Incidence and prevalence of basal cell

carcinoma (BCC) and locally advanced

BCC (LABCC) in a large commercially
insured population in the United States:
A retrospective cohort study
Gary Goldenberg, MD,a Tom Karagiannis, PharmD,b Jacqueline Blanche Palmer, PharmD,b Juzer Lotya, MSc,c
Caitriona ONeill, PhD,c Renata Kisa, MD,b Vivian Herrera, DDS,b and Daniel M. Siegel, MDd
New York, New York; East Hanover, New Jersey; and Dublin, Ireland

Background: Accurate evaluation of basal cell carcinoma (BCC) in the United States was not possible
before the 2011 release of BCC-specific International Classification of Diseases, Ninth Revision, Clinical
Modification codes.

Objective: We sought to describe BCC (including locally advanced BCC [LABCC]) incidence/prevalence
and the characteristics of patients in a commercially insured US population.

Methods: This retrospective cohort study used Truven Health MarketScan database insurance claims.
Patients, aged 18 years or older with 2 or more BCC claims at least 30 days apart from October 1, 2011, to
September 30, 2012, were continuously enrolled in medical and pharmacy benefits for 12 months before
and after the index claim. A specific algorithm was used to classify patients with LABCC.

Results: A total of 56,987 patients with BCC were identified (39,035 incident cases; 17,952 prevalent cases).
Age-adjusted BCC incidence and prevalence were 226.09 and 342.64 per 100,000 persons, respectively.
These values project to 542,782 patients (incidence) and 822,593 patients (prevalence) in the 2012 US
population. LABCC was uncommon (471 cases identified; projected US incidence and prevalence: 4399 and
7940 patients, respectively).

Limitations: Use of medical claims data and retrospective analysis are limitations.

Conclusion: In a study designed to distinguish patients with LABCC from the general BCC population
based on BCC-specific International Classification of Diseases, Ninth Revision, Clinical Modification
codes, 0.8% were found to have LABCC, the majority having pre-existing disease. ( J Am Acad Dermatol
2016;75:957-66.)

Key words: basal cell carcinoma; epidemiology; incidence; locally advanced disease; metastatic disease;
nonmelanoma skin cancer; prevalence; retrospective claims analysis.

From Mount Sinai School of Medicine, New Yorka; Novartis
Pharmaceuticals Corp, East Hanoverb; Novartis Ireland Ltd,
Dublinc; and State University of New York Downstate Medical
Center College of Medicine.d
Novartis Pharmaceuticals Corp supported this study and publica-
tion, and editorial support from BioScience Communications,
New York, NY.
Disclosure: Dr Goldenberg has served as a consultant to Genen-
tech, Novartis, and Xoft, and as a speaker for Genentech and
Novartis. Dr Karagiannis was a full-time employee of Novartis at
the time the study was conducted. Drs Palmer, ONeill, Kisa, and
Herrera, and Mr Lotya are full-time employees of Novartis. Dr
Siegel has served as an investigator, speaker, and consultant for
Genentech, LEO Pharma, and Valeant, and as a consultant for
Novartis.
Information contained in this article has been presented, in part, at
the following medical congresses: 73rd AAD Annual Meeting,
March 20-24, 2015, San Francisco, CA; AMCP 2015 Nexus,
October 26-29, 2015, Orlando, FL; and ISPOR 20th Annual
International Meeting, May 16-20, 2015, Philadelphia, PA.
Supplementary tables are online-only.
Accepted for publication June 12, 2016.
Reprint requests: Gary Goldenberg, MD, Mount Sinai School of
Medicine, 5 E 98th St, 5th Floor, New York, NY 10029. E-mail:
garygoldenbergmd@gmail.com.
Published online July 27, 2016.
0190-9622/$36.00
2016 by the American Academy of Dermatology, Inc.
http://dx.doi.org/10.1016/j.jaad.2016.06.020

