Guidelines o Modeling and Repo ing Heal h E ec s o Clima e Change
Mi iga ion Ac ions
Je emy J. Hess,
1
Nikhil Ranadi e,
2
Ch is Boye ,
1
Lukasz Aleksand owicz,
3
Susan C. Anenbe g,
4
K is in Aunan,
5
K is ine Beleso a,
6,7
Michelle L. Bell,
8
Sam Bicke s e h,
9
Ka h yn Bowen,
10
Ma ci Bu den,
1
Dia mid Campbell-Lend um,
11
Elizabe h Ca l on,
12
Guéladio Cissé,
13,14
F ancois Cohen,
15
Hancheng Dai,
16,17
Alan Da id Dangou ,
7
Pu nami a Dasgup a,
18
Howa d F umkin,
3
Peng Gong,
19
Robe J. Gould,
20
Andy Haines,
7
Simon Hales,
21
Ian Hamil on,
22
Tomoko Hasegawa,
23
Masahi o Hashizume,
24,25
Yasushi Honda,
26
Daniel E. Ho on,
27
Alexand a Ka ambelas,
28†
Ho Kim,
29
Sa byul Es ella Kim,
30
Pa ick L. Kinney,
31
Inza Kone,
32,33
Kim Knowl on,
34
Jos Lelie eld,
35
Vijay S. Limaye,
34
Qiyong Liu,
36
Lina Madaniyazi,
25,37
Micaela El i a Ma inez,
38,39
Denise L. Mauze all,
40
James Milne ,
6
Ta a Ne ille,
41
Ma k Nieuwenhuijsen,
42,43,44
Shonali Pachau i,
45
F ede ica Pe e a,
38
Helen Pineo,
46
Jus in V. Remais,
47
Rebecca K. Saa i,
48
Jon Samped o,
49‡
Pauline Scheelbeek,
7,50
Joel Schwa z,
51
D ew Shindell,
52
P iya Shyamsunda ,
53
Timo hy J. Taylo ,
54
Ca h yn Tonne,
42,43,44
De le Van Vuu en,
55
Can Wang,
56
Nicholas Wa s,
57
J. Jason Wes ,
58
Paul Wilkinson,
6
S ephen A. Wood,
8,59
James Woodcock,
60
Alis ai Woodwa d,
61
Yang Xie,
62,63
Ying Zhang,
64
and K is ie L. Ebi
1
1
Cen e o Heal h and he Global En i onmen , Uni e si y o Washing on, Sea le, Washing on, USA
2
Emo y Uni e si y School o Medicine, A lan a, Geo gia, USA
3
Ou Plane , Ou Heal h, Wellcome T us , London, UK
4
Milken Ins i u e School o Public Heal h, Geo ge Washing on Uni e si y, Washing on, Dis ic o Columbia, USA
5
CICERO Cen e o In e na ional Clima e Resea ch, Oslo, No way
6
Depa men o Public Heal h, En i onmen s, and Socie y, London School o Hygiene & T opical Medicine, London, UK
7
Cen e on Clima e Change and Plane a y Heal h, London School o Hygiene & T opical Medicine, London, UK
8
School o Fo es y and En i onmen al S udies, Yale Uni e si y, New Ha en, Connec icu , USA
9
Rocke elle Founda ion Economic Council on Plane a y Heal h, Ox o d, UK
10
Ri sumeikan Uni e si y, Kusa su, Japan
11
Depa men o En i onmen Clima e Change and Heal h, Wo ld Heal h O ganiza ion, Gene a, Swi ze land
12
Depa men o En i onmen al and Occupa ional Heal h, Colo ado School o Public Heal h, Uni e si y o Colo ado, Au o a, Colo ado, USA
13
Depa men o Epidemiology and Public Heal h, Swiss T opical and Public Heal h Ins i u e, Basel, Swi ze land
14
Uni e si y o Basel, Basel, Swi ze land
15
Smi h School o En e p ise and he En i onmen and Ins i u e o New Economic Thinking a he Ox o d Ma in School, Uni e si y o Ox o d, Ox o d, UK
16
Labo a o y o Ene gy & En i onmen al Economics and Policy (LEEEP), College o En i onmen al Sciences and Enginee ing, Peking Uni e si y, Beijing,China
17
College o En i onmen al Sciences and Enginee ing, Peking Uni e si y, Beijing, China
18
En i onmen al and Resou ce Economics Uni , Ins i u e o Economic G ow h, Delhi, India
19
Depa men o Ea h Sys em Science, Tsinghua Uni e si y, Beijing, China
20
Cen e o Clima e Change Communica ion, Geo ge Mason Uni e si y, Fai ax, Vi ginia, USA
21
Depa men o Public Heal h, Uni e si y o O ago, Welling on, New Zealand
22
UCL Ene gy Ins i u e, Uni e si y College London, London, UK
23
Na ional Ins i u e o En i onmen al S udies, Tsukuba, Japan
24
Ins i u e o T opical Medicine, Nagasaki Uni e si y, Nagasaki, Japan
25
School o T opical Medicine and Global Heal h, Nagasaki Uni e si y, Nagasaki, Japan
26
Facul y o Heal h and Spo Sciences, Uni e si y o Tsukuba, Tsukuba, Japan
27
Depa men o Ea h and Plane a y Sciences, No hwes e n Uni e si y, E ans on, Illinois, USA
28
Lamon -Dohe y Ea h Obse a o y, Columbia Uni e si y, Palisades, New Yo k USA
29
Depa men o Epidemiology and Bios a is ics, School o Public Heal h, Seoul Na ional Uni e si y, Seoul, Sou h Ko ea
30
Cen e o Clima e Change Adap a ion, Na ional Ins i u e o En i onmen al S udies, Tsukuba, Japan
31
Depa men o En i onmen al Heal h, Bos on Uni e si y School o Public Heal h, Bos on, USA
32
Cen e Suisse de Reche ches Scien i iques en Cô e d’I oi e, Abidjan, Cô e d’I oi e
33
Uni e si é Félix Houphoue -Boigny, Abidjan, Cô e d’I oi e
34
Na u al Resou ces De ense Council, New Yo k, New Yo k, USA
35
Max Planck Ins i u e o Chemis y, Dep . o A mosphe ic Chemis y, Mainz, Ge many
36
Na ional Ins i u e o Communicable Disease Con ol and P e en ion, Beijing, China
Add ess co espondence o Je emy J. Hess, [email p o ec ed] and K is ie L.
Ebi, [email p o ec ed]. 4225 Roose el Way NE #100, Sui e 2330, Box
354695, Sea le, WA, 98105
Supplemen al Ma e ial is a ailable online (h ps://doi.o g/10.1289/EHP6745).
†Cu en add ess: No heas S a es o Coo dina ed Ai Use Managemen ,
Bos on, Massachuse s USA
‡Cu en add ess: Join Global Change Resea ch Ins i u e, Pacific No hwes
Na ional Labo a o y, College Pa k, Ma yland
L.A. is employed by Wellcome T us , which unded his wo k. E.C. ecei ed
g an suppo p o ided o he Uni e si y o Colo ado om he U.S. EPA, NIH,
and Pan aleon, a majo suga manu ac u e in La in Ame ica. A. D. has been
awa ded compe i i e esea ch g an s om he Wellcome T us , he Bill and
Melinda Ga es Founda ion, and Uni ed Kingdom Resea ch and Inno a ion
(UKRI). A.K. is employed a No heas S a es o Coo dina ed Ai Use
Managemen (NESCAUM), 501(c)(3). K.K. and V.L. a e pa o a eam wi h
heal h cobenefi s wo k suppo ed by he Wellcome T us (g an No. 216093/Z/
19/Z). J.V.R. was suppo ed by he Na ional Science Founda ion Wa e ,
Sus ainabili y, and Clima e (g an s 1360330 and 1646708), he Na ional
Ins i u es o Heal h (g an s R01-TW010286 and R01-AI125842), and by he
Uni e si y o Cali o nia Mul icampus Resea ch P og ams and Ini ia i es
(awa d MRP-17-446315). J.W. has ecei ed unding o de elop anspo and
heal h models by he Eu opean Resea ch Council (ERC) unde he Ho izon
2020 esea ch and inno a ion p og am (g an ag eemen No. 817754). This ma-
e ial eflec s only he au ho ’s iews and he Commission is no liable o any
use ha may be made o he in o ma ion con ained he ein. J.W. also ecei es
unding o de elop anspo and heal h models om he UK MRC (METAHIT
and JIBE p ojec s) and om he UK D T (PCT p ojec ). The o he au ho s
decla e hey ha e no ac ual o po en ial compe ing financial in e es s.
