Mortality from nonneoplastic skin disease in the United States

ORIGINAL

ARTICLES

Mortality from nonneoplastic skin disease in the United States Jason P. Lott, MD, MSHP,a,c,d and Cary P. Gross, MDb,c New Haven, Connecticut Background: The mortality burden from nonneoplastic skin disease in the United States is unknown. Objective: We sought to estimate mortality from nonneoplastic skin disease as underlying and contributing causes of death. Methods: Population-based death certificate data detailing mortality from nonneoplastic skin disease for years 1999 to 2009 were used to calculate absolute numbers of death and age-adjusted mortality by year, patient demographics, and 10 most commonly reported diagnoses. Results: Nonneoplastic skin diseases were reported as underlying and contributing causes of mortality for approximately 3948 and 19,542 patients per year, respectively. Age-adjusted underlying cause mortality (per 100,000 persons) were significantly greater (P \.0001) for patients who were black/African American (3.4), women (1.4), and residing in the South (1.6). Most deaths occurred in patients ages 65 years and older (34,248 total deaths). Common underlying causes of death included chronic ulcers (1789 deaths/y) and cellulitis (1348 deaths/y). Limitations: Errors in death certificate data and inability to adjust for patient-level confounders may limit the accuracy and generalizability of our results. Conclusion: Mortality from nonneoplastic skin disease is uncommon yet potentially preventable. The elderly bear the greatest burden of mortality from nonneoplastic skin disease. Chronic ulcers and cellulitis constitute frequent causes of death. ( J Am Acad Dermatol 2014;70:47-54.) Key words: death certificate; dermatology; epidemiology; health disparities; mortality; outcomes; public health; skin disease; vital statistics.

M

ortality attributable to neoplastic skin disease, including cutaneous squamous cell carcinoma,1-3 basal cell carcinoma,3-5 and melanoma,6-8 has been previously described in US-based and internationally based studies. However, population-based studies examining death from nonneoplastic skin disorders in the United States have not been performed to our knowledge. Several factors may account for the lack of such analyses, including the low incidence of potentially fatal nonneoplastic skin disease and rarity of death from more commonly encountered

nonneoplastic dermatoses. In addition, advances in diagnosis and treatment of nonneoplastic skin disorders that may have once posed serious health threats, such as bullous dermatoses and autoimmune

From the Department of Dermatology,a Department of Internal Medicine,b and Robert Wood Johnson Foundation Clinical Scholars Program,c Yale University School of Medicine and the Veterans Administration (VA) Connecticut Healthcare System.d Funding sources: None. Conflicts of interest: None declared. Accepted for publication September 25, 2013.

Reprint requests: Jason P. Lott, MD, MSHP, Yale University School of Medicine, 333 Cedar St, SHM IE-61, PO Box 208088, New Haven, CT 06520. E-mail: [email protected]. Published online November 14, 2013. 0190-9622/$36.00 Ó 2013 by the American Academy of Dermatology, Inc. http://dx.doi.org/10.1016/j.jaad.2013.09.039

Abbreviations used: CDC: NCHS: WONDER:

Centers for Disease Control and Prevention National Center for Health Statistics Wide-ranging Online Data for Epidemiologic Research

47

J AM ACAD DERMATOL

48 Lott and Gross

JANUARY 2014

connective tissue diseases, have resulted in imnonneoplastic skin disease, as both underlying and proved survival for many of these conditions.9-14 contributing causes of death, using publicly available Nevertheless, estimating the burden of mortality death certificate data. We also sought to quantify from nonneoplastic skin disease remains important variation in this mortality burden by select patient for clinicians and policy makers. An accurate underdemographics, including age, gender, race, ethnicity, standing of this burden can allow for identification of and geographic residence. potentially preventable deaths and disparities in health care access and qualMETHODS ity. A recent study of inpaThe design of this study CAPSULE SUMMARY tient hospitalization records was cross-sectional and US and death certificates, for expopulation based. Our priThe mortality burden of nonneoplastic ample, found 87 deaths remary aim was to estimate skin disease is unknown. lated to psoriasis in the absolute numbers of death This study provides US population-based United States from 1999 to and age-adjusted mortality mortality estimates for nonneoplastic 2001, prompting the authors of nonneoplastic skin disskin disease, revealing variation by race to note that ‘‘psoriasis coneases reported as underlying and other patient demographics. tinues to be at least a factor, if and contributing causes of not the ultimate cause of, Mortality from nonneoplastic skin death. Our secondary aim patient deaths,’’ and reflect disease is unequally distributed across was to identify the 10 most that ‘‘it is distressing to think patients in the United States. common nonneoplastic derthat some of these deaths matoses reported as underlymight have been preing and contributing causes vented.’’15 Further studies, however, have not been of death. Our tertiary aim was to identify the 10 most conducted to quantify the overall magnitude of and common underlying causes of death (inclusive of all assess variation in mortality resulting from nonneodiseases) associated with nonneoplastic dermatoses plastic skin diseases across the United States. Given reported as contributing causes of mortality. In that death from nonneoplastic skin disease may be addition, we hypothesized that increased mortality potentially preventable, understanding the burden from nonneoplastic skin disease might be associated of mortality from these disorders may be important with increasing patient age, black/African American for optimization of clinical care. race, Hispanic ethnicity, residence in the South, and For the purposes of vital statistics reporting, skin residence in inner cities/rural counties. Our hypothdisease may be recorded as a cause of mortality in eses were motivated by well-known sociodemo2 ways. First, it may serve as the ‘‘underlying cause graphic parameters affecting overall health status of death,’’ defined, according to the World Health and outcomes,20,21 and geographic disparities in Organization, as ‘‘the disease or injury which initidermatology workforce distribution that may imated the train of morbid events leading directly to pede access to dermatologic care.22,23 death, or the circumstances of the accident or We examined death certificate data detailing violence which produced the fatal injury.’’16,17 nonneoplastic skin disease for years 1999 to 2009 Alternatively, skin disease may serve as a ‘‘multiple from the Centers for Disease Control and Prevention cause of death,’’ denoting ‘‘health conditions giving (CDC) Wide-ranging Online Data for Epidemiologic rise to the immediate cause of death and other Research (WONDER) database (http://wonder.cdc. conditions contributing to death.’’16,17 Diseases in gov/).16,24 CDC WONDER houses nationwide death the latter category can be conceptualized as ‘‘concertificate data from the National Center for Health tributing’’ causes of mortality. Consideration of Statistics (NCHS). Mortality data are obtained from nonneoplastic skin diseases as both underlying the 57 vital statistics jurisdictions of the Vital Statistics and contributing causes of mortality allows for a Cooperative Program (comprising the 50 US states, broad, clinically relevant contextualization of these Puerto Rico, the US Virgin Islands, Guam, American disorders, which may increase risk of death Samoa, the Northern Mariana Islands, the District of through various mechanisms, including proinflamColumbia, and New York City), and from death matory cytokine cascades in psoriasis,18,19 impaired certificates provided directly to NCHS by state govimmunity in blistering dermatoses, and visceral ernments.25 Death certification information is regisorgan damage in autoimmune connective tissue tered and recorded by physicians, certifying medical diseases. examiners, and/or coroners, which is then abstracted In this study, we sought to describe US and stored electronically using the Automated population-based mortality consequent to Classification of Medical Entities software program d