957
958 Goldenberg et al J AM ACAD DERMATOL
NOVEMBER 2016

Basal cell carcinoma (BCC), a subset of non- to classify these subgroups. The demographic and
melanoma skin cancer (NMSC), is the most common clinical characteristics of the overall BCC cohort
cancer in the United States, with an estimated 2.7 and these 2 subgroups are also described.
million cases diagnosed annually.1,2 Total costs for
skin cancer increased substantially from 2002 to METHODS
2011eto a greater extent than cost increases for all Study design
other cancer siteseowing to increases in the number This was a retrospective cohort study using health
of persons treated, and to insurance claims data from
per-person treatment costs.3 the Commercial Claims and
The overwhelming majority CAPSULE SUMMARY Encounters database within
of patients with BCC can be the Truven Health
dAccurate evaluation of basal cell
cured using a variety of treat- MarketScan database. These
carcinoma (BCC) epidemiology was not
ment modalities, including data contain claims from
possible before release of BCC-specific
curettage, excision, Mohs mi- approximately 100
International Classification of Diseases,
crographic surgery, radia- employers, health plans,
Ninth Revision, Clinical Modification
tion, and topical therapy; and government and public
codes.
however, a small proportion organizations, representing
have disease that is refrac- dIn a commercially insured US population, approximately 30 million
tory, inoperable, or metasta- 2012 BCC incidence and prevalence covered lives. All US census
tic. The term advanced estimates projected to 542,782 and regions are represented, with
BCC refers to locally inva- 822,593 patients, respectively; locally the most predominant being
sive BCC that is not amenable advanced BCC was uncommon (\1%). the South and North Central
to surgical or radiation d
Accurate BCC epidemiology estimates (Midwest) regions. Enrollees
therapy interventions, or will facilitate optimal resource allocation. in the database include em-
that has metastasized to ployees, dependents, and
regional lymph nodes or retirees with primary or
beyond.4-6 Medicare supplemental coverage through privately
Historically, BCC has been excluded as a category insured fee-for-service, point-of-service, or capitated
from cancer registry databases because the health plans. All study data were accessed using
number of patients with these generally localized techniques compliant with the Health Insurance
tumors greatly exceeds the number of patients with Portability and Accountability Act of 1996. No
other malignancies. Moreover, because administra- identifiable protected health information was
tive data could not distinguish among NMSC extracted, so this study did not require informed
subtypes, previous studies have focused on NMSC consent or institutional review board approval.
as a whole.1,7-9 The release of BCC-specific
International Classification of Diseases, Ninth Identification of patients with BCC
Revision, Clinical Modification (ICD-9-CM) codes Patients aged 18 years or older with at least 2
in October 2011 has allowed for more accurate claims and a diagnosis of BCC in any location during
epidemiologic analysis of this malignancy. the identification period (October 1, 2011, to
In a previous study, an algorithm was devel- September 30, 2012), with the claims separated by
oped to classify patients with commercial at least 30 days, were identified. A diagnosis of BCC
insurance and NMSC into 3 groups: patients with was based on any of the following ICD-9-CM codes:
metastatic disease, patients with locally advanced 173.01, 173.11, 173.21, 173.31, 173.41, 173.51,
disease, and all others.9 Although demographic 173.61, 173.71, 173.81, or 173.91. The date of first
data and insurance coverage were described, diagnosis of BCC during the identification period
information on patient comorbidities was not was designated as the index date. Included patients
provided. The study focused on NMSC, as were continuously enrolled in fee-for-service
BCC-specific ICD-9-CM codes were unavailable. medical and pharmacy benefit plans during the
In contrast, the current study used BCC-specific 12-month periods before and after the index date.
ICD-9-CM codes to estimate the incidence and For continuous enrollment, patients could not have a
prevalence of BCC, locally advanced BCC gap in enrollment of more than 40 days. Patients
(LABCC), and metastatic BCC (MBCC) in a were further characterized as having incident or
commercially insured US population. Because prevalent BCC according to the absence or presence,
specific ICD-9-CM codes are not available for respectively, of an NMSC claim in the 12-month
LABCC and MBCC, additional criteria were used period before the BCC index date (ICD-9-CM 173.x
J AM ACAD DERMATOL
VOLUME 75, NUMBER 5
Abbreviations used:
BCC: basal cell carcinoma
ICD-9-CM: International Classification of Dis-
eases, Ninth Revision, Clinical
Modification
LABCC: locally advanced basal cell carcinoma
MBCC: metastatic basal cell carcinoma
NMSC: nonmelanoma skin cancer
or 173.xx; allowing for the capture of both 4-digit
and 5-digit MarketScan records).
Patients with BCC were classified as having MBCC
if they had at least 2 diagnosis codes for metastatic
disease in any location, with the claims separated by
at least 30 days in the postindex period (ICD-9-CM
196.0, 196.1, 196.5, 196.6, 196.8, 196.9, 197.0, 197.4,
197.5, 197.6, 197.7, 197.8, 198.3, 198.4, 198.5, 198.6,
198.7, 198.89, 199.0, or 199.1). Classification as
incident or prevalent MBCC was based on the
absence or presence, respectively, of an NMSC claim
(ICD-9-CM 173.x or 173.xx) in the 12-month
period before the BCC index date. Patients with a
diagnosis of a primary malignancy other than NMSC
(ICD-9-CM 140.x-172.x, 174.x-195.8, 200.x-208.x, or
238.6) during the preindex or postindex period were
excluded from the MBCC cohort.
Patients with BCC were classified as having
LABCC if they did not have metastatic disease, had
a BCC diagnosis code in any location, and met 1 of
the following criteria: 2 medical oncology office
visits or outpatient visits on separate days; 2
Current Procedural Terminology codes for radiation
therapy on separate days; or at least 1 outpatient or
office visit to 2 or more physician specialties
including a medical oncology visit, otolaryngology
(ear, nose, and throat) visit, and head and neck
surgeon visit, or a radiation oncology Current
Procedural Terminology code. A diagnosis code for
BCC on the lip, eyelid, ear, face, scalp, or neck was
required for the radiation oncology Current
Procedural Terminology code, otolaryngology visit,
and head and neck surgeon visit. Classification as
incident or prevalent LABCC was based on the