Recei ed 15 Janua y 2020; Re ised 8 Sep embe 2020; Accep ed 13
Oc obe 2020; Published 10 No embe 2020.
No e o eade s wi h disabili ies: EHP s i es o ensu e ha all jou nal
con en is accessible o all eade s. Howe e , some figu es and Supplemen al
Ma e ial published in EHP a icles may no con o m o 508 s anda ds due o
he complexi y o he in o ma ion being p esen ed. I you need assis ance
accessing jou nal con en , please con ac [email protected] . Ou s aff
will wo k wi h you o assess and mee you accessibili y needs wi hin 3
wo king days.
En i onmen al Heal h Pe spec i es 115001-1 128(11) No embe 2020
A Sec ion 508–con o man HTML e sion o his a icle
is a ailable a h ps://doi.o g/10.1289/EHP6745.
Commen a y
37
Depa men o Paedia ic Diseases, Ins i u e o T opical Medicine, Nagasaki, Japan
38
Depa men o En i onmen al Heal h Sciences, Mailman School o Public Heal h, Columbia Uni e si y, New Yo k, New Yo k, USA
39
Facul y o Heal h and Medical Sciences, Uni e si y o Su ey, Guild o d, UK
40
Wood ow Wilson School o Public and In e na ional A ai s and he Depa men o Ci il and En i onmen al Enginee ing, P ince on Uni e si y, P ince on,
New Je sey, USA
41
Wo ld Heal h O ganiza ion, Gene a, Swi ze land
42
ISGlobal, Cen e o Resea ch in En i onmen al Epidemiology (CREAL), Ba celona, Spain
43
Uni e si a Pompeu Fab a (UPF), Ba celona, Spain
44
CIBER Epidemiologia y Salud Publica (CIBERESP), Ba celona, Spain
45
IIASA, Laxenbu g, Aus ia
46
Ba le Facul y o he Buil En i onmen , Uni e si y College London, London, UK
47
Di ision o En i onmen al Heal h Sciences, Uni e si y o Cali o nia, Be keley, Be keley, Cali o nia, USA
48
Ci il and En i onmen al Enginee ing, Uni e si y o Wa e loo, On a io, Canada
49
Basque Cen e o Clima e Change (BC3), Leioa, Spain
50
Depa men o Epidemiology & Popula ion Heal h, London School o Hygiene & T opical Medicine, London, UK
51
Depa men o En i onmen al Heal h, Ha a d T.H. Chan School o Public Hea h, Bos on, Massachuse s, USA
52
Nicholas School o he En i onmen , Duke Uni e si y, Du ham, No h Ca olina, USA
53
The Na u e Conse ancy, A ling on, Vi ginia USA
54
Eu opean Cen e o En i onmen and Human Heal h, Uni e si y o Exe e Medical School, T u o, Co nwall, UK
55
PBL Ne he lands En i onmen al Assessmen Agency, The Hague, Ne he lands
56
School o En i onmen , Tsinghua Uni e si y, Beijing, China
57
Ins i u e o Global Heal h, Uni e si y College London, London, UK
58
En i onmen al Sciences & Enginee ing, Uni e si y o No h Ca olina a Chapel Hill, Chapel Hill, No h Ca olina, USA
59
The Na u e Conse ancy, New Ha en, Connec icu , USA
60
MRC Epidemiology Uni , Uni e si y o Camb idge, Camb idge, UK
61
Epidemiology and Bios a is ics, Uni e si y o Auckland, Auckland, New Zealand
62
School o Economics and Managemen , Beihang Uni e si y, Beijing, China
63
Beijing Ad anced Inno a ion Cen e o Big Da a-based P ecision Medicine, Beihang Uni e si y, Beijing, China
64
School o Public Heal h, Uni e si y o Sydney, New Sou h Wales, Aus alia
BACKGROUND:Modeling sugges s ha clima e change mi iga ion ac ions can ha e subs an ial human heal h benefi s ha acc ue quickly and locally.
Documen ing he benefi s can help d i e mo e ambi ious and heal h-p o ec i e clima e change mi iga ion ac ions; howe e , documen ing he ad e se
heal h effec s can help o a oid hem. Es ima ing he heal h effec s o mi iga ion (HEM) ac ions can help policy make s p io i ize in es men s based
no only on mi iga ion po en ial bu also on expec ed heal h benefi s. To da e, howe e , he wide ange o incompa ible app oaches aken o de elop-
ing and epo ing HEM es ima es has limi ed hei compa abili y and use ulness o policymake s.
OBJECTIVE:The objec i e o his effo was o gene a e guidance o modeling s udies on scoping, es ima ing, and epo ing popula ion heal h effec s
om clima e change mi iga ion ac ions.
METHODS:An expe panel o HEM esea che s was ec ui ed o pa icipa e in de eloping guidance o conduc ing HEM s udies. The p ima y li e a-
u e and a syn hesis o HEM s udies we e p o ided o he panel. Panel membe s hen pa icipa ed in a modified Delphi exe cise o iden i y a eas o
consensus ega ding HEM es ima ion. Finally, he panel me o e iew and discuss consensus findings, esol e emaining diffe ences, and gene a e
guidance ega ding conduc ing HEM s udies.
RESULTS:The panel gene a ed a checklis o ecommenda ions ega ding s akeholde engagemen : HEM modeling, including model s uc u e, scope
and scale, demog aphics, ime ho izons, coun e ac uals, heal h esponse unc ions, and me ics; pa ame e iza ion and epo ing; app oaches o unce -
ain y and sensi i i y analysis; accoun ing o policy up ake; and discoun ing.
DISCUSSION:This checklis p o ides guidance o conduc ing and epo ing HEM es ima es o make hem mo e compa able and use ul o policy-
make s. Ha moniza ion o HEM es ima es has he po en ial o lead o ad ances in and imp o ed syn hesis o policy- ele an esea ch ha can in o m
e idence-based decision making and p ac ice. h ps://doi.o g/10.1289/EHP6745
In oduc ion
In 2015, 196 coun ies ou lined hei Na ionally De e mined
Con ibu ions (NDCs) as pa o he 2015 Pa is Ag eemen o he
Uni ed Na ions F amewo k Con en ion on Clima e Change
(UNFCCC), wi h he goal o a oiding dange ous clima e change
o 2°C o e p eindus ial le els (The Pa is Ag eemen ). The sci-
en ific communi y has since ei e a ed ha holding wa ming
below 1.5°C could a oid key en i onmen al ipping poin s and
p e en subs an ial clima e- ela ed isks o na u al and human
sys ems (Rogelj e al. 2018;UNFCCC 1992). Cu en clima e
change mi iga ion effo s, howe e , all a sho . G eenhouse gas
emissions (GHGE) eached a eco d high in 2018 globally, which
also saw con inuing expansion o ossil uel use and educed a es
o enewable ene gy ins alla ion (IEA 2019). Fu he , he ope a-
ional guidelines ou lined in he Ka owice Package (UNFCCC
n.d.b) se ou he unde aking o a Global S ock ake ha will
make use o he “bes -a ailable science” o mi iga ion (and
o he Pa is Ag eemen hema ic a eas) (UNFCCC n.d.a). Fo
s udies o he heal h effec s o mi iga ion (HEM) o make a
meaning ul con ibu ion o his global effo , common s anda ds
a e needed o acili a e syn hesis h ough me a-analysis and
o he app oaches (Chang e al. 2017).
Al hough clima e change mi iga ion is an u gen p io i y,
policy implemen a ion has been limi ed, o en due o associa ed
sho - e m financial cos s (Wo kman e al. 2018). Howe e , cos
assessmen s a ely accoun o concomi an impac s (gene ally
subdi ided in o co-benefi s and co-ha ms), i.e., “ he ... effec s ha
a policy o measu e aimed a one objec i e migh ha e on o he
objec i es”(Allwood e al. 2014). Comp ehensi e accoun ing o
co-benefi s and ad e se side effec s is essen ial (Zenghelis 2006)
because beneficial mi iga ion ex e nali ies may enhance he eco-
nomic case o pu suing agg essi e mi iga ion ac ion (Rogelj e al.
2018). (Haines e al. 2009). Sys ema ic e iews o HEM s udies
es ima ing ac ions, policies, in e en ions, and echnologies ha e
ound ha mi iga ion can lead o nea - e m changes in local and e-
gional ai quali y, anspo a ion beha io s, and die a y in ake ha
ha e significan heal h benefi s, wi h b oade benefi s acc uing in
he longe e m (Chang e al. 2017;Gao e al. 2018) (See Figu e 1;
no e his is no in ended o be comp ehensi e). In mos cases hese
heal h effec s a e beneficial, and hei economic alua ion could
En i onmen al Heal h Pe spec i es 115001-2 128(11) No embe 2020
offse a subs an ial p opo ion o mi iga ion cos s, pa icula ly in
eme ging economies (Smi h and Haigle 2008).