d

d

J AM ACAD DERMATOL

Lott and Gross 49

VOLUME 70, NUMBER 1

within NCHS.16,25,26 These data are subsequently aggregated into the online-accessible Multiple Cause of Death data file (http://wonder.cdc.gov/mcdicd10.html) available from CDC WONDER. Given that all US states and territories legally mandate registration of all deaths, it has been estimated that over 99% of deaths occurring within the United States are captured by the NCHS.16 In August 2013, we queried the Multiple Cause of Death data file to extract and compile absolute numbers of death and age-adjusted mortality for years 1999 to 2009 using International Statistical Classification of Diseases, 10th Revision codes L00 to L98 (diseases of skin and subcutaneous tissue), which excludes benign and malignant neoplasms of the skin and connective/soft tissue (International Statistical Classification of Diseases, 10th Revision codes C00-D48). We divided these data into 2 groups as delineated in the Multiple Cause of Death data file: nonneoplastic skin disease as the underlying cause of death, and nonneoplastic skin disease as a contributing cause of mortality. The latter was defined as any disease reported as a ‘‘multiple cause of death’’ recorded on standard US death certificate forms. We stratified mortality data by year, race (American Indian/Alaska native, Asian or Pacific Islander, black/African American, and white), Hispanic ethnicity, age group, census region (Northeast, Midwest, West, and South), and urbanization (large central metropolitan, large fringe metropolitan, medium metropolitan, small metropolitan, micropolitan, and noncoreeie, nonmetropolitan and nonmicropolitan counties). The NCHS developed the urbanization categorization in 2006 for epidemiologic purposes, recognizing that residents of inner cities and rural areas experience poorer health and greater health disparities compared with their suburban counterparts.27 We compared stratum-specific age-adjusted mortality using 2-sample or pairwise tests for binomial proportions as appropriate. Male was the referent group for gender, white was the referent group for race, non-Hispanic was the referent group for Hispanic origin, Northeast was the referent group for census region, and large metropolitan was the referent group for urbanization. We compared ageadjusted mortality across age categories using Cuzick28 nonparametric test of trend for ordered groups. In addition, we performed global tests of significance across racial, census region, and urbanization categories using the x 2 statistic. Finally, we combined the most common underlying and contributing causes of death into 10 disease categories based on clinical similarities (using International Statistical Classification of

Diseases, 10th Revision codes within the L00-L98 hierarchy) (Appendix; available at http://www.jaad. org). We then calculated absolute number of deaths and age-adjusted mortality per 1 million persons (to obtain stable estimates for infrequently occurring diseases) by disease group. Likewise, we calculated the absolute number of deaths and age-adjusted mortality (per 100,000 persons) for the 10 most common underlying causes of death inclusive of all diseases (collapsed by disease categories) (Appendix; available at http://www.jaad.org) associated with nonneoplastic skin diseases reported as contributing causes of death. Point estimates and confidence intervals for age-adjusted mortality for the combined disease categories were estimated using the binomial distribution. Statistical analyses were performed using software (Stata SE 12.1, StataCorp, College Station, TX). All statistical tests were 2-tailed with alpha equal to 0.05. After adjustment for multiple comparisons using Benjamini-Hochberg correction29 (assuming a false discovery rate of 0.2), P values less than .01 were considered statistically significant. The Yale University Human Research Protection Program exempted this study from institutional board review.

