www. helance .com/in ec ion Vol 18 Decembe 2018
1329
A icles
Global, egional, and na ional bu den o ube culosis,
1990–2016: esul s om he Global Bu den o Diseases,
Inju ies, and Risk Fac o s 2016 S udy
GBD Tube culosis Collabo a o s*
Summa y
Backg ound Al hough a p e en able and ea able disease, ube culosis causes mo e han a million dea hs each yea .
As coun ies wo k owa ds achie ing he Sus ainable De elopmen Goal (SDG) a ge o end he ube culosis epidemic
by 2030, obus assessmen s o he le els and ends o he bu den o ube culosis a e c ucial o in o m policy and
p og amme decision making. We assessed he le els and ends in he a al and non- a al bu den o ube culosis by
d ug esis ance and HIV s a us o 195 coun ies and e i o ies om 1990 o 2016.
Me hods We analysed 15 943 si e-yea s o i al egis a ion da a, 1710 si e-yea s o e bal au opsy da a, 764 si e-yea s o
sample-based i al egis a ion da a, and 361 si e-yea s o mo ali y su eillance da a o es ima e mo ali y due o
ube culosis using he Cause o Dea h Ensemble model. We analysed all a ailable da a sou ces, including annual case
no i ica ions, p e alence su eys, popula ion-based ube culin su eys, and es ima ed ube culosis cause-speci ic
mo ali y o gene a e in e nally consis en es ima es o incidence, p e alence, and mo ali y using DisMod-MR 2.1, a
Bayesian me a- eg ession ool. We assessed how he bu den o ube culosis di e ed om he bu den p edic ed by he
Socio-demog aphic Index (SDI), a composi e indica o o income pe capi a, a e age yea s o schooling, and o al
e ili y a e.
Findings Globally in 2016, among HIV-nega i e indi iduals, he numbe o inciden cases o ube culosis was
9·02 million (95% unce ain y in e al [UI] 8·05–10·16) and he numbe o ube culosis dea hs was 1·21 million
(1·16–1·27). Among HIV-posi i e indi iduals, he numbe o inciden cases was 1·40 million (1·01–1·89) and he
numbe o ube culosis dea hs was 0·24 million (0·16–0·31). Globally, among HIV-nega i e indi iduals he age-
s anda dised incidence o ube culosis dec eased annually a a slowe a e (–1·3% [–1·5 o –1·2]) han mo ali y did
(–4·5% [–5·0 o –4·1]) om 2006 o 2016. Among HIV-posi i e indi iduals du ing he same pe iod, he a e o change
in annualised age-s anda dised incidence was –4·0% (–4·5 o –3·7) and mo ali y was –8·9% (–9·5 o –8·4). Se e al
egions had highe a es o age-s anda dised incidence and mo ali y han expec ed on he basis o hei SDI le els in
2016. Fo d ug-suscep ible ube culosis, he highes obse ed- o-expec ed a ios we e in sou he n sub-Saha an A ica
(13·7 o incidence and 14·9 o mo ali y), and he lowes a ios we e in high-income No h Ame ica (0·4 o
incidence) and Oceania (0·3 o mo ali y). Fo mul id ug- esis an ube culosis, eas e n Eu ope had he highes
obse ed- o-expec ed a ios (67·3 o incidence and 73·0 o mo ali y), and high-income No h Ame ica had he
lowes a ios (0·4 o incidence and 0·5 o mo ali y).
In e p e a ion I cu en ends in ube culosis incidence con inue, ew coun ies a e likely o mee he SDG a ge o
end he ube culosis epidemic by 2030. P og ess needs o be accele a ed by imp o ing he quali y o and access o
ube culosis diagnosis and ca e, by de eloping new ools, scaling up in e en ions o p e en isk ac o s o
ube culosis, and in eg a ing con ol p og ammes o ube culosis and HIV.
Funding Bill & Melinda Ga es Founda ion.
Copy igh 2018 © The Au ho (s). Published by Else ie L d. This is an Open Access a icle unde he CC BY 4.0 license.
In oduc ion
Al hough ube culosis is a p e en able and ea able
disease, i is he cause o mo e han a million dea hs each
yea .1,2 Tube culosis was he leading cause o dea h om
a single in ec ious pa hogen in 2016.1 The ambi ious
Sus ainable De elopmen Goal (SDG) a ge 3 aims o
end he ube culosis epidemic by 2030, and nume i-
cal miles ones (eg, annual educ ion in global ube -
culosis incidence o 10% by 2025) ha e been se o
achie e his a ge .3 Robus assessmen , moni o ing, and
e alua ion o p og ess owa ds his SDG a ge a e
he e o e c ucial o in o m policy and p og amme
decision making.
Accu a ely assessing he ube culosis bu den o e
ime is di icul because o he pauci y o high-quali y
da a om many low-income and middle-income coun-
ies.2 The comple eness o i al egis a ion da a is
g adually imp o ing, bu many coun ies s ill do no
ha e good-quali y i al egis a ion sys ems.1 No i ica ion
da a can be o use as a p oxy o ube culosis incidence
Lance In ec Dis 2018;
18: 1329–49
See Commen page 1291
*Collabo a o s lis ed a he end
o he A icle
Co espondence o:
P o Ch is ophe J L Mu ay,
Ins i u e o Heal h Me ics
and E alua ion, Sea le,
WA 98121, USA
[email protected]
A icles
1330
www. helance .com/in ec ion Vol 18 Decembe 2018
in coun ies wi h high-quali y heal h and su eillance
sys ems whe e unde - epo ing is minimal;4 howe e , in
mos low-income and middle-income coun ies, hese
da a a e p one o unde - epo ing and canno be
in e p e ed wi hou addi ional in o ma ion on case
de ec ion a e.4,5 To deal wi h he lack o high-quali y da a
in hese coun ies, a ious me hods ha e been used o
es ima e ube culosis incidence (eg, adjus ing o unde -
epo ing in no i ica ion da a by use o expe opinion
case de ec ion a es,4 back-calcula ing incidence om
p e alence su ey da a by use o di e en assump ions
o he a e age du a ion o disease,4 o using a s a is ical
iangula ion app oach2,6). Fo he Global Bu den o
Diseases, Inju y, and Risk Fac o s S udy (GBD) 2015, we
used a s a is ical iangula ion app oach ha modelled
ube culosis incidence, p e alence, and mo ali y simul-
aneously o gene a e consis en es ima es o hese
pa ame e s.2
The bu den o ube culosis a ies by se e al ac o s
including age, sex, loca ion, HIV s a us, and d ug- esis ance
s a us. The e o e, hese ac o s should be aken in o
accoun when in es iga ing ube culosis ends. Addi-
ionally, he bu den o disease in many coun ies has
shi ed om communicable o non-communicable dis-
eases in line wi h sociodemog aphic de elopmen ( he
epidemiological ansi ion).7–9 As such, compa ing he
obse ed ube culosis bu den o ha expec ed on he basis
o a coun y’s socio-demog aphic le el could be use ul o
guiding in es men in esea ch and in e en ions.2 Fo
example, coun ies wi h a lowe ube culosis bu den han
expec ed ela i e o hei socio-demog aphic de elopmen
could p o ide insigh in o success ul p og amma ic
s a egies, and coun ies wi h a highe bu den han
expec ed migh need o in es iga e he easons why.
GBD 20152 examined he di e ence be ween he obse ed
and expec ed bu den o ube culosis bu did no p o ide a
de ailed assessmen by d ug- esis ance ype and HIV
s a us. Fo GBD 2016, we assessed he le els and ends in
he a al and non- a al bu den o ube culosis by d ug-
esis ance ype and HIV s a us om 1990 o 2016, o
195 coun ies and e i o ies. We also aimed o analyse he
associa ion be ween hese bu dens and he coun y o
e i o y’s Socio-demog aphic Index (SDI),1,10–12 which is a
composi e indica o o income, educa ion, and e ili y a e.
Me hods
O e iew
GBD is a sys ema ic, scien i ic e o o quan i y he
compa a i e magni ude o heal h loss due o diseases,
inju ies, and isk ac o s by age, sex, and loca ion o e
ime. The concep ual and analy ical amewo k o GBD
and de ailed me hods ha e been published elsewhe e.1,11,13
We desc ibe he e he me hods we used o he analysis o
he bu den o ube culosis o GBD 2016.
Selec ion o inpu da a
The inpu da a we used o model mo ali y due o
ube culosis among HIV-nega i e indi iduals included
Resea ch in con ex
E idence be o e his s udy
Tube culosis causes mo e han a million dea hs each yea and
was he leading cause o dea h om a single in ec ious
pa hogen in 2016. The global bu den o ube culosis has
been es ima ed by se e al g oups, including he WHO Global
Tube culosis P og amme and he Global Bu den o Diseases,
Inju ies, and Risk Fac o s S udy (GBD) 2015. Ne e heless,
ends in he bu den o d ug- esis an ube culosis by HIV
s a us and how he obse ed bu dens di e om he le els
expec ed on he basis o sociodemog aphic de elopmen ha e
no been comp ehensi ely assessed. We sea ched PubMed wi h
he sea ch e ms (“ ube culosis”[MeSH] AND “d ug-sensi i e”
OR “d ug-suscep ible”) OR “ ube culosis,
mul id ug- esis an ”[MeSH] AND (“bu den” OR “es ima es”)
AND “ end”, wi h no language es ic ions, o publica ions up
o June 7, 2018. We iden i ied en s udies ha p o ided
popula ion-based ime ends o he bu den o
mul id ug- esis an ube culosis (incidence, p e alence, o
dea hs). O hese s udies, he mos ecen pe iod assessed was
1999–2013 in Lebanon. None o hese s udies assessed he
ends in he bu den o d ug-suscep ible o mul id ug- esis an
ube culosis by HIV s a us and compa ed hese bu dens wi h
hose expec ed on he basis o a coun y’s socio-demog aphic
posi ion.
Added alue o his s udy
We ound ha , al hough HIV in ec ion and d ug- esis an
ube culosis ha e become he main challenges o ube culosis
con ol e o s, mo e han h ee-qua e s o inciden cases o
ube culosis and dea hs due o ube culosis in 2016 we e
es ima ed o occu in HIV-nega i e indi iduals who we e
suscep ible o i s -line ube culosis d ugs. Du ing he pas
decade, he global a e o decline o incidence o bo h
d ug-suscep ible and mul id ug- esis an ube culosis was
slowe han he co esponding a e o decline o mo ali y, o
HIV-posi i e and HIV-nega i e indi iduals alike. Many coun ies
had highe bu dens o d ug-suscep ible o mul id ug- esis an
ube culosis han expec ed on he basis o hei le el o
socio-demog aphic de elopmen .
Implica ions o all he a ailable e idence
I cu en ends in ube culosis incidence con inue, ew
coun ies will mee he Sus ainable De elopmen Goal a ge o
end he ube culosis epidemic by 2030. The pace o p og ess
needs o be inc eased h ough in e en ions including
imp o ing he quali y o and access o ube culosis diagnosis
and ca e, and in eg a ing con ol p og ammes o ube culosis
and HIV.
A icles
www. helance .com/in ec ion Vol 18 Decembe 2018
1331
15 943 si e-yea s o i al egis a ion da a, 1710 si e-yea s
o e bal au opsy da a, 764 si e-yea s o sample-based
i al egis a ion da a, and 361 si e-yea s o mo ali y
su eillance da a. We assessed and imp o ed he quali y
and compa abili y o da a on cause o dea h h ough
mul iple s eps,12 including edis ibu ion o ga bage
codes o unde lying causes o dea h using GBD
algo i hms and adjus ing o misclassi ied HIV dea hs
(ie, dea hs caused by HIV being assigned o o he
unde lying causes o dea h, such as ube culosis, because
o s igma o misdiagnosis). GBD 20161 also assessed he
o e all quali y o da a o each coun y (on he basis o
comple eness, ga bage coding, de ail o cause lis , and
ime pe iods co e ed), and assigned a quali y sco e om
ze o s a s (lowes ) o i e s a s (highes ); a sco e o ou o
i e is conside ed high quali y (quali y sco es by coun y
a e in he appendix p 19). We emo ed e bal au opsy
da a o coun ies wi h a high p e alence o HIV (using
an a bi a y cu o alue o 5% age-s anda dised p e-
alence o HIV), because e bal au opsy s udies ha e a
poo abili y o dis inguish dea hs due o HIV om dea hs
due o ube culosis among people who a e HIV posi i e
(HIV- ube culosis dea hs).2
Ou inpu da a o he es ima ion o mo ali y due o
HIV- ube culosis included 382 si e-yea s o high-quali y
i al- egis a ion da a om coun ies whe e da a on cause
o dea h di ec ly coded o HIV- ube culosis and ube -
culosis we e a ailable, and he numbe o ube culosis
cases (new and e- ea men ) eco ded as HIV-posi i e,
and he numbe o ube culosis cases (new and e-
ea men ) wi h an HIV es esul eco ded in he WHO
ube culosis egis e .
In GBD 2016, we included mul id ug- esis an ube -
culosis (wi hou ex ensi e d ug esis ance) and ex-
ensi ely d ug- esis an ube culosis by HIV s a us as
new ou comes (case de ini ions a e in he appendix p 3).
Inpu da a included he numbe o cases o ube culosis
ha we e mul id ug esis an , ex ensi ely d ug esis an ,
had a d ug-sensi i i y es ing esul o isoniazid and
i ampicin, and ha we e mul id ug esis an wi h a
d ug-sensi i i y esul o second-line d ugs om ou ine
su eillance and su eys epo ed o WHO ( o da a
a ailabili y by coun y see appendix p 16); ela i e isks o
mo ali y o cases o ube culosis ha we e mul id ug
esis an compa ed wi h cases ha we e d ug suscep ible,
and ela i e isks o cases ha we e ex ensi ely d ug
esis an compa ed wi h mul id ug esis an we e
ex ac ed om s udies iden i ied ia a sys ema ic e iew
( o de ails o sys ema ic e iew see appendix p 37); and
he isk o mul id ug- esis an ube culosis associa ed
wi h HIV in ec ion ex ac ed om a me a-analysis.14
Ou inpu da a o modelling non- a al ube culosis
included annual case no i ica ion da a, da a om p e-
alence su eys o ube culosis, da a om popula ion-
based ube culin su eys, and es ima ed cause-speci ic
mo ali y a es o ube culosis among indi iduals who
we e HIV posi i e and HIV nega i e. Links o da a
sou ces and code we used in analyses a e in he appendix
(pp 40–41).
