Revista de Saúde Pública
I SSN 0034- 8910 versão im pressaRev Saúde Pública 2003; 37(6)
Multiple- causes- of- death related to tuberculosis in the state of
São Paulo, Braz il, 1 9 9 8
August o Hasiak Sant oa, Celso Escobar Pinheirob and Margarete Silva Jordanic
aDepartam ento de Epidem iologia da Faculdade de Saúde Pública da Universidade de São Paulo. São Paulo, SP, Brasil. bDepart am ent o de I nform át ica do Sist em a Único de Saúde do Minist ério da Saúde. Rio de Janeiro, RJ, Brasil. cFundação Sist em a Est adual de Análise de Dados. São Paulo, SP, Brasil
ABSTRACT
Objectives
The goal of t his paper is t o invest igat e m ort alit y relat ed t o t uberculosis in t he st at e of São Paulo, sout heast ern Brazil, according t o m ult iple causes of deat h and t heir int errelat ion with other underlying causes.
Methods
associat ed causes of deat h. St at ist ical analysis included analysis of variance, St udent ’s t- distribution, and chi- squared t est s.
Results
TB was t he underlying cause of 1,644 deat hs, a 4.6/ 100,000 populat ion m ort alit y rat e. Main associated causes were respiratory failure ( 46,9% ) , pneum onias ( 16.5% ) , other specified sym ptom s and signs involving circulat ory and respirat ory syst em s ( 13.9% ) , cachexia ( 12.9% ) , diseases of the circulat ory syst em ( 10.3% ) , condit ions due t o alcohol use ( 8.4% ) , sept icem ias ( 7.2% ) and m alnut rit ion ( 7.1% ) . Tuberculosis occurred as an associat ed cause in anot her 1,388 deat hs. The m ort alit y rat e including TB as a bot h unde rlying and associat ed cause was 8.9/ 100,000 populat ion, pract ically t wice t he classical rat e. Deat hs whose associat ed cause was report ed as being TB had as underlying causes: AI DS ( 65.3% ) , diseases of t he circulat ory syst em ( 8.9% ) , neoplasm s ( 7.5% ) , and diseases of the digestive system ( 4.8% ) . Clinical form s of nervous system and m iliary TB were m ore frequent as a cause associat ed wit h AI DS t han wit h ot her underlying causes ( p< 0,001) .
Conclusions
Tot al report s of TB- relat ed deat h pract ically doubled it s m ort alit y rat e as an underlying cause. The increase in TB m ort alit y was dem onst rat ed t o be influenced by t he AI DS epidem ic.
Keywords
Tuberculosis. Mort alit y. Cause of deat h. Acquired im m unodeficiency syndrom e. Aids- relat ed opportunistic infections. Underlying cause of deat h. Mort alit y rat e. Mult iple - cause- of- death.
INTRODUCTION
Causes of deat h can be assessed t hrough t he inform at ion regist ered by physicians in deat h cer t ificat es.