RESULTS Nonneoplastic skin diseases as underlying causes of mortality For nonneoplastic skin disease as the underlying cause of mortality, absolute numbers of deaths ranged from 3697 (in 1999) to 4234 (in 2005). Approximately 43,424 total deaths occurred during this 11-year period, or approximately 3948 deaths per year. Age-adjusted mortality remained stable throughout the study period, ranging from 1.2 deaths per 100,000 persons to 1.4 deaths per 100,000 persons (Table I). Mortality varied significantly by race, with 33,009 deaths (age-adjusted mortality 1.1/ 100,000 persons) seen in whites compared with 9474 deaths (age-adjusted mortality 3.4/100,000 persons) seen in blacks/African Americans (P \ .0001) (Table II). Individuals of non-Hispanic origin experienced significantly more deaths (41,137 deaths; age-adjusted mortality 1.3/100,000 persons) compared with individuals of Hispanic descent (2169 deaths; age-adjusted mortality 1.1/100,000 persons). We observed 27,059 deaths in females (age-adjusted mortality 1.4/100,000 persons) compared with 16,365 deaths in males (age-adjusted mortality 1.2/ 100,000 persons; P \ .0001) (Table II). Elderly patients experienced a significantly larger number of deaths and higher incidence of death compared with younger age cohorts (P \.0001) (Table II), with

J AM ACAD DERMATOL

50 Lott and Gross

JANUARY 2014

Table I. Nonneoplastic skin disease deaths and age-adjusted mortality (per 100,000 persons) in the United States, by year: 1999 to 2009 Reported as underlying cause of mortality Year

Deaths

Age-adjusted rate [95% CI]

1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Total

3697 3753 3749 3937 4132 4180 4234 3862 3883 3983 4014 43,424

1.3 1.4 1.3 1.4 1.4 1.4 1.4 1.2 1.2 1.2 1.2 1.3

[1.3-1.4] [1.3-1.4] [1.3-1.4] [1.3-1.4] [1.4-1.4] [1.3-1.4] [1.3-1.4] [1.2-1.3] [1.2-1.3] [1.2-1.2] [1.3-1.3] [1.3-1.3]

Reported as contributing cause of mortality Deaths

18,350 18,312 18,290 18,351 18,900 18,996 19,441 21,401 21,348 21,062 20,508 214,959

Age-adjusted rate [95% CI]

6.8 6.6 6.4 6.3 6.4 6.3 6.3 6.8 6.7 6.4 6.2 6.5

[6.7-6.9] [6.5-6.7] [6.3-6.5] [6.2-6.4] [6.3-6.5] [6.2-6.4] [6.2-6.4] [6.7-6.9] [6.6-6.8] [6.3-6.5] [6.1-6.3] [6.4-6.5]

CI, Confidence interval.

34,248 deaths, or approximately 80%, occurring in individuals ages 65 years or older. Compared with the Northeast, the South had a significantly higher age-adjusted mortality (1.3/ 100,000 vs 1.6/100,000 persons; P \.0001), whereas age-adjusted mortality was significantly lower in the West (1.0/100,000 persons; P \.0001). Residence in large fringe and medium metropolitan counties (suburban communities and counties within metropolitan statistical areas containing 250,000-999,999 people, respectively) were associated with significantly fewer deaths compared with inner-city and rural areas of urbanization (P \.0001) (Table II). Decubitus and other ulcers were associated with 19,678 total deaths (age-adjusted mortality 6.0/1 million persons), or approximately 1789 deaths per year, and were the most frequently observed underlying cause of mortality, followed by cellulitis (14,825 total deaths or 1348 deaths/y, age-adjusted mortality 4.5/1 million persons) (Table III). The remaining 8 most common underlying causes of mortality included abscesses, cutaneous lupus erythematosus, toxic epidermal necrolysis and erythema multiforme, pemphigus, pemphigoid, psoriasis, dermatitis, and other/unspecified skin diseases (Table III). These 10 disease categories accounted for 42,923 deaths over the study period, or approximately 3902 deaths per year, and represented 99% of all underlying causes of death attributable to nonneoplastic skin disorders. Nonneoplastic skin diseases as contributing causes of mortality For nonneoplastic skin disease as a contributing cause of mortality, the absolute number of associated

deaths ranged from 18,290 (in 2001) to 21,401 (in 2006), with 214,959 total number of associated deaths, or approximately 19,542 associated deaths per year. Age-adjusted mortality ranged from 6.2 per 100,000 persons (in 2009) to 6.8 per 100,000 persons (in years 1999 and 2006) (Table I). Race, ethnicity, gender, age, census region, and urbanization categories showed similar variation as seen with underlying cause mortality (Table II). Many of the nonneoplastic skin diseases comprising the 10 most common underlying causes of mortality were also among the 10 most common contributing causes of death, with decubitus and other ulcers most frequently reported in conjunction with 119,930 deaths (age-adjusted mortality 36.8/ 1 million persons), or approximately 10,903 deaths per year. Of note, because multiple nonneoplastic skin diseases may be reported per death, the total number of the 10 most common nonneoplastic skin diseases as contributing causes of mortality (218,217) (Table III) exceeded the total number of actual deaths from nonneoplastic skin diseases (214,959) (Table I). Underlying causes of death associated with nonneoplastic skin diseases as contributing causes of mortality Atherosclerotic heart disease/acute myocardial infarction (21,728 deaths; age-adjusted mortality 0.7/100,000 persons), dementia (20,316 deaths; age-adjusted mortality 0.6/100,000 persons), and chronic ulcers (19,634 deaths; age-adjusted mortality 0.6/100,000 persons) were the most common underlying causes of death inclusive of all diseases reported in association with nonneoplastic skin diseases recorded as contributing causes of mortality (Table IV). The remaining 7 most common underlying causes of death included cellulitis, diabetes mellitus, stroke, renal failure, sepsis, chronic obstructive pulmonary disease/emphysema, and peripheral vascular disease. Together, these 10 underlying causes of mortality accounted for 123,351 deaths over the study period, or approximately 57% of the deaths associated with nonneoplastic skin diseases as contributing causes of mortality.

