20K 2 0 1 1 4 13 0 1 2 0 0 -1 0 Dynamic simulation periods are specified in Time's definition. This is usually a list of numbers or labels, typically in some unit of time (days, weeks, months, etc.). Use the ÒDynamic()Ó function in your variables to perform dynamic simulation. [0,1,2] 28.6.2004 Jouni Tuomisto This is the version 4 of the model calculating risks and benefits of farmed salmon. (c) Copyright KTL (National Public Health Institute, Finland). Jouni Tuomisto 9. tamta 2004 20:14 jtue 11. Janta 2005 12:48 48,24 1,40,5,856,569,17 2,102,90,553,461 Arial, 13 0,Model Risks_from_farmed_an,2,2,0,1,N:\YEP\publications\tuomisto\Viljelylohi\Web20050111\TuomistoFarmedSalmon2004_4_model.ana 97,1,1,0,2,1,2794,4312,0 <a href="http://ytoswww/yhteiset/Huippuyksikko/Tutkimus/Viljelylohi/Materiaali/Viljelylohi.rmd">Reference Manager database</a> <a href="http://ytoswww/yhteiset/Huippuyksikko/Tutkimus/Viljelylohi/">Directory for data and models</a> cases/a Pollutant health risk is calculated assuming additivity between the pollutants. However, dioxin risks are not considered because they were not considered in Hites. After unit conversion, numbers are calculated for Western Europe as cases per year. Note that negative numbers mean increased risk unlike in previous versions of the model. var a:= -Risk_slope*pollutant_exposure/1000*western_europe; var b:= (if pollutant1='Dioxin' then 0 else a); var c:= sum(b,pollutant1); c 256,328,1 48,24 2,291,123,476,224 2,70,18,628,450,0,MEAN Graphtool:0 Distresol:10 Diststeps:1 Cdfresol:5 Cdfsteps:1 Symbolsize:6 Baroverlap:0 Linestyle:1 Frame:1 Grid:1 Ticks:1 Mesh:1 Scales:1 Rotation:45 Tilt:0 Depth:70 Frameauto:1 Showkey:1 Xminimum:0 Xmaximum:1 Yminimum:0 Ymaximum:1 Zminimum:0 Zmaximum:1 Xintervals:0 Yintervals:0 Includexzero:0 Includeyzero:0 Includezzero:0 Statsselect:[1,1,1,1,1,0,0,0] Probindex:[0.05,0.25,0.5,0.75,0.95] [Reg_poll,Recommendation] cases/a Numbers are calculated for Western Europe as avoided deaths per year. Note that positive numbers mean increased benefit unlike in previous versions of the model. -Erf_h*min([N3_exposure,benefit_limit])*western_europe1 384,328,1 48,24 2,95,7,589,375,0,MEAN [Recommendation,Salmon] cases/a Net health effect of pollutant cancer and omega-3 cardiac benefit. Eff_fish+eff_poll 320,400,1 48,24 2,32,31,726,259,0,CONF ktluser 11. tamta 2004 9:20 48,24 56,200,1 48,24 1,40,0,610,544,17 100,1,1,0,2,9,4744,6798,7 Epa_model 288,40,1 56,32 However, defending his study in an interview with IntraFish yesterday afternoon, David Carpenter of the State University of New York at Albany said that the differences between wild and farmed salmon PCBs levels are not insignificant. ÒI donÕt agree with [Gallo] and I donÕt think others would agree with him either.Ó Carpenter said that the EPA risk assessment model that he and his colleagues used to determine that salmon posed a health risk is a well-respected and sound one. ÒIt is not just something that we made up. It is a time-tested measureÉitÕs a yard stick that we have.