Selecting indicators to monitor and assess environmental health in a Portuguese urban setting : a participatory approach

In e na ional Jou nal o En i onmen al Resea ch and Public Heal h A icle Selec ing Indica o s o Moni o and Assess En i onmen al Heal h in a Po uguese U ban Se ing: A Pa icipa o y App oach Ma a Salgado 1,*, Ana C. L. Viei a 2, Anália To es 3and Mónica D. Oli ei a 2 1Ins i u e o En i onmen al Heal h (ISAMB), Facul y o Medicine, Uni e si y o Lisbon, 1649-028 Lisboa, Po ugal 2Cen e o Managemen S udies o Ins i u o Supe io Técnico (CEG-IST), Ins i u o Supe io Técnico, Uni e sidade de Lisboa, 1049-001 Lisboa, Po ugal; [email p o ec ed] (A.C.L.V.); [email p o ec ed] (M.D.O.) 3Valo sul, Es ação Me cado ias Bobadela, Pla a o ma Ribei inha CP Lisboa, 2696-801 São João da Talha, Po ugal; [email p o ec ed] *Co espondence: [email p o ec ed] Recei ed: 13 Oc obe 2020; Accep ed: 17 No embe 2020; Published: 19 No embe 2020   Abs ac : En i onmen al heal h (EH) is in luenced by complex in e ac ions be ween heal h and he buil and na u al en i onmen s, he e being li le esea ch on i s speci ici ies in u ban se ings. The use o sui able indica o s o moni o and assess EH is undamen al in in o ming e idence-based in e en ions a he local le el. A pa icipa o y app oach o selec ing indica o s o in o m he moni o ing and assessmen o EH in Lisbon is he ein conside ed. E idence de i ed om a sys ema ic e iew o li e a u e and da a om Lisbon and Po uguese da abases we e analyzed by 12 Po uguese expe s in indi idual semi-s uc u ed in e iews. The in e iews aimed a iden i ying ele an indica o s and impo an eme ging issues in he Lisbon u ban se ing. The ou pu s om he in e iews we e alida ed by a wo- ound Web-Delphi p ocess in which panelis s (22 expe s) om di e en a eas o expe ise exp essed hei iews ega ding he ele ance o he indica o s o he analysis o EH in u ban se ings. Se en een indica o s we e alida ed in he Web-Delphi p ocess. High pa icipa ion achie ed along his p ocess suppo s he iew ha his pa icipa o y app oach was use ul o alida ion. Resul s om he adop ed pa icipa o y app oach poin ou gaps in he collec ion o noise and mobili y indica o s da a and aise eme ging issues on housing indica o s ha equi e u he esea ch. The esul s also sugges he need o local ac ion o imp o e indica o s and ools in o de o help he moni o iza ion o EH in u ban con ex s. The adop ed pa icipa o y app oach can be eplica ed o o he Po uguese and Eu opean u ban se ings. Keywo ds: en i onmen al heal h indica o s; u ban se ings; pa icipa o y app oach; in e iew; Web-Delphi 1. In oduc ion The Uni ed Na ions 2030 Agenda o sus ainable de elopmen goals (SDGs) ecall he impo ance o placing a he op o he poli ical agenda he pu sui o good heal h and wellbeing (SDG-3) and he p omo ion o sus ainable ci ies and communi ies (SDG-11) [ 1 ]. Pu suing such goals equi es he use o heal h in o ma ion sys ems ha enable he moni o ing and assessing o heal h in mul iple dimensions, wi h such sys ems en ailing speci ici ies ac oss geog aphical se ings [ 2 , 3 ]. Wi hin such se ings, i is ecognized ha heal h ou comes ha e a s ong ela ionship wi h he buil and na u al en i onmen , wi h his link becoming mo e e iden in u ban a eas [ 4 , 5 ]. Al hough he dispa i ies be ween u ban– u al se ings ha e been na owing, esiden s in u ban a eas a e mo e a ec ed by In . J. En i on. Res. Public Heal h 2020,17, 8597; doi:10.3390/ije ph17228597 www.mdpi.com/jou nal/ije ph In . J. En i on. Res. Public Heal h 2020,17, 8597 2 o 16 physical ac o s [ 6 , 7 ]. As ci ies a e home o mo e han 60% o he wo ld’s popula ion [ 5 , 8 ], he g owing unde s anding o hese ela ionships in he way ci ies a e planned and buil and how his impac s heal h has e ol ed o e he yea s [ 9 ]. Nowadays, sus ainable ci ies ha e come o he o e on o en i onmen al heal h (EH) p io i ies [10]. The en i onmen al heal h ield comp ises he moni o iza ion and su eillance o buil and na u al de e minan s like ai quali y, noise, g een spaces a ea, and walkabili y, and is conce ned wi h he p e en ion o disease and wi h he c ea ing o heal h-suppo i e en i onmen s [ 11 , 12 ]. One o he mos demanding challenges o esea che s and policymake s wi hin he ield is o isola e buil and na u al en i onmen dimensions and hei in e ac ions and o be able o de ine an app oach o managing hese in e ac ions, in o de o c ea e alue. Associa ed wi h his challenge, comes he need o do so while collec ing popula ion’ in elligence h ough collabo