DISCUSSION This study provides population-based estimates of mortality from nonneoplastic skin diseases from 1999 to 2009, and, to our knowledge, is the first assessment of the mortality burden of nonneoplastic skin diseases and its variation across patient demographics. These results complement the large body of literature focused on mortality from cutaneous

J AM ACAD DERMATOL

Lott and Gross 51

VOLUME 70, NUMBER 1

Table II. Nonneoplastic skin disease deaths in the United States and age-adjusted mortality (per 100,000 persons) by patient demographics, 1999 to 2009 Reported as underlying cause of mortality Patient demographic

Gender Male Female Racez White Black/African American Asian/Pacific Islander American Indian/Alaska Native Hispanic origin Hispanic or Latino Not Hispanic or Latino Not stated Age group, y \1 1-4 5-14 15-24 25-34 35-44 45-54 55-64 65-74 75-84 [85 Missing Census regionz Northeast Midwest South West Urbanizationz Large central metro Large fringe metro Medium metro Small metro Micropolitan (nonmetro) Noncore (nonmetro)

Deaths

Age-adjusted rate [95% CI]

P value*

16,365 27,059

1.2 [1.2-1.3] 1.4 [1.4-1.4]

33,009 9474 624 317

1.1 3.4 0.6 1.7

Reported as contributing cause of mortality Deaths

Age-adjusted rate [95% CI]

P value*

ref \.0001

84,984 129,975

6.4 [6.4-6.4] 6.5 [6.4-6.5]

ref ns

[1.1-1.2] [3.3-3.4] [0.6-0.7] [1.5-1.9]

ref \.0001 \.0001 \.0001

167,441 42,725 3307 1486

5.7 [5.7-5.8] 15.1 [14.9-15.2] 3.4 [3.3-3.6] 7.4 [7.0-7.8]

ref \.0001 \.0001 \.0001

2169 41,137 118

1.1 [1.0-1.1] 1.3 [1.3-1.3] na

ref \.0001 na

11,182 203,234 543

5.8 [5.6-5.9] 6.6 [6.6-6.6] na

ref \.0001 na

29 36 58 203 540 1327 2749 4233 6583 12,418 15,247 1

0.1 [0.0-0.1] 0.0 [0.0-0.0] 0.0 [0.0-0.0] 0.0 [0.0-0.1] 0.1 [0.1-0.1] 0.3 [0.3-0.3] 0.6 [0.6-0.6] 1.3 [1.3-1.4] 3.1 [3.1-3.2] 8.8 [8.6-9.0] 29.8 [29.3-30.2] na

\.0001y

155 111 180 740 2141 5971 13,231 21,218 33,241 63,538 74,424 9

0.4 [0.3-0.4] 0.1 [0.1-0.1] 0.0 [0.0-0.0] 0.2 [0.2-0.2] 0.5 [0.5-0.5] 1.2 [1.2-1.3] 2.9 [2.9-3.0] 6.6 [6.5-6.7] 15.8 [15.6-16.0] 45.0 [44.7-45.4] 145.2 [144.2-146.3] na

\.0001y

8617 10,123 17,979 6705

1.3 1.3 1.6 1.0

[1.2-1.3] [1.3-1.3] [1.6-1.6] [0.9-1.0]

ref ns \.0001 \.0001

39,137 49,763 84,557 41,502

5.8 6.5 7.4 6.2

[5.7-5.8] [6.4-6.5] [7.3-7.4] [6.1-6.2]

ref \.0001 \.0001 \.0001

12,080 9275 8636 4458 5280 3695

1.4 1.2 1.3 1.4 1.4 1.4

[1.3-1.4] [1.2-1.3] [1.3-1.3] [1.3-1.4] [1.3-1.4] [1.3-1.4]

ref \.0001 \.0001 ns ns ns

62,480 44,847 42,091 21,829 25,811 17,901

7.1 6.0 6.3 6.9 6.7 6.6

[7.0-7.1] [6.0-6.1] [6.3-6.4] [6.8-6.9] [6.7-6.8] [6.5-6.7]

ref \.0001 \.0001 \.001 \.0001 \.0001

CI, Confidence interval; na, not available; ns, not significant; ref, referent category. *P value derived from binomial test for proportions, except where otherwise indicated. y P value derived from Cuzick28 nonparametric test for trend across all age categories. z P value \.0001 using x 2 global test of significance across all categories.

