Sarcoidosis mortality in the United States, 1979–1991: An analysis of multiple-cause mortality data

Sarcoidosis mortality in the United States, 1979–1991: An analysis of multiple-cause mortality data

Sarcoidosis Mortality in the United States, 19794991: An Analysis of Multiple-Cause Mortality Data Nancy M. Gideon, MD, David M. Mannino, MD, Atlanta,...

585KB Sizes 0 Downloads 34 Views

Sarcoidosis Mortality in the United States, 19794991: An Analysis of Multiple-Cause Mortality Data Nancy M. Gideon, MD, David M. Mannino, MD, Atlanta, Georgia WRPOSE: We sought to describe

sarcoidosis morin the United States from 1979 through 1991. METHODS: We analyzed death certificate reports compiled by the National Center for Health Statistics for the period 1979 through 1991. RESULTS: Of the 26,866,6OC people who died during the study period, 9,014 had a diagnosis of sarcoidosis listed on their death certificates. We restricted our study group to 5,791 people who died because of sarcoidosis or one of its complications. Among men, age-adjusted mortality rates increased from 1.3 per l,OOO,OOO in 1979 to 1.6 per 1,000,000 in 1991, and among women, these rates increased from 1.9 per 1,000,000 in 1979 to 2.5 per l,OOO,OOO in 1991. Age-adjusted mortality rates were consistently higher among blacks than among whites. Age-adjusted mortality rates, stratified by race, varied by state. Among whites, the highest rates were in northern states, while among blacks, the highest rates were in the Middle Atlantic and northern Midwestern states. CONCLUSIONS: Reported mortality due to sarcoidosis varies by region, sex, and race. We cannot determine whether these differences are related to characteristics of the disease, or problems in death certification and coding. talii

arcoidosis is a multisystemic, chronic granulomS atous disease process of unknown cause.’ Many possible etiologic factors, including infectious, genetic, autoimmune, and environmental agents, in this disease or group of diseases have been proposed.‘12 Incidence rates for sarcoidosis range from 1.4 per 100,000 in Japan3 to 9.6 per 100,000 in the German Democratic Republic.4 Prevalence rates of sarcoidosis range from 0.04 per 100,000 in Spain to 64 per 100,000 in Sweden.4 Blacks in the United States have a prevalence of 64 per 100,000.5 Case mortality rates of sarcoidosis have been reported to From the Air Pollution and Respiratory Health Branch, Division of Environmental Hazards and Health Effects, National Center for Environmental Health. Centers for Disease Control and Prevention, Atlanta, Georgia. Reouests for reprints should be addressed to Dr. David M. Mannino, National Center for Environmental Health, Centers for Disease Control and Prevention (CDC), MSF39, 4770 Buford Hwy NE, Atlanta, Georgia 30341.3724. Manuscript submitted July 17, 1995 and accepted in revised form October 31, 1995.

vary from 2.29/06to 28.1%,7 although most authorities estimate the overall case mortality rate of sarcoidosis to be 5%.’ We have analyzed the US mortality from 1979 through 1991 by sex, race, age, and state of residence.

METHODS Using the vital records loom individual states, the National Center for Health Statistics (NCHS) annually compiles data from all death certificates filed in the United States (about 2,000,OOO each year). These data include demographic and geographic information on the decedent, as well as the Inlernational Classification of Diseases (ICD) codes for the underlying cause of death and up to 20 conditions listed on the death certificate. The International Classification of Diseases, Ninth Revision (ICD-9) was irqplemented in 1979,s and it was in effect during the 13year period we studied. The code for sarcoidosis ji ICD-9 is 135. The conditions are recorded in two :forms: the entity axis, which contains the conditions exactly as reported on the death certificate, and the record axis, which is edited by a computerized algorithm known as the translation of axes. The Automated Classification of Medical Entities algorithm determines the underlying cause of death from the conditions and their positions as listed on the death certificates.g Quality assurance of the data is maintained by trained nosologists who code conditions at the state level, and, in turn, by nosologists at NCHS who periodically review data from a loo/o sample of the submitted death certificates. The result of this process is the multiplecause mortality file (MCMF). We searched the record-axis portion of the 1979 to 1991 MCMFs for records containing ICD-91 code 135 (sarcoidosis). Because we were only interested in including deaths due to sarcoidosis or one of its complications, we included in our study group decedents for whom the following diseases were listed as the underlying cause of death: sarcoidosis (ICD-9 135); as pergillosis (ICD-9 117.3); cryptococcoses (ICD-9 117.5); cardiomyopathy (ICD-9 425 to 425.9); puImonary hypertension (ICD-9 416 to 416.9); pneumo nia (ICD-9 481,485,486); obstructive airways disease (ICD-9 492 to 494.9,496); or pulmonary fibrosis (ICD9 515 or 516.3).3t10 We analyzed the sarcoidosis group according to age, race, and sex. We used the 1980 and 1990 US census data (using linear interpolation to estimate population in intercensal years) to calculate rates. For racestratiiied, age-adjusted, state-specific April1996