Of t he various causes regist ered, prim ary deat h st at ist ics are based on t he “ underlying cause of deat h” , defined by t he World Healt h Organizat ion ( WHO) as “ ( a) t he disease or inj ury which init iat ed t he t rain of m orbid event s leading direct ly t o deat h, or ( b) t he circum st ances of t he accident or violence which produced t he fat al inj ury” .8 The underlying cause has been considered as t he m ost efficacious way of prevent ing deat h. During t he last half of t he Twent iet h Cent ury, however, t he growing dem and for full descriptions of deat h – including all causes m ent ioned in t he deat h cert ificat e – det erm ined t he present at ion of m ort alit y dat a using t he so called “ m ult iple causes of deat h” . Thus any com plicat ions of t he underlying cause, as well as cont ribut ing condit ions, designat ed as associat ed causes of deat h, were added t o t he underlying cause in m ort alit y st at ist ics as m ult iple causes. 1 1
cont rol and assist ance program s, and t o populat ion growt h.3 However, despit e t he increase regist ered, t he im port ance of TB t o populat ion- wide m ort alit y is not com plet ely reflect ed by st at ist ics present ed according t o underlying cause of deat h. TB occurs as an associat ed cause in a large num ber of deat hs in which dem ise is at t ribut ed t o a different underlying cause. I n t he Unit ed St at es, of all deat h cert ificat es t hat included m ent ions of TB, deat h was at t ribut ed t o anot her underlying cause in 56% in 1980 and 57% in 1990. 1 8 I n t he st at e of São Paulo, sout heast ern Brazil, in 1983, t his proportion was 22% .1 1
Co- infection by TB and AI DS is an im portant factor for prem ature m ortality and hum an suffering. St udies on multiple - cause m ort alit y provide wide- ranging inform ation about the m agnitude of the im pact of AI DS on the TB epidem ic as recom m ended by the World Health Assem bly 1 9. I n 1998, in the st at e of São Paulo, TB was m ent ioned as an associat ed cause of death in 19.6% (907/ 4,619) of AI DS deat hs.1 5
The Fundação Sist em a Est adual de Análise de Dados ( St at e Syst em for Dat a Analysis Foundat ion – SEADE) com piles m ultiple causes in the state of São Paulo since 1983, allowing for investigations of both underlying and associat ed causes of deat h. Wit h t he except ion of t he Unit ed St at es, t he st at e of São Paulo was t he first t o em ploy t he ACME syst em unt il 1995. Along wit h t he im plem ent at ion of t he Tent h Revision of t he I nt ernat ional St at ist ical Classificat ion of Diseases and Related Health Problem s ( I CD- 10) , in 1996, t he Sist em a Declarações de Óbit o de São Paulo ( São Paulo Deat h Cert ificat e Syst em – DOSP/ SCBX) , was also im plem ent ed, and t he first decision t ables for aut om at ic processing of AI DS- relat ed deat hs were const r uct ed.9
The present st udy is aim ed at invest igat ing t uberculosis - relat ed m ort alit y in t he st at e of São Paulo according t o m ult iple causes of deat h, showing it s im port ance as underlying and associat ed cause as well, and in t his lat t er form , t heir int errelat ion wit h ot her underlying causes, especially AI DS.
M ET HOD S
Dat a on deat hs and est im at ed populat ion in t he st at e of São Paulo in 1998 were obt ained from SEADE. Mort alit y files included fields which reproduced variables included in t he deat h cert ificat es used by t he Mort alit y I nform at ion Syst em , coordinat ed by t he Minist ry of Healt h. The underlying cause of deat h was regist ered in a separat e field. Anot her st ring field included all condit ions, inj uries, ext ernal causes, and m edical procedures m ent ioned in the I nt ernat ional Form of Medical Cert ificat e of Cause of Deat h. Addit ional fields were included in t he original deat h dat a file st ruct ure in order t o facilit at e processing. These included a num erical field for m ean age at deat h calculat ion. The proportion of m en and wom en over age 75 years, in five - year groupings, obt ained form t he recount carried out by t he Fundação I nst it ut o Brasileiro de Geografia e Est at íst ica ( Brazilian I nst it ut e for Geography and St at ist ics – I BGE Foundation) , was applied to the corre sponding population estim ated by SEADE in order to discrim inate TB m ortality am ong these age groups and avoid data distortion.
t o t he “ Sequelae of t uberculosis” cat egory, coded as B90.7 The expressions “ dea t hs by” or “ due t o” a given condit ion refer t o t he underlying cause of deat h, and t he expression “ deat hs wit h m ent ion of” refer t o t he presence of a given cause, be it underlying or associat e, in t he deat h cert ificat e.