Despi e his, HEM esea ch appea s o ha e had limi ed pol-
icy impac . E idence o mi iga ion co-benefi s ou side o heal h,
e.g., o de elopmen (economic and scien ific ad ancemen ) and
o bene olence (a mo e mo al and ca ing communi y) has mo i-
a ed suppo o mi iga ion ac ions (Bain e al. 2016). Bu HEM
esea ch has gained minimal poli ical ac ion and o da e has
done li le o influence clima e change policy (Wo kman e al.
2018). Influencing policy making wi h esea ch e idence is a
complex p ocess (Cai ney and Oli e 2017;Oli e and Cai ney
2019). Expe ience wi h o he popula ion heal h challenges such
as ai quali y and obacco abuse sugges s ha impo an elemen s
include de eloping a consis en and compelling body o e i-
dence, linking esea ch wi h a coali ion o ne wo k app oach o
ad ocacy, being p epa ed o ake ad an age o windows o op-
po uni y, and de eloping an unde s anding o how he ideologi-
cal belie s o decision make s influence he aming o policy
p oposals (Rose e al. 2017;Smi h 2013).
Inc easing he policy influence o HEM esea ch necessi a es
educing po en ial ba ie s and p omo ing po en ial acili a o s o
i s up ake. Two p e iously iden ified ba ie s a e esis ance om
powe ul es ed in e es s and s uc u al challenges ha limi c oss-
sec o al collabo a ion be ween he heal h sec o and clima e
change decision make s (Wo kman e al. 2018). Ano he ba ie is
he lack o a compelling, ha monized e idence base because cu -
en e idence consis s p ima ily o he e ogeneously modeled HEM
es ima es. Al hough he e is b oad ag eemen abou he gene al
app oach o modeling HEM (Remais e al. 2014;Smi h and
Haigle 2008), he b oad a ay o mi iga ion ac ions and wide a i-
e y o specific modeling app oaches aken ha e p ecluded me a-
analysis (Chang e al. 2017;Gao e al. 2018). The e is cu en ly no
clea amewo k o ad ancing HEM esea ch me hods and esul s
epo ing, and compelling examples o how concomi an impac s
on heal h (beneficial and ad e se) ha e acc ued om mi iga ion
ac i i ies a e s ill needed. In sum, basic guidance ha would
enhance he policy u ili y o HEM es ima es a e lacking, including
o scoping, es ima ing, and epo ing esea ch on HEM.
Wi h hese conce ns in mind, he Wellcome T us , he Wo ld
Heal h O ganiza ion (WHO), and he Uni e si y o Washing on
cocon ened a wo kshop o iden i y s a egies o enhancing he
policy u ili y o HEM es ima es, wi h a ocus on de eloping guid-
ance ha could help imp o e he applicabili y and compa abili y
o HEM e idence. Sho comings in cu en esea ch and epo -
ing p ac ices we e iden ified, and a consensus p ocess was unde -
aken o a i e a a minimal se o epo ing guidelines o
p omo e anspa ency, in e p e abili y, and up ake o HEM es i-
ma ion. He ein we p esen he me hods used o gene a e consen-
sus ega ding hese issues as well as he esul ing p ac ice and
epo ing guidance ha eme ged.
Me hods
Se e al me hods ac oss mul iple s ages we e used o de elop his
guidance, including a e iew and syn hesis o he li e a u e, as-
sembly o an expe panel, a modified Delphi consensus-building
p ocess, an expe wo kshop o de elop p elimina y guidance,
and pos wo kshop guidance efinemen (Figu e 2).
Li e a u e Re iew
A li e a u e e iew was conduc ed o p o ide ma e ial o he
expe panel and he modified Delphi p ocess. The e iew co e ed
Figu e 1. A concep ual amewo k ha p o ides sec o -specific examples o he mul iple mechanisms and pa hways h ough which clima e change mi iga ion
policies can affec human heal h. This figu e is no in ended o be comp ehensi e.
Figu e 2. Timeline and p ocess o de eloping guidance documen . R1, R2, and R3 we e successi e ounds.
En i onmen al Heal h Pe spec i es 115001-3 128(11) No embe 2020
HEM s udies published be ween 1 Feb ua y 2017 and 1 Feb ua y
2019, building on p io e iews o s udies conduc ed be o e ha
pe iod (Chang e al. 2017;Gao e al. 2018). Pape s we e classified
acco ding o mi iga ion scena io and sec o , geog aphic scale,
heal h me ics used, heal h ou comes es ima ed pe CO2equi alen
a e ed, mone ized benefi s, and ea men o unce ain y. The
pape s iden ified we e p o ided o he expe panel, along wi h a
na a i e syn hesis o he e iew findings.
Expe Panel Assembly
An expe panel was assembled based on esea che s iden ified in
he li e a u e e iew. Expe s we e con ac ed ia elec onic mail
and asked o pa icipa e in he panel. Fi s con ac was made wi h
fi s and senio au ho s; subs i u ions we e allowed upon mu ual
ag eemen be ween he o ganizing eam and he ini ially con-
ac ed expe . Panelis s we e asked o pa icipa e only i hey
could engage h ough he en i e p ocess. Six y- ou panelis s
we e ec ui ed. All panel membe s we e in i ed o he wo kshop
and o pa icipa e in edi ing he guidance documen . All pa ici-
pa ing panelis s a e included as au ho s.
Modi ied Delphi P ocess
The Delphi p ocess is a gene al me hod designed o measu e and
each consensus. Ou modified Delphi consis ed o h ee ounds
o anonymous, online su eys wi h wo ca ego ies o p edefined
ques ions, simila o hose o o he Delphi p ocesses (Boulkedid
e al. 2011). The fi s ca ego y consis ed o affi ma i ely ph ased
decla a i e s a emen s ela ed o HEM es ima ion p ac ices and o
he po en ial u ili y o guidance om he EQUATOR Ne wo k
(an in e na ional ini ia i e o p omo e epo ing guidelines o
published heal h esea ch) (EQUATOR Ne wo k n.d.) and else-
whe e on modeling s udies (Benne and Manuel 2012;S e ens
e al. 2016), obse a ional s udies ( on Elm e al. 2007), and
heal h impac assessmen s (HIAs) (Bha ia e al. 2014). Panelis s
we e asked o exp ess he deg ee o ag eemen on a scale o 1
o 9, wi h 1 indica ing s ong disag eemen and 9 indica ing
s ong ag eemen , and o p o ide a na a i e jus ifica ion o hei
esponses. The second ca ego y o ques ions was explo a o y and
mean o p io i ize opics o u he discussion and o gene a e
addi ional a eas in which o seek consensus. Be ween su ey
ounds, panelis s we e gi en desc ip i e analyses o esponses
and summa ies o na a i e commen s om he p io ound. Each
su ey is included in he Supplemen a y Ma e ials. Panelis s also
had ull, anonymized access o all aw da a and esponses.
Consensus was de e mined by measu es o cen al endency and
sp ead in he hi d ound o ques ioning, wi h a mean sco e and
in e qua ile ange (IQR) be ween 1 and 3 indica ing consensus
agains and a mean and IQR be ween 6 and 9 indica ing consen-
sus o a gi en s a emen .
Guidance Re inemen
Panelis s con ened in London in Ma ch 2019 in a wo kshop spon-
so ed by he Wellcome T us and cocon ened wi h WHO and he
Uni e si y o Washing on. The o e a ching goal was o each con-
sensus on key issues su ounding heal h co-benefi s esea ch.
Building om and guided by he modified Delphi p ocess, he 3D
wo kshop used small g oup discussions combined wi h epo -
back p esen a ions o discuss key elemen s o he su ey and, when
possible, each consensus. In o al, 53 esea che s encompassing a
ange o expe ise on clima e change mi iga ion heal h co-benefi s
esea ch a ended he wo kshop (Table S1). Small g oup discus-
sions we e mode a ed by p eselec ed acili a o s, and discussion
no es we e aken and summa ized by appo eu s.