absence or presence, respectively, of an NMSC claim
(ICD-9-CM 173.x or 173.xx) in the 12-month period
before the BCC index date. Patients with a diagnosis
of a primary malignancy other than NMSC (ICD-9-
CM 140.x-172.x, 174.x-195.8, 200.x-208.x, or 238.6)
during the preindex or postindex period were
excluded from the LABCC cohort.
Variables
Demographic variables included age at the index
date, gender, insurance type, and region. Clinical
Goldenberg et al 959
variables included myocardial infarction, congestive
heart failure, peripheral vascular disease, cerebro-
vascular disease, dementia, chronic pulmonary
disease, rheumatic disease, peptic ulcer disease,
diabetes with or without complications, hemiple-
gia/paraplegia, renal disease, other malignancy,
metastatic solid tumor, liver disease, and AIDS/HIV.
The Charlson Comorbidity Index score was
calculated from each patients comorbidities.10
Statistical analysis
Patient characteristics were descriptively evalu-
ated for the overall BCC cohort, and for the MBCC
and LABCC subgroups. We used x 2 tests to test for
differences in categorical variables between incident
and prevalent cases, and the Mann-Whitney U test
and unequal variance 2-sample t tests were used to
test for differences in continuous variables. The
incidence rates for BCC, LABCC, and MBCC were
calculated as the number of new cases in the
respective classification 3 100,000/person-time at
risk. The prevalence rates for BCC, LABCC, and
MBCC were calculated as the number of persons
with the respective classification 3 100,000/number
of persons in the population. Age-adjusted incidence
and prevalence rates were calculated per 100,000
persons continuously enrolled in the health plan
during the preindex and identification periods, and
were standardized to the 2012 US population. The
age-adjusted incidence rate was determined as the
crude incidence rate per age category/person-time at
risk 3 population weight. The term age-adjusted
incidence rate referred to the proportion of
persons in the corresponding age groups of the
2012 projections according to the 2010 US
census population.11 Age-adjusted prevalence was
calculated in the same manner as age-adjusted
incidence. Gender-adjusted incidence and preva-
lence rates were calculated in a similar manner to
the age-adjusted rates.
RESULTS
Demographic and clinical characteristics
A total of 56,987 patients were identified with BCC
during the study identification period, and had
continuous medical and pharmacy benefits during
the 12-month periods before and after the index
date. Of these, 39,035 (68.5%) were incident cases
and 17,952 (31.5%) were prevalent cases (Fig 1). The
mean age of the BCC cohort was 64.9 6 13.8 years;
patients with prevalent BCC were older than those
with incident BCC (P \.0001) (Tables I and II). The
majority of patients with BCC were male (57.6%);
prevalent cases had a higher proportion of male
patients (P \ .0001). The South had the highest
960 Goldenberg et al
Fig 1. Flow chart and attrition of study sample. BCC, Basal
cell carcinoma; NMSC, nonmelanoma skin cancer.
proportion of BCC cases overall (36.6%). The mean
Charlson Comorbidity Index score was 1.0 (0.9
for incident cases and 1.4 for prevalent cases).
Diabetes without chronic complications was the
most common comorbidity.
The LABCC cohort consisted of 471 patients (0.8%
of the BCC cohort); the mean age was
73.0 6 13.3 years and 58% were male (Table II). In
this cohort, there were 261 incident cases (55.4%)
and 210 prevalent cases (44.6%) (Table I). Age,
insurance type, and employment status did not differ
significantly between incident and prevalent cases.
However, there were significant differences in
gender (P = .021) and Charlson Comorbidity Index
score (P = .026). The mean Charlson Comorbidity
Index score was 1.2 6 1.6 (1.0 for incident cases and
1.3 for prevalent cases). Diabetes without chronic
complications was the most common comorbidity.
In all, 23 patients were categorized as having
MBCC during the study identification period (0.04%
of the BCC cohort), including 7 incident cases and 16
prevalent cases. The mean age was 71.9 6 11.8 years
and the majority of this subgroup (69.6%) was male
J AM ACAD DERMATOL
NOVEMBER 2016
(Table II). The mean Charlson Comorbidity Index
score was 5.2 6 3.7. Common comorbidities in the
MBCC cohort included metastatic solid tumor (52%),
diabetes with/without chronic complications (44%),
renal disease (26%), peripheral vascular disease
(26%), and chronic pulmonary disease (22%).
Incidence and prevalence rates
The unadjusted incidence rate for BCC was 191.34
per 100,000 person-years, which projects to 459,352
cases in the US population according to 2012 census
estimates (Table III). The age-adjusted incidence rate
was 226.09 per 100,000, which projects to 542,782
cases in the US population. Incidence rates were
highest among patients aged 65 years or older and
lowest among those aged 18 to 24 years, and they
were higher in males than in females. As expected,
prevalence rates for BCC were notably higher than
incidence rates (Table IV). The age-adjusted preva-
lence rate was 342.64 per 100,000, which equates to
822,593 projected cases in the United States.
Prevalence rates were highest in patients aged
65 years or older and lowest in those aged 18 to
24 years, and they were higher in males than in
females.
The unadjusted incidence rate for LABCC was 1.28
per 100,000, which projects to 3071 cases in the US
population according to 2012 census estimates. The
age-adjusted incidence rate for LABCC was 1.83,
which projects to 4399 cases in the US population
(Table V). Patients aged 65 years or older accounted
for most of the age-adjusted prevalence of LABCC
(Table VI); the age-adjusted prevalence rate was 3.31
per 100,000 (projecting to 7940 cases in the United
States).
The incidence and prevalence of MBCC were very
low: the age-adjusted incidence rate was 0.04 per
100,000 (projecting to 108 cases in the US popula-
tion) (Supplementary Table I) and the age-adjusted
prevalence rate was 0.16 per 100,000 (projecting to
384 cases) (Supplementary Table II). Both rates were
highest among patients aged 65 years or older, and
both were higher among males than females.
DISCUSSION
This database claims analysis describes the
epidemiology of BCC in a commercially insured US
population. The age-adjusted incidence of 226.09
per 100,000 persons projects to 542,782 cases in
the US based on 2012 census estimates, and the
age-adjusted prevalence of 342.64 per 100,000
projects to 822,593 cases. Patients with prevalent
BCC were more likely to be older and male, and to
have higher rates of comorbidities compared with
patients with incident disease. The incidence and
J AM ACAD DERMATOL Goldenberg et al 961
VOLUME 75, NUMBER 5