a i e p ac ices and o suppo inno a ion in EH in e en ions [ 13 , 14 ]. EH indica o s a e pa icula ly impo an in moni o ing and assessing he en i onmen s a us and i s ends in u ban se ings [ 15 ], in o de o imp o e in e en ions, alloca e esou ces, and in es iga e he e ec i eness o heal h policies [ 16 ]. I is known ha he collec ion o moni o ing and su eillance da a is a key ask o many go e nmen al o ganiza ions and indus ies cha ged wi h ensu ing he wellbeing o he popula ion [ 17 ]. En i onmen al in e ac ions a e di icul o quan i y and e alua e, and o ga he e idence-based in o ma ion o decision-making is a om s aigh o wa d [ 10 , 18 ]. As i is essen ial o build e idence-based ools o he de elopmen o EH in e en ions, he e mus be a a ionale on how o selec he dimensions and indica o s o assess EH [ 19 ]. Ne e heless, using his ype o e idence o in o m he design and e alua ion o policies in u ban se ings is an unde - esea ched a ea [ 20 ]. Mason and Lindbe g [ 21 ] and Pine and Glon i [ 22 ] epo ed ha esea che s and policymake s alue da a and expe ise in he local con ex . The li e a u e also ecalls ha buil and na u al en i onmen indica o s mus ha e a p o en e ec on heal h and ha hei selec ion should be based on a sound app oach, so ha hey can become a esou ce in iden i ying po en ial isks o heal h [23]. Me hods o selec ing indica o s a y om s a is ical and causal models [ 15 ] o sys ema ic e iews o he li e a u e [ 24 , 25 ] and pa icipa o y p ocesses [ 26 – 29 ]. The use o pa icipa o y p ocesses in EH esea ch is especially impo an because o he complexi y o he EH concep and o he in e ac ions be ween EH de e minan s and heal h ou comes, as well as because o he impo ance o engaging expe s wi h di e en knowledge in exp essing hei iews and discussing local EH p oblems [ 30 , 31 ]. The pe cep ion o he complexi y and speci ici ies o u ban se ings also makes i c ucial o conside a a ie y o expe s’ iews o inc ease he p obabili y ha he selec ed indica o s will be conside ed c edible, scien i ic, and ele an o EH in e en ions [ 32 ]. Mul iple quali a i e me hods like in e iews, su eys, o Delphi ha e been used o unde s anding complex issues ela ed o EH analysis. In e iewing is a way o collec ing da a and can be used in a ious si ua ions o co e a wide ange o heal h opics [ 33 ]. Ye B ewe on [ 34 ] a gues ha in e iews can ha e low eliabili y due o hei openness o di e en ypes o bias. To o e come issues o he alidi y and eliabili y o he indings o an in e iew, o he me hods mus be applied o se e as a gua an ee o he pa icipan s’ pe o mance. Ano he quali a i e me hod widely used in heal h esea ch is he Delphi p ocess which cap u es expe s’ opinion and can be used o p omo e consensus [ 35 ]. The Web-Delphi is being inc easingly explo ed due o he use o echnological pla o ms ha ing he ad an age o no equi ing ace o ace con ac , emo ing geog aphical ba ie s, and allowing he in ol emen o a la ge numbe o expe s [ 36 ]. Delphi p ocesses ha e been used o iden i y c i e ia o cha ac e ize and di e en ia e models o pallia i e ca e in heal hca e se ices [ 27 ], o de ine co e a eas o EH [ 30 ], and o compa e scena ios in o de o imp o e u ban sus ainabili y de elopmen [ 37 , 38 ]. In his a icle, we p opose he use o a Web-Delphi p ocess as a ool o alida e he in o ma ion collec ed o a se o buil and na u al indica o s. Lisbon has been ecen ly selec ed as Eu opean G een Capi al 2020 wi h an ambi ious agenda o he nex decade, unde he mo o “Choose o e ol e: measu es o 2030” [ 39 ]. This agenda is ully aligned wi h he SDGs-2030 Agenda and implemen ing i s goals equi es se ing ou he da a o be In . J. En i on. Res. Public Heal h 2020,17, 8597 3 o 16 collec ed in he u ban con ex , o moni o EH and o assess he ex en o which goals a e achie ed. Acco dingly, he e is an oppo uni y and need o de elop ools ( o ins ance, based on dashboa ds and/o on composi e indica o s) ha enable he moni o iza ion and imp o emen o EH by local au ho i ies and indus ies. Thus, wi h a iew o collec ing key in o ma ion o imp o ing EH, his s udy explo es an in eg a ed pa icipa o y p ocess o in ol e EH s akeholde s and expe s in he selec ion o he ele an dimensions and indica o s om buil and na u al en i onmen de e minan s, in o de o moni o and assess EH in Lisbon, Po ugal. 2. Ma e ials and Me hods Fo he pu pose o selec ing ele an dimensions and indica o s ele an o he e alua ion o EH in Lisbon—in o med by e idence, by da a, and by he iews o mul iple EH s akeholde s and