Small g oup discussions we e held on he ollowing issues:
s akeholde engagemen , modeling app oaches, model pa ame e -
iza ion and ea men o unce ain y, and epo ing. Due o he
cu en s a e o he li e a u e, which disp opo iona ely ocuses
on ai pollu ion and ac i e anspo , wo kshop discussions we e
la gely ocused on hese a eas, wi h a seconda y ocus on ood
sys ems, he buil en i onmen , and u ban o m. The wo kshop
concluded wi h a plena y discussion o consensus guidance s a e-
men s, including he ype o s udies o which he guidance
should be designed. The guidance s a emen s we e assembled
in o a d a documen a e he wo kshop and ci cula ed o he
g oup, including hose panelis s who we e unable o a end he
wo kshop, o e iew and efinemen p io o publica ion.
Discussion
He e we p esen he consensus findings o he au ho panel
eflec ing he ou pu s om he expe consensus p ocess ou lined
abo e. We fi s p esen “P ac ice and Repo ing Guidance”and
hen conclude wi h a “Discussion”sec ion eflec ing he pe spec-
i e o he panel ega ding he con ex o his guidance and i s
implica ions o HEM es ima ion and epo ing.
P ac ice and Repo ing Guidance
The ma e ial in his sec ion is he consensus ecommenda ion o
he panel, unless no ed in he ex o o he wise a ibu ed wi h a
ci a ion. This s a emen begins wi h a b ie discussion o guid-
ance scope and applica ion. Nex , he e a e h ee sec ions o ec-
ommended guidance: “S akeholde Engagemen ,”“Modeling
App oaches,”and “Pa ame e iza ion and Repo ing.”A sche-
ma ic o he o e all p ocess o which he guidance applies is p e-
sen ed in Figu e 3.
Scope and Applica ion
Gi en he wide ange o HEM esea ch, i is impo an o delin-
ea e he ac i i y o which his guidance is designed and how he
guidance is in ended o be implemen ed.
As no ed ea lie in he “In oduc ion,”p io esea ch on
HEM de e mined ha mi iga ion ac ions may esul in subs an-
ial heal h co-benefi s, bu he policy po en ial o his esea ch
has no been ully achie ed, pa ially as a esul o wide a i-
abili y in me hods and epo ing as well as he nonsys ema ic
na u e o p e ious app oaches (Remais e al. 2014;Gao e al.
2018). This guidance is mean o encou age specificmodeling
p ac ices, me hods, and esul s epo ing o maximize he likeli-
hood o implemen ing he mos beneficial mi iga ion ac ions
and maximize he likelihood o a oiding ad e se side effec s.
As such, his guidance is pa icula ly ele an o HEM esea ch
ocused on quan i a i e es ima ion o mi iga ion heal h co-
benefi s ia modeling o popula ion heal h impac s associa ed
wi h specific mi iga ion ac ions. I is likely o less ele ance
whe e he p ima y aim o a policy is no clima e change mi iga-
ion, bu bo h heal h and ca bon impac s a e modeled.
P ac ically, hese guidelines a e likely o be mos ele an o
modeling effo s scoped o na ional and in e na ional effo s,
bu hey could also be applied subna ionally in la ge coun ies
such as he Uni ed S a es, China, and India. I may be mo e
challenging o implemen he guidance a local le els due o
esou ce cons ain s o lack o ele an downscaled modeling
inpu s.
The guidance is mean o be comp ehensi e and p o ide a
checklis o s anda d p ac ices and analysis epo ing ha can be
used by unde s, esea che s, e iewe s, and jou nal edi o s in
scoping, p epa ing o , conduc ing, epo ing, and e iewing
HEM esea ch. Guidance ecommenda ions a e lis ed below,
En i onmen al Heal h Pe spec i es 115001-4 128(11) No embe 2020
p eceded by accompanying (and less p esc ip i e) backg ound
ex . Al hough we encou age ollowing he guidance o maximize
compa abili y ac oss s udies and policy impac , we a e no ad o-
ca ing ha adhe ence be equi ed as a es o s udy quali y o
asse ing ha s udies ha do no adhe e o he guidelines neces-
sa ily lack quali y o igo . Fu he , we acknowledge he need o
flexibili y in modeling app oaches, o a oid s ifling bo h inno a-
ion and esea ch ou pu om esou ce-poo se ings whe e ull
adhe ence o his checklis may no be easible. Finally, we
acknowledge he impo ance o good modeling p ac ice, includ-
ing he need o main ain a humble app oach o modeling complex
sys ems, o equen ly e isi modeling assump ions, and o con-
sis en ly seek o inco po a e new e idence.
S akeholde Engagemen
As no ed in he HIA li e a u e (Bha ia e al. 2014) and o he pe i-
nen wo k on en i onmen al heal h policy no ed abo e (Rose e al.
2017;Smi h 2013), a undamen al p e equisi e o policy ele-
ance is s akeholde engagemen , pa icula ly wi h policymake s,
du ing esea ch planning o imp o e use ulness o model inpu s,
imp o e ele ance o s udy design o policy con ex , and encou -
age policy up ake. The panel s ongly encou ages use o ansdis-
ciplina y and in e disciplina y s akeholde eams (including
nonheal h disciplines) o eflec he b oad ange o me hodologi-
cal, sociological, and o he conside a ions. Modeling effo s
should conside including go e nmen policy make s, he public
and ci il socie y, he p i a e sec o , and scien is s om a a ie y
o ele an disciplines (e.g., heal h, ma hema ics, s a is ics, ecol-
ogy, economics, and o he social sciences), as needed. Such
effo s could help lessen ba ie s o policy up ake and allow
esea che s, decision make s, and unde s o an icipa e policy e-
sis ance; encou age policy up ake; and accoun o a iable
up ake o policy scena ios in analysis. S akeholde engagemen
may ake mul iple o ms (Bha ia e al. 2014), employ a ange o
ools (e.g., mapping, e iew o specific cases), and occu o e a
a ie y o adminis a i e and geog aphic scales. Expe ise may
need o be added o he esea ch eam on effec i e app oaches o
acili a ing s akeholde engagemen .
Wi h hese conside a ions in mind, he panel ecommends he
ollowing Engagemen Guidelines:
1. Speci y he p ima y decision make (s) and/o a ge audi-
ence(s).
2. Lis and desc ibe s a egies used o acili a e s akeholde
engagemen and communica ion wi h he a ge audience
h oughou he HEM de elopmen and dissemina ion
p ocess.
3. As ele an , desc ibe in ol emen wi h knowledge b oke s,
esea ch up ake office s, policy ad iso s, and o he s ake-
holde s a mul iple s ages o planning and implemen a ion
(e.g., hei ole in selec ing model inpu s such as heal h
me ics).
4. As ele an , desc ibe collabo a ion(s) wi h s akeholde s
(whe he a he affec ed communi y le el, in academia, o
in go e nmen ) in heal h-de e mining sec o s (e.g., finance,
ene gy, anspo a ion, housing, indus y, ood sys ems,
and ag icul u e).
5. As ele an , desc ibe engagemen be ween mi iga ion and
adap a ion expe s ele an o a ious phases o effo , om
planning o implemen a ion and dissemina ion, including
any effo s o a oid unin ended consequences.
Modeling app oaches. The e is clea alue in diffe en mod-
eling app oaches, and ou guidance does no p esc ibe specific
modeling app oaches as long as he app oach allows o inco po-
a ion o baseline popula ion heal h in o ma ion, demog aphics,
and e alua ion o coun e ac uals (i.e., wha could happen wi h-
ou he mi iga ion scena io). HEM models gene ally ake he
Figu e 3. Engagemen , modeling, pa ame e iza ion, epo ing, and syn hesis conside a ions o heal h effec s o mi iga ion policies (HEM) s udies, building on
figu es o iginally published by Remais e al. (2014) and Gao e al. (2018).
En i onmen al Heal h Pe spec i es 115001-5 128(11) No embe 2020
app oach shown in Figu e 3 and d aw om a a ie y o me hods,
including compa a i e isk assessmen , complex mechanis ic
modeling, and mic oeconomic and beha io al modeling, hough
o he app oaches a e some imes employed (Remais e al. 2014).
Compa a i e isk assessmen app oaches allow o s anda diza-
ion and compa ison o isk, whe eas mechanis ic and beha io al
modeling can p o ide insigh s in o sys ems dynamics and ques-
ions o how indi iduals, households, and o he economic uni s
may espond o a ious policy op ions.
The panel ecommends ha modeling choices should be made
based on assessmen s o which app oach is mos app op ia e o
he esea ch ques ion o in e es and s akeholde inpu , as ele an .