Ó Epa_model 288,128,1 52,36 http://www.intrafish.com/articlea.php?articleID=41070&s=1 A set of four volumes that provides guidance for assessing health risks associated with the consumption of chemically contaminated non-commercial fish and wildlife. EPA developed the series of documents to help state, local, regional, and tribal environmental health officials who are responsible for developing and managing fish consumption advisories. Developed_for_spceci 528,192,1 56,28 http://www.epa.gov/waterscience/fish/guidance.html Developed_for_spceci 360,432,1 68,56 Applies_only_to_non_+Developed_for_spceci 528,296,1 64,56 200,296,1 48,38 [Constant Developed_for_spceci] Point_of_view_is_tha+This_is_not_a_public 528,432,1 64,56 2,102,90,476,373 Epa_model 528,80,1 72,64 Epa_model 360,296,1 84,56 [Constant Precautionary_princi] Epa_model+Fda_model 64,288,1 52,52 5 64,448,1 48,24 2,136,146,416,303,0,MIDM [] jtue 28. Junta 2004 18:03 48,24 64,80,1 48,29 1,40,0,517,300,17 U.S.EPA. Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisory. Volume 2: Risk Assessment and Fish Consumption Limits, 3rd Edition. 2000. Table 3-1. <a href="http://www.epa.gov/waterscience/fish/guidance.html">Open access Internet file</a> <a href="http://ytoswww/yhteiset/Huippuyksikko/Kirjallisuus/Fish_and_health/EPAFishAdvisory/">Intranet file</a> 2^(Floor(logten(Epa_model)/logten(2))) 56,176,1 48,24 2,48,219,743,303,1,MIDM [Location1,Undefined] meals/month CRmm variable in the U.S.EPA advisory model. index Effect:=['Cancer','Non-cancer']; var MS:= 0.227; var Tap:= 365.25/12; var CSF:=Potency[potency='Ca (CSF)']; var RfD:= Potency[Potency='Non-Ca (RfD)']; var CRlimCa:= ARL*BW/sum(CSF*in1*(Poll_salmon_hites/1000),Pollutant1); var a:= In1*RfD/(Poll_salmon_hites/1000); var b:= if isnan(a) then 1 else a; var c:= if b>0 then b else 1; var CRlimNonCa:= min(c,Pollutant1)*BW; var CRlim:= array(Effect,[CRlimCa,CRlimNonCa]); var CRmm:= CRlim*Tap/MS; CRmm 56,104,1 48,29 2,43,74,421,412 2,120,130,416,303,0,MIDM [] U.S.EPA. Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisory. Volume 2: Risk Assessment and Fish Consumption Limits, 3rd Edition. 2000. pages 3-5 - 3-24 <a href="http://www.epa.gov/waterscience/fish/guidance.html">Open access Internet file</a> <a href="http://ytoswww/yhteiset/Huippuyksikko/Kirjallisuus/Fish_and_health/EPAFishAdvisory/">Intranet file</a>. 1 56,32,1 48,24 Poll_salmon_hites Table(Pollutant1)( 1,1,0,1) 216,160,1 48,24 2,469,131,476,224 2,242,231,416,303,0,MIDM 1 216,104,1 48,24 Potency probability Acceptable risk level 10u 216,32,1 48,24 U.S.EPA. Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisory. Volume 2: Risk Assessment and Fish Consumption Limits, 3rd Edition. 2000. <a href="http://www.epa.gov/waterscience/fish/guidance.html">Open access Internet file</a> <a href="http://ytoswww/yhteiset/Huippuyksikko/Kirjallisuus/Fish_and_health/EPAFishAdvisory/">Intranet file</a> ktluser 11. tamta 2004 9:20 48,24 760,160,1 48,24 1,0,0,1,1,1,0,,0, 1,70,58,639,469,17 ['Dieldrin','Toxaphene','Dioxin','PCB'] 504,64,1 48,12 kg 70 504,392,1 48,24 1,1,1,1,1,1,0,0,0,0 ['Scotland','Faroe Islands','Frankfurt','Edinburgh','Norway','Paris','London','Oslo','East Canada','Boston','Maine','San Francisco','West Canada','Toronto','Los Angeles','Vancouver','Washington DC','Seattle','Chicago','New York','Washington St','Chile','SE AK Chinook','Denver','BC Chinook','BC Sockeye','Oregon Chinook','SE AK Sockeye','New Orleans','BC Coho','Kodiak AK Sockeye','SE AK Coho','Kodiak AK Coho','BC Pink','Kodiak