The American Journal of Medicine@ Volume 100

423

TABLE I Distribution of Decedents in the US Wiih Sarcoidosis or One of its Complications’ as the Underlying Cause of Death or as Any Cause of Death. Underlying Cause Any Cause of Death of Death nt nt % % Sarcoidosis 5,064 56.2 9,014 100.0 Fungal infections 87 1.0 175 1.9 Pneumonia 200 2.3 863 9.6 Cardiomyopathy 136 1.5 247 2.7 Pulmonary hypertension 52 0.6 86 1.0 Obstructive airways disease 178 2.4 953 10.6 Pulmonary fibrosis 74 0.8 597 6.6 Total 5,791 64.2 9,014 100.0 Table values are numbers of patients. ‘Sarcoidosrs (International Classification of Diseases, Ninth Revrsion [ICD91 1351, fungal infections fICD9 117 to 117.91, pneumonia (ICD-9 481, 485, 486), cardiomyopathy (ICD-9 4251, pulmonary hypertension WI!-9 4161, obstructive airways disease KD-9 492 to 494.9,496), or pulmonary fibrosrs KD9 515 or 516.3). The percentages are of the 9,014 death records that listed sarcoidosis as present, from the death certificates of 26,866,600 decedents in the United States, 1979 through 1991; MultipleCause Mortality Files. +Number of decedents.

TABLE II Age Stratification for Death in the US During 1979 to 1991, and the Percentage of Deaths by Sarcoidosis Within Each Age Stratum All Decedents With Decedents Sarcoidosis’ Age (Y) (%I
Files, National Center for Health

obstructive airways disease, or pulmonary fibrosis listed as the underlying cause of death (Table I). Thus our study group contained 5,791 decedents Age-Adjusted Mortality Rate per 1O,OOO,OOO (0.02% of total decedents). The age-adjusted mortality rate due to sarcoidosis Black Women or its complications increased slightly from 1.6 per million in 1979to 2.1 per million in 1991.Among men, deaths increased from 137 in 1979to 225 in 1991, and the age-adjusted mortality rate also increased from 1.3 per million in 1979 to 1.6 per million in 1991. Among women, deaths due to sarcoidosis; increased White Women from 220 in 1979 to 356 in 1991, and the age-adjusted mortality rate also increased from 1.9 per million in 1979 to 2.5 per million in 1991. White Men The age-adjusted mortality rates for sarcoidosis 2’ 1 I and its complications were consistently higher among blacks than whites (Figure 1). Within these 1979 80 81 82 88 84 86 88 87 88 89 98 91 racial strata, mortality rates were consistently 30%to Figure 1. Mortality rates of people who died because of 40% higher among women than among men. Age-adjusted mortality rates among people of races other sarcoidosis or one of its complications per 10,000,000 stratified by race and sex in the United States from 1979 through 1991 (12 than black or white were lower than those among deaths among people of other races are not included). From the whites, ranging from 0 to 0.5 per l,OOO,OOO during the Multiple-Cause Mortality Files, National Center for Health Statistics. study period. The mortality rates due to sarcoidosis and its comrates, we used the 1990 US population as the stan- plications varied by age group. The largest number of dard, using the race-specific populations in each state. deaths (1,283) and the highest percentage of total deaths (0.14%) occurred in the 35- to 44year age group RESULTS (Table II). Within age group, mortality r;&s varied During the 13-year period, 9,014 death records by race and sex (Table III). Among men and women listed sarcoidosis (ICD-9 code 135) either as the tm- of both races, the lowest age-specific mortality rates derlying cause of death or as a contributing cause of were in people younger than age 25 years. Among death among the 26,866,600decedents. Of those 9,014 black men and women, the highest mortality rates death records, 56.2% (5,064 records) had sarcoidosis were in the 45- to 54year age group, whereas among listed as the underlying cause of death. An additional white men and women, the highest mortality rates 727 decedents had aspergillosis, cryptococcoses, car- were in the 75-to 84-year agegroup. Among men, mordiomyopathy, pulmonary hypertension, pneumonia, tality rates in the 25 to 34yea.r age group remained 424