We also ecommend ha coun e ac ual scena ios, exposu e–
esponse unc ions (ERFs), heal h me ics, and s udy ime ames
should be selec ed based on s akeholde inpu , he esea ch ques-
ion posed, and heal h ocus. In addi ion, we encou age modele s
o be explici ega ding he a ionale o hese selec ions. Fo
ins ance, al hough disabili y-adjus ed li e-yea s (DALYs) may be
app op ia e o some in e na ional s udies, quali y-adjus ed li e-
yea s (QALYs), mo ali y a es, and loss o li e expec ancy (LLE)
may, depending on he con ex and s akeholde needs, be mo e ap-
plicable a na ional and subna ional scales. Addi ionally, he con-
e sion o heal h me ics in o mone a y measu es is a aluable
app oach i conduc ed anspa en ly because i may be pa icula ly
compelling o policymake s and allows o he compa ison o
impac s ac oss se e al dimensions (e.g., financial cos s s. heal h
benefi s).
Fo global and egional assessmen s, he panel ecommends
ha ERFs should gene ally be selec ed om sys ema ic e iews
and me a analyses. Howe e , when local e idence is o p ima y
impo ance o s akeholde s o when he e is eason o belie e ha
e idence gene a ed elsewhe e may no be gene alizable o he
s udy se ing (e.g., applying ERFs om high-income coun ies o
low-income coun ies o applying in a diffe en clima e egion),
ERFs may be based p ima ily on local e idence. The STROBE
epo ing guidelines o obse a ional s udies ( on Elm e al.
2007) and PRISMA s a emen o sys ema ic e iews (Mohe
e al. 2009) can be use ul in selec ing which s udies o inco po-
a e. Time ho izons should be simila ly jus ified and selec ed
based on con ex . Fo example, ime ho izons o na ional s udies
could be made cong uen wi h na ional go e nmen al policy
cycles, whe eas ime ho izons o egional s udies could align
wi h he sus ainable de elopmen goals (Uni ed Na ions n.d.). In
any case, esea che s should clea ly desc ibe and jus i y he
pa ame e iza ion choice.
We encou age, whe e possible and ele an , he use o ensem-
ble clima e model p ojec ions ha ollow he Rep esen a i e
Concen a ion Pa hways (RCPs), he Sha ed Socioeconomic
Pa hways (SSPs), o o he globally ag eed-upon scena ios. We
also encou age cha ac e iza ion o he unce ain ies inhe en in
scena io choice, empe a u e p ojec ions, model s uc u e, and in-
e nal a iabili y (Dese e al. 2012;Hawkins and Su on 2009).
The RCPs include ou emissions pa hways de eloped p ima ily
by he in eg a ed assessmen modeling (IAM) communi y ( an
Vuu en e al. 2011). The SSPs a e a se o fi e e e ence scena -
ios ha cap u e plausible ends in “ he e olu ion o socie y and
na u al sys ems o e he 21s cen u y”by inco po a ing a ange
o socioeconomic and en i onmen al condi ions (O’Neill e al.
2014). Taken oge he , he SSPs and RCPs can be used o explo e
unce ain y in ou comes and p ojec ions, wi h one axis ep esen -
ing plausible socioeconomic and en i onmen al condi ions and
he o he ep esen ing emissions pa hways (O’Neill e al. 2014).
The use o he RCPs and SSPs would align he effo s o HEM
esea che s wi h hose o he la ge clima e modeling communi y,
al hough some HEM s udies will be scoped and scaled such ha
hese scena ios will no be pa icula ly ele an . Fo ins ance, in a
s udy aimed a e alua ing he heal h impac s o changes in ans-
po a ion policy a a ci y scale o o e a ela i ely sho (i.e., dec-
adal) ime span, nei he he RCPs no he SSPs a e likely o add
pe spec i es ha will be use ul o s akeholde s. The SSPs can
also be used, whe e ele an , o inco po a e o he d i e s o
heal h ou comes, such as u baniza ion and inequi y.
Wi h hese conside a ions in mind, he panel ecommends he
ollowing Modeling App oach Guidelines:
Mi iga ion Policies
1. Desc ibe mi iga ion policies and scena ios and hei ele-
an sec o s (e.g., finance, ene gy, anspo a ion, housing,
indus y, ood sys ems including ag icul u e).
Geog aphic A ea and Scale
2. Speci y he geog aphic scale (in e na ional, egional,
na ional, subna ional, ci y) and geog aphic a ea o in e es .
Popula ion and Demog aphic Conside a ions
3. Desc ibe popula ions (size and o he cha ac e is ics) o
model baseline and coun e ac ual scena ios.
4. I applicable, accoun o equi y and desc ibe socially and
economically ma ginalized popula ions.
5. I applicable, accoun o and desc ibe popula ions ha a e
mos likely o expe ience ad e se consequences o benefi
he mos .
6. Lis and desc ibe da a sou ces used o popula ion and de-
mog aphic p ojec ions.
7. Desc ibe he deg ee o cong uence be ween popula ion and
demog aphic da a and emissions p ojec ions in e ms o
ime ame, unde lying assump ions, and o he ac o s
deemed ele an .
8. Desc ibe why a ge popula ion and demog aphic choices
o he baseline and coun e ac uals a e app op ia e o he
policy decision(s) being conside ed.
9. Desc ibe how he analyses will accoun o p ojec ed de-
mog aphic changes.
10. Desc ibe whe he and how he analyses will accoun o
eedbacks be ween and wi hin models and ac oss d i e s
o popula ion/demog aphic p ojec ions. I eedbacks a e
no add essed, s a e why.
Coun e ac ual Scena ios
11. Desc ibe how exposu e(s) o he mi iga ion ac ion(s) and
ela ed downs eam exposu e(s) is assigned, including, as
app op ia e, he p opo ion o he popula ion ha is
exposed o e ime as a unc ion o implemen a ion.
12. Lis and desc ibe coun e ac ual scena ios, ensu ing ha
hese eflec cu en eali ies in he absence o s ong mi i-
ga ion policies. As app op ia e, lis and desc ibe o he mi i-
ga ion policies used in he model.
13. Desc ibe any po en ial co espondence be ween mi iga ion
and coun e ac ual scena ios wi h SSPs, RCPs, and/o
o he globally ag eed-upon scena ios, ei he quali a i ely
h ough na a i e linkages (e.g., i he mi iga ion and coun-
e ac ual scena ios a e cha ac e is ic o pa icula scena -
ios) o quan i a i ely (e.g., using specific combina ions o
RCPs and SSPs wi h nume ical co ela es o emissions,
demog aphic shi s, e c.).
14. Desc ibe da a sou ces used o coun e ac ual scena ios
( o example, emission scena ios).
En i onmen al Heal h Pe spec i es 115001-6 128(11) No embe 2020
Time F ames and Ho izons
15. Speci y and jus i y he baseline yea o da a sou ces used
in he model.
16. Clea ly define and jus i y ime ames and p ojec ed ime
ho izons.
Exposu e–Response Func ions
17. Desc ibe and jus i y exposu e– esponse unc ions used.
18. I applicable, desc ibe how exposu e– esponse unc ions
may a y among ulne able and disad an aged popula ions.
19. Desc ibe heal h esponse s udies used o modeling (e.g.,
desc ip ion o sample size, loca ion, ime ame) and jus i y
i heal h esponses we e no ob ained om sys ema ic
e iews o me a-analyses (e.g., use o local s udies).
Heal h Me ics
20. Define and jus i y me ic(s) o measu ing he heal h o
popula ions (e.g., heal h me ics such as DALYs, yea s o
li e los (YLL), yea s li ed wi h disabili y (YLD), mo al-
i y, LLE, hospi aliza ions, eme gency depa men isi s,
among o he s) app op ia e o he specified causal pa h-
ways and a ge audience/decision make s.
21. Desc ibe and jus i y da a sou ces used o heal h me ics
(examples as abo e).
Baseline Heal h Es ima es
22. Desc ibe baseline popula ion heal h es ima es.
23. Desc ibe and jus i y sou ces used o baseline heal h es i-
ma es and demog aphics (e.g., na ional i al s a is ics, he
Global Bu den o Disease s udy, e c.).
Pa ame e iza ion and Repo ing
We ecommend quan i a i e analysis o unce ain y despi e he
complica ions in ol ed in doing so. In ins ances whe e sou ces o
unce ain y canno be quan ified, we ecommend addi ional quali-
a i e discussions. HEM esea che s should also conduc sensi i -
i y analyses o quan i y (i possible, using alue o in o ma ion o
o he me hods) how inpu unce ain ies, model pa ame e s, and
model s uc u e d i e a iabili y in model ou pu s. Bo h sensi i i y
and unce ain y analyses should be conduc ed o he ollowing:
exposu e, economic alua ion o heal h effec s, he ole o adap a-
ion in exposu e–ou come associa ions, he magni ude and unc-
ional o m o exposu e–ou come associa ions, and o he inpu s
such as demog aphy and policy up ake. Resea che s a e also
encou aged o conside discoun ing in e en ion cos s and impac s
i app op ia e o he decision and decision-making con ex .