AK Pink','SE AK Pink','SE AK Chum','BC Chum','Kodiak AK Chum'] 504,32,1 48,12 ['Farmed salmon','Wild salmon','Market salmon'] 504,96,1 48,12 20.1.2004 Jouni Tuomisto En ole ehtinyt aiemmin lokia kirjoittaa, joten nyt yleiskuvaus mallista. Viljelylohi on tehty arvioimaan, onko Hites et al (Science 9.1.2004) riskinarviointi hyvin tehty. Ensikommenttina KTL:ssŠ oli se, ettŠ kalan terveyshyšdyt on unohdettu. NiinpŠ rakensimme mallin, joka 1) kŠyttŠŠ samaa EPAn riskimallia saasteiden terveyshaittojen laskemiseen kuin Hites (PCB:n, dieldriinin ja toksafeenin (mutta ei dioksiinien) aiheuttama yhdistetty syšŠriski olettaen additiivisuuden ja linearised multistage-mallin eli suoraan kŠyttŠmŠllŠ EPAn CSF-arvoja) ja 2) laskee myšs omega-3-rasvahappojen tuoman hyšdyn sydŠnkuolemariskiin. Vertailu tehtiin 1) olettamalla lohensyšntiŠ 0.25 - 32 amerikkalaista annosta kuukaudessa ja laskemalla syšpŠriski ja/tai sydŠnhyšty sekŠ 2) olettamalla jokin lohensyšnti (esim. 20 g/d) ja lisŠksi jotain oletuksia muista omega-3-lŠhteistŠ sekŠ niiden muutoksista jos lohensyšnti muuttuisi. Vaikutuksen lisŠksi tehtiin argumenttianalyysi (oma moduli) jossa katsottiin importance analysis eli rank-korrelaatio lŠhtšmuuttujien ja lopputuleman vŠlille. TŠssŠ oli mukana erilaisia pŠŠtšksiŠ, mm. pitŠisikš katsoa saasteiden syšpŠhaittaa vai nettovaikutusta?, PitŠisikš katsoa terveysvastetta lainkaan vai pelkkŠŠ altistusta? ja MillŠ viljelty lohi pitŠisi korvata? PŠŠtškset otettiin mukaan analyysiin siten, ettŠ kullekin pŠŠtšsvaihtoehdolle oletettiin sama todennŠkšisyys, ja ne otettiin mukaan satunnaismuuttujina (ikŠŠn kuin me yrittŠisimme arvioida, mikŠ on ŠŠnestyksen tulos kun tŠstŠ ŠŠnestetŠŠn). TŠhŠn liittyen jŠin pohtimaan sitŠ, pitŠisikš meidŠn olettaa pienempi todennŠkšisyys huonoille vaihtoehdoille (kuten epŠtodennŠkšisille presidenttiehdokkaille annetaan vŠhemmŠn aikaa televisiossa) mutta en pŠŠtynyt tŠssŠ mihinkŠŠn lopputulokseen, ja niin tasajako jŠi malliin. LisŠksi on tehty VOI-analyysi (oma moduli). TŠssŠ yritin rakentaa VOI-funktiota, joka olisi suoraan laskenut mielenkiinnon kohteena olevan tuloksen (helpottaisi mallinrakennusta jatkossa ja tekisi erilaisten VOIn laskemisen kŠtevŠksi), mutta ongelmaksi muodostui se, ettŠ mean-funktio toimi oikein vain, kun se laskettiin variablesta. Jos yritti laskea sen tilapŠisestŠ, solmun sisŠllŠ olevasta muuttujasta, tuloksena oli yleensŠ mid. NiinpŠ tyydyttiin laskemaan homma kŠsipelillŠ kuten Particle VOI -mallissa. Conclusions from Hites 2004 sisŠltŠŠ sitaatteja ja argumentteja keskustelusta, joka on Hitesin myštŠ kŠynnistynyt. What should be the scope of the assessment oli aikeissa olla moduli, josta eri pŠŠtšsvaihtoehdot olisivat sinne kirjatun argumentoinnin seurauksena nousseet, mutta sitŠ ei ollut aikaa tyšstŠŠ kovin pitkŠlle. Confounder analysis -moduli sisŠltŠŠ pohdintaa siitŠ, millaiset tekijŠt voivat vaikuttaa Hitesin lopputulokseen ja kvalitatiivista argumentointia niiden mahdollisista vaikutuksista tulosten tulkintaan. Help on yksinkertaisesti kopio Help v4 mallista. 