. SW,
Rochester, MN 55905. E-mail: kudva.yogish@
mayo.edu.
© 2005 by the American Diabetes Association.
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
References
1. Biermann E, Bogner N, Rihl J, Standl E:
Telecare for patients with type 1 diabetes
and inadequate glycemic control (Letter).
L E T T E R S
DIABETES CARE, VOLUME 28, NUMBER 1, JANUARY 2005 229