expe s—we designed a mul i-me hodology wi h h ee s ages (see Figu e 1 o an o e iew). The i s s age comp ises a sys ema ic e iew o he li e a u e ha ocuses on en i onmen al de e minan s in u ban se ings and which is desc ibed in de ail in Salgado e al. [ 20 ], along wi h a sea ch and collec ion o da a om Po uguese and Lisbon da abases ele an o assess EH in u ban se ings. In he second s age, he ou pu s o he li e a u e e iew we e used as an inpu o semi-s uc u ed in e iews. In he in e iews, a es ic ed panel o Po uguese expe s om ele an EH backg ounds made a de ailed analysis and e alua ion o he collec ed e idence and da a and discussed which dimensions and indica o s should be used o moni o and assess EH in Lisbon. In e iews we e selec ed as hey a e powe ul o elici ing na a i e and complex da a and allow esea che s o in es iga e expe s’ iews in g ea e dep h [ 33 ]. The hi d s age depa ed om he knowledge collec ed du ing he semi-s uc u ed in e iews, and consis ed o a wo- ound Web-Delphi p ocess o p omo e a commi men among a la ge panel o na ional expe s owa ds consensus ega ding he inclusion o he inal se o dimensions and indica o s in a ool o moni o and assess EH in Lisbon, as well as o collec ing insigh s and issues ela ed o hese indica o s. The Web-Delphi was used as a alida ion s age wi h he aim o p o iding complemen a y comple eness, plausibili y, c edibili y, and con i ma ion o he knowledge acqui ed in he semi-s uc u ed in e iews [ 40 ], while enabling expe s o in e ac in an anonymous o ma . S ages 1, 2 and 3 a e de ailed below. In . J. En i on. Res. Public Heal h 2020, 17, x 3 o 16 and/o on composi e indica o s) ha enable he moni o iza ion and imp o emen o EH by local au ho i ies and indus ies. Thus, wi h a iew o collec ing key in o ma ion o imp o ing EH, his s udy explo es an in eg a ed pa icipa o y p ocess o in ol e EH s akeholde s and expe s in he selec ion o he ele an dimensions and indica o s om buil and na u al en i onmen de e minan s, in o de o moni o and assess EH in Lisbon, Po ugal. 2. Ma e ials and Me hods Fo he pu pose o selec ing ele an dimensions and indica o s ele an o he e alua ion o EH in Lisbon—in o med by e idence, by da a, and by he iews o mul iple EH s akeholde s and expe s—we designed a mul i-me hodology wi h h ee s ages (see Figu e 1 o an o e iew). The i s s age comp ises a sys ema ic e iew o he li e a u e ha ocuses on en i onmen al de e minan s in u ban se ings and which is desc ibed in de ail in Salgado e al. [20], along wi h a sea ch and collec ion o da a om Po uguese and Lisbon da abases ele an o assess EH in u ban se ings. In he second s age, he ou pu s o he li e a u e e iew we e used as an inpu o semi-s uc u ed in e iews. In he in e iews, a es ic ed panel o Po uguese expe s om ele an EH backg ounds made a de ailed analysis and e alua ion o he collec ed e idence and da a and discussed which dimensions and indica o s should be used o moni o and assess EH in Lisbon. In e iews we e selec ed as hey a e powe ul o elici ing na a i e and complex da a and allow esea che s o in es iga e expe s’ iews in g ea e dep h [33]. The hi d s age depa ed om he knowledge collec ed du ing he semi- s uc u ed in e iews, and consis ed o a wo- ound Web-Delphi p ocess o p omo e a commi men among a la ge panel o na ional expe s owa ds consensus ega ding he inclusion o he inal se o dimensions and indica o s in a ool o moni o and assess EH in Lisbon, as well as o collec ing insigh s and issues ela ed o hese indica o s. The Web-Delphi was used as a alida ion s age wi h he aim o p o iding complemen a y comple eness, plausibili y, c edibili y, and con i ma ion o he knowledge acqui ed in he semi-s uc u ed in e iews [40], while enabling expe s o in e ac in an anonymous o ma . S ages 1, 2 and 3 a e de ailed below. Figu e 1. O e iew o he app oach used o selec he key dimensions and indica o s, om buil and na u al en i onmen de e minan s, o e alua e en i onmen al heal h (EH) in he Lisbon u ban se ing. 2.1. E idence and Da a Collec ion: Sys ema ic Re iew o Li e a u e and Sea ch on Po uguese Da abases The sys ema ic e iew o li e a u e epo ed in Salgado e al.