Fa al ube culosis
We modelled ube culosis mo ali y among people who
a e HIV nega i e using he Cause o Dea h Ensemble
modelling (CODEm) s a egy,15–18 which e alua es a la ge
numbe o po en ial models ha apply di e en unc ional
o ms (mixed-e ec s models and spa io empo al Gaussian
p ocess eg ession models) o mo ali y o cause ac ions
wi h a ying combina ions o p edic i e co a ia es. These
co a ia es included alcohol (L pe capi a), diabe es ( as ing
plasma glucose in mmol/L), educa ion (yea s pe capi a),
lag-dis ibu ed income (LDI), indoo ai pollu ion, ou doo
ai pollu ion, popula ion densi y (people pe km²), smoking
p e alence, SDI, he summa y exposu e a iable scala
(which indica es exposu e o isk ac o s associa ed wi h
ube culosis; appendix p 9), and ou new co a ia es added
o GBD 2016 (ie, p e alence o ube culosis, p e alence o
la en ube culosis in ec ion, p opo ion o adul s who a e
unde weigh , and he Heal hca e Access and Quali y
[HAQ] Index19). We hen selec ed he ensemble o CODEm
models ha pe o med bes on ou -o -sample p edic -
i e alidi y es s (appendix pp 20–23). We es ima ed
HIV- ube culosis mo ali y using a popula ion-a ibu able
ac ion app oach, like in GBD 20152 (de ailed me hods and
equa ions a e in he appendix pp 34–36).
To spli ube culosis dea hs and HIV- ube culosis dea hs
by d ug- esis ance ype, we i s es ima ed he p opo ions
o ube culosis cases ha we e mul id ug esis an o all
loca ions and yea s using a spa io empo al Gaussian
p ocess eg ession. Second, we es ima ed he p opo ions o
ube culosis cases ha we e mul id ug esis an by HIV
s a us on he basis o he isk o mul id ug- esis an
ube culosis associa ed wi h HIV om a me a-analysis by
Mes in and colleagues.14 Thi d, we used he es ima ed
p opo ions o cases o ube culosis ha a e mul id ug
esis an by HIV s a us and he ela i e isk o dea h in
mul id ug- esis an cases compa ed wi h d ug-suscep ible
cases o calcula e he ac ion o ube culosis dea hs among
HIV-nega i e indi iduals a ibu able o mul id ug- esis an
ube culosis, and he ac ion o HIV- ube culosis dea hs
a ibu able o mul id ug- esis an ube culosis (de ailed
me hods and equa ions a e in he appendix pp 23–24,
35–36). Finally, we applied he ac ion o ube culosis
dea hs a ibu able o mul id ug- esis an ube culosis o
he numbe o ube culosis dea hs we es ima ed using
CODEm, and he ac ion o HIV- ube culosis dea hs
a ibu able o mul id ug- esis an ube culosis o ou
es ima ed numbe o HIV- ube culosis dea hs, o gene a e
he numbe o mul id ug- esis an ube culosis dea hs by
HIV s a us by loca ion, yea , age, and sex.
To dis inguish ex ensi ely d ug- esis an ube culosis
om mul id ug- esis an ube culosis, we agg ega ed
he cases o ex ensi ely d ug- esis an ube culosis and
mul id ug- esis an ube culosis (wi h d ug-sensi i i y
es ing o second-line d ugs) up o he GBD supe - egion
See Online o appendix
A icles
1332
www. helance .com/in ec ion Vol 18 Decembe 2018
le el ( o analy ical pu poses we g ouped 21 GBD egions
in o se en supe - egions:13 cen al Eu ope, eas e n Eu ope
and cen al Asia; high-income; La in Ame ica and
Ca ibbean; no h A ica and Middle Eas ; sou h Asia;
sou heas Asia, eas Asia, and Oceania; and sub-Saha an
A ica) and calcula ed he p opo ion o cases o ex ensi ely
d ug- esis an ube culosis among he cases o mul id ug-
esis an ube culosis a he supe - egion le el. We hen
used hese p opo ions and he ela i e isk o mo ali y
among people wi h ex ensi ely d ug- esis an ube culosis
compa ed wi h hose wi h mul id ug- esis an ube culosis
o calcula e he ac ion o ex ensi ely d ug- esis an
ube culosis dea hs among all mul id ug- esis an ube -
culosis dea hs a he supe - egion le el (de ailed me hods
and equa ions a e in he appendix p 24). These ac ions
we e hen applied o he es ima ed numbe o mul id ug-
esis an ube culosis dea hs and mul id ug- esis an
HIV- ube culosis dea hs in coun ies wi hin he supe -
egions o calcula e he numbe o dea hs due o
ex ensi ely d ug- esis an ube culosis by HIV s a us by
loca ion, yea , age, and sex.
We linea ly ex apola ed mo ali y o ex ensi ely d ug-
esis an ube culosis back om 2016 assuming mo ali y
was ze o in 1992, 1 yea be o e ex ensi ely d ug- esis an
ube culosis was i s eco ded in USA su eillance da a
in 1993.20 Nex , we sub ac ed he numbe o dea hs due o
ex ensi ely d ug- esis an ube culosis om he numbe
o dea hs due o mul id ug- esis an ube culosis o
gene a e he numbe o dea hs due o mul id ug- esis an
ube culosis (wi hou ex ensi e d ug esis ance) by loca-
ion, yea , age, and sex.
Non- a al ube culosis
We made se e al imp o emen s o he s a is ical ian-
gula ion app oach we used in GBD 20152 o model non-
a al ube culosis. Fi s , we es ima ed he p e alence o
la en ube culosis in ec ion by loca ion, yea , age, and sex
using da a om popula ion-based ube culin su eys and
coho s udies ha epo ed he isk o de eloping ac i e
ube culosis disease as a unc ion o indu a ion size.11 Nex ,
we di ided he inpu s on p e alence ( om ube culosis
p e alence su eys in low-income and middle-income
coun ies), incidence (no i ica ion da a om coun ies
wi h a a ing o ou o i e s a s and es ima ed incidence
om coun ies wi h a ings o ze o o h ee s a s), and
cause-speci ic mo ali y a e by he isk-weigh ed
p e alence o la en ube culosis in ec ion o model
ube culosis among indi iduals a isk in each coun y. A
de ailed explana ion o how we p epa ed each o hese
da a sou ces is in he appendix (pp 6–10).
To gene a e ini ial es ima es o incidence o coun ies
wi h a a ing o ze o o h ee s a s, we did a eg es-
sion analysis using mo ali y- o-incidence a ios (logi
ans o med) om loca ions wi h a a ing o ou o
i e s a s as inpu da a, wi h SDI as a co a ia e. We
calib a ed he lowes end o he SDI scale wi h a da apoin
om a communi y-based coho s udy,21 which epo ed
ha 49·2% o people wi h un ea ed pulmona y
ube culosis had died a he end o a 5 yea ollow-up
pe iod, o p edic mo ali y- o-incidence a ios as a unc ion
o SDI o all loca ions and yea s. We hen used he
p edic ed mo ali y- o-incidence a ios and es ima es o
cause-speci ic mo ali y o calcula e he age-sex speci ic
incidence inpu o modelling in DisMod-MR 2.1,22 he
GBD Bayesian me a- eg ession ool. In loca ions whe e
ou es ima ed mo ali y- o-incidence a ios we e g ea e
han no i ica ion-based mo ali y- o-incidence a ios, we
used he no i ica ion-based a ios o calcula e he incidence
inpu . We hen gene a ed a inal incidence es ima e ha
is consis en wi h p e alence da a and cause-speci ic
mo ali y es ima es using a Bayesian me a- eg ession.
We used DisMod-MR 2.1 o simul aneously model age-
sex speci ic ube culosis incidence, p e alence, and
mo ali y among he popula ion who a e la en ly in ec ed
and gene a e consis en ends in all pa ame e s. We hen
mul iplied he DisMod-MR 2.1 ou pu s by he p e alence
o la en ube culosis in ec ion o ge popula ion-le el
es ima es o incidence and p e alence. To dis inguish HIV-
ube culosis om all o ms o ube culosis, we applied he
p opo ion o cases o HIV- ube culosis among all cases o
ube culosis (es ima ed om a mixed-e ec s eg ession
using he adul HIV mo ali y a e co a ia e as in
GBD 20152) o he numbe o inciden and p e alen cases
o ube culosis. We hen applied he es ima ed p opo ion
o cases o ube culosis ha a e mul id ug esis an o ou
p edic ed numbe o cases o ube culosis, and he
es ima ed p opo ion o cases o HIV- ube culosis wi h
mul id ug- esis an ube culosis (as desc ibed ea lie o
a al ube culosis) o ou p edic ed numbe o HIV-
ube culosis cases, o gene a e he numbe o cases
o mul id ug- esis an ube culosis by HIV s a us. To
dis inguish ex ensi ely d ug- esis an ube culosis om
mul id ug- esis an ube culosis, we calcula ed he p o-
po ions o cases o ex ensi ely d ug- esis an ube culosis
among he cases o mul id ug- esis an ube culosis a
he supe - egion le el and applied hese p opo ions o
mul id ug- esis an ube culosis cases.
Simila o ou es ima ion o a al ube culosis wi h
ex ensi e d ug esis ance, we linea ly ex apola ed he
p e alence and incidence o ex ensi ely d ug- esis an
ube culosis back om 2016, assuming incidence and
p e alence we e ze o in 1992 and in ea lie yea s. Finally,
we sub ac ed he numbe o cases o ex ensi ely d ug-
esis an ube culosis om he numbe o cases o
mul id ug- esis an ube culosis o gene a e he numbe
o cases o mul id ug- esis an ube culosis (wi hou
ex ensi e d ug esis ance) by loca ion, yea , age, and sex.
We used he GBD wo ld popula ion age s anda d o
calcula e age-s anda dised a es.
SDI
SDI, ini ially de eloped o GBD 20159 and upda ed o
GBD 2016,1,10,11 was calcula ed on he basis o he
geome ic mean o h ee indica o s: income pe capi a,
A icles
www. helance .com/in ec ion Vol 18 Decembe 2018
1333
a e age yea s o schooling, and o al e ili y a es. SDI
sco es we e scaled om 0 (lowes income, lowes
a e age yea s o schooling, highes e ili y) o 1 (highes
income, highes a e age yea s o schooling, lowes
e ili y), and each loca ion was assigned an SDI sco e
o each yea . We es ima ed he a e age associa ion
be ween SDI and ube culosis incidence and mo ali y
using a Gaussian p ocess eg ession, and we hen used
hese associa ions o es ima e expec ed alues a each
SDI le el.
Role o he unding sou ce
The sponso o he s udy had no ole in s udy design,
da a collec ion, da a analysis, da a in e p e a ion, o
w i ing o he epo . The co esponding au ho
had ull access o all he da a in he s udy and had
inal esponsibili y o he decision o submi o
publica ion.
Resul s
Global bu den o ube culosis in 2016
Globally in 2016, among HIV-nega i e indi iduals, we
es ima ed 9·02 million (95% unce ain y in e al [UI]
8·05–10·16; igu e 1, able 1) inciden cases o ube culosis
and 1·21 million (1·16–1·27) dea hs due o ube culosis
( igu e 1, able 2). Among HIV-posi i e indi iduals, we
es ima ed 1·40 million (1·01–1·89; igu e 1, appendix
pp 75–87) inciden cases o ube culosis and 0·24 million
(0·16–0·31) dea hs due o ube culosis ( igu e 1, appendix
pp 88–100). Almos wo- hi ds o HIV-nega i e ube culosis
inciden cases (59·6% [59·2–59·7]) and dea hs (63·1%
[62·4–63·6]) we e in males ( igu e 2). Mos inciden cases
(89·5% [87·9–91·0]) and dea hs (64·3% [63·6–65·0]) we e
in people younge han 65 yea s o bo h sexes. Among
HIV-posi i e indi iduals, 53·8% (52·4–54·9) o inciden
cases o ube culosis and 56·9% (56·2–57·6) o dea hs due
o ube culosis we e in males (appendix p 46). Mos
Figu e 1: Global ube culosis incidence (A) and mo ali y (B) by d ug- esis ance ype and HIV s a us, 1990–2016
Da k lines a e es ima es and shaded a eas a e 95% unce ain y in e als. HIV- ube culosis= ube culosis in indi iduals wi h HIV/AIDS. Mul id ug- esis an
ube culosis=mul id ug- esis an ube culosis wi hou ex ensi e d ug esis ance.
1990
1995
2000
2005
2010
2015
2016
0
50
100
150
200
250
Incidence ( housands)
Yea
0
100
200
300
400
500
Incidence ( housands)
Ex ensi ely d ug- esis an ube culosis
Mul id ug- esis an ube culosis
0
2·5
5·0
7·5
10·0
Incidence (millions)
A
D ug-suscep ible ube culosis
D ug-suscep ible ube culosis
D ug-suscep ible HIV- ube culosis
Mul id ug- esis an ube culosis
Mul id ug- esis an HIV- ube culosis
Ex ensi ely d ug- esis an ube culosis
Ex ensi ely d ug- esis an HIV- ube culosis
1990
1995
2000
2005
2010
2015
2016
0
5
10
15
Dea hs ( housands)
Yea
0
50
100
150
200
250
Dea hs ( housands)
0
0·5
1·0
1·5
2·0
Dea hs (millions)
B

A icles
1334
www. helance .com/in ec ion Vol 18 Decembe 2018
inciden cases o HIV- ube culosis (82·4% [82·2–82·5])
and dea hs due o HIV- ube culosis (73·7% [72·7–74·8])
we e among people aged 20–54 yea s o bo h sexes
(appendix p 46).