malignancies and provide additional information regarding the impact of skin disease in the United States. These data suggest that although death from any particular nonneoplastic dermatosis is relatively infrequent, aggregate mortality for nonneoplastic skin diseases are similar to those of many other serious conditions reported as underlying causes of death in the CDC WONDER database over the same 11-year period, including Clostridium difficileeassociated

enterocolitis (44,915 deaths; age-adjusted mortality 1.4/100,000 persons), aortic dissection (37,331 deaths; age-adjusted mortality 1.2/100,000 persons), and acute pancreatitis (20,159 deaths; age-adjusted mortality 0.6/100,000 persons). By comparison, our age-adjusted mortality from nonneoplastic skin disease (as underlying cause of death) was approximately half that of melanoma, which ranged from 2.6/100,000 persons in 1999 to 2.8/100,000 persons in 2009 according to the Surveillance, Epidemiology,

J AM ACAD DERMATOL

52 Lott and Gross

JANUARY 2014

Table III. Nonneoplastic skin disease deaths in the United States and age-adjusted mortality (per 1 million persons), 1999 to 2009: 10 most common conditions Reported as underlying cause of mortality Disease category*

Deaths

Age-adjusted rate [95% CI]

Decubitus and other ulcers 19,678 6.0 [6.0-6.1] Cellulitis 14,825 4.5 [4.5-4.6] Abscesses 2202 0.7 [0.6-0.7] Cutaneous lupus erythematosus 2189 0.7 [0.6-0.7] Toxic epidermal necrolysis/ 1290 0.4 [0.4-0.4] erythema multiforme Pemphigoid 821 0.3 [0.2-0.3] Other/unspecified skin disease 742 0.2 [0.2-0.2] Pemphigus 568 0.2 [0.1-0.2] Psoriasis 369 0.1 [0.1-0.1] Dermatitis 239 0.1 [0.1-0.1] Total 42,923 13.2 [13.1-13.3] Reported as contributing cause of mortality Disease category*

Decubitus and other ulcers Cellulitis Cutaneous lupus erythematosus Abscesses Other/unspecified skin disease Psoriasis Pemphigoid Toxic epidermal necrolysis/ erythema multiforme Pemphigus Dermatitis Totaly

Deaths

Age-adjusted rate [95% CI]

119,930 59,075 11,236

36.8 [36.6-37.0] 18.1 [17.9-18.2] 3.4 [3.4-3.5]

9244 5602

2.8 [2.8-2.9] 1.7 [1.7-1.8]

3785 3268 2651

1.2 [1.1-1.2] 1.0 [1.0-1.0] 0.8 [0.8-0.8]

1856 1570 218,217

0.6 [0.6-0.6] 0.5 [0.4-0.5] 66.9 [66.6-67.1]

CI, Confidence interval. *See Appendix at http://www.jaad.org for a list of International Statistical Classification of Diseases, 10th Revision codes comprising disease categories. y Total reflects frequency of contributing causes of mortality and is greater than total number of deaths, because any single death record may report up to 20 contributing causes.

and End Results database (http://seer.cancer.gov/ faststats/). In particular, these results reveal significant mortality attributed to decubitus ulcers and cellulitis, and disproportionate mortality borne by patients ages 65 years and older. Also, age-adjusted mortality in blacks/African Americans was approximately triple that of whites in our study, and mortality was greater for individuals living in the South compared with other regions of the country. Although assessing the interdependence (eg, through regression modeling) among these variables and their impact on mortality

Table IV. Underlying causes of death and age-adjusted mortality (per 100,000 persons) associated with nonneoplastic skin diseases reported as contributing causes of death, 1999 to 2009: 10 most common conditions Disease category

Deaths

Age-adjusted rate [95% CI]

Atherosclerotic heart disease/ acute myocardial infarction Dementia Decubitus and other ulcers Cellulitis Diabetes mellitus Stroke Renal failure Sepsis Chronic obstructive pulmonary disease/emphysema Peripheral vascular disease Total

21,728

0.7 [0.6-0.7]

20,316 19,634 14,822 11,721 11,675 6102 6059 5695

0.6 0.6 0.5 0.4 0.4 0.2 0.2 0.2

5599 123,351

[0.6-0.6] [0.6-0.6] [0.4-0.5] [0.4-0.4] [0.4-0.4] [0.2-0.2] [0.2-0.2] [0.2-0.2]

0.2 [0.2-0.2] 4.0 [3.8-4.0]

CI, Confidence interval.

was not possible using the currently available, aggregated CDC WONDER data, we believe that potentially meaningful relationships exist within these results. Decubitus and other chronic cutaneous ulcers, for example, more commonly occur in older populations,30 in patients with dementia (who are often immobilized),30,31 and in patients with diabetes,31-33 and are known to be independent risk factors for mortality.30,34-37 In addition, the largest percentage of blacks/African Americans reside in the South and urban areas (http://www.census.gov/ population/race/), and blacks/African Americans are also at increased risk for diabetes.38 Similarly, cellulitis may be more prevalent in older individuals,39 blacks/African Americans,39 and diabetics.40 These associations are congruent with our results identifying cardiovascular disease, dementia, and diabetes mellitus (in addition to chronic ulcers and cellulitis) as frequent underlying causes of death (inclusive of all diseases) associated with nonneoplastic skin disease reported as contributing causes of mortality (Table IV). The concurrence of multiple disorders (eg, diabetes and cellulitis, dementia and pressure ulcers) in conjunction with greater disease incidence in select patient groups (ie, African Americans and the elderly) in select geographic areas (ie, the South and urban areas) may, therefore, partially explain the frequency of chronic ulcers and cellulitis as major contributing causes of mortality from nonneoplastic skin disease observed in this study. However, these results may also be interpreted to suggest potential barriers to dermatologic services