[20] aimed a iden i ying he dimensions and indica o s wi hin he socioeconomic, buil en i onmen , na u al en i onmen , heal hca e, and beha io de e minan s ha impac on heal h in u ban se ings. Fo ha pu pose, he e iew used a sea ch s a egy based on he keywo ds “popula ion heal h”, “ci y OR ci ies OR own OR “me opoli an a ea” OR “u ban en i onmen ”, “indica o s OR de e minan s”. The pe iod Figu e 1. O e iew o he app oach used o selec he key dimensions and indica o s, om buil and na u al en i onmen de e minan s, o e alua e en i onmen al heal h (EH) in he Lisbon u ban se ing. In . J. En i on. Res. Public Heal h 2020,17, 8597 4 o 16 2.1. E idence and Da a Collec ion: Sys ema ic Re iew o Li e a u e and Sea ch on Po uguese Da abases The sys ema ic e iew o li e a u e epo ed in Salgado e al. [ 20 ] aimed a iden i ying he dimensions and indica o s wi hin he socioeconomic, buil en i onmen , na u al en i onmen , heal hca e, and beha io de e minan s ha impac on heal h in u ban se ings. Fo ha pu pose, he e iew used a sea ch s a egy based on he keywo ds “popula ion heal h”, “ci y OR ci ies OR own OR “me opoli an a ea” OR “u ban en i onmen ”, “indica o s OR de e minan s”. The pe iod co e ed in he sea ch was om 2008 o 2018 and sea ches we e pe o med in PubMed, Web o Science, Scopus and SciELO Po ugal da abases. Th ough sc eening o he li e a u e, 94 s udies we e selec ed and included o analysis. The e iew de ec ed case-con ol, coho s, c oss-sec ional and ecological s udies quan i ying causal in e ences ha p o ided di ec e idence o link an en i onmen al isk ac o , measu ed by he indica o s, o a heal h ou come [ 41 ]. The esul s om he sys ema ic e iew ha ocused on he physical de e minan s, i.e., buil and na u al en i onmen , we e summa ized and used o build an ini ial lis o po en ial dimensions and indica o s ele an o e alua e EH in u ban se ings wi hin he scope o his s udy. A sea ch in Po uguese na ional and local open-access da abases, he Ins i u o Nacional de Es a í s ica (Na ional Ins i u e o S a is ics, INE, Po ugal) [ 42 ], Lisbon ci y council [ 15 ], and PORDATA [ 43 ], was pe o med o collec da a and analyze a ailable in o ma ion o each indica o p oposed in he ini ial lis , and o inqui e abou he spa io- empo al a ailabili y o hese indica o s o he ci y o Lisbon. Acco dingly, a lis o hi y- ou indica o s g ouped in o nine dimensions was ga he ed. The indica o s and dimensions we e o ganized acco ding o he ca ego ies o buil o na u al en i onmen . Along wi h he lis o indica o s, a desc ip ion o each indica o was de ined, oge he wi h in o ma ion conce ning geog aphical a ailabili y and measu emen pe iodici y. 2.2. E idence and Da a Analysis and Assessmen : Semi-S uc u ed In e iews Gi en he complexi y o he e idence and he na u e o da a collec ed, ace- o- ace semi-s uc u ed in e iews we e conduc ed o he indica o s and o da a analysis and assessmen . This me hod was selec ed o wo easons. Fi s , he associa ed bidi ec ional communica ion be ween in e iewe and in e iewee helps in ob aining mo e de ailed and in-dep h in o ma ion. Taci in his easoning is ha in e iewees canno sha e a common ocabula y and some aspec s migh ha e o be cla i ied conside ing he di e en a eas o expe ise, imp o ing he alidi y o he esul s [ 44 , 45 ]. Second, i allows he asking o o he ques ions ou side he semi-s uc u ed ques ionnai e o cla i y da a assessmen and o ask ollow-up ques ions. This lexibili y in sha ing knowledge was deemed app op ia e o he con ex and helped gene a ing unexpec ed insigh s. Twel e semi-s uc u ed in e iews we e conduc ed wi h Po uguese expe s wo king in Lisbon, chosen o ep esen equi ably he a ie y o pe spec i es conce ning EH: expe s in buil en i onmen , na u al en i onmen , EH, and om na ional egula ion ins i u ions (see Table 1). In ol ed in e iewees had o ul ill a leas one o he ollowing c i e ia: (1) mo e han i e yea s o expe ience in EH- ela ed esea ch p ojec s o he Lisbon a ea; (2) ha ing publica ions ela ed o ai quali y, u ban sus ainabili y, enewable ene gy consump ion and p oduc ion, in luence o g een a eas on heal h, wa e quali y; o (3) being supe io echnicians om na ional egula ion ins i u ions. The expe s ecei ed a o mal in i a ion ia email along wi h a desc ip ion o he s udy and he p ocess o in o med consen . The expe s who consen ed o be in e iewed ecei ed he ques ionnai e empla e in ad ance (see Supplemen a y Ma e ial) wi h he pu pose o explo ing he ques ions mo e sys ema ically and comp ehensi ely, as well as o keep he in e iew ocused on he desi ed line o ac ion. The ques ionnai e consis ed o h ee pa s. The i s pa collec ed in e iewee’s analysis o he esul s on EH dimensions om he li e a u e. The second pa included ques ions ela ed o he buil and na u al en i onmen de e minan indica o s collec ed in na ional and local da abases. The hi