Table I. Characteristics of BCC, LABCC, and MBCC cohorts (incident and prevalent cases)
Incident cases Prevalent cases
BCC LABCC MBCC BCC LABCC MBCC
Characteristic n = 39,035 n = 261 n=7 n = 17,952 n = 210 n = 16
Age at index date, y: mean (SD)* 63.2 (13.6) 72.2 (13.7) 71.3 (14.9) 68.3 (13.5) 74.0 (12.8) 72.1 (10.7)
Age group, y: n (%)*
18-24 39 (0.1) 0 0 6 (0.0) 0 0
25-34 463 (1.2) 1 (0.4) 0 103 (0.6) 0 0
35-44 2490 (6.4) 7 (2.7) 0 598 (3.3) 3 (1.4) 0
45-54 7563 (19.4) 21 (8.0) 1 (14.3) 2300 (12.8) 10 (4.8) 2 (12.5)
55-64 12,263 (31.4) 48 (18.4) 2 (28.6) 4489 (25.0) 48 (22.9) 2 (12.5)
$65 16,217 (41.5) 184 (70.5) 4 (57.1) 10,456 (58.2) 149 (71.0) 12 (75.0)
Gender, n (%)yz
Male 21,093 (54.0) 139 (53.3) 6 (85.7) 11,724 (65.3) 134 (63.8) 10 (62.5)
Female 17,942 (46.0) 122 (46.7) 1 (14.3) 6228 (34.7) 76 (36.2) 6 (37.5)
Region, n (%)yz
Northeast 7096 (18.2) 23 (8.8) 0 3404 (19.0) 24 (11.4) 3 (18.8)
North Central 9502 (24.3) 86 (33.0) 1 (14.3) 3890 (21.7) 46 (21.9) 5 (31.3)
South 14,211 (36.4) 101 (38.7) 3 (42.9) 6664 (37.1) 100 (47.6) 2 (12.5)
West 7868 (20.2) 51 (19.5) 3 (42.9) 3865 (21.5) 36 (17.1) 6 (37.5)
Unknown 358 (0.9) 0 0 129 (0.7) 4 (1.9) 0
Plan type, n (%)y
Comprehensive 8682 (22.2) 94 (36.0) 2 (28.6) 5532 (30.8) 96 (45.7) 5 (31.3)
EPO 328 (0.8) 2 (0.8) 0 153 (0.9) 1 (0.5) 0
HMO 4636 (11.9) 28 (10.7) 1 (14.3) 1862 (10.4) 18 (8.6) 6 (37.5)
POS 2391 (6.1) 9 (3.4) 0 1000 (5.6) 7 (3.3) 1 (6.3)
PPO 19,077 (48.9) 115 (44.1) 4 (57.1) 8078 (45.0) 75 (35.7) 3 (18.8)
POS with capitation 170 (0.4) 1 (0.4) 0 88 (0.5) 0 0
CDHP 1448 (3.7) 3 (1.1) 0 475 (2.6) 4 (1.9) 0
HDHP 786 (2.0) 4 (1.5) 0 238 (1.3) 1 (0.5) 0
Missing/unknown 1517 (3.9) 5 (1.9) 0 526 (2.9) 8 (3.8) 1 (6.3)
Employment status, n (%)y
Active full-time 11,655 (29.9) 42 (16.1) 0 3531 (19.7) 20 (9.5) 1 (6.3)
Active part-time or seasonal 319 (0.8) 0 0 109 (0.6) 2 (1.0) 0
Early retiree 2832 (7.3) 12 (4.6) 1 (14.3) 1040 (5.8) 18 (8.6) 0
Medicare-eligible retiree 11,500 (29.5) 119 (45.6) 3 (42.9) 7476 (41.6) 107 (51.0) 6 (37.5)
Retiree (status unknown) 2091 (5.4) 11 (4.2) 2 (28.6) 1302 (7.3) 11 (5.2) 1 (6.3)
COBRA continue 71 (0.2) 0 0 19 (0.1) 0 0
Long-term disability 52 (0.1) 1 (0.4) 0 29 (0.2) 0 0
Surviving spouse/dependent 1061 (2.7) 24 (9.2) 0 477 (2.7) 17 (8.1) 1 (6.3)
Other/unknown 9454 (24.2) 52 (19.9) 1 (14.3) 3969 (22.1) 35 (16.7) 7 (43.8)
Comorbidities, n (%)
Myocardial infarctiony 661 (1.7) 10 (3.8) 0 396 (2.2) 5 (2.4) 0
Congestive heart failurey 1608 (4.1) 17 (6.5) 0 1093 (6.1) 17 (8.1) 1 (6.3)
Peripheral vascular diseasey 2211 (5.7) 22 (8.4) 3 (42.9) 1600 (8.9) 20 (9.5) 3 (18.8)
Cerebrovascular diseasey 2577 (6.6) 33 (12.6) 0 1740 (9.7) 31 (14.8) 2 (12.5)
Dementiax 251 (0.6) 7 (2.7) 0 145 (0.8) 4 (1.9) 0
Chronic pulmonary diseasey 3962 (10.1) 32 (12.3) 1 (14.3) 2255 (12.6) 35 (16.7) 4 (25.0)
Rheumatic diseasey 801 (2.1) 6 (2.3) 0 475 (2.6) 7 (3.3) 1 (6.3)
Peptic ulcer disease 220 (0.6) 1 (0.4) 0 121 (0.7) 0 0
Mild liver disease 882 (2.3) 5 (1.9) 0 443 (2.5) 5 (2.4) 3 (18.8)
Diabetes without CCy 5221 (13.4) 60 (23.0) 2 (28.6) 2819 (15.7) 47 (22.4) 4 (25.0)
Diabetes with CCy 1305 (3.3) 11 (4.2) 1 (14.3) 765 (4.3) 13 (6.2) 3 (18.8)
Hemiplegia or paraplegiay 121 (0.3) 3 (1.1) 0 90 (0.5) 3 (1.4) 0
Renal diseasey 1528 (3.9) 16 (6.1) 0 1239 (6.9) 16 (7.6) 6 (37.5)
Any malignancyy{ 4260 (10.9) 1 (0.4) 0 3525 (19.6) 4 (1.9) 2 (12.5)
Moderate/severe liver diseasex 54 (0.1) 1 (0.4) 0 38 (0.2) 0 0
Continued
962 Goldenberg et al
Table I. Contd
Incident cases
BCC LABCC
Characteristic n = 39,035 n = 261
Metastatic solid tumory 280 (0.7) 0 2 (28.6)
AIDS/HIVy 64 (0.2) 1 (0.4)
Charlson Comorbidity Index score, 0.9 (1.6) 1.0 (1.5)
mean (SD)*k#
BCC, Basal cell carcinoma; CC, chronic complications; CDHP, consumer-directed health plan; COBRA, Consolidated Omnibus Budget
Reconciliation Act; EPO, exclusive provider organization; HDHP, high-deductible health plan; HMO, health maintenance organization; LABCC,
locally advanced BCC; MBCC, metastatic BCC; POS, point of service; PPO, preferred provider organization; SD, standard deviation.