April 1996 The American Journal of Medicine@ Volume 100

SARCOIDOSIS MORTALIN/GIDEON TABLE III Annual US Mortality Rates During 1979 to 1991 per 10,000,000 Population for Death by Sarcoidosis,’ Stratified by Age Group, Race, and Sex Black Black White White Males Females Males Females Age (~1
Files, National Center for Health

stable through the study period, whereas those in the 35 to 74year age groups increased slightly through the study period (Figure 2); rates are not shown for men or women older than 75 because the rates were highly variable. Among women, there was a similar pattern of mortality rates; rates were stable through the study period in the 25 to 34-year age group, but they increased slightly in the 35 to 74year age groups (Figure 3). The age-adjusted annual mortality rate of sarcoidosis and its complications from 1979 through 1991varied by state of residence. These rates ranged from a low of 0.2 per l,OOO,OOO in Hawaii to a high of 15.2 per l,OOO,OOO in the District of Columbia. After we stratified the data by race and recalculated ageadjusted mortality rates by state using race-specific populations, a different pattern emerged. Among whites, the highest mortality rates were in northern states, whereas among blacks, the highest mortality rates were in Midwestern and Middle Atlantic states (Figures 4 and 5).

COMMENTS We have determined that the overall age-adjusted mortality rate of sarcoidosis in the United States has increased from 1.6 per l,OOO,OOO in 1979 to 2.1 per l,OOO,OOO in 1991. This mortality rate has increased from 1958,when it was 1.0 per million.” We have also determined that rates are higher among blacks than among whites and among women than among men. Mortality rates varied by state and race, where the highest rates among whites were found in northern states, whereas the highest rates among blacks were found in Middle Atlantic and Midwestern states. Our study group included only those deaths due to sarcoidosis or its complications. These cases comprised 64.2% of the 9,014 death records in which sar-

AND MANNlND

Mortality Rates per 10,000,000 Men 70 c

40 30

2

20 10 1979

60

61

62

63

64

66

66

67

66

69 90

91

Age Group in Years 26-34 36-44 -----c--f--

46-64

66-64

66-7’4

I

Figure 2. Mortality rates of men who died because of sarcoidosis or one of its complications per 10,000,000 men, stratified by age, in the United States from 1979 through 1991. From the MultipleCause Mortality Files, National Center for Health Statistics. Mortality Rates per 1 O,OOO,OOOWomen 70 60

Figure 3. Mortality rates of women who died because of sarcoidosis or one of its complications per 10,000,000 women, stratified by age, in the United States from 1979 through 1991. From the MultipleCause Mortality Files, National Center for Health Statistics.

coidosis was listed as a cause of death. This finding is similar to findings in studies done in Japan3 and the United States.l” In the Japanese study,3 60.6% of 320 deaths where sarcoidosis was present were due to sarcoidosis and its complications. That study also showed that most (77%) of those deaths were due to the cardiac complications of sarcoidosis. This fmding can be contrasted to findings in the study in the United States,l” which showed that 68%of tlhe deaths in which sarcoidosis was present were due to sarcoidosis, and 87% of these deaths were secondary to pulmonary complications. We found that pulmonary complications of sarcoidosis, including pneumonia, pulmonary fibrosis, and obstructive airways disease, were listed as a cause of death more frequently than was cardiomyopathy. Of course, it is possible that the April

1996 The

American

Journal

of Medicine@

Volume 100

425

SARCOIDOSIS

MORTALITY/GIDEON

Age-adjusted

Age-adjusted 0

4

per 10 million

Mortality

Mortality

Rates

AND MANNINO

Rates

among

by State (1979

6 to 9 per 10 million

n

Whites

through

Age-adjusted

1991)

Age-adjusted

9 to 24 per 10 million

0

~70 per 10 million

Mortality

Mortality

Rates

Rates

among

by State (1979 through

70 to 90 per 10 million

n

Ellacks

1991)

90 to 124 per 10 million J

Figure 4.