d pa asked o opinions ega ding missing in o ma ion on indica o s ha should be included in o de o e alua e EH. In . J. En i on. Res. Public Heal h 2020,17, 8597 5 o 16 Table 1. P o ile o he pa icipan s in s ages 2 and 3 o he app oach used o selec he key dimensions and indica o s o e alua e EH in he Lisbon u ban se ing. S age 2 S age 3 I em In e iews 1s Round Web-Delphi 2nd Round Web-Delphi Gende , n(%) Female 6 (50%) 14 (48%) 10 (45%) Male 6 (50%) 15 (51%) 12 (55%) Main a ea o expe ise, n(%) Buil en i onmen 3 (25%) 5 (17%) 5 (23%) Na u al en i onmen 3 (25%) 9 (31%) 5 (23%) En i onmen al heal h 3 (25%) 10 (35%) 8 (36%) Na ional heal h egula ion ins i u ions 3 (25%) 5 (17%) 4 (18%) In he i s pa , he in e iewees a ed hei le el o ag eemen wi h he inclusion o each dimension used o e alua e EH in Lisbon making use o a i e-le el Like scale (SA: S ongly disag ee, D: Disag ee, N: Nei he ag ee no disag ee, A: Ag ee, SA: S ongly ag ee), wi h an ex a op ion o DNK/DNWA: do no know/do no wan o answe . In he second pa , he same a ionale was ollowed o a ing he le el o ag eemen , wi h he inclusion o he indica o s p e iously collec ed om he na ional da abases. The in e iewees we e also asked o analyze he ca ego iza ion o indica o s wi hin he dimensions, and o indica e which indica o s could be seen as edundan , i.e., which could be conside ed as cap u ing he same phenomena. In he hi d pa , he pa icipan s we e in i ed o s a e i any indica o s conside ed ele an o moni o and assess EH we e missing, ega dless o he a ailabili y o da a o Lisbon, and i he e was a possible s a egy o measu e hose indica o s. In he end, o each indica o o which da a was only a ailable a he na ional le el, in e iewees we e asked whe he he na ional da a could be used as a p oxy o local da a. The answe s we e w i en by he in e iewe on in e iew shee s. When p ocessing he in o ma ion ga he ed in all he in e iews, he dimensions and indica o s we e conside ed ele an i mo e han 50% o he in e iewees s a ed “s ongly ag ee” o “ag ee” and, a he same ime, did no ha e mo e han 25% o “s ongly disag ee” o “disag ee” esponses. Con e sely, indica o s wi h mo e han 75% o ei he “s ongly disag ee” o “disag ee” we e ejec ed. I any o he in e iewees choose “DNK/DNWA” o any ques ion, he answe was no conside ed o ha dimension o indica o . 2.3. Valida ion: Web-Delphi P ocess The use o he Web-Delphi me hod had a wo- old objec i e. Fi s ly, i aimed a ga he ing opinions o a wide ange o expe s wi h di e se backg ounds, and coming om a ious geog aphical loca ions, ega ding he ele ance o he indica o s o assessing EH in u ban se ings, in o de o alida e he esul s om he in e iews. Secondly, i exposed expe s o al e na i e iewpoin s om o he expe s, while allowing hem o in e ac and econside hei esponses. The wo- ound Web-Delphi was conduc ed be ween Feb ua y and Ma ch 2020. 2.3.1. Delphi Panel E idence sugges s ha knowledge gene a ion p ocesses can be imp o ed by ensu ing di e si y among he ec ui ed pa icipan s [ 46 ]. Fo he pu pose o his s udy, he Delphi panel membe s we e de ined as a g oup o membe s knowledgeable o and wi h wo k expe ience in he a eas being s udied and able o pa icipa e in he p ocess [ 47 ]. The panel was chosen based on a “snowball” e ec [ 48 ]. The in e iewees we e no included in he panel bu ecommended a leas wo o he expe s o pa icipa e in he Delphi p ocess. The panel included a a ie y o expe ise in ai quali y, In . J. En i on. Res. Public Heal h 2020,17, 8597 6 o 16 epidemiology, en i onmen al heal h, na ional ins i u ions o heal h egula ion, noise, public heal h, and u ban sus ainabili y. Based on hese c i e ia, 55 Po uguese expe s we e in i ed o pa icipa e in he Web-Delphi. The e is no gene al o speci ic ule o an op imal numbe o panelis s, wi h his numbe a ying in many Delphi s udies [ 49 – 51 ]. In his s udy, he ocus was on in ol ing pa icipan s wi h he e ogenous expe ise and co e ing all ele an EH pe spec i es, as well as including expe s om na ional and local ins i u ions [ 52 , 53 ]. The main a eas o expe ise o Delphi pa icipan s a e p esen ed in Table 1. 