*P \ .0001, incident vs prevalent cases for BCC (Wilcoxon rank sum test).
y
P # .0005, incident vs prevalent cases for BCC (x 2 test).
z
P \ .05, incident vs prevalent cases for LABCC (x 2 test).
x
P # .05, incident vs prevalent cases for BCC (x 2 test).
{
Including lymphoma and leukemia, except malignant neoplasm of skin.
k
P \ .05, incident vs prevalent cases for LABCC (Wilcoxon rank sum test).
#
P \ .05, incident vs prevalent cases for MBCC (t test).
prevalence estimates are somewhat lower than those
in previously published retrospective claims ana-
lyses of incident and prevalent NMSC cases, which
were conducted before availability of BCC-specific
ICD-9-CM codes.1,7,9,12 Other factors may also ac-
count for the lower estimates reported in our study,
as compared with previous analyses. A recent study
by Rogers et al,1 for example, was based on a
Medicare fee-for-service population, as compared
with the younger, commercially insured population
in our study.
Our study was also designed to distinguish
patients with LABCC and MBCC from the general
BCC population based on BCC-specific ICD-9-CM
codes; LABCC accounted for 0.8% of all BCC cases
and MBCC was exceedingly rare at 0.04% of the BCC
cohort. The age-adjusted incidence rates of LABCC
and MBCC project to only 4399 and 108 cases in the
United States, respectively, and the age-adjusted
prevalence rates project to 7940 and 384 cases,
respectively. Compared with the overall BCC cohort,
patients with LABCC and MBCC were more likely to
be older, and those with MBCC were more likely to
have comorbidities.
Several factors may explain the differences
between the BCC and LABCC estimates derived in
the current study and those derived from previous
studies. The most common difference is the use of
NMSC claims to identify patients with BCC, but other
factors may be relevant as well. Dacosta Byfield
et al,9 for example, used diagnostic imaging to
identify patients with LABCC, whereas we used a
more conservative definition of LABCC. Despite the
differences in estimates across studies, our results
confirm the rare nature of these disease subtypes.
J AM ACAD DERMATOL
NOVEMBER 2016
Prevalent cases
MBCC BCC LABCC MBCC
n=7 n = 17,952 n = 210 n = 16
328 (1.8) 3 (1.4) 10 (62.5)
0 57 (0.3) 3 (1.4) 0
2.9 (2.5) 1.4 (2.0) 1.3 (1.8) 6.3 (3.7)
More specifically, we confirmed that the incidence
and prevalence of BCC, LABCC, and MBCC are
highest in the older age categories (particularly in
patients aged 65 years or older) and higher in males
than in females.9 Future studies might focus on
treatment patterns and health care use and costs for
these patient populations.
Several study limitations should be noted. First,
administrative claims data, such as those used in the
current study, are not collected for research
purposes; the diagnostic coding on administrative
claims is recorded by physicians for the purpose of
supporting reimbursement. As such, these data are
never complete or detailed enough to provide a
clinically precise description of patients, and the
impact of disease severity and other descriptive
parameters (eg, lesion size, sun exposure) cannot
be assessed. Second, the generalizability of the study
results is limited by the use of claims data from a
commercially insured US population. The reported
findings may not apply, for example, to patients in
other countries, to uninsured patients, or to those
insured by Medicare and Medicaid (we would note,
however, that the database used is generally
representative of the US population with respect to
age and gender, according to estimates from the US
Census Bureau). Third, medical claims databases
may not capture all the services that patients receive.
They may omit out-of-pocket transactions, changes
in health care insurance enrollment, and changes in
claims submissions made by physicians. Fourth, the
inclusion criterion requiring that patients have 2
claims for BCC separated by at least 30 days may
have led to underestimation of the number of cases;
however, it should be noted that this type of
J AM ACAD DERMATOL Goldenberg et al 963
VOLUME 75, NUMBER 5

