P\.01) and see fewer patients (123 vs 202.3, P\.

01)
than their male counterparts (Table II).
Osteopathic dermatologists spent the majority
of their time practicing medical dermatology (21
hours per week). The average amount of time spent
performing noncosmetic surgical care was 9.2 hours
per week followed by cosmetic care, 4.3 hours per
week. Male and female osteopathic dermatologists
practiced the most number of hours in medical
dermatology followed by noncosmetic and cosmetic
dermatology (Table II). When queried about their
future scope of practice, 85% of males and females
responded that they would practice the same
amount or more of medical dermatology.
Despite philosophical differences in their training,
osteopathic dermatologists appear to have practice
proles that are substantially similar to allopathic
dermatologists when using the data from the parallel
survey administered to allopathic dermatologists in
2002 and 2005. Signicant differences are that oste-
opathic dermatologists appear more likely to work in
solo practices (58% vs 44%) and see more patients
per week (200 vs 142).
3
Limitations of this study include potential re-
sponse bias given the response rate of 36%.
However, this rate is consistent with the expected
response for this kind of survey and the baseline
demographic distribution at least for gender is
representative.
Osteopathic physicians in the United States in
dermatology remain a minority of practitioners, repre-
senting only 3% of the dermatologic workforce.
4
However, as the demand for dermatological services
remains high in the United States, the growth in
osteopathic dermatologists will most likely continue
in response to this demand. The growth of osteo-
pathic dermatologists will have a substantial impact
in regions where they represent a significant
proportion of the workforce.
Jane Y. Yoo, BS, MPP,
a
Hang Lee, PhD,
b
and Alexa
B. Kimball, MD, MPH
c
Medical College of Virginia, Richmond,
a
Harvard
Medical School, Massachusetts General Hospital
Biostatistics Center, Boston,
b
and Clinical Unit
for Research Trials in Skin, Department of Der-
matology, Harvard Medical School, Boston
c
Funding sources: None
Conicts of interest: None
This work presented as a poster at the 64th Annual
Meeting of the American Academy of Dermatol-
ogy, 2005
Reprint requests: Alexa B. Kimball, MD, MPH,
Director, Clinical Unit for Research Trials in
Skin (CURTIS), Department of Dermatology,
Harvard Medical School, 50 Staniford St, Suite
240, Boston, MA 02114
E-mail: [email protected]
REFERENCES
1. Jacobson CC, Resneck JS Jr, Kimball AB. Generational differ-
ences in practice patterns of dermatologists in the United
States: implications for workforce planning. Arch Dermatol
2004;140:1477-82.
2. Jacobson CC, Nguyen JC, Kimball AB. Gender and parenting
significantly affect work hours of recent dermatology program
graduates. Arch Dermatol 2004;140:191-6.
3. 2005 AADA practice profile surveys administered to allopathic
derms in 2002 and 2005. Dermatology World 2006;16(6):1.
4. Physician characteristics and distribution in the US, 2006.
Department of Data Quality and Measurement, Division of
Data Operations, American Medical Association, 2006.
doi:10.1016/j.jaad.2006.10.958
Costs of pulse versus continuous terbinane
for onychomycosis
To the Editor: Onychomycosis, a fungal infection of
the nail, affects 40% of persons older than 61 years
of age.
1
Results from our recently published double-
blind, controlled trial demonstrated the superiority
of continuous over pulse-dose terbinafine for the
treatment of onychomycosis.
2
Rates of complete
Table II. Productivity
All (N = 88) Males (n = 68) Females (n = 20) P value
Patient hours/wk 34.1 6 9.2* 36.4 6 8.7 26.5 6 7.1 \.01
No. of patients/wk 184 6 95.3 202 6 97.1 123 6 57 \.01
Days/wk 4.2 6 0.8 4.3 6 0.5 3.6 6 1.1 \.01
Scope of practice
Medical 20.6 6 9.2
y
21.7 6 9.8 17 6 7.1 \.05
Noncosmetic 9.2 6 7.8 10.3 6 8.1 5.9 6 5.3 \.01
Cosmetic 4.3 6 7.8 4.5 6 8.6 3.6 6 4.2 NS
NS, Not statistically significant.
*All data represent mean 6 standard deviation.
y
Hours.