2.3.2. Web-Delphi Design The in o ma ion esul ing om he in e iews was aken as he inpu in o he Web-Delphi p ocess. I an indica o had been ag eed o inclusion by a majo i y (mo e han 50%) o in e iewees, ha indica o was conside ed in he Web-Delphi ounds. The ques ionnai es we e implemen ed ia a web pla o m (a ailable a www.welphi.com) [ 54 ]. As e idence sugges s ha a high numbe o ounds can esul in a lowe esponse a e [ 47 , 55 ], and Dalkey and B own [ 56 ] no ice ha he bes accu acy in esponse a es is ob ained wi h wo- ound Delphi i e a ions wi h he d op-ou inc easing a e wo i e a ions, a wo- ound p ocess design was adop ed. The p ocess s a ed wi h he online in i a ion o each panelis , con aining ins uc ions o access he Welphi pla o m, he in o med consen , in o ma ion ega ding he s udy, and he i s ound ques ionnai e. The pa icipan s we e asked o answe wi hin en wo kdays o ecei ing he email. Two eminde s we e sen o hose pa icipan s who did no comple e he su ey. The Welphi pla o m was used o communica e wi h he panelis s. Pa icipan s who did no comple e he ound du ing he deadline we e no in i ed o he second ound. In he i s ound, panel membe s we e asked o indica e hei le el o ag eemen o disag eemen wi h each indica o on a i e-le el Like scale (SD, D, N, A, SA), wi h he op ion o answe ing DNK/DNWA. Speci ic ules we e applied o deal wi h di e ences in opinion and o measu e he le el o ag eemen . The ule adop ed o he inclusion o an indica o , i.e., o ha e mo e han 50% o he panel membe s s a ing “s ongly ag ee” and “ag ee” (ag eemen ), and a he same ime ha ing less han 33% o answe s a ing as “s ongly disag ee” and “disag ee” (SA+A>50% and SD+D<33.3% ), we e app o ed by an absolu e majo i y. Addi ionally, an indica o ecei ing mo e han 75% o “SA” and “A” esponses was hen app o ed by quali ied majo i y. Con e sely, indica o s wi h mo e han 50% o ei he “s ongly disag ee” o “disag ee” esponses we e immedia ely ejec ed by absolu e majo i y. In he second ound, he ules o app o al and ejec ion, as applied in he i s ound, we e kep . The Web-Delphi ques ionnai e included he lis o 17 indica o s which esul ed om he semi-s uc u ed in e iews, along wi h in o ma ion ega ding he espec i e dimensions and spa io- empo al a ailabili y o Lisbon. The pa icipan s also had he oppo uni y o make commen s and sugges ions abou each indica o in a commen sec ion p o ided in he Welphi pla o m. In he second ound, panel membe s had he oppo uni y o see he esul s om he i s ound (bo h pa icipan s’ choices and commen s) and we e in i ed o keep o e ise hei answe s i hey conside ed his app op ia e. 2.3.3. S a is ical Analysis The Web-Delphi p ocess aimed a alida ing he esul s om he semi-s uc u ed in e iews by p omo ing a highe le el o ag eemen among he expe s owa ds a consensus, he e being ne e heless an expec a ion o di e gence o opinion ega ding ei he he inclusion o exclusion o some o he indica o s. To unde s and such di e ences in opinion, C onbach’s α was calcula ed o unde s and he ex en o which he e was alignmen among all panelis s. C onbach’s α has been used p e iously as an associa ion index o measu e he le el o ag eemen among panelis s [ 57 ] and was used du ing each ound o he Web-Delphi o es ima e he associa ion o he a ings in he unidimensional Like scale [ 58 ]. The meaning o he alue o C onbach’s α depends on he con ex , bu when C onbach’s α is close o 1.0, i means ha he pa icipan s’ choices a e s ongly associa ed, while alues close o 0 sugges ha he pa icipan s’ choices a e highly un ela ed o one ano he . A cu o o he C onbach’s In . J. En i on. Res. Public Heal h 2020,17, 8597 7 o 16 α o 0.70 was adop ed, aking he iew ha alues abo e 0.7 a e aligned wi h he inclusion o he indica o wi hin an en i onmen al heal h moni o iza ion ool [16]. 3. Resul s 3.1. E idence and Da a Collec ion: Sys ema ic Re iew o Li e a u e and Sea ch o Po uguese Da abases Based upon he e iew o he li e a u e, a summa y o he 34 indica o s ela ed o buil and na u al en i onmen de e minan s and wi h he po en ial o impac heal h in u ban se ings was selec ed (see Table 2). The selec ion was based on he causal in e ences epo ed mainly in he c oss-sec ional and coho s udies included in he e iew o he li e a u e, o ensu e he alidi y o he e idence wi hd awn. To complemen his e idence, an addi ional sea ch in na ional and local da abases was pe o med o acqui e spa io- empo al a ailabili y in o ma ion (see Supplemen a y Ma e ial) o indica o s measu ed in he Lisbon egion and in Po ugal. A desc ip ion wi h de ails ega ding he pe iodici y o measu emen , geog aphical le el and uni s o measu e was made o each indica o (see Supplemen a y Ma e ial). The indica o s o which in o ma ion is cu en ly a ailable o he na ional le el we e iden i ied, as well as he indica o s lacking da a. The e idence and da a collec ion p ocess iden i ied i e dimensions (g een