Table II. Characteristics of BCC, LABCC, and MBCC cohorts (total cases)
Total cases
BCC LABCC MBCC
Characteristic n = 56,987 n = 471 n = 23
Age at index date, y: mean (SD) 64.9 (13.8) 73.0 (13.3) 71.9 (11.8)
Age group, y: n (%)
18-24 45 (0.1) 0 0
25-34 566 (1.0) 1 (0.2) 0
35-44 3088 (5.4) 10 (2.1) 0
45-54 9863 (17.3) 31 (6.6) 3 (13.0)
55-64 16,752 (29.4) 96 (20.4) 4 (17.4)
$65 26,673 (46.8) 333 (70.7) 16 (69.6)
Gender, n (%)
Male 32,817 (57.6) 273 (58.0) 16 (69.6)
Female 24,170 (42.4) 198 (42.0) 7 (30.4)
Region, n (%)
Northeast 10,500 (18.4) 47 (10.0) 3 (13.0)
North Central 13,392 (23.5) 132 (28.0) 6 (26.1)
South 20,875 (36.6) 201 (42.7) 5 (21.7)
West 11,733 (20.6) 87 (18.5) 9 (39.1)
Unknown 487 (0.9) 4 (0.8) 0
Plan type, n (%)
Comprehensive 14,214 (24.9) 190 (40.3) 7 (30.4)
EPO 481 (0.8) 3 (0.6) 0
HMO 6498 (11.4) 46 (9.8) 7 (30.4)
POS 3391 (6.0) 16 (3.4) 1 (4.3)
PPO 27,155 (47.7) 190 (40.3) 7 (30.4)
POS with capitation 258 (0.5) 1 (0.2) 0
CDHP 1923 (3.4) 7 (1.5) 0
HDHP 1024 (1.8) 5 (1.1) 0
Missing/unknown 2043 (3.6) 13 (2.8) 1 (4.3)
Employment status, n (%)
Active full-time 15,186 (26.6) 62 (13.2) 1 (4.3)
Active part-time or seasonal 428 (0.8) 2 (0.4) 0
Early retiree 3872 (6.8) 30 (6.4) 1 (4.3)
Medicare-eligible retiree 18,976 (33.3) 226 (48.0) 9 (39.1)
Retiree (status unknown) 3393 (6.0) 22 (4.7) 3 (13.0)
COBRA continue 90 (0.2) 0 0
Long-term disability 81 (0.1) 1 (0.2) 0
Surviving spouse/dependent 1538 (2.7) 41 (8.7) 1 (4.3)
Other/unknown 13,423 (23.6) 87 (18.5) 8 (34.8)
Comorbidities, n (%)
Myocardial infarction 1057 (1.9) 15 (3.2) 0
Congestive heart failure 2701 (4.7) 34 (7.2) 1 (4.3)
Peripheral vascular disease 3811 (6.7) 42 (8.9) 6 (26.1)
Cerebrovascular disease 4317 (7.6) 64 (13.6) 2 (8.7)
Dementia 396 (0.7) 11 (2.3) 0
Chronic pulmonary disease 6217 (10.9) 67 (14.2) 5 (21.7)
Rheumatic disease 1276 (2.2) 13 (2.8) 1 (4.3)
Peptic ulcer disease 341 (0.6) 1 (0.2) 0
Mild liver disease 1325 (2.3) 10 (2.1) 3 (13.0)
Diabetes without CC 8040 (14.1) 107 (22.7) 6 (26.1)
Diabetes with CC 2070 (3.6) 24 (5.1) 4 (17.4)
Hemiplegia or paraplegia 211 (0.4) 6 (1.3) 0
Renal disease 2767 (4.9) 32 (6.8) 6 (26.1)
Any malignancy* 7785 (13.7) 5 (1.1) 2 (8.7)
Moderate/severe liver disease 92 (0.2) 1 (0.2) 0
Continued
964 Goldenberg et al J AM ACAD DERMATOL
NOVEMBER 2016