Dem ographic and m edical variables were processed using dBASE I I I Plus, version 1.1 ( Ashton - Tat e, 1985,1986) , Epi I nfo ,version 6.04b/ c, 1997, and Excel 97 SR- 1. Com puter program “Tabulador de Causas Múlt iplas de Mort e ( Mult iple Cause of Deat h Tabulat or – TCM) ( version 2.0, 25/ 08/ 99) was used for t he st udy of m ent ions, associat ions, and m ean num ber of causes of deat h.1 4 Special t abulat ion list s were prepared in order t o ident ify causes of deat h relat ed t o t he nat ural hist ories of TB and AI DS, and also ot her causes m ost frequent ly m ent ioned. Cause duplicat ions, present ed according t o abbreviat ed t abulat ion list s, were elim inat ed. The num ber of associat ed causes depends on t he am plit ude of t he class in which t hese causes are included. This was done in order t o elim inat e from t he count cause duplicat ion/ m ult iplicat ion. Only one cause was included per class ( cat egory, group, or chapt er of I CD - 10) in case t wo or m ore causes belonging t o t he sam e class were recorded in t he sam e deat h cert ificat e.1 4 Com puter program Separador de Registros de Mortalidade ( Mort alit y record separat or – SRM_DBF) was used for recovering records for which t here was an int erest in a given associat ed cause, in order t o st udy event ual relat ionships wit h TB.
Mort alit y rat es per 100,000 populat ion were calculat ed, according t o sex and age groups, for classical underlying causes, and for t ot al num ber of deat hs wit h m ent ion of TB as bot h an underlying and an associat ed cause of deat h.
Using Epi I nfo soft ware, variance analysis was carried t o t est differences in m ean age at deat h, St udent ’s t t est for differences in m ean num ber of causes inform ed in deat h cert ificat es, and chi-squared t est for differences in proport ions of associat ed causes of deat h when TB and AI DS were present ed as t he underlying cause. Deat hs in which pat ient age was ignored were not included in m ean age at deat h calculat ions. Significance level was est ablished at 1% . Mean age at deat h and num ber of causes are present ed wit h t heir respect ive st andard deviat ions.
RESULT S
Figure 1 - TB mortality rates, underlying causes, and total mentions, by age and sex,
São Paulo State, Brazil, 1998.
Of t he 1,644 deat hs by TB as an underlying cause, 1,326 ( 82.5% ) were due t o TB of t he respirat ory tract, 26 ( 1.6% ) of the nervous system , 23 ( 1.4% ) of other organs, 124 ( 7.5% ) to m iliary TB, and 115 ( 7.0% ) t o TB sequelae.
Associat ed causes in TB deat hs are present ed in Table 1. Most of t hese causes m ay be considered as t erm inal. This cat egory includes respirat ory failure, pneum onias, condit ions included in cat egory R09 – concerned wit h ot her sym pt om s and signs involving the circulatory and respiratory system s in the chapter “ Sym ptom s, signs, and abnorm al clinical and laboratory findings” –, septicem ias, and renal failure. Cachexia and m alnutrition proportions together add up to as m uch as 19.7% of deaths by TB. The great er proport ions of circulat ory syst em diseases am ong wom en, and of m ent al disorders due t o psychoact ive subst ance use am ong m en were significant ( p< 0.001) . Of t he 138 deat hs associat ed t o psychoact ive subst ance use, 130 ( 94.2% ) were alcohol relat ed.
Table 1 – Number and percentage of deaths due to TB*, according to associated causes of death and sex, São Paulo State, Brazil, 1988.