Globally in 2016, among HIV-nega i e indi iduals, we
es ima ed ha 0·30 million (95% UI 0·26–0·34) inciden
cases o ube culosis we e mul id ug esis an ( able 1) and
0·10 million (0·08–0·11) dea hs we e due o mul id ug-
esis an ube culosis ( able 2). In indi id uals who a e HIV
posi i e, we es ima ed ha 35 815 (23 524–51 741) inciden
cases o ube culosis we e mul id ug- esis an (appendix
pp 75–87) and 18 375 (11 208–27 747) dea hs we e due
D ug-suscep ible ube culosis Mul id ug- esis an ube culosis Ex ensi ely d ug- esis an
ube culosis
All HIV-nega i e ube culosis
Numbe o
inciden
cases, 2016
Annualised a e o
change o
age-s anda dised
incidence
Numbe o
inciden
cases, 2016
Annualised a e o
change o
age-s anda dised
incidence
Numbe o
inciden
cases, 2016
Annualised a e o
change o
age-s anda dised
incidence
Numbe o
inciden
cases, 2016
Annualised a e o
change o
age-s anda dised
incidence
1990–2006 2006–16 1990–2006 2006–16 1990–2006 2006–16 1990–2006 2006–16
Global 8 705 207
(7 754 638 o
9 817 655)
–1·8
(–2·0 o
–1·5)
–1·3
(–1·5 o
–1·1)
295 637
(261 369 o
335 586)
11·5
(10·8 o
12·3)
–2·1
(–2·9 o
–1·3)
18 452
(16 087 o
21 187)
43·9
(43·1 o
44·8)
7·9
(6·6 o
9·1)
9 019 296
(8 051 800 o
10 156 811)
–1·6
(–1·8 o
–1·3)
–1·3
(–1·5 o
–1·2)
High-income 122 022
(107 600 o
138 800)
–5·3 (–5·7
o
–5·0)
–2·6
(–2·8 o
–2·4)
1715
(1352 o
2225)
1·6
(0·1 o
3·0)
–3·2
(–5·3 o
–0·6)
216
(170 o
280)
29·7 (28·3
o
31·1)
4·4
(2·3 o
7·0)
123 952
(109 362 o
140 956)
–5·3
(–5·6 o
–5·0)
–2·6
(–2·8 o
–2·4)
High-income No h
Ame ica
12 179
(11 023 o
13 394)
–2·5
(–3·3 o
–1·7)
–2·2
(–2·6 o
–1·8)
139
(124 o
155)
–3·5
(–4·4 o
–2·6)
–2·2
(–2·8 o
–1·5)
17
(16 o
20)
20·4
(19·6 o
21·0)
5·5
(4·8 o
6·1)
12 335
(11 163 o
13 563)
–2·5
(–3·3 o
–1·7)
–2·2
(–2·6 o
–1·8)
Aus alasia 1396
(1194 o
1603)
–3·5
(–4·3 o
–2·6)
–1·5
(–2·1 o
–0·9)
27
(12 o
49)
6·5
(–1·8 o
14·5)
–0·2
(–11·0 o
11·3)
3
(2 o
6)
25·0
(18·4 o
30·6)
7·5
(–3·4 o
18·9)
1426
(1228 o
1640)
–3·4
(–4·2 o
–2·5)
–1·4
(–2·0 o
–0·9)
High-income Asia
Paci ic
69 366
(61 928 o
77 057)
–6·4
(–6·8 o
–6·1)
–3·1
(–3·5 o
–2·7)
872
(627 o
1210)
2·5
(0·1 o
4·8)
–5·2
(–8·6 o
–1·6)
110
(79 o
152)
36·7
(34·8 o
38·5)
2·5
(–0·9 o
6·0)
70 347
(62 801 o
78 134)
–6·4
(–6·7 o
–6·1)
–3·1
(–3·5 o
–2·7)
Wes e n Eu ope 30 930
(24 845 o
38 614)
–3·6
(–4·2 o
–3·1)
–1·8
(–2·2 o
–1·4)
512
(401 o
656)
1·0
(0·0 o
2·2)
–0·8
(–2·0 o
0·4)
64
(50 o
83)
27·1
(25·6 o
28·5)
6·9
(5·6 o
8·1)
31 506
(25 320 o
39 267)
–3·6
(–4·2 o
–3·0)
–1·8
(–2·2 o
–1·4)
Sou he n La in Ame ica 8152
(6651 o
9877)
–4·6
(–5·3 o
–3·9)
–0·9
(–1·7 o
–0·3)
165
(39 o
519)
8·3
(0·2 o
18·2)
–0·7
(–15·2 o
11·4)
21
(5 o
65)
31·6
(26·4 o
35·8)
7·0
(–7·6 o
19·1)
8338
(6857 o
10 068)
–4·4
(–5·2 o
–3·8)
–0·9
(–1·6 o
–0·3)
Cen al Eu ope,
eas e n Eu ope, and
cen al Asia
190 982
(167 765 o
218 563)
0·4 (0 o
0·8)
–3·3
(–3·9 o
–2·7)
34 818
(27 860 o
42 503)
14·2 (11·2
o
17·5)
–2·3
(–5·0 o
0·2)
7629
(6104 o
9312)
55·4 (54·1
o
56·7)
9·0
(6·2 o
11·5)
233 428
(206 001 o
263 867)
1·3 (0·9 o
1·7)
–2·9
(–3·2 o
–2·6)
Eas e n Eu ope 102 960
(88 487 o
119 678)
1·3
(0·7 o
1·8)
–3·1
(–3·9 o
–2·3)
20 668
(15 832 o
26 250)
10·9
(7·7 o
14·4)
–1·1
(–4·2 o
1·6)
4529
(3469 o
5752)
55·3
(53·9 o
56·8)
10·1
(7·1 o
12·9)
128 157
(110 861 o
146 360)
2·1
(1·5 o
2·6)
–2·5
(–3·0 o
–2·1)
Cen al Eu ope 26 337
(23 435 o
29 369)
–1·4
(–1·9 o
–1·0)
–3·8
(–4·1 o
–3·4)
440
(252 o
757)
7·6
(1·0 o
14·2)
–5·1
(–12·9 o
2·6)
97
(55 o
166)
37·5
(33·7 o
41·1)
6·2
(–1·6 o
13·8)
26 873
(23 937 o
29 909)
–1·3
(–1·8 o
–0·9)
–3·8
(–4·1 o
–3·5)
Cen al Asia 61 685
(53 120 o
71 399)
–0·7
(–1·3 o
–0·2)
–4·1
(–5·3 o
–2·8)
13 709
(10 092 o
18 011)
31·6
(24·9 o
38·6)
–4·5
(–9·2 o
0·1)
3004
(2211 o
3947)
61·0
(58·9 o
63·0)
6·7
(2·1 o
11·4)
78 397
(69 544 o
88 584)
0·7
(0·4 o
1·0)
–3·9
(–4·2 o
–3·7)
La in Ame ica and
Ca ibbean
161 862
(140 835 o
184 477)
–3·2 (–3·5
o
–3·0)
–2·3
(–2·5 o
–2·1)
3491
(2856 o
4329)
10·3 (7·1 o
14·0)
–3·3
(–5·0 o
–1·5)
276
(226 o
342)
34·5 (33·3
o
35·8)
6·5
(4·8 o
8·3)
165 629
(143 934 o
188 555)
–3·1
(–3·4 o
–2·8)
–2·3
(–2·5 o
–2·1)
Cen al La in Ame ica 48 806
(41 842 o
56 204)
–2·4
(–2·7 o
–2·2)
–1·8
(–2·0 o
–1·5)
991
(776 o
1261)
17·5
(15·7 o
19·3)
–2·9
(–4·8 o
–1·0)
78
(61 o
100)
31·8
(30·5 o
33·1)
6·9
(5·0 o
8·8)
49 875
(42 743 o
57 514)
–2·3
(–2·5 o
–2·1)
–1·8
(–2·0 o
–1·6)
Andean La in Ame ica 31 412
(27 884 o
35 278)
–6·0
(–6·3 o
–5·7)
–3·8
(–4·3 o
–3·4)
1417
(1007 o
2070)
6·3
(2·0 o
11·3)
–3·7
(–7·6 o
–0·1)
112
(80 o
163)
43·4
(41·3 o
45·6)
6·0
(2·2 o
9·7)
32 940
(29 161 o
36 850)
–5·8
(–6·0 o
–5·5)
–3·8
(–4·2 o
–3·4)
Ca ibbean 16 519
(13 885 o
19 541)
–3·0
(–3·7 o
–2·5)
–0·9
(–1·6 o
–0·2)
94
(30 o
277)
2·7
(–5·2 o
10·6)
–3·7
(–13·8 o
7·9)
7
(2 o
22)
27·9
(20·8 o
33·6)
6·1
(–4·0 o
17·7)
16 620
(13 957 o
19 656)
–3·0
(–3·7 o
–2·4)
–1·0
(–1·6 o
–0·2)
T opical La in Ame ica 65 126
(56 191 o
74 662)
–1·8
(–2·3 o
–1·3)
–2·2
(–2·4 o
–2·0)
989
(832 o
1150)
22·8
(22·1 o
23·5)
–3·2
(–3·9 o
–2·5)
78
(66 o
91)
32·5
(31·5 o
33·4)
6·6
(5·9 o
7·3)
66 193
(57 040 o
75 915)
–1·7
(–2·2 o
–1·2)
–2·2
(–2·4 o
–2·0)
(Table 1 con inues on nex page)
A icles
www. helance .com/in ec ion Vol 18 Decembe 2018
1335
o mul id ug- esis an ube culosis (appendix pp 88–100).
Among HIV-nega i e indi iduals in 2016, we es ima ed
18 452 (16 087–21 187) inciden cases o ube culosis we e
ex ensi ely d ug esis an and 10 920 (8896–13 162) dea hs
we e due o ex ensi ely d ug- esis an ube culosis
( ables 1 and 2). Among HIV-posi i e indi iduals in 2016,
we es ima ed ha 1303 (793–2019) inciden cases o ube -
culosis we e ex ensi ely d ug esis an and 1151 (689–1802)
dea hs we e due o ex ensi ely d ug- esis an ube culosis
(appendix pp 75–100). Es im a ed ube culosis p e alence by
d ug- esis ance ype and HIV s a us a e a ailable online.
Changes in he bu den o ube culosis o e ime
Globally, he annualised a e o change in age-s anda dised
incidence o ube culosis among HIV-nega i e indi iduals
was –1·3% (95% UI –1·5 o –1·2) om 2006 o 2016
( able 1), which is a slowe a e o change han in
1990–2006 (–1·6% [–1·8 o –1·3]; able 1). These a es o
change a e small compa ed wi h he dec ease in he
annualised a e o change in age-s anda dised
ube culosis mo ali y (–4·5% [–5·0 o –4·1]) om 2006
o 2016, which is la ge han he annualised a e o change
om he pe iod 1990–2006 (–3·2% [–3·7 o –2·9]; able 2).
Globally, he annualised a e o change in age-s an-
da dised incidence o ube culosis among HIV-posi i e
indi iduals dec eased om 2006 o 2016 (–4·0% [95% UI
–4·5 o –3·7]), whe eas in 1990–2006 he a e o change
inc eased (8·1% [7·5–8·8]; appendix pp 75–87). Mo ali y
among HIV-posi i e indi iduals has dec eased, wi h an
annualised a e o change o –8·9% (–9·5 o –8·4) o
D ug-suscep ible ube culosis Mul id ug- esis an ube culosis Ex ensi ely d ug- esis an
ube culosis
All HIV-nega i e ube culosis
Numbe o
inciden
cases, 2016
Annualised a e o
change o
age-s anda dised
incidence
Numbe o
inciden
cases, 2016
Annualised a e o
change o
age-s anda dised
incidence
Numbe o
inciden
cases, 2016
Annualised a e o
change o
age-s anda dised
incidence
Numbe o
inciden
cases, 2016
Annualised a e o
change o
age-s anda dised
incidence
1990–2006 2006–16 1990–2006 2006–16 1990–2006 2006–16 1990–2006 2006–16
(Con inued om p e ious page)
Sou heas Asia, eas Asia,
and Oceania
2 381 270
(2 144 141 o
2 637 352)
–3·6 (–3·9
o
–3·3)
–1·9
(–2·1 o
–1·8)
71 140
(59 963 o
86 643)
5·9
(4·8 o
6·9)
–4·0
(–5·9 o
–1·8)
6487
(5468 o
7900)
45·0 (44·1
o
46·0)
7·9
(6·0 o
10·1)
2 458 896
(2 215 080 o
2 722 674)
–3·4
(–3·7 o
–3·2)
–2·0
(–2·2 o
–1·8)
Eas Asia 1 207 570
(1 136 842 o
1 285 280)
–4·1
(–4·2 o
–3·9)
–2·0
(–2·2 o
–1·8)
50 864
(44 680 o
58 560)
4·8
(3·9 o
5·7)
–4·2
(–5·4 o
–2·9)
4638
(4074 o
5339)
45·1
(44·3 o
45·9)
7·7
(6·6 o
9·0)
1 263 072
(1 188 367 o
1 344 259)
–3·8
(–4·0 o
–3·6)
–2·1
(–2·3 o
–1·9)
Sou heas Asia 1 161 747
(993 911 o
1 345 961)
–3·3
(–3·9 o
–2·9)
–2·2
(–2·5 o
–2·0)
19 831
(12 086 o
32 586)
11·8
(7·5 o
16·2)
–3·3
(–9·8 o
2·9)
1808
(1102 o
2971)
44·4
(42·0 o
47·0)
8·6
(2·1 o
14·8)
1 183 386
(1 015 931 o
1 372 615)
–3·2
(–3·8 o
–2·8)
–2·3
(–2·5 o
–2·0)
Oceania 11 953
(10 254 o
13 622)
–1·3
(–3·4 o
–0·7)
–2·0
(–3·2 o
1·5)
444
(84 o
1373)
22·1
(6·4 o
39·3)
–7·1
(–29·7 o
23·5)
40
(8 o
125)
48·5
(30·5 o
58·9)
4·9
(–17·8 o
35·4)
12 438
(10 886 o
14 063)
–0·9
(–1·4 o
–0·6)
–2·2
(–2·6 o
–1·7)
No h A ica and
Middle Eas
264 890
(215 268 o
327 017)
–2·6
(–2·9 o
–2·3)
–2·2
(–2·6 o
–1·8)
7721
(4118 o
14 403)
15·3
(9·4 o
21·1)
–1·2
(–10·0 o
7·4)
273
(145 o
508)
34·5
(30·7 o
38·0)
7·0
(–1·8 o
15·6)
272 884
(221 272 o
336 385)
–2·4
(–2·7 o
–2·1)
–2·1
(–2·5 o
–1·8)
Sou h Asia 3 460 801
(3 161 059 o
3 815 348)
–1·7
(–1·8 o
–1·5)
–1·8
(–1·9 o
–1·6)
134 673
(121 505 o
149 109)
24·9
(24·4 o
25·4)
–1·5
(–2·1 o
–0·9)
3301
(2978 o
3654)
42·6
(41·9 o
43·2)
8·2
(7·6 o
8·8)
3 598 775
(3 290 577 o
3 966 147)
–1·4
(–1·6 o
–1·2)
–1·7
(–1·9 o
–1·6)
Sub-Saha an A ica 2 123 380
(1 768 596 o
2 579 273)
–0·9 (–1·4
o
–0·4)
–1·9
(–2·4 o
–1·4)
42 079
(30 717 o
57 455)
15·9 (13·8
o
18·0)
–2·2
(–5·4 o
1·1)
271
(198 o
371)
30·4 (28·4
o
32·5)
9·0
(5·7 o
12·2)
2 165 731
(1 80 1974 o
2 625 879)
–0·8
(–1·3 o
–0·2)
–1·9
(–2·3 o
–1·4)
Sou he n sub-Saha an
A ica
378 324
(290 890 o
490 853)
0·9
(–0·1 o
1·9)
–1·0
(–2·0 o
–0·2)
9732
(5716 o
15 799)
13·9
(10·1 o
17·9)
–0·5
(–4·8 o
4·0)
63
(37 o
102)
34·4
(31·2 o
37·3)
10·6
(6·4 o
15·2)
388 119
(299 438 o
504 411)
1·1
(0·0 o
2·0)
–1·0
(–2·0 o
–0·2)
Wes e n sub-Saha an
A ica
545 758
(444 535 o
672 322)
–1·6
(–2·3 o
–1·0)
–2·9
(–3·4 o
–2·3)
13 358
(7551 o
22 859)
17·4
(13·8 o
21·1)
–3·9
(–11·0 o
3·3)
86
(49 o
147)
30·6
(27·3 o
34·0)
7·3
(0·1 o
14·4)
559 202
(457 442 o
686 867)
–1·5
(–2·1 o
–0·9)
–2·9
(–3·4 o
–2·3)
Eas e n sub-Saha an
A ica
809 654
(678 742 o
983 108)
–0·8
(–1·2 o
–0·3)
–1·5
(–2·1 o
–0·8)
16 166
(10 561 o
24 674)
21·8
(17·9 o
25·5)
–0·8
(–4·9 o
3·5)
104
(68 o
159)
29·4
(27·1 o
31·6)
10·3
(6·2 o
14·7)
825 924
(693 500 o
1 004 959)
–0·6
(–1·1 o
–0·2)
–1·5
(–2·1 o
–0·8)
Cen al sub-Saha an
A ica
389 644
(329 875 o
455 126)
–0·9
(–1·3 o
–0·6)
–1·6
(–2·0 o
–1·1)
2824
(1782 o
4300)
8·6
(5·5 o
11·2)
–2·1
(–6·2 o
2·1)
18
(11 o
28)
28·1
(25·3 o
30·7)
9·0
(4·9 o
13·3)
392 486
(332 258 o
458 256)
–0·9
(–1·2 o
–0·6)
–1·6
(–2·0 o
–1·1)
Da a in pa en heses a e 95% unce ain y in e als. Mul id ug- esis an ube culosis=mul id ug- esis an ube culosis wi hou ex ensi e d ug esis ance.