Table II. Contd

Total cases
BCC LABCC MBCC
Characteristic n = 56,987 n = 471 n = 23
Metastatic solid tumor 608 (1.1) 3 (0.6) 12 (52.2)
AIDS/HIV 121 (0.2) 4 (0.8) 0
Charlson Comorbidity Index score, mean (SD) 1.0 (1.7) 1.2 (1.6) 5.2 (3.7)

BCC, Basal cell carcinoma; CC, chronic complications; CDHP, consumer-directed health plan; COBRA, Consolidated Omnibus Budget
Reconciliation Act; EPO, exclusive provider organization; HDHP, high-deductible health plan; HMO, health maintenance organization; LABCC,
locally advanced BCC; MBCC, metastatic BCC; POS, point of service; PPO, preferred provider organization; SD, standard deviation.
*Including lymphoma and leukemia, except malignant neoplasm of skin.

Table III. Incidence of basal cell carcinoma

US 2012 standard population Projected for US population
No. of Population Crude Adjusted
Category cases at risk incidence rate* No. Distribution incidence rate* No. Percentage
All 39,035 20,401,092 191.34 240,073,659 1.00 191.34 459,352 100
Age group, y
18-24 39 2,744,940 1.42 31,318,550 0.13 0.19 445 0.08
25-34 463 2,839,612 16.31 42,276,195 0.18 2.87 6893 1.27
35-44 2490 3,861,670 64.48 40,496,919 0.17 10.88 26,112 4.81
45-54 7563 4,675,649 161.75 44,249,580 0.18 29.81 71,575 13.19
55-64 12,263 4,110,234 298.35 38,577,181 0.16 47.94 115,096 21.20
$65 16,217 2,168,987 747.68 43,155,234 0.18 134.40 322,661 59.45
Age-adjusted 226.09 542,782 100
Gender
Male 21,093 9,688,722 217.71 116,751,886 0.49 105.87 254,177 55.17
Female 17,942 10,712,370 167.49 123,321,773 0.51 86.04 206,550 44.83
Gender-adjusted 191.91 460,727 100

*Per 100,000 persons.

Table IV. Prevalence of basal cell carcinoma

US 2012 Projected for
Crude Adjusted
standard population US population
Population prevalence prevalence
Category No. of cases at risk rate* No. Distribution rate* No. of cases Percentage
All 56,987 (39,035 20,401,092 279.33 240,073,659 1.00 279.33 670,605 100
incident 1 17,952
prevalent)
Age group, y
18-24 45 2,744,940 1.64 31,318,550 0.13 0.21 513 0.06
25-34 566 2,839,612 19.93 42,276,195 0.18 3.51 8427 1.02
35-44 3088 3,861,670 79.97 40,496,919 0.17 13.49 32,384 3.94
45-54 9863 4,675,649 210.94 44,249,580 0.18 38.88 93,342 11.35
55-64 16,752 4,110,234 407.57 38,577,181 0.16 65.49 157,228 19.11
$65 26,673 2,168,987 1229.74 43,155,234 0.18 221.06 530,699 64.52
Age-adjusted 342.64 822,593 100
Gender
Male 32,817 9,688,722 338.71 116,751,886 0.49 164.72 395,454 58.70
Female 24,170 10,712,370 225.63 123,321,773 0.51 115.90 278,247 41.30
Gender-adjusted 280.62 673,701 100

*Per 100,000 persons.

J AM ACAD DERMATOL Goldenberg et al 965
VOLUME 75, NUMBER 5

Table V. Incidence of locally advanced basal cell carcinoma

US 2012 Projected for
Crude Adjusted
standard population US population
No. of Population incidence incidence
Category cases at risk rate* No. Distribution rate* No. Percentage
All 261 20,401,092 1.28 240,073,659 1.00 1.28 3071 100
Age group, y
18-24 0 2,744,940 0.00 31,318,550 0.13 0.00 0 0
25-34 1 2,839,612 0.04 42,276,195 0.18 0.01 15 0.34
35-44 7 3,861,670 0.18 40,496,919 0.17 0.03 73 1.66
45-54 21 4,675,649 0.45 44,249,580 0.18 0.08 199 4.52
55-64 48 4,110,234 1.17 38,577,181 0.16 0.19 451 10.25
$65 184 2,168,987 8.48 43,155,234 0.18 1.52 3661 83.22
Age-adjusted 1.83 4399 100
Gender
Male 139 9,688,722 1.43 116,751,886 0.49 0.70 1675 54.40
Female 122 10,712,370 1.14 123,321,773 0.51 0.59 1404 45.60
Gender-adjusted 1.28 3079 100

*Per 100,000 persons.

Table VI. Prevalence of locally advanced basal cell carcinoma

US 2012 Projected for
Crude Adjusted
standard population US population
Population prevalence prevalence
Category No. of cases at risk rate* No. Distribution rate* No. of cases Percentage
All 471 (261 incident 1 20,401,092 2.31 240,073,659 1.00 2.31 5543 100
210 prevalent)
Age group, y
18-24 0 2,744,940 0.00 31,318,550 0.13 0.00 0 0.00
25-34 1 2,839,612 0.04 42,276,195 0.18 0.01 15 0.19
35-44 10 3,861,670 0.26 40,496,919 0.17 0.04 105 1.32
45-54 31 4,675,649 0.66 44,249,580 0.18 0.12 293 3.69
55-64 96 4,110,234 2.34 38,577,181 0.16 0.38 901 11.35
$65 333 2,168,987 15.35 43,155,234 0.18 2.76 6626 83.45
Age-adjusted 3.31 7940 100
Gender
Male 273 9,688,722 2.82 116,751,886 0.49 1.37 3290 59.08
Female 198 10,712,370 1.85 123,321,773 0.51 0.95 2279 40.92
Gender-adjusted 2.32 5569 100

*Per 100,000 persons.