20.1. alkoi ongelmaksi tulla se, ettŠ malli ršnsysi liikaa, ja oli hankalaa saada indeksit tŠsmŠŠmŠŠn lŠhtšarvojen ja lopputuloksen kesken. NiinpŠ pŠŠtin tehdŠ uuden version 3, josta kaikki ršnsyt on poistettu ja jonka tarkoituksena on toimia mallina Science-artikkelia varten. Kaikki laajemmat tarkastelut siis sŠŠstetŠŠn mallin seuraaviin versioihin. NiinpŠ versio 2:een jŠtetŠŠn kaikki ršnsyt, josta niitŠ sitten voi tarpeen mukaan kopioida takaisin kŠytšssŠ olevaan malliversioon. NŠin ehkŠ pysyy selvŠnŠ se, mitŠ Science-vastineessa on ja mitŠ ei ole. Fishing and farming, Arguments on fish pollutants, Total pollutant exposure, What should be the scope of risk assessment?, Conclusions from Hites 2004, ja Confounder analysis ovat semmoiset modulit jotka nyt poistetaan versiosta 3. Samoin poistetaan pŠŠmodulista solmut Risk or net health effect?, Acceptable risk ja Health effects or exposures? sekŠ nŠiden input nodet. 0 504,128,1 48,12 2,463,67,476,399 65535,54067,19661 20.1.2003 Jouni Tuomisto Versiosta 3 siis on tehty riisuttu versio Science-juttua varten. Lue tarkemmin Loki v 2:sta. Nyt versiosta 3 poistetaan aiemmin kuvatun lisŠksi Other parts -modulista indeksejŠ, joita ei kŠytetŠ missŠŠn. NŠmŠ ovat Viljelyalue, Kalastusalue, Ostokaupunki, Lohilaji ja Saaste. Argument analysis -modulista poistetaan solmut We should not consider concentrations..., Acceptable exposure, Va2, Health or exposure?, What is salmon replacement?, Va5, Va3, Va3 inputs, Va3 importance eli kaikki solmut. TŠrkeyssolmu luodaan uudelleen, mutta nyt se voidaan tehdŠ suoraan Outcome-solmulle ilman indeksimuunnoksia. NiinpŠ koko Argument analysis -moduli poistetaan ja asia siirretŠŠn VOIs-moduliin, joka nimetŠŠn uudelleen VOI and importance analysis. Food intake -modulista poistetaan Salmon consumption, Salmon replacement, Food change, Salmon amount, Oil increase, Food intake, Wild salmon compensation, Va1, Food_rec, Source1. Eli kaikki solmut keskittyvŠt nyt vain loheen, eikŠ muita omega-3-lŠhteitŠ huomioida. Ne tulevat mukaan malliin raja-arvossa, joka kuvaa hyšdyllisen lisŠsaannin rajaa ja siis sisŠltŠŠ absoluuttisen fysiologisen rajan, josta on vŠhennetty muusta ravinnosta tuleva mŠŠrŠ. TŠmŠ solmu tehdŠŠn Annosvastemoduliin. VOIs-modulista poistetaan Va16, Va12, VOI, VOI1 ja VOI-laskenta tehdŠŠn suoraan Outcome-solmusta. 21.1.2004 Jouni Tuomisto Malli muuttui eilen siten, ettŠ nyt lasketaan VOI kahdelle eri kysymykselle: pitŠisikš suositella viljellyn lohen enimmŠissaanniksi 1 annos/kk ja pitŠisikš rajoittaa enemmŠn kalanrehun saastepitoisuuksia. NŠmŠ kaksi nostetaan esiin, koska edellinen on suora vastine Hitesin ym. argumenttiin, ja jŠlkimmŠinen on korostamassa sitŠ, ettŠ asetettu kysymys mŠŠrŠŠ sen, mikŠ tieto on tŠrkeŠŠ ja mikŠ ei. Other parts -modulista poistetaan solmut Acceptable exposure increase ja Amount or replacement, ja ARL siirretŠŠn Fish advirories -moduliin sekŠ Potency Exposure-response function for pollutant risk -moduliin. NŠistŠ moduleista poistetaan vastaavat aliakset. Unit- ja Description-kentŠt pŠivitetŠŠn koko mallissa, ja viitteitŠ listŠtŠŠn sikŠli kuin ne ovat helposti kŠsillŠ. Kuitenkin viitteet on vielŠ pistettŠvŠ kuntoon, nyt muotoilut eivŠt ole kunnossa. 