We encou age wo king wi h IAM and o he modeling com-
muni ies o de elop ealis ic a es o implemen a ion o key poli-
cies. IAM is a commonly used app oach in he clima e science
communi y o in eg a e aspec s o human sys ems (demog aphy,
ene gy use, and he economy, among o he ea u es) wi h aspec s
o he clima e sys em (Moss e al. 2010). I has allowed o he
de elopmen o emissions scena ios such as he RCPs ( an
Vuu en e al. 2011) and can be used o simula e eedback loops,
e alua e unce ain ies, and p o ide insigh in o he benefi s o cli-
ma e change mi iga ion (Moss e al. 2010).
Wi h hese conside a ions in mind, he panel ecommends he
ollowing Pa ame e iza ion and Repo ing Guidelines:
Heal h Ou come Repo ing
1. Repo elemen s o componen heal h me ics used (e.g.,
numbe o dea hs o disease cases, baseline a es, pe cen a ibu -
able, heal h e en s pe capi a, change in li e expec ancy).
2. Desc ibe ex en o which en i onmen al exposu es (e.g., ai
pollu an s), s. demog aphic change, esul ed in heal h impac s.
3. Conduc a sensi i i y analysis and desc ibe me hods used.
4. Desc ibe gene al s udy limi a ions quali a i ely (e.g.,
desc ibe po en ial sou ces o bias and likely impac on findings)
and quan i a i ely (e.g., desc ibe limi a ions in alid pa ame e -
iza ion o model).
5. Whe e easible, conduc quan i a i e unce ain y analysis,
pa icula ly alue o in o ma ion me hods o p io i ize da a and
esea ch needs.
6. Discuss sou ces o unce ain y ha could no be add essed
quan i a i ely.
7. Discuss po en ial and/o ac ual (social, poli ical) esis ance
o policy up ake ei he quali a i ely o quan i a i ely (e.g.,
h ough inco po a ion o a s ep unc ion o policy implemen a-
ion o e ime).
8. Discuss ad e se consequences o mi iga ion ac ions (po en-
ial, hypo he ical, o hose obse ed in esul s) ei he quali a i ely
o quan i a i ely.
Accoun ing o Va iable Policy Up ake
9. Diffe en ia e scena ios be ween hough expe imen s (“fi s
bes wo ld”), maximum c edible a es o implemen a ion (“sec-
ond bes wo ld”), and ealis ic a es o implemen a ion.
10. Desc ibe equi y impac s o policy up ake.
11. I possible, quali a i ely discuss nonlinea dynamics
(e.g., delays and h esholds) associa ed wi h in e en ion imple-
men a ion and expec ed ou comes using empi ical da a.
Discoun ing
12. Fo discoun ed alua ion es ima es, jus i y choice o fixed
s. a iable (e.g., ce ain y-equi alen ) a es, specific a es cho-
sen, and conduc sensi i i y analyses, including a leas a es o
0% (wi h a 100-y ime ho izon) and 3%.
Da a and Code T anspa ency
13. When legally and e hically possible, openly and publicly
sha e da a and code used o HEM s udies o acili a e model al-
ida ion o models by ex e nal esea che s, collabo a ion be ween
in es iga o s, and he p oduc ion o me a-analyses (including
h ough e ospec i e ha moniza ion), and new s udies. I he e
a e legal o e hical limi a ions o da a sha ing, hese should be ex-
plici ly s a ed.
Discussion
The guidance, which add esses s akeholde engagemen , model-
ing app oaches, pa ame e iza ion, and epo ing, is aimed a
quan i a i e modeling scoped o p o ide policy- ele an insigh s
ha acili a e he selec ion and implemen a ion o mi iga ion poli-
cies based bo h on mi iga ion po en ial and on an icipa ed heal h
impac s. This guidance may be mos ele an o effo s a he
na ional o in e na ional le el, al hough he guidance is also
likely ele an o subna ional effo s in la ge coun ies and o
se ings such as megaci ies. The ul ima e aim o he guidance is
o gene a e es ima es ha a e esponsi e o s akeholde needs and
ha can be compa ed and combined o suppo policy decisions.
This guidance includes se e al depa u es om p e ious
wo k. Fi s among hem is he emphasis on s akeholde engage-
men , which is a fix u e o o he policy-o ien ed ac i i ies such as
HIA (Lock 2000). This guidance eflec s and unde sco es he p i-
o i y o policy ele ance in HEM esea ch. Al hough we s op
sho o s ipula ing ha s akeholde s always be engaged in HEM
effo s as s ipula ed in HIA guidance (Bha ia e al. 2014),
En i onmen al Heal h Pe spec i es 115001-7 128(11) No embe 2020
consensus none heless eme ged among he panel membe s ha
s akeholde engagemen , including engagemen o po en ially
affec ed ci il socie y and communi y membe s and sec o s ou -
side heal h, would likely enhance e idence up ake in o policy.
Resou ces om he HIA p ac ice communi y (SOPHIA
S akeholde Engagemen Wo king G oup 2015;SOPHIA n.d.)
may be ele an o s akeholde engagemen in u u e HEM
effo s.
Ano he depa u e is he emphasis on using common scena -
ios in modeling effo s. Socioeconomic and de elopmen ends
will affec baseline mo ali y a es o e he coming decades, and
adop ing he SSPs as a s anda d se o e e ence pa hways can
enhance compa abili y ac oss s udies. Rele an quan ifica ions o
he SSPs a e a ailable (Dellink e al. 2015;Jiang and O’Neill
2017;Kc and Lu z 2017;Li e al. 2019;Ma angoni e al. 2017;
Nepal e al. 2019;Rao e al. 2019), and he e is inc easing appli-
ca ion in he heal h sec o (Ma kandya e al. 2018;Selle s and
Ebi 2017). Using he SSPs, Selle s p ojec s diffe en bu dens o
noncommunicable disease (NCD) mo ali y as a p opo ion o all
dea hs, pa icula ly in low- and middle-income coun ies
(LMICs), depending on de elopmen pa hways (Selle s 2020), a
ele an end o u u e baseline popula ion heal h es ima es in
HEM s udies.
A hi d depa u e is he encou agemen o inco po a e demo-
g aphic p ojec ions and mo e ealis ic assump ions ega ding pol-
icy implemen a ion. Many HEM s udies made he simpli ying
and un ealis ic assump ion ha he mi iga ion ac ion o in e es is
imposed on oday’s wo ld immedia ely and uni o mly wo ldwide,
wi h a cons an popula ion, despi e he p esence o a wide ange
o possible popula ion dis ibu ions o e coming decades, bo h in
e ms o numbe s o people (6.9–12.6 billion in he yea 2100,
depending on socioeconomic de elopmen pa hway) and in e ms
o he age s uc u es and heal h s a us o popula ions (Kc and
Lu z 2017).
Less o a depa u e, bu s ill an impo an p ac ice inno a ion,
is he emphasis on a) s anda dized me ics o heal h impac s, b)
mo e comp ehensi e causal pa hways linking exposu es and
heal h impac s, and c) inco po a ion o clima e change in o p o-
jec ed changes in exposu e. Ai pollu ion p o ides an example o
he impo ance o hese issues. As unde s anding o he wide
ange o heal h impac s o exposu e o fine pa icula e ma e
(PM2:5) con inues o inc ease, i has become clea ha some
ecen HEM s udies unde es ima ed he ancilla y heal h benefi s
o educing PM2:5exposu e. A 2019 e iew by fi e na ional aca-
demies o science and medicine concluded ha associa ions we e
unequi ocal be ween PM2:5and hea disease, s oke, ch onic ob-
s uc i e lung disease, lung cance , p ema u e bi h, demen ia,
and b ain de elopmen (Academy o Science o Sou h A ica,
B azilian Academy o Sciences, Ge man Na ional Academy o
Sciences Leopoldina, U.S. Na ional Academy o Medicine, U.S.
Na ional Academy o Sciences 2019), wi h addi ional ad e se
heal h ou comes epo ed in indi idual epidemiology s udies. A
2017 e iew o p ojec ions o heal h isks om ozone and pa ic-
ula e ma e unde diffe en clima e change scena ios concluded
ha clima e change is expec ed o wo sen ai quali y by changing
a mosphe ic p ocesses and chemis y, esul ing in inc eases in
mo ali y, wi h esul s diffe ing by egion, scena io, and o he
ac o s (O u e al. 2017). Since he p eindus ial pe iod, clima e
change al eady inc eased he global popula ion-weigh ed fine
pa icle (PM2:5) concen a ions by 5% and he nea -su ace ozone
concen a ions by 2% (Fang e al. 2013). P ojec ions should
ake in o accoun he complex in e ac ions be ween clima e
change and ai pollu ion, such as clima e change cons aining
imp o emen s in ai quali y in some egions (T ail e al. 2014).