Table 1: Inciden cases o ube culosis, d ug-suscep ible ube culosis, mul id ug- esis an ube culosis, and ex ensi ely d ug- esis an ube culosis in HIV-nega i e indi iduals in 2016,
and annualised a es o change o age-s anda dised incidence du ing he 1990–2006 and 2006–16 o 21 Global Bu den o Disease egions o bo h sexes combined
Fo isualisa ion o da a see
h p:// izhub.heal hda a.o g/
gbd-compa e
A icles
1336
www. helance .com/in ec ion Vol 18 Decembe 2018
2006–16, which is a subs an ial change om he annualised
inc ease o 9·2% (8·3–10·1) o 1990–2006 (appendix
pp 88–100).
Globally om 2006 o 2016, he annualised a e o
change in age-s anda dised incidence o mul id ug-
esis an ube culosis among HIV-nega i e indi iduals
was –2·1% (95% UI –2·9 o –1·3; able 1), and he a e o
change in mo ali y was –5·5% (–6·5 o –4·5; able 2). By
con as , we es ima ed ha he bu den o ex ensi ely
d ug- esis an ube culosis has inc eased globally. F om
2006 o 2016, he annualised age-s anda dised a e o
change in he incidence o ex ensi ely d ug- esis an
D ug-suscep ible ube culosis Mul id ug- esis an ube culosis Ex ensi ely d ug- esis an
ube culosis
All HIV-nega i e ube culosis
Numbe o
dea hs, 2016
Annualised a e o
change o
age-s anda dised
mo ali y
Numbe o
dea hs,
2016
Annualised a e o
change o
age-s anda dised
mo ali y
Numbe
o dea hs,
2016
Annualised a e o
change o
age-s anda dised
mo ali y
Numbe o
dea hs, 2016
Annualised a e o
change o
age-s anda dised
mo ali y
1990–2006 2006–16 1990–2006 2006–16 1990–2006 2006–16 1990–2006 2006–16
Global 1 105 898
(1 055 638 o
1 158 544)
–3·7
(–4·3 o
–3·4)
–4·4
(–4·9 o
–4·1)
96 238
(79 994 o
113 348)
11·9
(10·9 o
12·7)
–5·5
(–6·5 o
–4·5)
10 920
(8896 o
13 162)
43·9
(42·7 o
45·1)
3·1
(1·8 o
4·5)
1 213 057
(1 161 548 o
1 265 425)
–3·2
(–3·7 o
–2·9)
–4·5
(–5·0 o
–4·1)
High-income 12 759
(11 633 o
13 971)
–6·3
(–6·6 o
–5·9)
–3·5
(–4·4 o
–2·7)
457
(347 o
600)
–0·6
(–1·8 o
0·5)
–5·1
(–7·2 o
–2·7)
152
(115 o
202)
25·9
(24·4 o
27·3)
2·3
(0·3 o
4·7)
13 367
(12 202 o
14 591)
–6·1
(–6·4 o
–5·7)
–3·5
(–4·4 o
–2·7)
High-income
No h Ame ica
1041
(991 o
1095)
–7·0
(–7·1 o
–6·7)
–2·5
(–2·9 o
–2·0)
31
(26 o
38)
–8·7
(–9·1 o
–8·4)
–3·2
(–3·7 o
–2·6)
10
(8 o
13)
16·0
(14·8 o
17·3)
4·3
(3·7 o
4·8)
1082
(1033 o
1136)
–7·0
(–7·2 o
–6·8)
–2·5
(–2·9 o
–2·0)
Aus alasia 83
(74 o
93)
–5·0
(–5·7 o
–4·4)
–3·6
(–4·8 o
–2·4)
4
(2 o
8)
4·3
(–3·9 o
12·2)
–3·4
(–13·7 o
8·1)
1
(1 o
3)
19·1
(12·1 o
24·8)
4·1
(–6·3 o
15·6)
89
(80 o
99)
–4·7
(–5·3 o
–4·2)
–3·5
(–4·4 o
–2·5)
High-income Asia
Paci ic
7092
(6222 o
8028)
–7·1
(–7·6 o
–6·3)
–4·0
(–5·5 o
–2·5)
217
(156 o
304)
0·6
(–1·4 o
2·6)
–6·6
(–9·7 o
–3·4)
72
(50 o
102)
31·6
(29·8 o
33·4)
0·8
(–2·2 o
4·0)
7381
(6469 o
8328)
–6·9
(–7·4 o
–6·1)
–4·0
(–5·6 o
–2·6)
Wes e n Eu ope 3428
(3123 o
3777)
–6·3
(–6·6 o
–6·0)
–4·4
(–5·2 o
–3·6)
154
(121 o
191)
–1·5
(–2·5 o
–0·5)
–4·6
(–6·0 o
–3·2)
51
(40 o
65)
23·7
(22·4 o
24·9)
2·9
(1·5 o
4·3)
3633
(3322 o
4005)
–6·1
(–6·4 o
–5·8)
–4·3
(–5·1 o
–3·5)
Sou he n La in
Ame ica
1116
(955 o
1273)
–6·1
(–6·7 o
–5·6)
–3·7
(–5·2 o
–2·3)
50
(15 o
141)
5·7
(–2·2 o
14·9)
–5·1
(–16·6 o
5·0)
17
(5 o
46)
32·6
(27·9 o
36·2)
2·4
(–9·2 o
12·4)
1183
(1049 o
1333)
–5·8
(–6·3 o
–5·3)
–3·7
(–4·9 o
–2·4)
Cen al Eu ope,
eas e n Eu ope, and
cen al Asia
15 636
(12 358 o
19 998)
1·0
(0·1 o
1·7)
–7·8
(–10·2 o
–5·2)
6536
(5013 o
8358)
13·5
(10·8 o
16·6)
–8·2
(–10·9 o
–5·5)
3780
(2873 o
4801)
54·1
(52·8 o
55·1)
2·9
(0·2 o
5·6)
25 952
(21 354 o
31 882)
3·2
(2·6 o
3·8)
–6·9
(–8·9 o
–4·6)
Eas e n Eu ope 10 165
(7287 o
14 072)
3·1
(1·9 o
4·1)
–8·1
(–11·4 o
–4·7)
4660
(3317 o
6321)
11·7
(8·8 o
15·1)
–7·4
(–11·1 o
–4·0)
2695
(1893 o
3686)
55·3
(54·1 o
56·6)
3·7
(0·1 o
7·1)
17 520
(13 136 o
23 086)
5·0
(4·0 o
5·9)
–6·8
(–9·7 o
–3·7)
Cen al Eu ope 2156
(1944 o
2435)
–4·2
(–4·7 o
–3·9)
–6·2
(–7·4 o
–5·0)
79
(45 o
135)
4·3
(–1·6 o
10·6)
–8·6
(–16·2 o
–1·5)
46
(26 o
80)
33·8
(30·3 o
37·2)
2·5
(–5·1 o
9·6)
2281
(2075 o
2569)
–4·0
(–4·3 o
–3·6)
–6·2
(–7·2 o
–5·0)
Cen al Asia 3315
(2564 o
4188)
–1·3
(–2·8 o
–0·1)
–8·3
(–11·4 o
–5·0)
1797
(1330 o
2282)
30·5
(24·1 o
37·6)
–10·6
(–14·5 o
–7·0)
1039
(766 o
1322)
58·7
(56·9 o
60·1)
0·5
(–3·4 o
4·1)
6150
(5598 o
6926)
2·4
(1·8 o
2·9)
–8·1
(–9·2 o
–6·8)
La in Ame ica and
Ca ibbean
15 076
(14 239 o
16 158)
–7·1
(–7·5 o
–6·7)
–4·5
(–5·0 o
–4·1)
865
(693 o
1085)
6·7
(3·6 o
10·3)
–6·6
(–8·2 o
–4·8)
180
(140 o
232)
34·8
(33·3 o
36·4)
3·1
(1·4 o
4·9)
16 121
(15 263 o
17 292)
–6·7
(–7·1 o
–6·3)
–4·6
(–5·0 o
–4·2)
Cen al La in
Ame ica
4988
(4703 o
5312)
–7·7
(–8·0 o
–7·5)
–4·4
(–5·0 o
–3·9)
277
(219 o
349)
13·5
(11·8 o
15·1)
–6·2
(–7·9 o
–4·5)
58
(45 o
74)
33·1
(31·6 o
34·6)
3·4
(1·7 o
5·1)
5323
(5030 o
5657)
–7·3
(–7·6 o
–7·0)
–4·5
(–5·0 o
–4·0)
Andean La in
Ame ica
2800
(2299 o
3434)
–9·8
(–10·7 o
–8·7)
–5·9
(–7·4 o
–4·3)
335
(214 o
504)
2·8
(–1·3 o
7·7)
–7·1
(–10·9 o
–3·3)
70
(44 o
107)
43·9
(41·6 o
46·4)
2·5
(–1·3 o
6·3)
3205
(2723 o
3929)
–9·1
(–9·9 o
–7·9)
–5·9
(–7·3 o
–4·5)
Ca ibbean 1890
(1490 o
2306)
–5·5
(–6·4 o
–4·3)
–3·1
(–4·4 o
–1·7)
28
(7 o
86)
0·4
(–8·7 o
10·0)
–7·4
(–18·1 o
4·1)
6
(2 o
18)
28·9
(21·5 o
35·5)
2·2
(–8·5 o
13·7)
1923
(1522 o
2335)
–5·4
(–6·3 o
–4·2)
–3·1
(–4·4 o
–1·8)
T opical La in
Ame ica
5315
(5065 o
5662)
–4·4
(–4·7 o
–4·1)
–4·3
(–4·8 o
–3·8)
224
(189 o
263)
19·9
(19·3 o
20·5)
–6·1
(–6·9 o
–5·3)
47
(39 o
56)
31·8
(30·5 o
33·0)
3·5
(2·7 o
4·3)
5586
(5330 o
5940)
–4·1
(–4·3 o
–3·8)
–4·4
(–4·8 o
–3·9)
(Table 2 con inues on nex page)
A icles
www. helance .com/in ec ion Vol 18 Decembe 2018
1337
ube culosis was 7·9% (6·6–9·1; able 1), and he a e o
change o mo ali y was 3·1% (1·8–4·5; able 2). Fo he
same ime pe iod, he annualised age-s anda dised a es
o change o mul id ug- esis an ube culosis among
HIV-posi i e indi iduals was –4·6% (–6·6 o –2·7)
o incidence and –9·1% (–11·2 o –7·2) o mo ali y.
Fo ex ensi ely d ug- esis an ube culosis among
HIV-posi i e in di iduals, he annualised a e o change
was 7·2% (5·4 o 8·7) o incidence and 2·3% (0·9 o
3·7) o mo ali y (appendix pp 75–100).
Region-speci ic and coun y-speci ic ube culosis
incidence and mo ali y
Al hough we obse ed a global dec ease in he bu den o
ube culosis, his end was no uni o m ac oss all
egions and coun ies. Du ing 2006–16, among HIV-
nega i e indi iduals he annual pe cen age change in he
incidence o ube culosis a ied om –6·2% (95% UI
–6·7 o –5·6) in Kazakhs an, o 1·2% (0·7–1·8) in he
Philippines, and 1·3% (0·6–2·1) in U uguay (appendix
pp 48–61). In 2016, he age-s anda dised incidence a e
(pe 100 000 popula ion) o ube culosis in HIV-nega i e
indi iduals a ied om 3·1 (2·8–3·4) in he USA and
3·8 (2·5–5·4) in Pales ine, o 729·6 (537·3–1013·1) in
Leso ho and 842·7 (670·9–1040·0) in Cen al A ican
Republic ( igu e 3; appendix pp 101–07).