0 504,152,1 48,12 2,212,144,476,344 65535,54067,19661 g/d eff_poll/Erf_h/western_europe 504,224,1 48,24 [Recommendation,Salmon] var a:= sum(eff_poll,salmon); Probability(a[recommendation='BAU']+0.0001>a[recommendation='Restrict farmed salmon use']) 504,336,1 48,24 2,115,372,476,291 var a:= sum(eff_net,salmon); Probability(a[recommendation='BAU']>a[recommendation='Change farmed to wild salmon']) 504,280,1 48,24 index a:= ['Net effect of salmon recommendation','Net effect of feed regulation','Cancer effect of recommendation']; var b:= array(a,[eff_net,0,eff_poll]); var c:= b[recommendation='Restrict farmed salmon use']-b[recommendation='BAU']; var d:= (if regulate_pollutants_=1 then c[reg_poll='More actions'] else c[reg_poll='BAU']); var e:= eff_net[reg_poll='More actions'] - eff_net[reg_poll='BAU']; var f:= (if recommend= 1 then e[recommendation='Restrict farmed salmon use'] else e[recommendation='BAU']); var g:= (if a= 'Net effect of feed regulation' then f else d); var h:= sum(g,salmon); h 232,264,1 48,24 2,452,264,476,337 2,767,202,367,474,0,MEAN probability A chance node that collapses the decision about whether the proper endpoint metric is pollutant risk or net health effect. Bernoulli( .5 ) 336,56,1 48,29 1,1,1,1,1,1,0,0,0,0 2,585,196,416,303,0,MIDM 2,168,178,416,303,0,SAMP cases/a Net health effect indexed by only consumption recommendation. if Regulate_pollutants_=1 then outcome3[reg_poll='More actions'] else outcome3[reg_poll='BAU'] 168,152,1 48,32 2,102,90,476,293 2,499,269,416,303,0,MIDM 2,415,126,518,378,0,MEAN [Alias Outcome1] [Salmon,Recommendation] probability A chance node that collapses the decision about regulating fish feed. bernoulli(0.5) 56,152,1 48,24 cases/a Net health effect indexed by only fish feed regulation. if recommend = 1 then outcome3[recommendation='Restrict farmed salmon use'] else outcome3[recommendation='BAU'] 280,152,1 48,32 2,499,269,416,303,0,MIDM 2,77,202,518,378,0,MEAN [Alias Lifetime_cancer_chd_] [Salmon,Recommendation] probability A chance node that collapses the decision about consumption recommendations for farmed salmon. bernoulli(0.5) 384,152,1 48,24 1,1,1,1,1,1,0,0,0,0 2,298,233,476,224 2,80,145,416,303,0,SAMP cases/a Net health effect indexed by the two decisions of concern: a) whether to recommend salmon consumption restrictions and b) whether to apply stricter regulations for fish feed. The definition contained also this row: if a>ARL then 1 else 0 But it was removed when we cut the acceptable concentration concept out of the model. var a:= (if Poll_or_net = 1 then Eff_poll else Eff_net); sum(a,salmon) 224,56,1 48,46 2,102,90,476,277 2,499,269,416,303,0,MIDM 2,469,277,518,176,0,MEAN [Reg_poll,Recommendation] 28.6.2004 Jouni Tuomisto This version is an update of the model that was used for calculating the results for Tuomisto paper submitted to Science on January 28, 2004. Only argumentation and comments have been clarified and added. No substantive changes have been made to definitions. The descriptions of the variables described in the Table 1 in 'Supportive online material' have not been changed. Therefore, this version produces the results that were presented in the manuscript. The main conclusions of this study were added as arguments on the top level of the model. The argumentation about the pollutant model selection was clarified (see module Fish advisories). 