approach is commonly used in database studies for
other therapeutic areas to ensure that the patient
actually has the disease in question. In addition, a
primary focus of our study was to identify patients
with advanced BCC to derive incidence and
prevalence estimates for this subpopulation, and it
was thought that advanced patients, even if
treated within a month, would have follow-up past
this period. Fifth, because there were no BCC
codes available in the year preceding the time
frame of the study, we defined the prevalent
population as those patients with an NMSC claim in
the previous year. This conservative approach
could have led to underestimation of the
incident population/overestimation of the prevalent
population. Sixth, we did not use surgical and
destruction codes as part of our BCC identification
protocol, as it was believed this would have
broadened the scope of the study too much and
compromised the sensitivity of our estimates,
especially given our emphasis on advanced BCC;
this approach may also have led to underestimation
of the overall number of BCC cases. Finally, patients
treatment-seeking behavior may have affected the
incidence of LABCC identified by the algorithm used
in this study.
In summary, BCC incidence and prevalence
estimates from our study project to 542,782
and 822,593 patients, respectively. LABCC was
uncommon, being identified in only 0.8% of BCC
966 Goldenberg et al
cases; nevertheless, it is an important health concern
given the high incidence and prevalence of BCC and
likely contributes disproportionately to the health
burden of BCC. MBCC was also rare in this study
population (0.04% of the BCC cohort).
REFERENCES
1. Rogers HW, Weinstock MA, Feldman SR, Coldiron BM.
Incidence estimate of nonmelanoma skin cancer (keratinocyte
carcinomas) in the US population, 2012. JAMA Dermatol. 2015;
151:1081-1086.
2. Skin Cancer Foundation. Skin cancer facts. Available from: URL
http://www.skincancer.org/skin-cancer-information/skin-cancer-
facts. Accessed April 22, 2015.
3. Guy GP Jr, Machlin SR, Ekwueme DU, Yabroff KR. Prevalence
and costs of skin cancer treatment in the U.S., 2002-2006 and
2007-2011. Am J Prev Med. 2015;48:183-187.
4. Cowey CL. Targeted therapy for advanced basal-cell
carcinoma: vismodegib and beyond. Dermatol Ther (Heidelb).
2013;3:17-31.
5. Lear JT, Corner C, Dziewulski P, et al. Challenges and
new horizons in the management of advanced basal cell
carcinoma: a UK perspective. Br J Cancer. 2014;111:1476-1481.
J AM ACAD DERMATOL
NOVEMBER 2016
6. Mohan SV, Chang AL. Advanced basal cell carcinoma:
epidemiology and therapeutic innovations. Curr Dermatol
Rep. 2014;3:40-45.
7. Eide MJ, Krajenta R, Johnson D, et al. Identification of
patients with nonmelanoma skin cancer using health
maintenance organization claims data. Am J Epidemiol. 2010;
171:123-128.
8. Eide MJ, Tuthill JM, Krajenta RJ, Jacobsen GR, Levine M,
Johnson CC. Validation of claims data algorithms to identify
nonmelanoma skin cancer. J Invest Dermatol. 2012;132:
2005-2009.
9. Dacosta Byfield S, Chen D, Yim YM, Reyes C. Age distribution
of patients with advanced non-melanoma skin cancer in the
United States. Arch Dermatol Res. 2013;305:845-850.
10. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method
of classifying prognostic comorbidity in longitudinal studies:
development and validation. J Chronic Dis. 1987;40:373-383.
11. US Census Bureau. 2012 National population projections:
downloadable files. Available from: URL http://www.census.
gov/population/projections/data/national/2012/download
ablefiles.html. Accessed April 22, 2015.
12. Rogers HW, Weinstock MA, Harris AR, et al. Incidence estimate
of nonmelanoma skin cancer in the United States, 2006. Arch
Dermatol. 2010;146:283-287.
J AM ACAD DERMATOL Goldenberg et al 966.e1
VOLUME 75, NUMBER 5

Supplementary Table I. Incidence of metastatic basal cell carcinoma

Crude US 2012 standard population Adjusted Projected for US population
No. of Population incidence incidence
Category cases at risk rate* No. Distribution rate* No. Percentage
All 7 20,401,092 0.03 240,073,659 1.00 0.03 82 100
Age group, y
18-24 0 2,744,940 0.00 31,318,550 0.13 0.00 0 0.00
25-34 0 2,839,612 0.00 42,276,195 0.18 0.00 0 0.00
35-44 0 3,861,670 0.00 40,496,919 0.17 0.00 0 0.00
45-54 1 4,675,649 0.02 44,249,580 0.18 0.00 9 8.33
55-64 2 4,110,234 0.05 38,577,181 0.16 0.01 19 17.59
$65 4 2,168,987 0.18 43,155,234 0.18 0.03 80 74.07
Age-adjusted 0.04 108 100
Gender
Male 6 9,688,722 0.06 116,751,886 0.49 0.03 72 85.71
Female 1 10,712,370 0.01 123,321,773 0.51 0.00 12 14.29
Gender-adjusted 0.03 84 100

*Per 100,000 persons.

966.e2 Goldenberg et al J AM ACAD DERMATOL
NOVEMBER 2016

Supplementary Table II. Prevalence of metastatic basal cell carcinoma

US 2012 Projected for
Crude Adjusted
standard population US population
Population prevalence prevalence
Category No. of cases at risk rate* No. Distribution rate* No. of cases Percentage
All 23 (7 incident 1 20,401,092 0.11 240,073,659 1.00 0.11 271 100
16 prevalent)
Age group, y
18-24 0 2,744,940 0.00 31,318,550 0.13 0.00 0 0.00
25-34 0 2,839,612 0.00 42,276,195 0.18 0.00 0 0.00
35-44 0 3,861,670 0.00 40,496,919 0.17 0.00 0 0.00
45-54 3 4,675,649 0.06 44,249,580 0.18 0.01 28 7.29
55-64 4 4,110,234 0.10 38,577,181 0.16 0.02 38 9.90
$65 16 2,168,987 0.74 43,155,234 0.18 0.13 318 82.81
Age-adjusted 0.16 384 100
Gender
Male 16 9,688,722 0.17 116,751,886 0.49 0.08 193 70.44
Female 7 10,712,370 0.07 123,321,773 0.51 0.03 81 29.56
Gender-adjusted 0.11 274 100

*Per 100,000 persons.