Inco po a ing he in e ac ions among ai pollu ion, clima e
change, heal h, and demog aphy in o HEM analyses and exp ess-
ing hese impac s using s anda dized me ics inclusi e o mo bid-
i y and mo ali y would inc ease he accu acy and compa abili y
o he es ima es.
I implemen ed widely, we expec ha he ecommended guid-
ance would lead o inc eased me hodological ha moniza ion in
HEM s udies and he eby enhance compa abili y and syn hesis. As
se e al e iews o mi iga ion heal h co-benefi s li e a u e high-
ligh ed (Chang e al. 2017;Gao e al. 2018;Haines e al. 2009;
Remais e al. 2014;Smi h and Haigle 2008), he a ie y o me h-
ods used o quan i y mi iga ion heal h co-benefi s has limi ed he
ex en o which esul s can be syn hesized, pa icula ly h ough
me a-analysis. These e iews codified he HEM p ocess bu
s opped sho o p o iding p esc ip i e guidance. Implemen a ion
o his guidance should acili a e combining u u e HEM esea ch
es ima es using me a-analyses and o he app oaches such as mul i-
c i e ia decision analysis (Huang e al. 2011;Linko e al. 2006;
Wilson 2015;Yoko a and Thompson 2004;Zhang e al. 2016). We
ecognize ha he selec ion o modeling echniques will be
con ex -specific, a he disc e ion o s udy g oups, and scaled o a
le el ha is ele an o decision make s. Howe e , ha moniza ion
o modeling me hods whe e app op ia e and explici ly epo ing
he a ionale o modeling app oaches will acili a e syn hesis,
modeling a global scales, and g ea e policy impac , and allow o
in eg a ion wi h o he e idence om con olled and obse a ional
s udies. As a co olla y, he emphasis on open sou ce da a and code
sha ing o modeling, echoing o he ele an EQUATOR Ne wo k
guidance, will also acili a e compa ison, e ospec i e ha moniza-
ion, and unde s anding o he implica ions o me hodological
choices and un angling o diffe ences due o use o diffe en da a
sou ces, assump ions, and scena ios.
The panel no es ha his guidance will likely equi e addi ional
esou ces o implemen and build global capaci y o suppo s ake-
holde engagemen , inco po a ion o addi ional eam membe s,
b oade unce ain y analyses, and anspa ency ega ding sou ce
da a and model code. Iden i ying he needed esou ces may be a
challenge because HEM esea ch is al eady inadequa ely unded
(Ebi e al. 2009) and mo e limi ed in scope han models o he eco-
nomic cos s and benefi s o clima e change mi iga ion. An implica-
ion o he guidance ela ed o engagemen is ha addi ional
in es men s a e needed specifically o suppo ansdisciplina y and
in e disciplina y eams (e.g., unding o economis s on eams o
heal h esea che s; imp o ed s akeholde /decision make engage-
men ; s eng hening pa ne ships and capaci y o LMICs; inc easing
access o da a sou ces, pa icula ly in LMICs; and esea ch on
highly exposed and/o highly suscep ible popula ions wi h lowe
adap i e capaci y). Al hough some ede al unding suppo s such
effo s, such as he Na ional Science Founda ion’s Con e gence
Resea ch effo s, u he in es men s a e needed o upda e models
as new insigh s a e gained, o ins ance, in o causal pa hways.
In es men s by na ional unding agencies, ounda ions, and nongo-
e nmen al o ganiza ions would lead o be e -in o med mi iga ion
decisions, building a mo e esilien and sus ainable u u e.
Afinal poin bea s men ion. This guidance is pa icula ly ele-
an o la ge-scale HEM modeling effo s aimed a p io i izing
mi iga ion in es men s and highligh s he impo ance o conside -
ing implemen a ion in modeling mi iga ion ac ions. Ou expe
panel was unanimous in highligh ing he need o concomi an ,
pa allel discussion o implemen a ion, o include de eloping
implemen a ion case s udies and o he app oaches o building he
e idence base ela ed o ecogni ion, documen a ion, maximiza-
ion, and publiciza ion o mi iga ion heal h co-benefi s—bo h o
in o m HEM modeling effo s and o acili a e apid, effec i e
implemen a ion, o he policies ha ha e been assessed o effec-
i ely p omo e op imal popula ion heal h.
En i onmen al Heal h Pe spec i es 115001-8 128(11) No embe 2020
The guidance p o ided he e can, wi h modifica ion, be
applied o o he ques ions, such as he heal h effec s o s a egies
o educe u ban hea islands h ough g eening p og ams. As no ed
abo e, he po en ial ele ance o such in e en ions o local pop-
ula ion heal h and policy may jus i y hei s udy despi e a likely
smalle global mi iga ion impac .
The need o be e unde s anding, es ima ion, and epo ing o
he heal h effec s o mi iga ion is clea . These guidelines ha e been
de eloped o help imp o e he quali y o modeling s udies o
in o m mi iga ion policies. Imp o ing his e idence base is c i ical
o ensu ing heal h is ully conside ed in mi iga ion s a egies,
encou aging bo h mo e mi iga ion and be e design o mi iga ion
o minimize nega i e heal h effec s and maximize posi i e heal h
effec s. I is ou hope ha ha moniza ion o heal h co-benefi s s ud-
ies will help in o m be e policy and hus imp o e public heal h
ou comes.
Acknowledgmen s
This wo k was unded by he Ou Plane , Ou Heal h P og am
a Wellcome T us . The au ho s hank ou collabo a o s a
Wellcome T us and he WHO o hei suppo in con ening he
wo kshop ha culmina ed in his manusc ip .
Re e ences
Academy o Science o Sou h A ica, B azilian Academy o Sciences, Ge man
Na ional Academy o Sciences Leopoldina, U.S. Na ional Academy o Medicine,
U.S. Na ional Academy o Sciences. 2019. Ai Pollu ion and Heal h –A Science-
Policy Ini ia i e. Ann Glob Heal h 85(1):140, PMID: 31871903,h ps://doi.o g/10.
5334/aogh.2656.
Allwood JM, Bose i V, Dubash NK, Gómez-Eche e i L, on S echow C. 2014.
Glossa y. In: Clima e Change 2014: Mi iga ion o Clima e Change. Wo king
G oup III Con ibu ion o he Fi h Assessmen Repo o he In e go e nmen al
Panel on Clima e Change. Edenho e O, Pichs-Mad uga R, Sokona Y, Fa ahani
E, Kadne S, Seybo h K, Adle A, Baum I, B unne S, Eickemeie P, K iemann B,
Sa olainen J, Schlöme S, on S echow C, Zwickel T ,Minx JC, eds. New Yo k,
NY: Camb idge Uni e si y P ess.
Bain PG, Mil on TL, Kashima Y, Bilewicz M, Do on G, Ga ða sdó i RB, e al. 2016.
Co-bene i s o add essing clima e change can mo i a e ac ion a ound he
wo ld. Na u e Clim Change 6(2):154–157, h ps://doi.o g/10.1038/nclima e2814.
Benne C, Manuel DG. 2012. Repo ing guidelines o modelling s udies. BMC
Med Res Me hodol 12:168, PMID: 23134698,h ps://doi.o g/10.1186/1471-2288-
12-168.
Bha ia R, Fa hang L, Helle J, Lee M, O ens ein M, Richa dson M, e al. 2014.
Minimum elemen s and p ac ice s anda ds o heal h impac assessmen ,
Ve sion 3. Oakland, CA: HIA P ac ice S anda ds Wo king G oup. h ps://www.
co.cowli z.wa.us/Documen Cen e /View/13420/Minimum-Elemen s-and-P ac ice-
S anda ds- o -Heal h-Impac -Assessmen ?bidId= [accessed 2 No embe 2020].
Boulkedid R, Abdoul H, Lous au M, Sibony O, Albe i C. 2011. Using and epo ing
he Delphi me hod o selec ing heal hca e quali y indica o s: a sys ema ic
e iew. PLoS One 6(6):e20476, PMID: 21694759,h ps://doi.o g/10.1371/jou nal.
pone.0020476.
Cai ney P, Oli e K. 2017. E idence-based policymaking is no like e idence-based
medicine, so how a should you go o b idge he di ide be ween e idence and
policy? Heal h Res Policy Sys 15(1):35, PMID: 28446185,h ps://doi.o g/10.1186/
s12961-017-0192-x.