11.1.2005 Jouni Tuomisto The model was given a unified resource name (URN) and metadata. The only addition since 28.6.2004 is the node Urn and this note. 0 504,184,1 48,12 2,306,93,476,540 [Alias Log_v4] 65535,54067,19661 µg/kg/d Pollutant exposure per body weight per day. Poll_salmon*Salmon_intake/1000/BW 256,264,1 48,24 2,287,149,476,224 2,72,47,653,399,0,MIDM [Salmon,Pollutant1] ktluser 11. Janta 2004 12:08 48,24 168,136,1 48,24 1,57,156,516,377,17 2,40,50,576,600 Rideout said that major feed companies have been able reduce toxin levels in their fish meal over the past several years by using substitute ingredients and less contaminated fish. ÒIt has been an issue that the industry is responding to. Feed companies have been working overtime to use high quality meals that are very low in contaminants and have twice the amount of omega-3sÉThe line is trending downward. We donÕt like having this in our product.Ó Feed_backgr 328,48,1 64,36 http://www.intrafish.com/articlea.php?articleID=41061&s=1 var a:= (if reg_poll='More actions' then impr_in_feed else 0); Feed_backgr*(1-a) 192,128,1 48,29 Feed_poll 192,240,1 60,44 Impr_in_feed+Recommendation+Reg_poll 64,248,1 48,55 [Alias Should_we_change_fi1] - Pollutant concentrations in fish feed can be further reduced from current levels. The estimate of 0-100 % with a gradual decrease in probability density is based on author judgement. It reflects rather a theoretical range of improvement than a realistic estimate. Triangular( 0, 0, 1 ) 64,128,1 48,38 2,102,90,476,409 2,515,277,416,303,0,MIDM 2,216,226,416,303,1,CDFP 52425,39321,65535 - This is a dummy variable only, because the actual concentrations in fish feed are not needed in the current model. 1 192,48,1 48,24 52425,39321,65535 Two options are assumed for fish feed regulations: 1) business as usual (BAU) with current legislation, and 2) More restrictive regulations for fish feed, resulting in reduction of pollutant levels in feed and consequently in salmon. (This is irrespective of any trends unrelated to the decision). ['BAU','More actions'] 280,136,1 48,29 jtue 12. Janta 2004 8:51 48,24 504,328,1 48,38 1,76,122,586,359,17 probability/(g/d) Exposure-response function where also the uncertainty about the population that benefits from omega-3 is taken into account. Erf_hcrude*(if All_or_chd=1 then 1 else F_chd_pati) 200,248,1 52,44 2,291,175,476,224 2,136,146,489,288,0,MEAN Graphtool:0 Distresol:10 Diststeps:1 Cdfresol:5 Cdfsteps:1 Symbolsize:6 Baroverlap:0 Linestyle:1 Frame:1 Grid:1 Ticks:1 Mesh:1 Scales:1 Rotation:45 Tilt:0 Depth:70 Frameauto:1 Showkey:1 Xminimum:0 Xmaximum:1 Yminimum:0 Ymaximum:1 Zminimum:0 Zmaximum:1 Xintervals:0 Yintervals:0 Includexzero:0 Includeyzero:0 Includezzero:0 Statsselect:[1,1,1,1,1,0,0,0] Probindex:[0,0.01,0.05,0.25,0.5,0.75,0.95,0.99,1] Erf_hcrude 336,152,1 48,55 All_or_chd 64,56,1 48,46 probability A large part of omega-3 benefit literature is based on studies on cardiac patients. This node reflects the uncertainty whether there is cardiac health benefit for everyone or only coronary heart disease (CHD) patients. The estimate is not based on data but the aim is to maximise uncertainty. Bernoulli( 0.5 ) 64,160,1 48,38 2,102,90,476,333 52425,39321,65535 fraction Fraction of coronary heart disease patients among the deaths. Current estimate is based on the fraction of cardiac deaths from total deaths in EEA countries, although there are cardiac deaths among non-CHD patients, and there are CHD patients with other causes of death. 