J AM ACAD DERMATOL VOLUME 70, NUMBER 1

resulting in significant mortality differences across sociodemographic parameters. Our findings are concordant, for example, with previous research demonstrating ongoing disparities in access to dermatologic care and worse skin-related outcomes faced by racial minorities in the United States.41-44 Variation in mortality by gender, region, and urbanization of residence may further indicate an unmet need for interventions that increase access to dermatologic expertise and interdisciplinary care. Given the mortality burden observed from chronic ulcers, cellulitis, and abscesseseconditions that are often managed by internists, wound care specialists, and other nondermatologistsebetter integration of dermatologists in the provision of team-based, coordinated care may help minimize death from these and other nonneoplastic skin diseases. These results also highlight mortality resulting from conditions commonly managed by dermatologists, such as cutaneous lupus erythematosus, pemphigoid, pemphigus, and psoriasis, despite improvements in the diagnosis and medical management of these disorders. Although we are uncertain which proportion (if any) of these deaths may have been preventable, we question whether at least a subset of these deaths may have resulted from disparities in access to timely and appropriate dermatologic care. Policies aimed at optimizing dermatology workforce distribution, improving coordination and integration of dermatology with primary care, and enhancing diffusion of teledermatology technologies may help ameliorate these disparities in the future. This study has important limitations. Death certificate data inherently reflect priorities and subjectivities of health care professionals, who are prone to errors in documentation, coding, and misattribution of underlying and contributing causes of death, all of which intrinsically limit the fidelity of our data and demand that our results be interpreted with appropriate caution.45-48 A recent validation study comparing death certificate data to hospital records, for example, revealed overestimation of herpes zoster as an underlying or contributing cause of death,49 and thus death certificate reporting of nonneoplastic cutaneous disease may suffer from similar inaccuracy as noncutaneous disease. Arguably, however, designation of many of these conditions, such as psoriasis, pemphigoid, and pemphigus, might not be mistakenly coded by physicians, given their distinctive nature and infrequent occurrence (as opposed to, say, ill-defined diagnoses such as cardiac arrhythmia, cardiopulmonary arrest, respiratory failure, and other ‘‘catch-all’’ categories50), though in the absence of autopsy

Lott and Gross 53

confirmation definitive cause of death ultimately cannot be known. As such, our findings may underestimate or overestimate the actual mortality burden of nonneoplastic skin disease. In addition, our study, by design, cannot account for underlying clinical risk factors that invariably impact prognosis from nonneoplastic skin disease, nor can our analysis identify potential pitfalls in diagnosis, management, or other confounders that may influence mortality. Future research using patient-level health information and autopsy data is needed to provide more precise and risk-adjusted estimates of mortality from nonneoplastic skin disease. Nonetheless, we believe this study underscores the clinical and public health relevance of nonneoplastic skin disorders as potentially preventable causes of death in the United States. We hope our findings may help inform ongoing efforts in health care system reform and reorganization,51 which should consider expanded access to dermatologic care a priority in our pursuit of exceptional population-based health. REFERENCES 1. Hollestein LM, de Vries E, Nijsten T. Trends of cutaneous squamous cell carcinoma in the Netherlands: increased incidence rates, but stable relative survival and mortality 1989-2008. Eur J Cancer 2012;48:2046-53. 2. Clayman GL, Lee JJ, Holsinger FC, Zhou X, Duvic M, EI-Naggar AK, et al. Mortality risk from squamous cell skin cancer. J Clin Oncol 2005;23:759-65. 3. Hannuksela-Svahn A, Pukkala E, Karvonen J. Basal cell skin carcinoma and other nonmelanoma skin cancers in Finland from 1956 through 1995. Arch Dermatol 1999;135:781-6. 4. Farmer ER, Helwig EB. Metastatic basal cell carcinoma: a clinicopathologic study of seventeen cases. Cancer 1980;46: 748-57. 5. Wong C, Strange R, Lear J. Basal cell carcinoma. BMJ 2003; 327:794-8. 6. Pollack LA, Li J, Berkowitz Z, Weir HK, Wu XC, Ajani UA, et al. Melanoma survival in the United States, 1992 to 2005. J Am Acad Dermatol 2011;65(Suppl):S78-86. 7. Weir HK, Marrett LD, Cokkinides V, Barnholtz-Sloan J, Patel P, Tai E, et al. Melanoma in adolescents and young adults (ages 15-39 years): United States, 1999-2006. J Am Acad Dermatol 2011;65(Suppl):S38-49. 8. Jemal A, Saraiya M, Patel P, Cherala SS, Barnholtz-Sloan J, Kim J, et al. Recent trends in cutaneous melanoma incidence and death rates in the United States, 1992-2006. J Am Acad Dermatol 2011;65(Suppl):S17-25, e1-3. 9. Steen VD, Medsger TA. Changes in causes of death in systemic sclerosis, 1972-2002. Ann Rheum Dis 2007;66:940-4. 10. Condliffe R, Kiely DG, Peacock AJ, Corris PA, Gibbs JS, Vrapi F, et al. Connective tissue disease-associated pulmonary arterial hypertension in the modern treatment era. Am J Respir Crit Care Med 2009;179:151-7. 11. Bernatsky S, Boivin JF, Joseph L, Manzi S, Ginzler E, Gladman DD, et al. Mortality in systemic lupus erythematosus. Arthritis Rheum 2006;54:2550-7. 12. Joly P, Roujeau JC, Benichou J, Delaporte E, D’Incan M, Dreno B, et al. A comparison of two regimens of topical corticosteroids in the treatment of patients with bullous pemphigoid: a

54 Lott and Gross

13.