J AM ACAD DERMATOL
VOLUME 56, NUMBER 3
Letters 525
cure (mycological and clinical cure) of the target toe
were relatively low (40.5% and 28.0% for continuous
and pulse-dose, respectively). Although not as ef-
fective in the overall study group, pulse therapy may
be more cost-effective, especially in certain sub-
groups, such as those with mild disease. The purpose
of this report is to describe calculated medication
costs as well as total costs per cure using data from
our published trial for baseline disease severity
subgroups.
The average wholesale cost for terbinane, in
US dollars, was obtained from the Red Book listing.
3
For each treatment regimen (continuous and pulse-
dose), the total cost of medication was calculated by
multiplying the cost of medication by the dosing
schedule. Costs per complete cure were calculated
by dividing the cost of medication by the complete
cure rates for the target toenail. Total costs were
calculated using the Toenail Onychomycosis Eco-
nomic Model,
4
a computerized economic model that
includes not only the cost of medication but also
costs of physician visits, laboratory tests, procedures,
and management of adverse events and which is
very similar to Cost-of-Regimen Analysis.
5
We also
calculated medication and total cost per cure using
a downstream model that assumed that one-half
of patients who failed treatment will complete a
second course of treatment with continuous-dose
terbinafine regardless of the previous treatment
schedule (continuous or pulse-dose).
Both continuous and pulse-dose terbinane
were statistically more effective in patients with mild
disease as compared to those with moderate/severe
disease (P # .04; Table I). In the group with only
mild disease, though the cure rate for continuous-
dose was higher than that of pulse-dose, this differ-
ence was not statistically significant (P = .30). Table II
summarizes medication and total cost per cure for
continuous and pulse-dose terbinafine. In all cases,
the medication costs, as well as total costs per cure,
of continuous-dose therapy were greater than that
for pulse-dose therapy. The medication costs, as well
as total costs per cure, for treatment of moderate/
severe disease were markedly higher than that for
mild disease (Table II).
The use of the average wholesale cost, whichoften
does not reect the actual price paid for medication,
limits the generalizability of this study. Cost is only
one factor in the decision to treat a patient with
onychomycosis; other important factors to consider
are patient preference, clinical severity, onychomy-
cosis pattern, treatment side effects, risk of reinfec-
tion, and comorbidities. Although this trial was not
powered for subgroup analysis, this data may help
physicians counsel patients with onychomycosis,
especially those with mild disease and those whose
insurance does not cover medication costs.
Sarah E. Schram, MD,
a
and Erin M. Warshaw,
MD, MS
a,b
Departments of Dermatology at the Minneapolis
Veterans Affairs Medical Center
a
and the
University of Minnesota, School of Medicine,
b
Minneapolis, Minn
Supported by the Minneapolis Veterans Affairs
Medical Center and its Health Services Research
Center of Excellence for Chronic Disease Out-
comes Research.
Conicts of interest: E. W. has conducted clinical
trials and served as a consultant for several
Table I. Target toenail complete cure rates for
continuous- and pulse-dose terbinafine
Disease severity Continuous Pulse P value
Mild (n = 49)* 60.9%
y
46.2%
z
.30
Moderate/severe
(n = 240)
37.4% 25.0% .02
Overall (n = 289) 40.5% 28.0% .02
*Mild disease is defined as \75% involvement of the target
toenail.
y
Statistically significantly higher than in moderate/severe disease
(P = .04).
z
Statistically significantly higher than in moderate/severe disease
(P = .02).
Table II. Costs per complete cure (target toenail) for continuous- and pulse-dose terbinafine
with and without downstream modeling
Medication costs/cure
Medication costs/cure
with downstream model* Total costs/cure
Total costs/cure
with downstream model*
Disease severity Continuous Pulse Continuous Pulse Continuous Pulse Continuous Pulse
Mild $1,671 $1,102 $3,342 $2,773 $2,439 $2,114 $4,878 $4,553
Moderate/severe $2,720 $2,034 $5,440 $4,754 $3,972 $3,908 $7,944 $7,880
Overall $2,509 $1,818 $5,018 $4,328 $3,668 $3,489 $7,336 $7,157
*Downstream model assumes that half of treatment failures will take a second course of continuous-dose terbinafine (250 mg/day for
3 months).
J AM ACAD DERMATOL
MARCH 2007
526 Letters
pharmaceutical companies which produce anti-
fungal medications, including Dermik, Colo-
plast, Clay Park, Paddock, Novartis, and Barrier.