Age-s anda dised a es o ube culosis mo ali y among
HIV-nega i e indi iduals dec eased a a ying a es ac oss
coun ies and e i o ies om 2006 o 2016, wi h he
highes annual dec eases seen in Kazakhs an (–13·7%
D ug-suscep ible ube culosis Mul id ug- esis an ube culosis Ex ensi ely d ug- esis an
ube culosis
All HIV-nega i e ube culosis
Numbe o
dea hs, 2016
Annualised a e o
change o
age-s anda dised
mo ali y
Numbe o
dea hs,
2016
Annualised a e o
change o
age-s anda dised
mo ali y
Numbe
o dea hs,
2016
Annualised a e o
change o
age-s anda dised
mo ali y
Numbe o
dea hs, 2016
Annualised a e o
change o
age-s anda dised
mo ali y
1990–2006 2006–16 1990–2006 2006–16 1990–2006 2006–16 1990–2006 2006–16
(Con inued om p e ious page)
Sou heas Asia, eas
Asia, and Oceania
194 147
(183 457 o
205 231)
–6·1
(–6·7 o
–5·6)
–6·4
(–7·0 o
–5·7)
11 293
(8191 o
15 536)
3·3
(1·5 o
5·1)
–9·5
(–12·9 o
–5·9)
2720
(1951 o
3788)
43·7
(41·8 o
45·5)
2·3
(–1·2 o
5·8)
208 159
(198 135 o
219 944)
–5·7
(–6·3 o
–5·3)
–6·5
(–7·1 o
–6·0)
Eas Asia 38 192
(35 902 o
42 306)
–8·5
(–9·0 o
–8·1)
–8·3
(–8·9 o
–7·4)
4297
(3483 o
5259)
–0·2
(–1·1 o
0·8)
–11·3
(–12·6 o
–10·0)
1035
(822 o
1293)
40·4
(38·9 o
41·8)
0·5
(–0·8 o
1·8)
43 523
(41 277 o
48 120)
–7·8
(–8·2 o
–7·4)
–8·5
(–9·1 o
–7·6)
Sou heas Asia 155 084
(145 733 o
165 219)
–5·1
(–5·8 o
–4·5)
–5·9
(–6·6 o
–5·1)
6938
(4218 o
10 905)
9·1
(4·4 o
14·0)
–8·1
(–13·9 o
–2·3)
1671
(1012 o
2659)
48·7
(46·0 o
51·4)
3·6
(–2·1 o
9·5)
163 693
(154 881 o
173 618)
–4·8
(–5·5 o
–4·1)
–5·9
(–6·6 o
–5·2)
Oceania 573
(457 o
678)
–3·4
(–6·1 o
–2·2)
–4·1
(–6·7 o
0·4)
44
(10 o
119)
16·9
(1·4 o
33·3)
–8·6
(–28·3 o
18·6)
11
(2 o
29)
44·2
(27·9 o
52·5)
3·2
(–16·5 o
30·3)
628
(547 o
721)
–2·7
(–3·6 o
–1·9)
–4·4
(–5·6 o
–3·4)
No h A ica and
Middle Eas
35 401
(23 520 o
52 168)
–4·4
(–5·1 o
–3·7)
–4·7
(–6·0 o
–3·5)
3392
(1305 o
7231)
16·5
(10·7 o
22·3)
–4·8
(–14·9 o
4·8)
316
(121 o
684)
39·8
(33·8 o
44·1)
3·2
(–6·9 o
12·8)
39 109
(25 473 o
57 682)
–3·9
(–4·4 o
–3·3)
–4·7
(–5·3 o
–3·9)
Sou h Asia 451 816
(420 616 o
480 353)
–4·7
(–5·2 o
–4·3)
–4·9
(–5·6 o
–4·3)
52 768
(43 568 o
62 706)
22·0
(21·3 o
22·7)
–5·0
(–5·9 o
–4·2)
3416
(2777 o
4170)
46·8
(45·5 o
48·2)
4·5
(3·6 o
5·4)
508 000
(474 024 o
538 330)
–4·0
(–4·5 o
–3·6)
–4·9
(–5·6 o
–4·2)
Sub-Saha an A ica 381 032
(353 681 o
415 691)
–1·3
(–2·2 o
–0·5)
–3·8
(–4·4 o
–3·3)
20 924
(15 874 o
27 461)
16·1
(14·1 o
18·1)
–4·5
(–7·1 o
–1·8)
357
(269 o
473)
37·3
(35·6 o
38·9)
6·7
(4·1 o
9·3)
402 312
(374 030 o
438 809)
–1·0
(–1·8 o
–0·2)
–3·8
(–4·4 o
–3·4)
Sou he n sub-
Saha an A ica
35 286
(31 658 o
38 511)
2·9
(0·1 o
4·0)
–5·7
(–6·8 o
–4·0)
2604
(1717 o
4034)
15·7
(11·6 o
19·7)
–5·6
(–9·6 o
–1·1)
44
(29 o
68)
38·2
(35·6 o
40·9)
5·6
(1·6 o
10·1)
37 935
(34 180 o
41 177)
3·3
(0·5 o
4·3)
–5·7
(–6·8 o
–4·1)
Wes e n sub-
Saha an A ica
85 821
(75 972 o
100 174)
–2·8
(–3·5 o
–2·1)
–4·4
(–5·6 o
–3·3)
5677
(3510 o
8928)
15·7
(12·5 o
19·0)
–6·4
(–12·2 o
–0·5)
97
(59 o
154)
36·7
(33·8 o
39·5)
4·8
(–1·0 o
10·7)
91 594
(80 902 o
106 901)
–2·3
(–3·0 o
–1·7)
–4·6
(–5·6 o
–3·5)
Eas e n sub-
Saha an A ica
164 799
(147 918 o
184 516)
–1·7
(–2·8 o
0·2)
–3·7
(–4·5 o
–3·0)
10 563
(7041 o
15 603)
21·2
(16·6 o
25·6)
–3·0
(–6·9 o
0·9)
180
(118 o
268)
38·0
(35·9 o
40·0)
8·1
(4·2 o
12·0)
175 543
(157 583 o
196 713)
–1·3
(–2·4 o
0·6)
–3·6
(–4·4 o
–3·0)
Cen al sub-
Saha an A ica
95 003
(75 102 o
114 664)
–0·6
(–1·8 o
0·7)
–2·9
(–3·8 o
–2·1)
2072
(1209 o
3262)
8·7
(5·5 o
11·9)
–3·5
(–7·4 o
0·7)
35
(21 o
56)
36·1
(32·8 o
39·3)
7·7
(3·8 o
11·8)
97 110
(76 824 o
117 021)
–0·4
(–1·6 o
0·8)
–3·0
(–3·8 o
–2·1)
Da a in pa en heses a e 95% unce ain y in e als. Mul id ug- esis an ube culosis=mul id ug- esis an ube culosis wi hou ex ensi e d ug esis ance.
Table 2: Mo ali y om ube culosis o d ug-suscep ible ube culosis, mul id ug- esis an ube culosis, and ex ensi ely d ug- esis an ube culosis in HIV-nega i e indi iduals in 2016,
and annualised a es o change o age-s anda dised mo ali y du ing he pe iods 1990–2006 and 2006–16 o 21 Global Bu den o Disease egions o bo h sexes
A icles
1344
www. helance .com/in ec ion Vol 18 Decembe 2018
in incidence was obse ed in Kazakhs an, whe e im-
p o emen s we e a ibu able o ad ances in diagnos ics
and e ec i e ea men o newly diagnosed ube culosis
cases.23 We saw li le o no imp o emen in some coun ies,
including he Philippines and U uguay. In he Philippines,
a high p opo ion o smea -nega i e indi iduals who a e
posi i e o ube culosis by Xpe MTB/RIF assay (Cepheid,
USA) has been documen ed in high- isk popula ions,
including p ison inma es and indigenous popula ions,
sugges ing ha spu um-smea mic oscopy alone as a
ou ine diagnos ic es is inadequa e.24 In U uguay, a
dec ease in ea men success a e o new cases o
ube culosis ( om 84% in 2010 o 77% in 2015)25 p obably
con ibu es o he coun y’s he lack o p og ess.
Despi e imp o emen s in socio-demog aphic condi ions,
se e al egions ha e allen behind in hei p og ess o
educe he bu den o ube culosis. In 2016, mos coun ies
in Asia, sub-Saha an A ica, and eas e n Eu ope had a
highe bu den o ube culosis (bo h d ug suscep ible and
mul id ug esis an ) han expec ed gi en hei le el o
socio-demog aphic de elopmen . In many coun ies,
p o iding ea men se ices o mul id ug- esis an
ube culosis emains a challenge, pa ly because o he
high cos o second-line d ugs and poo adhe ence o
egimens.26,27 Globally in 2016, only 22% o people
wi h newly diagnosed d ug- esis an ube culosis we e
es ima ed o begin ea men , wi h a ea men success
a e o 54%.28 E idence sugges s ha alcohol abuse and
HIV in ec ion a e associa ed wi h inc eased isk o
unsuccess ul ou comes in pa ien s wi h mul id ug-
esis an ube culosis, bu he associa ion is unclea o
o he ac o s and como bidi ies such as smoking and
ch onic kidney disease.29 A mo e comp ehensi e unde -
s anding o he key d i e s o unsuccess ul ea men
ou comes in hese pa ien s is c ucial o imp o ing hei
ea men ou comes.
Al hough he bu den o ube culosis emains a o
om he expec ed le el in many coun ies wi h a high
bu den o ube culosis, he p e alence o diabe es—an
impo an isk ac o o bo h ube culosis and ad e se
ou comes om ube culosis ea men —has inc eased
o e ime wo ldwide as a consequence o se e al ac o s
including popula ion ageing and exposu e o li es yle-
ela ed isk ac o s,11,30 c ea ing addi ional challenges
o ube culosis ca e and p e en ion. Because o he
in e ac ion o ube culosis wi h diabe es and HIV/AIDS,
in eg a ing con ol p og ammes o he h ee diseases
could help p e en ube culosis among people wi h
HIV/AIDS and diabe es and educe he bu den o all
h ee diseases.31 Addi ionally, e o s o p e en o he isk
ac o s o ube culosis, including smoking and alcohol
misuse, could ha e a complemen a y e ec on he bu den
o ube culosis.2
In coun ies wi h a highe bu den o ube culosis han
expec ed on he basis o SDI, an impo an i s s ep is o
iden i y he easons o alling behind so ha app op ia e
measu es can be aken. Gaps in case de ec ion and delays
in diagnosis and ea men mos likely con ibu e o he
bu den being highe han expec ed, and coun y-speci ic
easons should also be in es iga ed. Al hough na ional
ube culosis p og ammes we e no i ied o abou
6·3 million new o elapsed cases o ube culosis globally
in 2016,28 we es ima ed ha he numbe o inciden cases
in 2016 was 10·4 million, implying a global case de ec ion
a e o 61%. E idence sugges s ha a ou ine passi e case-
inding s a egy is insu icien o de ec ing all ube culosis
cases.32,33 Ac i e case inding has been ecommended as a
complemen a y s a egy o passi e case inding o inc ease
case de ec ion; ne e heless, he e ec o ac i e case
inding on ea men ou comes and a e o ansmission,
and longe - e m e ec s on he epidemiology o
ube culosis, ha e ye o be de e mined.32
Also, despi e ad ances in ube culosis diagnos ics,
smea mic oscopy emains he mos commonly used
diagnos ic es in many coun ies ha a e endemic o
ube culosis.34 The Xpe MTB/RIF assay has highe
sensi i i y o he de ec ion o ube culosis han smea
mic oscopy,35 bu ew coun ies use Xpe o gene al
ube culosis case inding.36 Policies on he use o Xpe
MTB/RIF a y la gely be ween coun ies, wi h only
a subse o pa ien s wi h ube culosis being eligible o
he es (eg, pa ien s wi h suspec ed d ug esis ance,
HIV-posi i e indi iduals).36 A scale-up o Xpe MTB/RIF
could help in de ec ing addi ional cases, bu i has been
impeded by se e al ac o s, including high cos s, eliance
on unding om in e na ional dono s, and he lack o
subsidised p icing in he p i a e sec o , which is elied
on o mos ube culosis cases in some coun ies.26
O e all, e en wi h di e ences in me hods used, bo h
GBD 2016 and WHO es ima ed 10·4 million inciden
cases o ube culosis in 2016, al hough ou es ima ed
numbe o all ube culosis dea hs (1·45 million) is lowe
han WHO’s es ima e (1·7 million) o 2016.28 The
20 coun ies wi h he highes bu den, as assessed by he
numbe o inciden cases, di e be ween ou es ima es
and WHO’s: WHO includes Angola, B azil, and Thailand;
ins ead, we include Uganda, Zambia, and Zimbabwe. The
mos no able di e ence be ween ou and WHO’s es ima es
is be ween he es ima ed numbe s o ube culosis dea hs
among child en. We es ima ed 39 311 dea hs (95% UI
34 415–44 847) among child en who a e HIV nega i e and
younge han 15 yea s o 2016, which is subs an ially
lowe han he es ima es om WHO (201 000 dea hs)28 and
Dodd and colleagues (200 000 dea hs).37 The inpu da a and
he me hods used o gene a e es ima es o dea hs in
child en due o ube culosis a e e y di e en be ween
s udies: we used i al egis a ion and e bal au opsy da a
and he CODEm s a egy o es ima e ube culosis dea hs
in child en and WHO used he me hod o Dodd and
colleagues om hei 2017 s udy37 in which child mo ali y
due o ube culosis was back-calcula ed om he incidence
and case- a ali y a io. WHO es ima ed he incidence
o ube culosis in child en by combining esul s
om wo app oaches: he case de ec ion a e adjus men

A icles
www. helance .com/in ec ion Vol 18 Decembe 2018
1345
app oach (ie, incidence=no i ica ions/es ima ed case
de ec ion a e); and he me hod o Dodd and colleagues
om hei 2014 s udy38 in which incidence was es ima ed
om he annual isk o in ec ion in child en, WHO adul
smea -posi i e ube culosis p e alence da a, and
demog aphic in o ma ion by use o a ma hema ical model.
Bo h ou me hod and he me hod used by WHO and Dodd
and colleagues ha e limi a ions. Speci ically, conce ns
ha e been aised abou he misclassi ica ion o ube culosis
dea hs in child en as dea hs due o pneumonia in coun ies
wi h a high bu den o ube culosis.39 In his s udy, we did
no edis ibu e pneumonia dea hs o ube culosis dea hs
because o a lack o e idence on whe he ube culosis is a
cause o como bidi y o acu e se e e pneumonia in
child en.40 The back-calcula ion app oach used by WHO
and Dodd and colleagues mos likely has subs an ial
unce ain y due o assump ions in he p ocess o
es ima ing annual isk o in ec ion, he p e alence o adul
ube culosis, and case de ec ion a es.