Chang KM, Hess JJ, Balbus JM, Buonoco e JJ, Cle eland DA, G abow ML, e al.
2017. Ancilla y heal h e ec s o clima e mi iga ion scena ios as d i e s o pol-
icy up ake: a e iew o ai quali y, anspo a ion and die co-bene i s modeling
s udies. En i on Res Le 12(11), h ps://doi.o g/10.1088/1748-9326/aa8 7b.
Dellink R, Cha eau J, Lanzi E, Magné B. 2015. Long- e m economic g ow h p ojec-
ions in he Sha ed Socioeconomic Pa hways. Glob En i on Change,
h ps://doi.o g/10.1016/j.gloen cha.2015.06.004.
Dese C, Knu i R, Solomon S, Phillips AS. 2012. Communica ion o he ole o na u-
al a iabili y in u u e No h Ame ican clima e. Na u e Clim Change 2(11):775–
779, h ps://doi.o g/10.1038/nclima e1562.
Ebi KL, Balbus J, Kinney PL, Lipp E, Mills D, O’Neill MS, e al. 2009. U.S. unding is
insu icien o add ess he human heal h impac s o and public heal h esponses
o clima e a iabili y and change. En i on Heal h Pe spec 117(6):857–862, PMID:
19590674,h ps://doi.o g/10.1289/ehp.0800088.
EQUATOR Ne wo k. n.d. Abou Us. The EQUATOR Ne wo k. h ps://www.equa o -
ne wo k.o g/abou -us/ [accessed 19 Decembe 2019].
Fang Y, Naik V, Ho owi z LW, Mauze all DL. 2013. Ai pollu ion and associa ed
human mo ali y: he ole o ai pollu an emissions, clima e change and me h-
ane concen a ion inc eases om he p eindus ial pe iod o p esen . A mos
Chem Phys 13(3):1377–1394, h ps://doi.o g/10.5194/acp-13-1377-2013.
Gao J, Ko a s S, Va doulakis S, Wilkinson P, Woodwa d A, Li J, e al. 2018. Public
heal h co-bene i s o g eenhouse gas emissions educ ion: a sys ema ic
e iew. Sci To al En i on 627:388–402, PMID: 29426161,h ps://doi.o g/10.1016/j.
sci o en .2018.01.193.
Haines A, McMichael AJ, Smi h KR, Robe s I, Woodcock J, Ma kandya A, e al.
2009. Public heal h bene i s o s a egies o educe g eenhouse-gas emissions:
o e iew and implica ions o policy make s. Lance 374(9707):2104–2114,
PMID: 19942281,h ps://doi.o g/10.1016/S0140-6736(09)61759-1.
Hawkins E, Su on R. 2009. The po en ial o na ow unce ain y in egional clima e
p edic ions. Bull Ame Me eo Soc 90(8):1095–1108, h ps://doi.o g/10.1175/
2009BAMS2607.1.
Huang IB, Keisle J, Linko I. 2011. Mul i-c i e ia decision analysis in en i onmen al
sciences: en yea s o applica ions and ends. Sci To al En i on 409(19):3578–
3594, PMID: 21764422,h ps://doi.o g/10.1016/j.sci o en .2011.06.022.
IEA (In e na ional Ene gy Adminis a ion). 2019. Renewable ene gy g ow h
wo ldwide s alled in 2018 a e wo decades o s ong expansion. h ps://
www.iea.o g/news/ enewable-capaci y-g ow h-wo ldwide-s alled-in-2018-
a e - wo-decades-o -s ong-expansion [accessed 2 No embe 2020].
Jiang L, O’Neill BC. 2017. Global u baniza ion p ojec ions o he Sha ed
Socioeconomic Pa hways. Glob En i on Change 42:193–199, h ps://doi.o g/
10.1016/j.gloen cha.2015.03.008.
Kc S, Lu z W. 2017. The human co e o he sha ed socioeconomic pa hways: popu-
la ion scena ios by age, sex and le el o educa ion o all coun ies o 2100.
Glob En i on Change 42:181–192, PMID: 28239237,h ps://doi.o g/10.1016/j.
gloen cha.2014.06.004.
Li X, Zhou Y, Eom J, Yu S, As a GR. 2019. P ojec ing global u ban a ea g ow h h ough
2100 based on his o ical ime se ies da a and u u e sha ed socioeconomic pa h-
ways. Ea h’s Fu u e 7(4):351–362, h ps://doi.o g/10.1029/2019EF001152.
Linko I, Sa e s om FK, Kike G, Ba chelo C, B idges T, Fe guson E. 2006. F om
compa a i e isk assessmen o mul i-c i e ia decision analysis and adap i e
managemen : ecen de elopmen s and applica ions. En i on In 32(8):1072–
1093, PMID: 16905190,h ps://doi.o g/10.1016/j.en in .2006.06.013.
Lock K. 2000. Heal h impac assessmen . BMJ 320(7246):1395–1398, PMID: 10818037,
h ps://doi.o g/10.1136/bmj.320.7246.1395.
Ma angoni G, Ta oni M, Bose i V, Bo gono o E, Cap os P, F icko O, e al. 2017.
Sensi i i y o p ojec ed long- e m CO
2
emissions ac oss he Sha ed
Socioeconomic Pa hways. Na u e Clim Change 7(2):113–117, h ps://doi.o g/
10.1038/nclima e3199.
Ma kandya A, Samped o J, Smi h SJ, Van Dingenen R, Piza o-I iza C, A o I, e al.
2018. Heal h co-bene i s om ai pollu ion and mi iga ion cos s o he Pa is
Ag eemen : a modelling s udy. Lance Plane Heal h 2(3):e126–e133, PMID:
29615227,h ps://doi.o g/10.1016/S2542-5196(18)30029-9.
Mohe D, Libe a i A, Te zla J, Al man DG, P isma G oup. 2009. P e e ed epo ing
i ems o sys ema ic e iews and me a-analyses: he PRISMA s a emen . PLoS
Med 6(7):e1000097, PMID: 19621072,h ps://doi.o g/10.1371/jou nal.pmed.1000097.
Moss RH, Edmonds JA, Hibba d KA, Manning MR, Rose SK, an Vuu en DP, e al.
2010. The nex gene a ion o scena ios o clima e change esea ch and
assessmen . Na u e 463(7282):747–756, PMID: 20148028,h ps://doi.o g/10.1038/
na u e08823.
Nepal P, Ko honen J, P es emon JP, Cubbage FW. 2019. P ojec ing global and e-
gional o es a ea unde he Sha ed Socioeconomic Pa hways using an
upda ed En i onmen al Kuzne s Cu e model. Fo T ees Li elihoods 10:387,
h ps://doi.o g/10.3390/ 10050387.
O’Neill BC, K iegle E, Riahi K, Ebi KL, Hallega e S, Ca e TR, e al. 2014. A new
scena io amewo k o clima e change esea ch: he concep o sha ed socio-
economic pa hways. Clim Change 122(3):387–400, h ps://doi.o g/10.1007/
s10584-013-0905-2.
Oli e K, Cai ney P. 2019. The dos and don’ s o in luencing policy: a sys ema ic
e iew o ad ice o academics. Palg a e Commun 5(1):21, h ps://doi.o g/10.
1057/s41599-019-0232-y.
O u H, Ebi KL, Fo sbe g B. 2017. The in e play o clima e change and ai pollu ion
on heal h. Cu En i on Heal h Rep 4(4):504–513, PMID: 29080073,h ps://doi.o g/
10.1007/s40572-017-0168-6.
Rao ND, Saue P, Gidden M, Riahi K. 2019. Income inequali y p ojec ions o he
Sha ed Socioeconomic Pa hways (SSPs). Fu u es 105:27–39, h ps://doi.o g/10.
1016/j. u u es.2018.07.001.
Remais JV, Hess JJ, Ebi KL, Ma kandya A, Balbus JM, Wilkinson P, e al. 2014.
Es ima ing he heal h e ec s o g eenhouse gas mi iga ion s a egies: add ess-
ing pa ame ic, model, and alua ion challenges. En i on Heal h Pe spec
122(5):447–455, PMID: 24583270,h ps://doi.o g/10.1289/ehp.1306744.
Rogelj J, Shindell D, Jiang K, Fi i a S, Fo s e P, Ginzbu g V, e al. 2018. Chap e 2:
Mi iga ion pa hways compa ible wi h 1.5°C in he con ex o sus ainable
En i onmen al Heal h Pe spec i es 115001-9 128(11) No embe 2020