14. 15.

16.

17.

18.

19.

20.

21.

22. 23.

24.

25. 26. 27. 28. 29.

30.

31.

32.

multicenter randomized study. J Invest Dermatol 2009;129: 1681-7. Carson PJ, Hameed A, Ahmed AR. Influence of treatment on the clinical course of pemphigus vulgaris. J Am Acad Dermatol 1996;34:645-52. Jessop S, Khumalo NP. Pemphigus: a treatment update. Am J Clin Dermatol 2008;9:147-54. Pearce DJ, Lucas J, Wood B, Chen J, Balkrishnan R, Feldman SR. Death from psoriasis: representative US data. J Dermatolog Treat 2006;17:302-3. National Center for Health Statistics. Vital statistics of the United States, 1999: technical appendix. Hyattsville (MD): National Center for Health Statistics; 2004. Available from http://wonder.cdc.gov/wonder/help/CMF/TechnicalAppendix 1999.pdf. Accessed August 8, 2013. Calcaterra S, Glanz J, Binswanger IA. National trends in pharmaceutical opioid-related overdose deaths compared to other substance-related overdose deaths: 1999-2009. Drug Alcohol Depend 2013;131:263-70. Pietrzak A, Bartosinska J, Chodorowska G, Szepietowski JC, Paluszkiewicz P, Schwartz RA. Cardiovascular aspects of psoriasis: an updated review. Int J Dermatol 2013;52:153-62. Boehncke WH. Epidemiology and potential pathomechanisms of cardiovascular comorbidities in psoriasis: a report from the GRAPPA 2010 annual meeting. J Rheumatol 2012; 39:441-4. Lillie-Blanton M, Parsons PE, Gayle H, Dievler A. Racial differences in health: not just black and white, but shades of gray. Annu Rev Pub Health 1996;17:411-48. Lantz PM, House JS, Lepkowski JM, Williams DR, Mero RP, Chen J. Socioeconomic factors, health behaviors, and mortality: results from a nationally representative prospective study of US adults. JAMA 1998;279:1703-8. Yoo JY, Rigel DS. Trends in dermatology: geographic density of US dermatologists. Arch Dermatol 2010;146:779. Uhlenhake E, Brodell R, Mostow E. The dermatology work force: a focus on urban versus rural wait times. J Am Acad Dermatol 2009;61:17-22. Marwick C. ‘CDC WONDER’ computer use revolutionizes acquiring, applying reams of public health data. JAMA 1993;269:2063. Weed JA. Vital statistics in the United States: preparing for the next century. Popul Index 1995;61:527-39. Hoyert DL, Arias E, Smith BL, Murphy S, Kochanek K. Deaths: final data for 1999. Natl Vital Stat Rep 2001;49:1-113. Ingram DD, Franco SJ. NCHS urban-rural classification scheme for counties. Vital Health Stat 2 2012;154:1-65. Cuzick J. A Wilcoxon-type test for trend. Stat Med 1985;4:543-7. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Series B 1995;57:289-300. Flattau A, Blank AE. Risk factors for 90-day and 180-day mortality in hospitalised patients with pressure ulcers. Int Wound J doi: 10.1111/j.1742-481X.2012.01032.x. Published online June 28, 2012. Coleman S, Gorecki C, Nelson EA, Closs SJ, Defloor T, Halfens R, et al. Patient risk factors for pressure ulcer development: systematic review. Int J Nurs Stud 2013;50:974-1003. Bours GJ, Halfens RJ, Abu-Saad HH, Grol RT. Prevalence, prevention, and treatment of pressure ulcers: descriptive study in 89 institutions in the Netherlands. Res Nurs Health 2002;25:99-110.