Associated causes of death ** Men
(deaths =1,226)
Women
(deaths =418) Total
(óbitos =1,644)
N % N % N° %
Respiratory failure (J96) 575 46.9 196 46.9 771 46.9
Pneumonias (J12-J18) 186 15.2 86 20.6 272 16.5
Cachexia (R64) 160 13.1 52 12.4 212 12.9 Diseases of the circulatory system (I00-I99)*** 109 8.9 61 14.6 170 10.3 Rem. resp. tract (J00-J11, J20-J39, J60-J68, J70-J95, J98) 115 9.4 44 10.5 159 9.7 Mental disord. Due to psychoactive subst. use (F10-F19)*** 128 10.4 10 2.4 138 8.4
Septicemias (A40-A41) 88 7.2 30 7.2 118 7.2
Chronic lower respiratory diseases (J40-J47) 81 6.6 36 8.6 117 7.1
Malnutrition (E40-E46) 87 7.1 25 6.0 112 6.8
Diseases of liver (K70-K77) 67 5.5 16 3.8 83 5.0
Diabetes mellitus (E10-E14) 52 4.2 20 4.8 72 4.4
Hemorrhage from respiratory passages (R04) 38 3.1 13 3.1 51 3.1
Remaining dis. of the digestive system (K00-K66, K80-K92) 36 2.9 11 2.6 47 2.9
Anemia, unspecified (D649) 23 1.9 14 3.3 37 2.3
Renal failure (N17-N19) 25 2.0 9 2.2 34 2.1
Remaining associated causes of death 182 14.8 75 17.9 257 15.6
Total 2,122 NC 757 NC 2,879 NC
Source: Fundação Sistema Estadual de Análise de Dados database
*Active tuberculosis and its late effects (A15-A19, B90)
**Rubrics and codes of the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision.
***Difference in proportions between men and women (p<0.001)
% Percentages calculated in relation to the number of deaths.
NC = not calculated.
AI DS was the m ain underlying cause in 65.3% ( 907/ 1,388) of deaths in which TB was m entioned as an associat ed cause ( Table 2) and in 85.9% ( 828/ 964) of t hose in t he 20- 49 years age group. As t o t he t ot al num ber of deat hs wit h m ent ion of TB, corresponding percent ages were 29.9% , ( 907/ 3,032) and 47.6% ( 828/ 1,738) . Diseases of t he circulat ory syst em , neoplasm s – including m align neoplasm of the bronchi and lungs –, and diseases of the respiratory system followed AI DS in order of im port ance as underlying causes ( Table 2) .
Table 2 – Number and percentage of deaths in which TB is an associated cause, according to underlying cause of death and sex, São Paulo State, Brazil, 1998.
Underlying causes of death* Men Women Total
N % N % N %
Certain infectious and parasitic diseases (A00-B99) 678 67.3 262 68.9 940 67.7 Human immunodeficiency virus [HIV] disease (B20-B24) 658 65.3 249 65.5 907 65.3
Neoplasms (C00-D48) 76 7.5 28 7.4 104 7.5
Malign neoplasm of bronchus and lung (C34) 21 2.1 9 2.4 30 2.2
Endocrine, nutritional, and metabolic dis. (E00-E90) 18 1.8 7 1.8 25 1.8
Diabetes mellitus (E10-E14) 12 1.2 6 1.6 18 1.3
Mental disorders ... due to use of alcohol (F10) 9 0.9 2 0.5 11 0.8
Diseases of the circulatory system (I00-I99) 88 8.7 36 9.5 124 8.9
Ischemic heart diseases (I20-I25) 19 1.9 12 3.2 31 2.2
Heart failure (I50) 16 1.6 6 1.6 22 1.6
Diseases of the respiratory system (J00-J99) 68 6.7 17 4.5 85 6.1
Chronic lower respiratory diseases (J40-J47) 46 4.6 11 2.9 57 4.1
Diseases of the digestive system (K00-K93) 50 5.0 17 4.5 67 4.8
Alcoholic liver disease (K70) 12 1.2 0 0.0 12 0.9
Other liver disease (K71-K77) 22 2.2 9 2.4 31 2.2
Other causes (**) 19 1.9 10 2.6 29 2.1
Total 1,008 100.0 380 100.0 1,388 100.0
Source: Fundação Sistema Estadual de Análise de Dados database
*Rubrics and codes of the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision.