Ou s udy has se e al limi a ions. Fi s , ou assessmen
o he ends in he bu den o mul id ug- esis an
ube culosis was es ic ed by a pauci y o ime-se ies da a
o many coun ies in Asia and A ica. We assumed ha
hese coun ies ha e a simila age-sex dis ibu ion o
mul id ug- esis an ube culosis o o he coun ies in he
same egion and used his common dis ibu ion o
gene a e end es ima es o coun ies and yea s wi h li le
da a; he lack o da a in a pa icula coun y is e lec ed in
wide unce ain y in e als. Second, e bal au opsy s udies
ha e modes sensi i i y in iden i ying ube culosis
dea hs.41–43 Howe e , a he ypical ange o he cause
ac ion o dea hs due o ube culosis in India and
sub-Saha an A ica (3–5%),43 and a he epo ed le el o
sensi i i y and speci ici y o a ibu ing ube culosis as he
cause o dea h in a la ge, mul icen e, e bal au opsy
alida ion s udy,43 we es ima e ha he alse posi i es and
alse nega i es la gely cancel ou . Thi d, as no ed in ou
p e ious publica ion,2 he main challenge in ou s a is ical
iangula ion app oach has been he sho age o da a om
su eys on cause o dea h and p e alence, pa icula ly
om coun ies in sub-Saha an A ica wi h a high
p e alence o HIV. We applied sophis ica ed modelling
me hods and co a ia es, using dis ibu ions ac oss geo-
g aphies and ime o help p edic o hose loca ions.
Acco dingly, he es ima es o a loca ion wi h spa se da a
a e coupled wi h wide unce ain y in e als. Fou h, o
in o m he case- a ali y a io among pa ien s wi h
un ea ed ube culosis o ou mo ali y- o-incidence a io
eg ession, we used da a om a single communi y-based
ollow-up s udy done in Bangalo e, India;21 wo o he
communi y-based s udies,44 done in India45 and he USA, 46,47
we e no included because o a lack o in o ma ion abou
he ea men o ube culosis o any sys ema ic ollow-up
o cases.
Despi e hese limi a ions, we made se e al imp o e-
men s in ou me hods compa ed wi h GBD 2015. Fi s ,
we no longe used case-de ec ion a es based on expe
opinion in he p ocess o es ima ing he incidence o
ube culosis. Ins ead, we used a mo ali y- o-incidence
a io app oach o be e e lec highe mo ali y and
incidence in low-income and middle-income coun ies.
Second, we s eng hened ou s a is ical iangula ion
app oach by inco po a ing popula ion-based su eys o
la en ube culosis in ec ion, and modelling incidence,
p e alence, and mo ali y simul an eously among he
popula ion who a e la en ly in ec ed o enhance
compa abili y ac oss coun ies. Because we used
Bayesian me a- eg ession o gene a e an incidence
es ima e ha is consis en wi h p e alence da a o cause-
speci ic mo ali y es ima es, ou es ima ed incidence
migh di e om coun ies’ o icial s a is ics (e en om
hose wi h a ou s a o i e s a quali y a ings). Thi d,
we imp o ed ou es ima es o ube culosis mo ali y by
including addi ional co a ia es ha ha e p oximal o
s ong associa ions wi h ube culosis mo ali y (ie,
p e alence o la en ube culosis in ec ion, p e alence o
ac i e ube culosis disease, p opo ion o adul s who a e
unde weigh , and HAQ Index). These imp o emen s,
oge he wi h subs an ial e o s o colla e da a o he
es ima ion o ube culosis bu den, ha e esul ed in
changes in GBD 2016 compa ed wi h GBD 2015,
especially in es ima es o mo ali y. The global numbe o
dea hs in 2016 due o ube culosis was 11% highe han
he GBD 2015 es ima e o 2015. The inc ease mainly
occu ed in some A ican coun ies—no ably, Bu undi,
Cen al A ican Republic, Congo, Democ a ic Republic
o he Congo, Gabon, Nige ia, Uganda, and Zambia had
mo e han wice he numbe o es ima ed dea hs in 2016
compa ed wi h in 2015. Fou h, in GBD 2015, we did
no sepa a ely examine he bu den o mul id ug-
esis an ube culosis. Gi en hei epidemiological and
clinical impo ance, we included es ima es o mul id ug-
esis an and ex ensi ely d ug- esis an ube culosis in
GBD 2016. Fu he es im a ion and mapping o he
bu den o ube culosis by d ug- esis ance ype and
HIV s a us a a ine spa ial esolu ion could be e
in o m su eillance and he a ge ing o esou ces o
in e en ions.48
As coun ies wo k owa ds achie ing he SDG a ge
o end he ube culosis epidemic by 2030, con empo a y
in o ma ion on he le els and ends o he bu den o
ube culosis is essen ial o ack and moni o he
p og ess o con ol e o s in indi idual coun ies.
Loca ions wi h he g ea es imp o emen s in con olling
ube culosis could p o ide insigh in o success ul
p og amma ic s a egies o coun ies wi h s agnan
p og ess. Ou indings sugges ha , i cu en ends in
ube culosis incidence con inue, ew coun ies will
mee he SDG a ge . P og ess needs o be accele a ed
by imp o ing he quali y o and access o ube culosis
diagnosis and ca e, scaling up o in e en ions o
p e en isk ac o s o ube culosis, and in eg a ing
con ol p og ammes o ube culosis, HIV, and
diabe es.
A icles
1346
www. helance .com/in ec ion Vol 18 Decembe 2018
GBD Tube culosis collabo a o s
Hmwe Hmwe Kyu, Emilie R Maddison, Na haniel J Hen y,
Jo ge R Ledesma, Ki s en E Wiens, Robe Reine J , Molly H Biehl,
Chloe Shields, Aa on Osgood-Zimme man, Jenni e M Ross,
Aus in Ca e , Tah i D F ank, Haidong Wang, Vinay S ini asan,
Zegeye Abebe, Sanjay Kuma Aga wal, Fa es Alahdab,
Ke yalew Addis Alene, Be iwan Abdulqadi Ali, Nelson Al is-Guzman,
Jason R And ews, Ca l Abela do T An onio, Suleman A ique,
Sachin R A e, Ashish Awas hi, Henok Tadesse Ayele, Hamid Badali,
Alaa Badawi, Aleksand a Ba ac, Nee aj Bedi, Masoud Behzadi a ,
Meysam Behzadi a , Bayu Begashaw Bekele, Saba Ab aham Belay,
Isabela M Benseno , Zahid A Bu , Félix Ca alho, Kelly Ce cy,
De asahayam J Ch is ophe , Alemneh Kabe a Daba, Lali Dandona,
Rakhi Dandona, Ahmad Da yani, Feleke Mekonnen Demeke,
Kebede De ibe, Sama h Dhamminda Dha ma a ne, Da id Teye Doku,
Manisha Dubey, Dumessa Edessa, Ziad El-Kha ib, Shymaa Enany,
Edua da Fe nandes, Flo ian Fische , Albe o L Ga cia-Bas ei o,
Abadi Kahsu Geb e, Geb emedhin Be he Geb ege gs,
Teklu Geb ehiwo Geb emichael, Tilayie Fe o Gelano, Demeke Ge emew,
Philimon N Gona, Amado Good idge, Rahul Gup a,
Hassan Haghpa as Bidgoli, Gessessew Bugssa Hailu,
Hamid Yimam Hassen, Mohammad T Tadesse Hedaya i,
Andualem Henok, So in Hos iuc, Mamusha Aman Hussen,
Olayinka S ephen Ilesanmi, Seyed Sina Naghibi I ani,
Ka h yn H Jacobsen, Sa ah C Johnson, Jos B Jonas, Amaha Kahsay,
Su ya Kan , Ami Kasaeian, Tes aye Dessale Kassa, Youse Saleh Khade ,
Mo eza Abdulla i Kha aie, Ib ahim Khalil, Ejaz Ahmad Khan,
Young-Ho Khang, Yun Jin Kim, Sonali Kochha , Ai Koyanagi,
K is ophe J K ohn, G Anil Kuma , Ayenew Molla Lakew,
Che u Tesema Lesha gie, Rakesh Lodha, E lyn Rachelle King Maca ayan,
Reza Majdzadeh, F ancisco Roge lândio Ma ins-Melo, Addisu Melese,
Ziad A Memish, Wal e Mendoza, Desalegn Tadese Mengis u,
Ge ne Mengis u, Tomisla Mes o ic, Babak Moazen,
Ka zan Abdulmuhsin Mohammad, Sha iu Mohammed, Ali H Mokdad,
Mahmood Moosazadeh, Seyyed Meysam Mousa i, Ghulam Mus a a,
Jean B Nachega, Long Hoang Nguyen, Son Hoang Nguyen,
T ang Huyen Nguyen, Dina Nu Angg aini Ning um,
Yi ga Legesse Ni ayo, Vuong Minh Nong, Richa d O o i-Asenso,
Felix Akpojene Ogbo, In-Hwan Oh, Olan ewaju Oladimeji,
And ew T Olagunju, Eyal O en, Da id M Pe ei a, Swayam P akash,
Mos a a Qo bani, Anwa Ra ay, Rajesh Kuma Rai, Usha Ram,
Sal a o e Rubino, Saeid Sa i i, Joshua A Salomon, Abdallah M Samy,
Benn Sa o ius, Maheswa Sa pa hy, Seyedmoj aba Seyedmousa i,
Mehdi Sha i , João Ped o Sil a, Dayane Gab iele Al es Sil ei a,
Jas inde A Singh, Chand ashekha T S ee ama eddy, Bach Xuan T an,
A ewe ki Geb emeskel Tsadik, Kingsley Nnanna Ukwaja, I an Ullah,
Olalekan A U hman, Vasily Vlasso , S ein Emil Vollse , Giang Vu,
Fi sum Weldegeb eal, And ea We decke , Eb ahim M Yime , Naohi o
Yonemo o, Ma cel Yo ebieng, Mohsen Nagha i, Theo Vos, Simon I Hay,
Ch is ophe J L Mu ay.
A ilia ions
Depa men o Heal h Me ics Sciences (H H Kyu PhD,
R Reine J PhD, H Wang PhD, I Khalil MD, P o A H Mokdad PhD,
P o S E Vollse D PH, P o M Nagha i PhD, P o T Vos PhD,
P o S I Hay DSc, P o C J L Mu ay DPhil), Ins i u e o Heal h Me ics
and E alua ion (H H Kyu PhD, E R Maddison BS, N J Hen y BS,
J R Ledesma BA, K E Wiens PhD, R Reine J PhD, M H Biehl MPH,
C Shields BA, A Osgood-Zimme man MS, A Ca e MPH,
T D F ank BS, H Wang PhD, V S ini asan BA, K Ce cy BS,
P o L Dandona MD, P o R Dandona PhD, S D Dha ma a ne MD,
P o S I Hay DSc, S C Johnson MSc, I Khalil MD, K J K ohn MPH,
P o A H Mokdad PhD, P o M Nagha i PhD, P o S E Vollse D PH,
P o T Vos PhD, P o C J L Mu ay DPhil), Depa men o Global Heal h
(J M Ross MD, S Kochha MD), Uni e si y o Washing on, Sea le, WA,
USA (P o E O en PhD); Depa men o Neph ology
(P o S K Aga wal MD), Depa men o Paedia ics (P o R Lodha MD),
All India Ins i u e o Medical Sciences, New Delhi, India; E idence
Based P ac ice Cen e , Mayo Clinic Founda ion o Medical Educa ion
and Resea ch, Roches e , MN, USA (F Alahdab MD); Ins i u e o Public
Heal h (K A Alene MPH, B B Bekele MPH), Depa men o Medical
Mic obiology (F M Demeke MSc, D Ge emew MSc), Epidemiology and
Bios a is ics (A M Lakew MPH), Human Nu i ion Depa men
(Z Abebe MSc), Uni e si y o Gonda , Gonda , E hiopia; Resea ch School
o Popula ion Heal h, Aus alian Na ional Uni e si y, Canbe a, ACT,
Aus alia (K A Alene MPH); Medical Technical Ins i u e, E bil
Poly echnic Uni e si y, E bil, I aq (B A Ali PhD); Facul y o Pha macy
(B A Ali PhD), E bil (K A Mohammad PhD), Ishik Uni e si y, E bil, I aq;
Resea ch G oup on Heal h Economics, Uni e si y o Ca agena,
Ca agena, Colombia (P o N Al is-Guzman PhD); Resea ch G oup in
Hospi al Managemen and Heal h Policies, Uni e si y o he Coas ,
Ba anquilla, Colombia (P o N Al is-Guzman PhD); Di ision o
In ec ious Disease & Geog aphic Medicine (J R And ews MD), Cen e
o Heal h Policy & Cen e o P ima y Ca e and Ou comes Resea ch
(P o J A Salomon PhD), S an o d Uni e si y, S an o d, CA, USA;
Depa men o Heal h Policy and Adminis a ion (C A T An onio MD),
De elopmen and Communica ion S udies (E K Maca ayan PhD),
Uni e si y o he Philippines Manila, Manila, Philippines; Uni e si y
Ins i u e o Public Heal h, The Uni e si y o Laho e, Laho e, Pakis an
(S A ique PhD); Uni e si y o Hail, Hail, Saudi A abia (S A ique PhD);
Cen e o Clinical Global Heal h Educa ion (S R A e PhD), Depa men
o Epidemiology (P o J B Nachega PhD), Johns Hopkins Uni e si y,
Bal imo e, MD, USA; D . D.y. Pa il Medical College,
D . D.y. Pa il Vidyapee h, Pune, India (S R A e PhD); Indian Ins i u e o
Public Heal h, Gandhinaga , India (A Awas hi PhD); Public Heal h
Founda ion o India, Gu ug am, India (A Awas hi PhD,
P o L Dandona MD, P o R Dandona PhD, G Kuma PhD); Depa men
o Epidemiology, Bios a is ics, and Occupa ional Heal h, McGill
Uni e si y, Mon eal, QC, Canada (H T Ayele PhD); Public Heal h
Depa men , Dilla Uni e si y, Dilla, E hiopia (H T Ayele PhD);
Depa men o Medical Mycology (H Badali PhD,
P o M T Hedaya i PhD), Toxoplasmosis Resea ch Cen e
(P o A Da yani PhD), Heal h Sciences Resea ch Cen e
(M Moosazadeh PhD), In asi e Fungi Resea ch Cen e
(P o S Seyedmousa i PhD, P o M T Hedaya i PhD), Mazanda an
Uni e si y o Medical Sciences, Sa i, I an; Public Heal h Risk Sciences
Di ision, Public Heal h Agency o Canada, To on o, ON, Canada
(A Badawi PhD); Depa men o Nu i ional Sciences, Uni e si y o
To on o, To on o, ON, Canada (A Badawi PhD); Clinic o In ec ious and
T opical Diseases, Clinical Cen e o Se bia, Belg ade, Se bia
(A Ba ac PhD); Facul y o Medicine, Uni e si y o Belg ade, Belg ade,
Se bia (A Ba ac PhD); Depa men o Communi y Medicine, Gandhi