Age-adjusted annual mortality rate of sarcoidosis by state among whites in the United States from 1979 through 1991. From the Multiple-Cause Mortality Files, National Center for Health Statistics.

Figure 5. Age-adjusted annual mortality rate of s8arcoidosrs by state among blacks in the United States from 1979 through 1991. From the Multiple-Cause Mortality Files, National Center for Health Statistics.

cardiac complications of sarcoidosis, which might netic or environmental factors. We found that the racause cardiac conduction disturbances and sudden tio of mortality rates among blacks to t.hat among death, might not be identified as such, especially in whites varied by age group: among men, it ranged an older population. l2 from 3.1 in the 85-year-old and older ag;e group to Overall, we found mortality rates of sarcoidosis and 35.0 in the 15-through-24-year-old age group; among its complications to be higher among women than women it ranged from 2.0 in the 85-year-old and among men in both blacks and whites. Similar fmd- older age group to 38.3 in the 15-through-2Pyearings have been reported in earlier studies from the old age group. Sarcoidosis prevalence varies by geographic reUnited State&‘” and Japan.‘” We also found that the ratio of male-to-female mortality rates varied by age gion, both between countries and within counThe reported prevalence ranges from 0.04 groups: among blacks this ratio ranged from 0.55 in tries.1,“,5,1g the 55through-64-year-old age group to 1.21 in the per 100,000in Spain to 64 per 100,000in Sweden The 15through-%&year-old age group; among whites this reasons for this wide range are likely related to a difference in case ascertainment among different counratio ranged from 0.58 in the 55-through-64-year-old age group to 1.33 in the 15-through-24-year-old age tries, in addition to any actual differences that exist. group. Many studies have reported a higher preva- Within individual countries, where one would expect lence of sarcoidosis among women than among case ascertainment biases to be less of a factor, sarmen.1*4,15 A study in the United States #at evaluated coidosis prevalence varies by geographic area sarcoidosis prevalence in a community-based white Results of studies in Japan, Denmark, and Italy have population, however, found that rates were similar shown a higher rate of sarcoidosis in the northern among men and women. I6Unpublished data from the part of the country (in the northern hemisphere) comThis finding Mayo Clinic suggest that the higher reported preva- pared with the rate in southern part~+.~~‘~ of a higher prevalence in the northern region of a lences of sarcoidosis in women may be related to country, however, has been disputed as being artiwomen having a higher incidence of symptoms and factualzo In the United States, sarcoidosis prevalence seeking more medical attention. li This bias of women has been reported to be higher in the Middle Atlantic seeking more medical attention than men, however, would not have a large effect on mortality, unless sar- and South Atlantic states for both whites and blacks, coidosis was totally unrecognized in men who died and in rural areas compared with urban area.s.1~18 Results of a previous analysis of sarcoidosis mortalbecause of sarcoidosis or one of its complications. Both the prevalence of and mortality rate due to ity in the United States revealed the highest mortalsarcoidosis in the United States is higher among ity rates to be in the South Atlantic states, although blacks than whites.j,1”,13v18 The reasons for this racial that study did not control for the racial distribution difference are not clear, but may be related to ge- of the population. l l We determined the lowest rate of 426

April

1996

The American

Journal

of Medicine@

Volume

100

SARCOIDOSIS MORTALITY/GIDEON

AND MANNINO

sarcoidosis mortality to be in Hawaii, where this disease is known to be rare.21 The distribution of the age-adjusted sarcoidosis mortality in our study varied dramatically by race. Among whites, the highest rates were found in northem states, whereas among blacks, the highest rates were found in the Middle Atlantic states and the northern Midwestern states. We were not able to control for rural residence in our analysis. A limitation of our study is our dependence on data from death certificates. Sarcoidosis may be unrecognized in some decedents and diagnosed when it is not present in others. Although we included in our analysis decedents in whom the underlying cause of death was not sarcoidosis, we did need to have sarcoidosis mentioned on the death record for inclusion in this study. We included people with sarcoidosis and an additional diagnosis of cryptococcosis in this study, although it is possible that some people with only cryptococcosis may have been misdiagnosed as also having sarcoidosis. Sarcoidosis is a syndrome with no known causethat may well represent several different diseases.Several different etiologic factors that people encounter in their environments, such asinfectious agents, farm animals, and wood smoke1 may be important in the development of sarcoidosis. Our analysis of death certificate data reveals that the geographic distribution and risk of mortality due to this syndrome differ among blacks and whites. The reasons for these findings are not clear, and may represent characteristics of the disease,differences in the diagnosis of the disease, or problems in death certification and coding.