J AM ACAD DERMATOL

JANUARY 2014

33. Leigh IH, Bennett G. Pressure ulcers: prevalence, etiology, and treatment modalities: a review. Am J Surg 1994;167: S25-30. 34. Sullivan N, Schoelles KM. Preventing in-facility pressure ulcers as a patient safety strategy: a systematic review. Ann Intern Med 2013;158:410-6. 35. Maida V, Ennis M, Kuziemsky C, Corban J. Wounds and survival in cancer patients. Eur J Cancer 2009;45:3237-44. 36. Maida V, Ennis M, Kuziemsky C, Corban J. Wounds and survival in noncancer patients. J Palliat Med 2010;13:453-9. 37. Barba R, Martinez JM, Zapatero A, Plaza S, Losa JE, Canora J, et al. Mortality and complications in very old patients (901) admitted to departments of internal medicine in Spain. Eur J Intern Med 2011;22:49-52. 38. Harris MI. Diabetes in America: epidemiology and scope of the problem. Diabetes Care 1998;21:C11-4. 39. Klevens RM, Morrison MA, Nadle J, Petit S, Gershman K, Ray S, et al. Invasive methicillin-resistant Staphylococcus aureus infections in the United States. JAMA 2007;298:1763-71. 40. Muller L, Gorter K, Hak E, Goudzwaard WL, Schellevis FG, Hoepelman AI, et al. Increased risk of common infections in patients with type 1 and type 2 diabetes mellitus. Clin Infect Dis 2005;41:281-8. 41. Wich LG, Ma MW, Price LS, Sidash S, Berman RS, Pavlick AC, et al. Impact of socioeconomic status and sociodemographic factors on melanoma presentation among ethnic minorities. J Community Health 2011;36:461-8. 42. Hu S, Parmet Y, Allen G, Parker DF, Ma F, Rouhani P, et al. Disparity in melanoma: a trend analysis of melanoma incidence and stage at diagnosis among whites, Hispanics, and blacks in Florida. Arch Dermatol 2009;145:1369-74. 43. Taylor SC, Kelly AP, Dupree NE, Kimball AB, Lawrence RC. Health disparities in arthritis and musculoskeletal and skin diseasesethe dermatology session: National Institute of Arthritis and Musculoskeletal and Skin Diseases, Bethesda, Maryland, December 15-16, 2000. J Am Acad Dermatol 2002; 47:770-3. 44. Buster KJ, Stevens EI, Elmets CA. Dermatologic health disparities. Dermatol Clin 2012;30:53-9. 45. Landes SD, Peek CW. Death by mental retardation? The influence of ambiguity on death certificate coding error for adults with intellectual disability. J Intellect Disabil Res 2013; 57:1183-90. 46. Kircher T, Nelson J, Burdo H. The autopsy as a measure of accuracy of the death certificate. N Engl J Med 1985;313: 1263-9. 47. Yin D, Morris CR, Bates JH, German RR. Effect of misclassified underlying cause of death on survival estimates of colon and rectal cancer. J Natl Cancer Inst 2011;103:1130-3. 48. Drummond MB, Wise RA, John M, Zvarich MT, McGarvey LP. Accuracy of death certificates in COPD: analysis from the TORCH trial. COPD 2010;7:179-85. 49. Mahamud A, Marin M, Nickell SP, Shoemaker T, Zhang JX, Bialek SR. Herpes zoster-related deaths in the United States: validity of death certificates and mortality rates, 1979-2007. Clin Infect Dis 2012;55:960-6. 50. Ravakhah K. Death certificates are not reliable: revivification of the autopsy. South Med J 2006;99:728-33. 51. Resneck JS Jr. An analysis of health system reform for dermatologists: elements and implications of the Patient Protection and Affordable Care Act. J Am Acad Dermatol 2010;63:706-15.

J AM ACAD DERMATOL

Lott and Gross 54.e1

VOLUME 70, NUMBER 1

Appendix. Disease categories and corresponding International Statistical Classification of Diseases, 10th Revision codes Disease category

Abscess Atherosclerotic heart disease/acute myocardial infarction Cellulitis Chronic obstructive pulmonary disease/emphysema Cutaneous lupus erythematosus Decubitus and other ulcers Dementia Dermatitis Diabetes mellitus Other/unspecified skin disease Pemphigoid Pemphigus Peripheral vascular disease Psoriasis Renal failure

Sepsis Stroke

Toxic epidermal necrolysis/erythema multiforme

ICD-10 codes

L02.0-L02.4, L02.8, L02.9 I21.3, I21.4, I21.9, I22.9, I24.9, I25.0, I25.1, I25.2, I25.9, I51.6, I70.9 L03.0-L03.3, L03.8, L03.9, L08.0, L08.8, L08.9 J43.2, J43.9, J44.0, J44.1, J44.8, J44.9, J98.2 L93.0-L93.2 L89.0, L97.0, L98.4 F01.1, F01.9, F03.0, F05.1, G30.1, G30.9 L20.8, L20.9, L21.1, L21.8, L21.9, L22.0, L23.7, L23.9, L24.5, L24.8, L24.9, L25.1, L25.5 L25.8, E10.0-E10.7, E10.9, E11.0-E11.9, E14.0-E14.9, P70.2 L11.9, L13.8, L13.9, L53.8, L53.9, L56.8, L57.8, L57.9, L59.8, L72.8, L73.8, L73.9, L85.8, L85.9, L92.8, L92.9, L98.8, L98.9 L12.0, L12.1, L12.8, L12.9 L10.0-L10.2, L10.4, L10.5, L10.8, L10.9 I73.8, I73.9 L40.0-L40.5, L40.8, L40.9 I12.0, I13.0-I13.2, I13.9, K76.7, N00.9, N03.9, N04.9, N05.1, N05.2, N05.5, N05.7, N05.8, N05.9, N10.0, N11.0, N13.2, N13.6, N15.9, N17.0, N17.9, N18.0, N18.8, N18.9, N19.0, N25.8, N25.9, N28.0, P96.0, Y84.1 A02.1, A26.7, A39.4, A40.0-A40.3, A40.8, A40.9, A41.0-A41.5, A41.8, A41.9, B37.7, P36.9 G45.9, I60.0, I60.1, I60.7-I60.9, I61.0-I61.5, I61.8, I61.9, I62.0, I62.9, I63.0, I63.2-I63.5, I63.8, I63.9, I64.0, I67.2, I67.9, I69.0-I69.4, I69.8, P52.3, P52.4 L51.1, L51.2, L51.8, L51.9

ICD-10, International Statistical Classification of Diseases, 10th Revision.