Medical College Bhopal, Bhopal, India (P o N Bedi MD); Jazan
Uni e si y, Jazan, Saudi A abia (P o N Bedi MD); Heal h Managemen
and Economics Resea ch Cen e , I an Uni e si y o Medical Sciences,
Teh an, I an (M Behzadi a PhD); Social De e minan s o Heal h
Resea ch Cen e (M Behzadi a PhD), Lo es an Uni e si y o Medical
Sciences, Kho amabad, I an (M Behzadi a MSc); Public Heal h
Depa men (B B Bekele MPH, H Y Hassen MPH), Mizan-Tepi
Uni e si y, Teppi, E hiopia (A Henok MPH); D . Tewelde Legesse Heal h
Sciences College, Mekelle, E hiopia (S A Belay MPH); Depa men o
In e nal Medicine, Uni e si y o São Paulo, São Paulo, B azil
(I M Benseno PhD); School o Popula ion and Public Heal h, Uni e si y
o B i ish Columbia, Vancou e , BC, Canada (Z A Bu PhD); Al Shi a
School o Public Heal h, Al Shi a T us Eye Hospi al, Rawalpindi,
Pakis an (Z A Bu PhD); Ins i u e o Public Heal h
(P o F Ca alho PhD), REQUIMTE/LAQV (P o E Fe nandes PhD,
P o D M Pe ei a PhD), Ucibio (J P Sil a PhD), Applied Molecula
Biosciences Uni (P o F Ca alho PhD), Uni e si y o Po o, Po o,
Po ugal; Depa men o Pulmona y Medicine, Ch is ian Medical
College and Hospi al (CMC), Vello e, India (P o D J Ch is ophe MD);
College o Medicine and Heal h Sciences, Hawassa Uni e si y, Hawassa,
E hiopia (A K Daba MSc); Depa men o Global Heal h and In ec ion,
B igh on and Sussex Medical School, B igh on, UK (K De ibe PhD);
School o Public Heal h, Addis Ababa Uni e si y, Addis Ababa, E hiopia
(K De ibe PhD); Depa men o Communi y Medicine, Uni e si y o
Pe adeniya, Pe adeniya, S i Lanka (S D Dha ma a ne MD); Depa men
o Popula ion and Heal h, Uni e si y o Cape Coas , Cape Coas , Ghana
(D T Doku PhD); Facul y o Social Sciences, Heal h Sciences, Uni e si y
o Tampe e, Tampe e, Finland (D T Doku PhD); Uni ed Na ions Wo ld
Food P og amme, New Delhi, India (M Dubey PhD); School o
Pha macy (D Edessa MSc, G Mengis u MSc), Depa men o Medical
Labo a o y Science (F Weldegeb eal MSc), Ha amaya Uni e si y, Ha a ,
A icles
www. helance .com/in ec ion Vol 18 Decembe 2018
1347
E hiopia (T F Gelano MSc); Depa men o Public Heal h Sciences,
Ka olinska Ins i u e , S ockholm, Sweden (Z El-Kha ib PhD);
Depa men o Mic obiology and Immunology, Suez Canal Uni e si y,
Ismailia, Egyp (S Enany PhD); Depa men o Public Heal h Medicine,
Biele eld Uni e si y, Biele eld, Ge many (F Fische PhD); Tube culosis,
Manhiça Heal h Resea ch Cen e , Manhiça, Mozambique
(A L Ga cia-Bas ei o MD); Tube culosis Depa men , Ba celona Ins i u e
o Global Heal h, Ba celona, Spain (A L Ga cia-Bas ei o MD); School o
Pha macy (A K Geb e MSc, T G Geb emichael MSc, A G Tsadik MSc,
E M Yime MSc), School o Public Heal h (G B Geb ege gs MPH),
Biomedical Sciences Di ision (G B Hailu MSc), Depa men o Nu i ion
and Die e ics (A Kahsay MPH), Clinical Pha macy Uni (T D Kassa MSc,
Y L Ni ayo MSc), School o Medicine (D T Mengis u MSc), Mekelle
Uni e si y, Mekelle, E hiopia; Nu sing and Heal h Sciences Depa men ,
Uni e si y o Massachuse s Bos on, Bos on, MA, USA (P N Gona PhD);
Tube culosis Bioma ke Resea ch Uni , Ins i u e o Scien i ic Resea ch
and High Technology Se ices, Ci y O Knowledge, Panama
(A Good idge PhD); Commissione o Public Heal h, Wes Vi ginia
Bu eau o Public Heal h, Cha les on, WV, USA (P o R Gup a MD);
Depa men o Heal h Policy, Managemen & Leade ship, Wes Vi ginia
Uni e si y School o Public Heal h, Mo gan own, WV, USA
(P o R Gup a MD); Ins i u e o Global Heal h, Uni e si y College
London, London, UK (H Haghpa as Bidgoli PhD); Uni o
Epidemiology and Social Medicine, Uni e si y Hospi al An we p,
Wil ijk, Belgium (H Y Hassen MPH); Facul y o Den is y, Dep o Legal
Medicine and Bioe hics, Ca ol Da ila Uni e si y o Medicine and
Pha macy, Bucha es , Romania (S Hos iuc PhD); Depa men o Heal h
Educa ion & Beha io al Sciences, Jimma Uni e si y, Jimma, E hiopia
(M A Hussen MPH); Depa men o Public Heal h and Communi y
Medicine, Uni e si y o Libe ia, Mon o ia, Libe ia (O S Ilesanmi PhD);
Resea ch Ins i u e o Endoc ine Sciences, Shahid Behesh i Uni e si y
o Medical Sciences, Teh an, I an (S N I ani MD); Non-Communicable
Diseases Resea ch Cen e (S N I ani MD), Hema ology-Oncology and
S em Cell T ansplan a ion Resea ch Cen e (A Kasaeian PhD),
Communi y-Based Pa icipa o y-Resea ch Cen e
(CBPR; P o R Majdzadeh PhD), Depa men o Heal h Managemen
and Economics (S Mousa i PhD), Cen e o Expe ise in Mic obiology
(P o S Seyedmousa i PhD), Hema ologic Malignancies Resea ch Cen e
(A Kasaeian PhD), Knowledge U iliza ion Resea ch Cen e
(P o R Majdzadeh PhD), Teh an Uni e si y o Medical Sciences,
Teh an, I an; Depa men o Global and Communi y Heal h, Geo ge
Mason Uni e si y, Fai ax, VA, USA (K H Jacobsen PhD); Depa men o
Oph halmology (P o J B Jonas MD), Ins i u e o Public Heal h
(B Moazen MSc, S Mohammed PhD), Heidelbe g Uni e si y,
Mannheim, Ge many; Beijing Ins i u e o Oph halmology, Beijing
Tong en Hospi al, Beijing, China (P o J B Jonas MD); Depa men o
Respi a o y Medicine, King Geo ge’s Medical Uni e si y, Lucknow, India
(P o S Kan MD); Depa men o Public Heal h and Communi y
Medicine, Jo dan Uni e si y o Science and Technology, Ram ha,
Jo dan (P o Y S Khade PhD); Depa men o Public Heal h, Ah az
Jundishapu Uni e si y o Medical Sciences, Ah az, I an
(M A Kha aie PhD); Epidemiology and Bios a is ics Depa men , Heal h
Se ices Academy, Islamabad, Pakis an (E A Khan MPH); Ins i u e o
Heal h Policy and Managemen (P o Y Khang MD), Depa men o
Heal h Policy and Managemen (P o Y Khang MD), Seoul Na ional
Uni e si y, Seoul, Sou h Ko ea; School o Medicine, Xiamen Uni e si y
Malaysia, Sepang, Malaysia (Y Kim PhD); Depa men o Public Heal h,
E asmus Uni e si y Medical Cen e , Ro e ham, Ne he lands
(S Kochha MD); Resea ch and De elopmen Uni , San Juan de Dios
Sani a y Pa k, San Boi De Llob ega , Spain (A Koyanagi MD);
Depa men o Public Heal h, Deb e Ma kos Uni e si y, Deb e Ma kos,
E hiopia (C T Lesha gie MPH); A iadne Labs, Ha a d Uni e si y,
Bos on, MA, USA (E K Maca ayan PhD); Campus Caucaia, Fede al
Ins i u e o Educa ion, Science and Technology o Cea á, Caucaia, B azil
(F R Ma ins-Melo PhD); College o Heal h Sciences, Deb e Tabo
Uni e si y, Deb e Tabo , E hiopia (A Melese MSc); Resea ch Depa men
P ince Mohammed Bin Abdulaziz Hospi al, Minis y o Heal h, Riyadh,
Saudi A abia (P o Z A Memish MD); College o Medicine, Al aisal
Uni e si y, Riyadh, Saudi A abia (P o Z A Memish MD); Pe u Coun y
O ice, Uni ed Na ions Popula ion Fund (UNFPA), Lima, Pe u
(W Mendoza MD); Depa men o Pha macy, Wollo Uni e si y, Dessie,
E hiopia (G Mengis u MSc); Clinical Mic obiology and Pa asi ology
Uni , D . Zo a P o ozic Polyclinic, Zag eb, C oa ia (T Mes o ic PhD);
Uni e si y Cen e Va azdin, Uni e si y No h, Va azdin, C oa ia
(T Mes o ic PhD); Ins i u e o Addic ion Resea ch, F ank u Uni e si y
o Applied Sciences, F ank u , Ge many (B Moazen MSc); Depa men
o Biology, Salahaddin Uni e si y, E bil, I aq (K A Mohammad PhD);
Heal h Sys ems and Policy Resea ch Uni , Ahmadu Bello Uni e si y,
Za ia, Nige ia (S Mohammed PhD); Pedia ics Depa men , Nish a
Medical Uni e si y, Mul an, Pakis an (P o G Mus a a MD); Pedia ics &
Pedia ic Pulmonology, Ins i u e o Mo he & Child Ca e, Mul an,
Pakis an (P o G Mus a a MD); Depa men o Epidemiology, Uni e si y
o Pi sbu gh, Pi sbu gh, PA, USA (P o J B Nachega PhD); Cen e o
Excellence in Beha io al Heal h (L H Nguyen PhD,
T H Nguyen B Med Sc), Cen e o Excellence in Beha io al Medicine
(S H Nguyen BS, G Vu MSc), Nguyen Ta Thanh Uni e si y, Ho Chi
Minh Ci y, Vie nam; Public Heal h Science Depa men , S a e Uni e si y
o Sema ang, Ko a Sema ang, Indonesia (D N A Ning um MPH);
G adua e Ins i u e o Biomedical In o ma ics, Taipei Medical Uni e si y,
Taipei Ci y, Taiwan (D N A Ning um MPH); Ins i u e o Global Heal h
Inno a ions, Duy Tan Uni e si y, Hanoi, Vie nam (V M Nong MPH);
Cen e o Ca dio ascula Resea ch and Educa ion in The apeu ics,
Monash Uni e si y, Melbou ne, VIC, Aus alia (R O o i-Asenso MSc);
Independen Consul an , Acc a, Ghana (R O o i-Asenso MSc,
A We decke PhD); Wes e n Sydney Uni e si y, Pen i h, NSW, Aus alia
(F A Ogbo PhD); Depa men o P e en i e Medicine, Kyung Hee
Uni e si y, Dongdaemun-gu, Sou h Ko ea (I Oh PhD); Human Sciences
Resea ch Council, Du ban, Sou h A ica (O Oladimeji MD); School o
Public Heal h, Facul y o Heal h Sciences, Uni e si y o Namibia,
Osakha i, Namibia (O Oladimeji MD); Uni e si y o Adelaide, Adelaide,
SA, Aus alia (A T Olagunju MD); Depa men o Psychia y, Uni e si y
o Lagos, Lagos, Nige ia (A T Olagunju MD); G adua e School o Public
Heal h, San Diego S a e Uni e si y, San Diego, CA, USA
(P o E O en PhD); Ca agena Uni e si y, Ca agena, Colombia
(P o D M Pe ei a PhD); Depa men o Neph ology, Sanjay Gandhi
Pos g adua e Ins i u e o Medical Sciences, Lucknow, India
(S P akash PhD); Non-communicable Diseases Resea ch Cen e , Albo z
Uni e si y o Medical Sciences, Ka aj, I an (M Qo bani PhD);
Depa men o Epidemiology & Bios a is ics, Con ech School o Public
Heal h, Laho e, Pakis an (A Ra ay MS); Socie y o Heal h and
Demog aphic Su eillance, Su i, India (R Rai MPH); Depa men o
Economics, Uni e si y o Goe ingen, Gö ingen, Ge many
(R Rai MPH); Depa men o Public Heal h & Mo ali y S udies,
In e na ional Ins i u e o Popula ion Sciences, Mumbai, India
(P o U Ram PhD); Depa men o Biomedical Science, Uni e si y o
Sassa i, Sassa i, I aly (S Rubino MD); Manage ial Epidemiology
Resea ch Cen e , Ma agheh Uni e si y o Medical Sciences, Ma agheh,
I an (S Sa i i PhD); Depa men o En omology, Ain Shams Uni e si y,
Cai o, Egyp (A M Samy PhD); Depa men o Public Heal h Medicine,
Uni e si y o Kwazulu-na al, Du ban, Sou h A ica (P o B Sa o ius
PhD); Ugc Cen e o Ad anced S udy in Psychology, U kal Uni e si y,
Bhubaneswa , India (M Sa pa hy PhD); Udyam-global Associa ion o
Sus ainable De elopmen , Bhubaneswa , India (M Sa pa hy PhD);
Depa men o Basic Sciences, Islamic Azad Uni e si y, Sa i, I an
(M Sha i PhD); B asília Uni e si y, B asília, B azil
(P o D A Sil ei a MD); Depa men o he Heal h Indus ial Complex
and Inno a ion in Heal h, Fede al Minis y o Heal h, B asília, B azil
(P o D A Sil ei a MD); Depa men o Epidemiology (J A Singh MD),
Depa men o Medicine (J A Singh MD), Uni e si y o Alabama a
Bi mingham, Bi mingham, AL, USA; Di ision o Communi y Medicine,
In e na ional Medical Uni e si y, Kuala Lumpu , Malaysia
(C T S ee ama eddy MD); Depa men o Heal h Economics, Hanoi
Medical Uni e si y, Hanoi, Vie nam (B X T an PhD); Depa men o
In e nal Medicine, Fede al Teaching Hospi al, Abakaliki, Nige ia
(K N Ukwaja MD); Gomal Uni e si y, De a Ismail Khan, Pakis an
(I Ullah PhD); Tb Cul u e Labo a o y, Mu i Mehmood Memo ial
Teaching Hospi al, De a Ismail Khan, Pakis an (I Ullah PhD); Di ision
o Heal h Sciences, Uni e si y o Wa wick, Co en y, UK
(O A U hman PhD); Depa men o Heal h Ca e Adminis a ion and
Economy, Na ional Resea ch Uni e si y Highe School o Economics,
Moscow, Russia (P o V Vlasso MD); Demog aphic Change and Ageing
Resea ch A ea, Fede al Ins i u e o Popula ion Resea ch, Wiesbaden,
A icles
1348
www. helance .com/in ec ion Vol 18 Decembe 2018
Ge many (A We decke PhD); Depa men o Psychopha macology,
Na ional Cen e o Neu ology and Psychia y, Tokyo, Japan
(N Yonemo o MPH); College o Public Heal h, Ohio S a e Uni e si y,
Columbus, OH, USA (M Yo ebieng PhD); and School o Public Heal h,
Uni e si y o Kinshasa, Kinshasa, Democ a ic Republic o he Congo
(M Yo ebieng PhD).