The views expressed in this article are those of the
authors and do not necessarily reect the position
or policy of the Department of Veterans Affairs.
Reprints not available from the authors.
Correspondence to: Erin Warshaw, MD, MS, Derma-
tology Department 111K, Veterans Affairs Medical
Center, One Veterans Dr, Minneapolis, MN 55417
E-mail: [email protected]
REFERENCES
1. Elweski BE, Charif MA. Prevalence of onychomycosis in patients
attending a dermatology clinic in northeastern Ohio for other
conditions. Arch Dermatol 1997;133:1172-3.
2. Warshaw EM, Fett DD, Bloomfield HE, Grill JP, Nelson DB,
Quintero V, et al. Pulse versus continuous terbinafine for
onychomycosis: a randomized, double-blind, controlled trial.
J Am Acad Dermatol 2005;53:578-84.
3. Fleming T, editor. 2006 Red Book: pharmacys fundamental
reference. Montvale (NJ): Thomson Healthcare; 2006.
4. Omar MA, Kahler KH. Cost-effectiveness analysis of therapies
for the treatment of toenail onychomycosis. Presented at the
61st annual meeting of the American Academy of Dermatol-
ogy, San Francisco, Calif, March 21-26, 2003.
5. Casciano J, Amaya K, Doyle J, Arikian S, Shear N, Haspel M, et al.
Economic analysis of oral and topical therapies for onychomy-
cosis of the toenails and fingernails. Managed Care 2003;12:
47-57.
doi:10.1016/j.jaad.2006.11.027
Sorafenib-induced erythema multiforme
To the Editor: Cutaneous toxicity is common in
patients receiving sorafenib (Nexavar), a new oral
multikinase inhibitor used to delay disease prog-
ression in advanced solid organ malignancies
and metastatic melanoma. Rash and hand-foot
syndrome is reported in the majority of patients
and appears to represent a toxic, dose-dependent
reaction.
1,2
We report a case of sorafenib-induced
erythema multiforme mimicking this common
eruption.
A 50-year-old woman had pulmonary metastases
from a 5-mm amelanotic melanoma excised in 2003.
She failed a right lobectomy, interleukin 2 therapy,
and chemotherapy (carboplatin and docetaxel).
Two weeks before our evaluation, sorafenib was
initiated at 200 mg orally twice a day. Four days after
a dose increase to 400 mg twice a day, she devel-
oped a generalized rash and a burning sensation in
her lips, palms, and soles. Physical examination
revealed tender erythematous papules and plaques
with dusky or pseudovesicular centers over her
face, trunk, extremities, palms, and soles (Fig 1).
Her lips were hyperemic and fissured. There were
no mucosal lesions. Biopsy specimen revealed a
superficial, perivascular lymphocytic infiltrate and
necrotic keratinocytes, consistent with erythema
multiforme (Fig 2). Three weeks later she was
rechallenged with a single 100-mg dose and devel-
oped a tender erythematous eruption within 24
hours. Prednisone (60 mg/d) and topical triamcino-
lone 0.1% cream resulted in significant improvement
within days.
Sorafenib, or BAY 43-9006, has multiple actions
in vitro. It inhibits Raf-1, wild-type B-Raf, V599E
b-raf kinases, proangiogenic vascular endothelial
growth factor receptor-2, vascular endothelial growth
factor receptor-3, and platelet-derived growth factor
receptor-b tyrosine kinases.
3
It also inhibits phos-
phorylation of Flt3, c-KIT, and p38aa member of
the mitogen-activated protein kinase family.
3
In vivo
action is incompletely understood. It has been
shown to delay time to progression in advanced
refractory renal and hepatocellular carcinoma
4
;
more than 70 clinical trials are in progress or
recruiting patients to evaluate the antineoplastic
potential of sorafenib in other solid organ tumors
and metastatic melanoma. The most frequently
reported toxicities attributed to sorafenib include
rash, hypertension, fatigue, anorexia, and diarrhea.
Up to 93% of patients receiving sorafenib as mono-
therapy experience cutaneous effects including: rash
Fig 1. Erythematous papules and plaques with dusky
central necrosis.
J AM ACAD DERMATOL
VOLUME 56, NUMBER 3
Letters 527