spaces, housing, mobili y, sa e y, and sani a ion) 18 indica o s om he buil en i onmen ; and ou dimensions (ai quali y, noise, soil, and wa e quali y) and 16 indica o s o he na u al en i onmen . The indica o s we e g ouped in o hese dimensions and an ini ial lis o nine po en ial dimensions and 34 indica o s was p oposed o analysis in he nex s age (see Supplemen a y Ma e ial and Table 2). Table 2. Lis o he 34 indica o s e ie ed om he li e a u e and om da abases sea ch, and syn hesis o expe s’ opinions collec ed in he semi-s uc u ed in e iews ega ding he ele ance o indica o s in moni o ing and assessing EH in he Lisbon u ban se ing. Buil En i onmen De e minan Dimension Indica o Disag eemen (%) Ag eemen (%) Included G een Spaces Numbe o zoos, bo anic pa ks, and aqua iums * 75.0 0 No P o ec ed a ea su ace (km2)83.4 0 No P o ec ed a ea p opo ion (km2)75.0 0 No Numbe o ga dens and pa ks 75.0 0 No A ea o leisu e pa ks and ga dens (m2)0 83.3 Yes A ea o communi y ga dens (m2)0 83.3 Yes Housing Numbe o classic amily housing buildings * 83.3 0 No Numbe o buildings by geog aphic loca ion and ype * 75.0 0 No Mobili y Buil -up a eas (km2) * 66.7 0 No Road ehicles by ype and uel * 16.7 58.4 Yes T a ic acciden s wi h ic ims/1000 inhabi an s 66.7 0 No Cycle oads (km) 0 75.0 Yes Road ne wo k (km) 0 75.0 Yes P opo ion o esiden popula ion using indi idual mode o anspo while commu ing * 66.7 0 No P opo ion o esiden popula ion using public anspo mode du ing commu ing * 58.3 0 No Sa e y Police- epo ed c imes/1000 inhabi an s * 58.3 0 No Sani a ion U ban was e collec ion * 58.3 0 No Popula ion se ed by was ewa e d ainage sys ems * 58.3 0 No In . J. En i on. Res. Public Heal h 2020,17, 8597 8 o 16 Table 2. Con . Buil En i onmen De e minan Na u al En i onmen De e minan Dimension Indica o Disag eemen (%) Ag eemen (%) Included Ai quali y PM2.5 (pa icula e ma e ) 0 83.3 Yes PM10 (pa icula e ma e ) 0 83.3 Yes O3(ozone) 0 83.3 Yes NO2(ni ogen dioxide) 0 83.3 Yes NOx(ni ogen oxides) 8.33 75.0 Yes SO2(sul u dioxide) 0 83.3 Yes CO (ca bon monoxide) 0 83.3 Yes C6H6(benzene) 8.33 75.0 Yes Noise Exposu e o high noise 75.0 16.7 No Su e ing om noise * 66.7 25.0 No Lden (day-e ening-nigh equi alen sound le el) 0 83.3 Yes Ln(day-nigh equi alen sound le el) 0 83.3 Yes Wa e quali y “Água segu a” *,a 0 91.7 Yes Public wa e supply * 58.3 0 No Soil Soil o ganic ca bon con en *,b 100 0 No Soil sealing index *,c 75.0 16.7 No Legend: * indica o s o which in o ma ion is cu en ly a ailable only o he na ional le el; a: “ Á gua segu a” is he Po uguese indica o ha measu es he pe cen age o compliance wi h he pa ame ic alues es ablished in he legisla ion; b: amoun o o ganic ca bon capable o a ec ing g oundwa e ; c: pe cen age o unsealed soil. 3.2. E idence and Da a Analysis and Assessmen : Semi-S uc u ed In e iews The 12 semi-s uc u ed in e iews we e pe o med wi h expe s whose main expe ise ela ed o he buil en i onmen , na u al en i onmen , EH, and na ional heal h egula ion ins i u ions, as shown in Table 1. Each in e iew ook a median o 55 min ( ange 45–70 min). The i s pa o he in e iew ga he ed iews om pa icipan s ega ding EH dimensions o be used o moni o and assess EH in Lisbon. F om he se o nine dimensions, six dimensions eached a high le el o ag eemen —mo e han a 50% ag eemen a e—abou hei ele ance in e alua ing EH in Lisbon. O e hal o he in e iewees disag eed wi h he inclusion o he sa e y and sani a ion dimensions. Soil dimension was indica ed o be impo an in e alua ing EH in u al se ings, bu o ci ies, i was no seen as ele an o inclusion. In he second pa , he in e iewees s a ed hei ag eemen ega ding he indica o s collec ed in he na ional and local da abases sea ch. An ea ly inding a his s age was ha expe s we e e y conce ned abou he a ailabili y o he da a when s a ing hei ag eemen ega ding he inclusion and/o exclusion o he indica o s, e en i hey had been ins uc ed o conside indica o s o which da a migh no exis . Fo some indica o s, he in e iewees chose no o s a e hei opinion (DNK/DNWA). F om he ini ial lis o 34 indica o s, a o al o 16 (47%) we e conside ed as ele an o moni o ing and assessing EH in Lisbon by he majo i y o he in e iewees. The indica o “Public wa e supply” in he “Wa e quali y” dimension was conside ed as impo an only o ci ies a ec ed by d ough episodes, which is no he case o Lisbon. The “Build-up a ea” indica o was conside ed edundan in he ligh o o he indica o s included in he “G een spaces” and “Mobili y” dimensions. The “P o ec ed a ea su ace”, “P o ec ed a ea p opo ion” and “Numbe o ga dens and pa ks” indica o s we e also conside ed edundan . Indica o s measu ing he a ea o leisu e pa ks and communi y ga dens achie ed a highe ag eemen han he indica o s conside ed edundan (see Table 2). In . J. En i on. Res. Public Heal h 2020,17, 8597 9 o 16 The las pa o he in e iew ga he ed opinions ega ding missing in o ma ion ha should be included o moni o and assess EH, wi h six in e iewees p oposing an addi ional indica o — “ene gy po e y ulne abili y index”— o he housing dimension. Based on he e idence and in o ma ion collec ed du ing he semi-s uc u ed in e iews, a inal se o indica o s and dimensions was o ganized. Hence he buil en i onmen de e minan was di ided in o h ee dimensions—g een spaces, housing, and mobili y- and he na u al en i onmen de e minan in o ano he h ee dimensions—ai quali y, noise, and wa e quali y. These six dimensions encompassed a o al o 17 indica o s, six in he buil en i onmen de e minan s, and 11 in he na u al en i onmen de e minan s. This se o dimensions and indica o s was he inpu o he Web-Delphi p ocess in he alida ion s age. 3.3. Valida ion: Web-Delphi P ocess 3.3.1. Panel Pa icipa ion F om he 55 expe s in i ed o he Web-Delphi p ocess, 29 accep ed he in i a ion and comple ed he i s ound o he Web-Delphi ques ionnai e (53% esponse a e). In he second ound, 22 expe s comple ed he ques ionnai e (76% esponse a e). A high esponse a e was achie ed om he buil en i onmen expe s whe e all pa icipan s comple ed bo h ounds. The g oup o esponden s in he inal ound had a balanced numbe o expe s om all a eas o expe ise (see Table 1). 3.3.2. Indica o s A lis o 17 indica o s, g ouped in six dimensions, esul ed om he in e iews and was alida ed as ele an in moni o ing and assessing EH in Lisbon ci y, as cap u ed by he high ag eemen a e among expe s in he Web-Delphi. In summa y, en indica o s we e alida ed by absolu e majo i y (SA >50% and SD+D<33.3%): nine om he na u al en i onmen de e minan and one om he g een spaces dimension (“A ea o leisu e pa ks and ga dens”). Se en indica o s we e alida ed by quali ied majo i y (SA+A>50%): i e om he buil en i onmen and wo om he na u al en i onmen . The i e indica o s om he buil en i onmen included “A ea o communi y ga dens” om he g een spaces dimension, and ou indica o s in he housing and mobili y dimensions. The wo indica o s included om he na u al en i onmen we e he SO 2 and L den indica o s. Resul s show ha he la ge g oup alida ed he wo k aken om he e idence and da a collec ion and om he in e iews. The “Ene gy po e y ulne abili y index” and he “ Á gua segu a” indica o s eached he highes le el o ag eemen (Table 3). Table 3. Indica o s app o ed by majo i y decision ules a e he second ound, g ouped by buil and na u al en i onmen de e minan s. De e minan Dimension Indica o s Ag eemen Ra e (%) Inclusion By Buil en i onmen G een spaces A ea o leisu e pa ks and ga dens 63.6 Absolu e A ea o communi y ga dens 54.5 Quali ied Housing Ene gy po e y ulne abili y index 95.5 Quali ied Mobili y Cycle oads 86.4 Quali ied Numbe o oad ehicles 81.8 Quali ied Road ne wo k 77.3 Quali ied Na u al en i onmen Ai quali y PM2.5 86.4 Absolu e PM10 63.6 Absolu e O363.6 Absolu e NO263.6 Absolu e NOx59.1 Absolu e SO286.4 Quali ied CO 59.1 Absolu e C6H654.5 Absolu e Noise Lden 90.9 Quali ied Ln63.6 Absolu e Wa e quali y “Água segu a” 95.5 Absolu e In . J. En i on. Res. Public Heal h 2020,17, 8597 16 o 16 69. S osnide , H.; Kennedy, C.; Mon i, M.; Yip, F. Ru al and U ban Di e ences in Ai Quali y, 2008–2012, and Communi y D inking Wa e Quali y, 2010–2015—Uni ed S a es. MMWR. Su eill. Summ. 2017 ,66, 1–10. [C ossRe ] 70. Da Sil a, R.T.; Fleskens, L.; Van Delden, H.; Van De Ploeg, M. Inco po a ing soil ecosys em se ices in o u ban planning: S a us, challenges and oppo uni ies. Landsc. Ecol. 2018,33, 1087–1102. [C ossRe ] 71. Calzola i, C.; Ta occo, P.; Lomba do, N.; Ma chi, N.; Unga o, F. Assessing soil ecosys em se ices in u ban and pe i-u ban a eas: F om u ban soils su ey o p o iding suppo ool o u ban planning. Land Use Policy 2020,99, 105037. [C ossRe ] Publishe ’s No e: MDPI s ays neu al wi h ega d o ju isdic ional claims in published maps and ins i u ional a ilia ions. © 2020 by he au ho s. Licensee MDPI, Basel, Swi ze land. This a icle is an open access a icle dis ibu ed unde he e ms and condi ions o he C ea i e Commons A ibu ion (CC BY) license (h p://c ea i ecommons.o g/licenses/by/4.0/).