3. lwai K, Tachibana T, Takemura T, et al. PathologIcal studie!; on sarcoidosis autopsy. 1. Epidemiological features of 320 cases in Japan. Acta Pathol Japan.

REFERENCES

Asia and Africa: analytic epidemiology. Ann NY Acad Sci. 1976;278:355-367. 20. Levinsky L, Cummlskey J, Romer FK, et al. Sarcoidosis in Europe: cooperative study. Ann NY Acad Sci. 1976;278:335-346.

1. Bresmtz EA, Strom 1983;5:124-156. 2. Thomas sarcoidoas.

BL. Epidemiology

PO, Hunninghake Am Rev Respir

of sarcoidosis.

GW. Current concepts Dis. 1987;135:747-760.

Epidemiol

of the pathogenesls

Rev. of

1993;43:372-376. 4.Teirstein AS, Lesser M. Worldwide dtstributlon and epidemiology of sarcoidosis. In: Fanburg BL, ed. Sarcoldosis and Other Granulomatous Diseases of the Lung. Vol. 20. 1983:101-134. 5. Young RC, Hackney RL, Harden KA. Epidemiology of sarcoidosis: ethnic and geographic considerations. JNMA. 1974;66:386-388. 6. James PG, Nevil e E, Stiltzbach LE, et al. A worldwlde review of sarcoidosis. Proceedings of the Vllth International Conference on Sarcoidosis and Other Granulomatous 7. Longcope investigation

Disorders. Ann NY Acad SCI. 1976;278:321-334. WT, Freiman DG. A study of sarcoidosis. Based on a combined of 160 cases including 30 autopsies from the Johns Hopkins

Hospital and Massachusetts General Hospital. Medrcine. 1952:31:1-132. 8. U.S. Public Health Service. lnternabonal Classification of Diseases, Ninth Revision. DHHS Publication No. (PHS) 80-1260, Washiqton, DC: U.S. Government Printing OffIce; 1980. 9. Centers for Disease Control and Prevention. Advance report of final mortaiity statistics, 1991. In: Monthly Vital Statistics Report. Atlanta: Centers for Disease Control and Prevention; 1993;42:56-61. 10. Huang CT, Heurich AE, Sutton AL, Lyons HA. Mortality In sarcoidosis. A changing pattern of the causes of death. Eur J Respir Dis. 1981;62: 231-238. 11. Morlyama IM. Mortality Resprr Dis. 1961;84isuppl):I

from

sarcoidosls

in the Untted

Am Rev

States.

16-119.

12. Matsui Y, lwal K, Tachibana T, et al. Cltnlcopathological myocardlal sarcoidosis. Ann NY Acad SC;. 1976;278:455-469. 13. Gillum RF. Sarcoidosis in the United States-1968

study

to 1984.

Hospitalization and death. JNMA. 1988;80:1179-1184. 14. Yamaguchi M, Hosoda Y, Sasaki R, Aokl K. Epidemlological sarcoldosis in Japan: recent trends In mcldence and prevalence changes in epldemiologtc features. Sarcoidous. 1989;6:138-146. 15. Milman N, Selroos 0. Pulmonary sarcoidosls in the Nordic

1950-1982. Epidemiology and cllnlcal picture. Sarcoidosis. 16. Henke CE, Henke G, Elveback LR, et al. The epidemiology

of fatal

study on rates and countries

1990;7:5C-57.

of sarcoidosis in Rochester, Minnesota: a population-based study of lnctdence and survival. Am J Epidemlol. 1986;123:840445. 17. DeRemee RA. Sarcoldosls. Mayo Ckn Proc. 1995;70:177-1.81. 18. Sartwell PE, Edwards LB. Epldemlology of sarcoldosls in the U.S. Navy. Am J Epldemiol. 1974;99:250-257. 19. Hosoda Y, Hiraga Y, Odaka M, et al. A cooperabve study of sarcoidosis in

21. Charles 204-210.

April

1996

J, Elpem

DJ. Sarcoldosis:

The American

Journal

a Hawaii

rarity.

of Medicine@

Hawaii

a

ivIed J. 1991;50:

Volume

100

427