**Diseases of the blood and blood & blood forming organs and certain immune disorders (D50-D89), Diseases of the nervous system (G00-G99), Diseases of the musculoskeletal system and connective tissue (M00-M99), Diseases of the genitourinary system (N00-N99), Pregnancy, childbirth, and puerperium (O00-O99), and external causes of morbidity and mortality (V00-Y98).
Clinical form s of nervous syst em and m iliary TB were m ore frequent as associat ed causes in deat hs due t o AI DS, whereas TB sequelae were m ost ly associat ed wit h ot her causes of deat h ( p< 0.001) ( Table 3) .
Table 3 – Number and percentage of deaths due to AIDS and other causes, according to clinical forms of TB as associated causes of death, São Paulo State, Brazil, 1998.
Associated causes of death* AIDS
(deaths =907)
Other causes
(deaths =481) p-value
N % N %
Respiratory TB (A15-A16) 708 78,1 363 75,5 0,274 TB of nervous system (A17) 65 7,2 2 0,4 <0,001 TB of other organs (A18) 36 4,0 16 3,3 0,548
Miliary TB (A19) 127 14,0 21 4,4 <0,001
Sequelae of TB (B90) 1 0,1 81 16,8 <0,001
Total 937 NC 483 NC
Source: Fundação Sistema Estadual de Análise de Dados database
*Rubrics and codes of the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision.
% Percentages calculated in relation to the number of deaths.
Considering all deat hs wit h m ent ion of TB, and according t o t he respect ive underlying causes of deat h, t he differences in m ean age of deat h by AI DS ( 36,2± 9.3 years) , TB ( 51.1± 18.0 years) , and ot her causes ( 58.6± 9.3 years) were st at ist ically significant ( p< 0.001) . Figure 2 shows m ort alit y rat es by age group for TB as an underlying cause, and for AI DS and ot her underlying causes wit h TB as an associat ed cause. Mean age at deat h in t he 137 deat hs by TB associat ed wit h psychoact ive subst ance use was approxim at ely 5.6 years lower ( 46.4± 11.7 years) t han t he 1,490 deat hs by TB wit hout t his association ( 52.1± 17.6 years) ( p< 0.001) .
Figure 2 - Mortality rates for TB as an underlying cause and for AIDS and OTHER causes with TB as an associated cause, by age, São Paulo State, Brazil, 1998.
Mean num ber of causes regist ered per deat h cert ificat e was significant ly different bet ween t he 1,644 deat hs due t o TB ( 2.8± 1.1) , t he 907 deat hs due t o AI DS ( 3.7± 1.0) , t he 481 deat hs due t o ot her causes ( 4.0± 1.1) , and t he t ot al 232,806 deat hs regist ered in t he st at e of São Paulo in 1998 ( 2.7± 1.3) (p< 0.001). Approxim ately 59.2% (937/ 1,644) of deaths by TB had three or m ore causes recorded in t he deat h cer t ificat e.
D ISCUSSION
deat hs in which AI DS was present . For t his reason, in t he st at e of São Paulo in 1998, AI DS was identified as the underlying cause in 98.2% ( 4,619/ 4,702) of deaths in it was m entioned.1 5
Associat ed causes for deat hs in which TB was report ed as t he underlying cause included charact erist ics of consum pt ive disease and of t he nat ural hist ory of TB. Respirat ory failure, when t he m ain reason for adm ission to specialized hospitals, is related to greater fatality.5 The nat ural hist ory of TB includes, as consequent ial causes, pneum onias, septicem ia, m alnutrition, anem ia, and, generally speaking, t he condit ions included in t he I CD- 10 chapter on ill- defined causes. Am ong cont ribut ing causes, t he associat ion of TB wit h alcoholism is im port ant in Brazil, especially am ong m en, as a determ inant for non - adherence t o t reat m ent and disciplinarian hospit al discharge.5 These associat ed causes m ust be t aken int o considerat ion so as t o prevent t heir influence on t he det erm inat ion of deat h by TB.