Con ibu o s
HHK p epa ed he i s d a o he manusc ip wi h con ibu ions
om KEW, ERM, and JRL in d a ing he esul s sec ion. ERM
cons uc ed he ables and igu es. CJLM, TV, SIH, MN, and HHK
p o ided o e all guidance. CS and MHB managed he p ojec . HHK
analysed he da a. ERM and NJH p o ided da a analys suppo . AOZ
and RR ini ia ed he modelling in as uc u e o es ima ing
p opo ions o mul id ug- esis an ube culosis. JRL and VS analysed
he associa ion be ween ube culosis bu den and Socio-demog aphic
Index. CJLM, HHK, ERM, JRL, and MHB inalised he manusc ip on
he basis o commen s om o he au ho s and e iewe eedback. All
o he au ho s p o ided da a, e iewed esul s, o e iewed he
manusc ip and p o ided commen s.
Decla a ion o in e es s
Wal e Mendoza, a collabo a o and co-au ho , is cu en ly a P og am
Analys o Popula ion and De elopmen a he Uni ed Na ions
Popula ion Fund-UNFPA Coun y O ice in Pe u, which does no
necessa ily endo se his s udy. We decla e no compe ing in e es s.
Da a sha ing
To download he da a used in hese analyses, please isi he Global
Heal h Da a Exchange a h p://ghdx.heal hda a.o g/node/373720.
Acknowledgmen s
The Bill & Melinda Ga es Founda ion unded he s udy.
We hank he WHO Global P ojec on An i-Tube culosis D ug
Resis ance Su eillance o p o iding he d ug- esis an ube culosis
da a. B en Bell, Sabina S Bloom, Ad ienne Chew, Selina Deipa ine,
Ka e Mulle , and Molly R Nixon con ibu ed o p oduc ion o he
manusc ip . We hank all con ibu o s o he Global Bu den o
Diseases 2016 S udy.
Re e ences
1 GBD 2016 Causes o Dea h Collabo a o s. Global, egional,
and na ional age-sex speci ic mo ali y o 264 causes o dea h,
1980–2016: a sys ema ic analysis o he Global Bu den o Disease
S udy 2016. Lance 2017; 390: 1151–210.
2 GBD Tube culosis Collabo a o s. The global bu den o ube culosis:
esul s om he Global Bu den o Disease S udy 2015.
Lance In ec Dis 2018; 18: 261–84.
3 Floyd K, Glaziou P, Houben RMGJ, Sumne T, Whi e RG,
Ra iglione M. Global ube culosis a ge s and miles ones se o
2016–2035: de ini ion and a ionale. In J Tube c Lung Dis 2018;
22: 723–30.
4 Glaziou P, Sismanidis C, Dodd PJ, Zignol M, Floyd K.
Me hods used by WHO o es ima e he global bu den o TB
disease. Gene a: Wo ld Heal h O ganiza ion, 2017. h p://www.
who.in / b/publica ions/global_ epo /g b 2017_online_ echnical_
appendix_disease_bu den_es ima ion.pd ?ua=1 (accessed
Oc 12, 2018).
5 Obe meye Z, Abbo -Kla e J, Mu ay CJ. Has he DOTS s a egy
imp o ed case inding o ea men success? An empi ical
assessmen . PLoS One 2008; 3: e1721.
6 Mu ay CJ, O blad KF, Guino a C, e al. Global, egional,
and na ional incidence and mo ali y o HIV, ube culosis, and
mala ia du ing 1990–2013: a sys ema ic analysis o he Global
Bu den o Disease S udy 2013. Lance 2014; 384: 1005–70.
7 GBD 2015 Mo ali y and Cause o Dea h Collabo a o s.
Global, egional, and na ional li e expec ancy, all-cause mo ali y,
and cause-speci ic mo ali y o 249 causes o dea h, 1980–2015:
a sys ema ic analysis o he Global Bu den o Disease S udy 2015.
Lance 2016; 388: 1459–544.
8 GBD 2015 Disease and Inju y Incidence and P e alence
Collabo a o s. Global, egional, and na ional incidence, p e alence,
and yea s li ed wi h disabili y o 310 diseases and inju ies,
1990–2015: a sys ema ic analysis o he Global Bu den o Disease
S udy 2015. Lance 2016; 388: 1545–602.
9 GBD 2015 DALYs and HALE Collabo a o s. Global, egional,
and na ional disabili y-adjus ed li e-yea s (DALYs) o 315 diseases
and inju ies and heal hy li e expec ancy (HALE), 1990–2015:
a sys ema ic analysis o he Global Bu den o Disease S udy 2015.
Lance 2016; 388: 1603–58.
10 GBD 2016 DALYs and HALE Collabo a o s. Global, egional,
and na ional disabili y-adjus ed li e-yea s (DALYs) o 333 diseases
and inju ies and heal hy li e expec ancy (HALE) o 195 coun ies
and e i o ies, 1990–2016: a sys ema ic analysis o he Global
Bu den o Disease S udy 2016. Lance 2017; 390: 1260–344.
11 GBD 2016 Disease and Inju y Incidence and P e alence
Collabo a o s. Global, egional, and na ional incidence, p e alence,
and yea s li ed wi h disabili y o 328 diseases and inju ies o
195 coun ies, 1990–2016: a sys ema ic analysis o he Global
Bu den o Disease S udy 2016. Lance 2017; 390: 1211–59.
12 GBD 2016 Mo ali y Collabo a o s. Global, egional, and na ional
unde -5 mo ali y, adul mo ali y, age-speci ic mo ali y, and li e
expec ancy, 1970–2016: a sys ema ic analysis o he Global Bu den
o Disease S udy 2016. Lance 2017; 390: 1084–150.
13 Mu ay CJ, Ezza i M, Flaxman AD, e al. GBD 2010:
design, de ini ions, and me ics. Lance 2012; 380: 2063–66.
14 Mes in YM, Hailema iam D, Biadgilign S, Kib e KT.
Associa ion be ween HIV/AIDS and mul i-d ug esis ance
ube culosis: a sys ema ic e iew and me a-analysis.
PLoS One 2014; 9: e82235.
15 GBD 2013 Mo ali y and Causes o Dea h Collabo a o s.
Global, egional, and na ional age–sex speci ic all-cause and
cause-speci ic mo ali y o 240 causes o dea h, 1990–2013:
a sys ema ic analysis o he Global Bu den o Disease S udy 2013.
Lance 2015; 385: 117–71.
16 Lozano R, Nagha i M, Fo eman K, e al. Global and egional
mo ali y om 235 causes o dea h o 20 age g oups in 1990 and
2010: a sys ema ic analysis o he Global Bu den o Disease S udy
2010. Lance 2012; 380: 2095–128.
17 Fo eman KJ, Lozano R, Lopez AD, Mu ay CJ. Modeling causes o
dea h: an in eg a ed app oach using CODEm.
Popul Heal h Me 2012; 10: 1.
18 O blad KF, Lozano R, Mu ay CJ. The bu den o HIV: insigh s om
he Global Bu den o Disease S udy 2010. AIDS 2013; 27: 2003–17.
19 GBD 2015 Heal hca e Access and Quali y Collabo a o s.
Heal hca e Access and Quali y Index based on mo ali y om causes
amenable o pe sonal heal h ca e in 195 coun ies and e i o ies,
1990–2015: a no el analysis om he Global Bu den o Disease
S udy 2015. Lance 2017; 390: 231–66.
20 Cen e s o Disease Con ol and P e en ion (CDC). Ex ensi ely
d ug- esis an ube culosis—Uni ed S a es, 1993–2006.
MMWR Mo b Mo al Wkly Rep 2007; 56: 250–53.
21 Na ional Tube culosis Ins i u e, Bangalo e. Tube culosis in a u al
popula ion o Sou h India: a i e-yea epidemiological s udy.
Bull Wo ld Heal h O gan 1974; 51: 473–88.
22 Flaxman AD, Vos T, Mu ay CJL. An in eg a i e me a eg ession
amewo k o desc ip i e epidemiology, 1s edn. Sea le, WA:
Uni e si y o Washing on P ess, 2015.
23 Minis y o Heal h and Social De elopmen o he Republic o
Kazakhs an. Incidence o ube culosis dec eased wice in
Kazakhs an o 10 yea s. The P ime Minis e o Kazakhs an O icial
websi e: Jan 25, 2016. h p://p imeminis e .kz/en/
news/10/ -kazahs ane-za-10-le -u o en-zabole aemos i-
ube kulezom-snizilsja- -22- aza-mzs - k (accessed Oc 12, 2018).
24 Mo ishi a F, Ga in AM, Lew W, e al. B inging s a e-o - he-a
diagnos ics o ulne able popula ions: he use o a mobile sc eening
uni in ac i e case inding o ube culosis in Palawan, he
Philippines. PLoS One 2017; 12: e0171310.
25 WHO. Tube culosis (TB). Da a p o ided by coun ies and e i o ies:
ea men ou comes. Gene a: Wo ld Heal h O ganiza ion.
h p://www.who.in / b/coun y/da a/download/en/ (accessed
Dec 14, 2017).
26 Albe H, Na ha i ha ana RR, Isaacs C, Pai M, Denkinge CM,
Boehme CC. De elopmen , oll-ou and impac o Xpe MTB/RIF
o ube culosis: wha lessons ha e we lea n and how can we do
be e ? Eu Respi J 2016; 48: 516–25.
27 O’Donnell MR, Da a y A, F ick M, e al. Re-in en ing adhe ence:
owa d a pa ien -cen e ed model o ca e o d ug- esis an
ube culosis and HIV. In J Tube c Lung Dis 2016; 20: 430–34.
A icles
www. helance .com/in ec ion Vol 18 Decembe 2018
1349
28 WHO. Global ube culosis epo 2017. Gene a: Wo ld Heal h
O ganiza ion, 2017.
29 Samuels JP, Sood A, Campbell JR, Ahmad Khan F, Johns on JC.
Como bidi ies and ea men ou comes in mul id ug esis an
ube culosis: a sys ema ic e iew and me a-analysis. Sci Rep 2018;
8: 4980.
30 Lonn o h K, Roglic G, Ha ies AD. Imp o ing ube culosis
p e en ion and ca e h ough add essing he global diabe es
epidemic: om e idence o policy and p ac ice.
Lance Diabe es Endoc inol 2014; 2: 730–39.
31 Seka R, My h eyee M. Tube culosis, HIV/AIDS and diabe es–is i
ime o hink oge he ? J Med Mic obiol Diagnosis 2012; 1: e103.
32 Ho J, Fox GJ, Ma ais BJ. Passi e case inding o ube culosis is no
enough. In J Mycobac e iol 2016; 5: 374–78.
33 Fox GJ, Nhung NV, Sy DN, e al. Household-con ac in es iga ion
o de ec ion o ube culosis in Vie nam. N Engl J Med 2018;
378: 221–29.
34 Pi aino F, Selimo ić Š. Diagnos ic de ices wi h mic o luidics,
1s edn. CRC P ess, 2017.
35 WHO. Au oma ed eal- ime nucleic acid ampli ica ion echnology
o apid and simul aneous de ec ion o ube culosis and i ampicin
esis ance: Xpe MTB/RIF sys em: policy s a emen . Gene a:
Wo ld Heal h O ganiza ion, 2011.
36 Qin ZZ, Pai M, Van Geme W, Sahu S, Ghiasi M, C eswell J.
How is Xpe MTB/RIF being implemen ed in 22 high ube culosis
bu den coun ies? Eu Respi J 2015; 45: 549–54.
37 Dodd PJ, Yuen CM, Sismanidis C, Seddon JA, Jenkins HE.
The global bu den o ube culosis mo ali y in child en:
a ma hema ical modelling s udy. Lance Glob Heal h 2017;
5: e898–06.
38 Dodd PJ, Ga dine E, Coghlan R, Seddon JA. Bu den o childhood
ube culosis in 22 high-bu den coun ies: a ma hema ical
modelling s udy. Lance Glob Heal h 2014; 2: e453–59.
39 G aham SM, Sismanidis C, Menzies HJ, Ma ais BJ, De jen AK,
Black RE. Impo ance o ube culosis con ol o add ess child
su i al. Lance 2014; 383: 1605–07.
40 Oliwa JN, Ka umbi JM, Ma ais BJ, Madhi SA, G aham SM.
Tube culosis as a cause o como bidi y o childhood pneumonia in
ube culosis-endemic a eas: a sys ema ic e iew.
Lance Respi Med 2015; 3: 235–43.
41 James SL, Flaxman AD, Mu ay CJ. Pe o mance o he Ta i
Me hod: alida ion o a simple addi i e algo i hm o analysis o
e bal au opsies. Popul Heal h Me 2011; 9: 31.
42 Lozano R, Lopez AD, A kinson C, Nagha i M, Flaxman AD,
Mu ay CJ. Pe o mance o physician-ce i ied e bal au opsies:
mul isi e alida ion s udy using clinical diagnos ic gold s anda ds.
Popul Heal h Me 2011; 9: 32.
43 Mu ay CJ, Lozano R, Flaxman AD, e al. Using e bal au opsy o
measu e causes o dea h: he compa a i e pe o mance o exis ing
me hods. BMC Med 2014; 12: 5.
44 Tieme sma EW, an de We MJ, Bo gdo MW, Williams BG,
Nagelke ke NJ. Na u al his o y o ube culosis: du a ion and a ali y
o un ea ed pulmona y ube culosis in HIV nega i e pa ien s:
a sys ema ic e iew. PLoS One 2011; 6: e17601.
45 Pam a SP, Goyal SS, Ma hu GP. Changes in p e alence and
incidence o pulmona y ube culosis in Delhi in ecen yea s.
Ind J Tube c 1973; 20: 57–64.
46 A ms ong DB. Fou yea s o he F amingham Demons a ion.
Am Re Tube c 1921; 4: 908–19.
47 A ms ong DB. Fou yea s o he F amingham Communi y Heal h
and Tube culosis Demons a ion. Am Re Tube c 1918; 2: 195–206.
48 Dowell SF, Blazes D, Desmond-Hellmann S. Fou s eps o p ecision
public heal h. Na u e 2016; 540: 189–91.