Any event ual differences in associat ed cause dist ribut ion, in deat hs for which TB was ident ified as t he underlying cause ( Table 1) when com pared to the 1983 distribution, m ay be explained, am ong other fact ors, by t he updat e in t he I CD version used at t he t im e. However, t his is not t he case wit h underly ing cause dist ribut ion in deat hs for which TB was an associat ed cause ( Table 2) . I n 1983, only 5.4% of t hese deat hs were at t ribut ed t o infect ious and parasit ic diseases,1 1 whereas t his value rose t o 67.7% in 1998, and 65.3% were related to AI DS ( Table 2) . The relative im portance of the rem aining underlying causes decreased not iceably, as in t he case of diseases of t he circulat ory syst em – from 26.9% to 8.9% – and neoplasm s – from 15.3 to 7.5% –, bet ween 1983 and 1998, respect ively.1 1
The associat ion bet ween AI DS and TB m ay occur t hrough t he react ivat ion of t he lat t er, t hrough rapid progression of a prim ary infect ion, and t hrough reinfect ion ( exogenous) at any st age of HI V infect ion. The risk of prim ary and reinfection am ong HI V- posit ive pat ient s is considera bly high. The progression of HI V disease is aggravat ed by TB. Co- infect ion by HI V and TB duplicat es t he risk of dem ise in relat ion t o infect ion by HI V alone.
The im por t ance of co- infect ion was charact erized by t he clinical form s of TB associat ed wit h AI DS.1 6 Nervous system and m iliary clinical form s of TB are found with greater frequency in HI V- posit ive pat ient s t han in ot hers. Cent ral nervous syst em TB occurs in 5- 10% of HIV- posit ive pat ient s.1 6
I n 1990, of all deat hs in pat ient s infect ed by TB, 4.6% were at t ribut ed t o HI V infect ion. I n t he present st udy, t his value was 29.9% ( 907/ 3032) of all deat hs wit h m ent ion of TB, m ore t han doubling t he 14% est im at ed for t he year 2000.1 0
Graphic represent at ion of TB m ort alit y dist ribut ion by age before t he AI DS epidemic was unim odal in shape, rising along wit h age. The first st udies t o invest igat e t his pat t ern for t he AI DS/ TB co- infect ion describe a bim odal curve, wit h peaks bet ween 24 and 45, and above 60 years.2 , 1 8 These st udies, however, st udied TB and AI DS deat hs joint ly, wit hout a hierarchical dist inct ion bet ween underlying and associat ed causes. This was due t o t heir being conduct ed at a t im e when t he defining diseases of AI DS m ight have been ident ified as t he underlying cause inst ead of AI DS it self. I n t he present st udy, only curves plot t ed according t o t ot al m ent ions had a slight t rend t owards elevat ion bet ween 30 and 44 years am ong m en and 25 and 39 years am ong wom en. Graphic representation of deaths due to underlying causes wit h TB as an associat ed cause shows t hat t he elevat ion det ect ed at t his younger age is due t o AI DS deat hs, a hypot hesis corroborat ed by t he respect ive m ean age at deat h. On t he ot her hand, t he slight rise det ect ed in t he TB- related m ortality curve am ong younger age groups was part ly ascribed t o deat hs due t o alcohol or anot her psychoact ive subst ance abuse. 1 8 The pr esent st udy det ect ed an associat ion bet ween alcohol use and TB m ort alit y at younger- t han- norm al age.
Par t I of any correct ly filled deat h cert ificat e, in which t he pat hological sequence leading t o deat h includes underlying, int ervening, and t erm inal causes. When necessary, one or m ore cont ribut ing causes m ay be added in Part I I of t he deat h cert ificat e. The great er t he num ber of causes regist ered, t he great er our knowledge about t he nat ural hist ory of t he disease, and consequent ly t he great er t he possibilit ies of prevent ing sim ilar deat hs. Mean num ber of causes per cert ificat e for deat hs by TB was great er t han t he correspondent num ber for t he t ot al deat hs in t he st at e. The 2.8 causes per cert ificat e obt ained w as adequat e for t he study of m ultiple - cause m ort alit y. However, it was lower t han t he m ean num ber of causes regist ered for deat hs for which AI DS was select ed as t he underlying cause, which m ay reflect the greater im portance given to this syndrom e by physicians when filling deat h cert ificat es. The 907 deat hs due t o AI DS wit h m ent ion of TB had a great er m ean num ber of causes ( 3.7± 1.0) t han t he 3,712 deat hs wit hout ( 3.1± 1.1) in t he St at e of São Paulo, in 1998.1 3
The present st udy present s t he usual lim it at ions of m ort alit y st ud ies based on secondary dat a from deat h cert ificat es forwarded t o t he Civil Regist er.1 1 Although quantitative lim itations im posed by under- report ing of deat hs in t he st at e are negligible, qualit at ive lim it at ions should not be neglect ed, especially t hose relat ed t o t he inadequat e filling of deat h cert ificat e it em s. This sit uat ion has been im proving in t he last few years, however, due t o great er knowledge on t he part of physicians concerning t he st at ist ical and legal im port ance of t his docum ent for public healt h and for t he deceased’s fam ily. The 6.6% share of deat hs at t ribut ed t o ill- defined causes in São Paulo St at e in 1998 was one of the lowest in Brazil. Such im provem ent is partly due to the progressive increase in t he m ean num ber of causes report ed per deat h cert ificat e, which is at t ribut ed t o t he addit ion of a fourt h line t o Part I of t he I nt ernat ional Form of Medical Cert ificat e of Cause of Deat h.8 On t he ot her hand, t he advant ages of t his t ype of st udy are great er t han t he above- m ent ioned lim it at ions. TB deat h coverage is great er t han t hat of epidem iological surveillance case not ificat ion. A com parison bet ween m ort alit y dat a and com pulsory case not ificat ion showed t hat in São Paulo St at e in 1980, and in São Paulo Cit y, from 1986 t o 1995, respect ively 67.5% and 64.2% of TB cases were not report ed.4,6 I n addition, m ultiple- cause m ort alit y dat a unit e, in a single regist er, deat hs relat ed t o TB, AI DS, and ot her causes, allowing for t he evaluat ion of t he im pact of any given cause and of it s let hal associat ions.
The st udy of m ort alit y adds a valuable cont ribut ion t o TB epidem iology. The result s of t he present st udy show t hat m ort alit y rat es alm ost doubled when deat hs wit h m ent ion of TB as an associat ed cause – of which, in roughly t wo- t hirds of cases, AI DS was report ed as t he underlying cause – were added. They also show t hat inform at ion including all regist ered causes of deat h allow for a part ial reconst ruct ion of t he nat ural hist ory of TB, and for t he recom m endat ion of adequat e prevent ive and t herapeut ic m easures.
REFEREN CES
1. Antunes JLF, Waldm an EA. Tuberculosis in the twentieth century: tim e - series m ortality in Sao Paulo, Brazil, 1900- 97. Cad Saúde Pública 1999; 15: 463 - 76.
3. Dye C, Scheele S, Dolin P, Pathania V, Raviglione MC. Global burden of tuberculosis. estim ated incidence, prevalence, and m ort alit y by count ry. JAMA 1999; 282: 677 - 86.
4. Galesi VMN. Mortalidade por tuberculose no m unicípio de São Paulo: análise de um a década, 1986 a 1995 [ Dissert ação de Mest rado] . São Paulo: Faculdade de Saúde Pública da Universidade de São Paulo; 1999.
5. Nogueira PA. Mot ivos e t em po de int ernação e o t ipo de saída em hospit ais de t uberculose no Estado de São Paulo, Brasil – 1981 a 1995. J Pneum ol 2001; 27: 123- 9.
6. Nogueira PA. Tuber culose com o causa de óbit o em adult os r esident es no m unicípio de São Paulo em 1980 [ Tese de Dout orado] . São Paulo: Faculdade de Saúde Pública da Universidade de São Paulo; 1984.
7. Organização Mundial da Saúde. Classificação est at íst ica int er nacional de doenças e pr oblem as r elacionados à saúde: 10a revisão. São Paulo: Cent ro Colaborador da OMS para a Classificação de Doenças em Port uguês. EDUSP; 1993. v. 1.
8. Organização Mundial da Saúde. Classificação est at íst ica int er nacional de doenças e pr oblem as r elacionados à saúde: 10a revisão. São Paulo: Cent ro Colaborador da OMS para a Classificação de Doenças em Português. EDUSP; 1994. v. 2. ( Manual de inst rução) .
9. Pinheiro CE, Sant o AH. Processam ent o de causas de m ort e em lot e pelo sist em a de seleção de causa básica. Rev Saúde Pública 1998; 32: 72 - 3.
10. Raviglione MC, Snider DE, Kochi A. Global epidem iology of tuberculosis. Morbidity and m ortality of a worldwide epidem ic. JAMA 1995; 273: 220- 6.
11. Sant o, AH. Causas m últ iplas de m or t e: for m as de apr esent ação e m ét odos de análise [ Tese de Dout orado] . São Paulo: Faculdade de Saúde Pública da Universidade de São Paulo; 1988.
12. Sant o AH. Equivalência ent re revisões da classificação int ernacional de doenças: causas de m ort e. Rev Saúde Pública 2000; 34: 21 - 8.
13. Santo AH, Pinheiro CE. Opport unist ic infect ions associat ed t o deat hs from AI DS wit h and wit hout tuberculosis. I n: 32nd World Confere nce on Lung Health of the I nternational Union Against Tuberculosis and Lung Diseases ( I UATLD) , 2001. Abst ract book. I nt J Tuberc Lung Dis 2001; 5(11 Suppl 1): S173.
14. Santo AH, Pinheiro CE. Tabulador de causas m últiplas de m orte. Rev Bras Epidem iol 1999; 2: 90- 7.
15. Sant o AH, Pinheiro CE, Jordani MS. Causas básicas e associadas de m ort e por AI DS, Est ado de São Paulo, Brasil, 1998. Rev Saúde Pública 2000; 34: 581- 8.
16. Shafer RW, Edlin BR. Tuberculosis in patients infected with hum an im m unodeficiency virus: perspect ive on t he past decade. Clin I nfect Dis 1996; 22: 683- 704.
17. Tuberculosis deat hs on t he increase. World Healt h Forum 1996; 17: 422.
19. World Healt h Assem bly. St op t uberculosis init iat ive. Disponível em < URL: ht t p./ / www.st opt b.org/ WHA53.1.ht m l>
Address to correspondence
August o Hasiak Sant o Av. Dr. Arnaldo, 715
01246 - 904 São Paulo, SP, Brasil E- mail: auhsant o@usp.br
Prelim inary version present ed at t he I For um and I I Confer ence of Hor izont al Technical Cooper at ion on HI V/ Aids and STD in Lat in Am erica and t he Caribbean, Rio de Janeiro, Brazil, 2000.
Received on 25/ 9/ 2002. Reviewed on 12/ 5/ 2003. Approved on 12/ 6/ 2003.
© 2 0 0 3 Fa cu lda de de Sa ú de Pú blica da U n iv e r sida de de Sã o Pa u lo
Ave nida D r . Ar na ldo, 7 1 5 0 1 2 4 6 - 9 0 4 Sã o Pa ulo SP Br a zil
Te l./ Fa x : + 5 5 1 1 3 0 6 8 - 0 5 3 9