1.5717M/3.8664M 64,248,1 48,38 2,102,90,476,274 <a href="http://www.who.int">WHO data</a> probability/(g/d) Dose-response function comes from secondary prevention trials reviewed by Din 2004 Table 1. The relative risk reductions are divided by the omega3 exposure in each study. A continuous distribution is used, and each study result is used as a quintile point for the distribution. Another review is Marckmann and Gronbaek 1999 that concluded that 0.6-0.9 g/d of omega-3 results in 40-60 % decrease in coronary heart disease mortality. The low estimate from this result was used (40% per 0.9 g/d). -Fractiles( [0/3.5,.325/1.5,.482/1.8,.297/0.85, 0.4/0.9 ]) 200,152,1 52,32 2,102,90,512,527 2,72,82,416,303,1,PDFP 52425,39321,65535 Din JN, Newby DE, Flapan AD. Science, medicine, and the future - Omega 3 fatty acids and cardiovascular disease - fishing for a natural treatment. British Medical Journal 2004; 328(7430):30-35. <a href="http://ytoswww/yhteiset/Huippuyksikko/Kirjallisuus/Fish_and_health/Din_Omega3andCVD_BMJ2004.pdf>Intranet file</a> Marckmann P, Gronbaek M. Fish consumption and coronary heart disease mortality. A systematic review of prospective cohort studies. European Journal of Clinical Nutrition 1999; 53(8):585-590. g/d Describes the amount of fish that is still beneficial when added to diet. After this limit, no extra benefit is assumed from omega-3 fatty acids. The value reflects both the physiological need of omega-3 and the current intake of omega-3 from other sources than salmon. Both parts of the estimate are complicated, and the latter varies from country to country. This might have implications to the decision if we could give country-wise recommendations of feed regulations. The estimate is based on author judgement. A rough idea about the magnitude comes from Din 2004, where the trials had 0.85 - 1.8 g/d of omega-3 with benefit but 3.5 g/d showed no benefit in a small trial where the population used reasonable amount of fish anyway. If the physiological limit is lower, the slope of the exposure-response function should be steeper. Another data comes from Albert 1998: the benefit may be limited to omega-3 doses <4.9 g/mo = 0.16 g/d. Markmann and Gronbaek concluded that 0.6-0.9 g/d is beneficial. Triangular( .2, .5, 1 ) 496,64,1 48,38 2,102,90,476,598 52425,39321,65535 Din JN, Newby DE, Flapan AD. Science, medicine, and the future - Omega 3 fatty acids and cardiovascular disease - fishing for a natural treatment. British Medical Journal 2004; 328(7430):30-35. <a href="http://ytoswww/yhteiset/Huippuyksikko/Kirjallisuus/Fish_and_health/Din_Omega3andCVD_BMJ2004.pdf>Intranet file</a> Albert CM, Hennekens CH, O'Donnell CJ, Ajani UA, Carey VJ, Willett WC et al. Fish consumption and risk of sudden cardiac death. Jama-Journal of the American Medical Association 1998; 279(1):23-28. Marckmann P, Gronbaek M. Fish consumption and coronary heart disease mortality. A systematic review of prospective cohort studies. European Journal of Clinical Nutrition 1999; 53(8):585-590. 28.6.2004 Jouni Tuomisto This module contains brief description about pyrkilo diagram method in Analytica platform. There are explanations for node usage and colours. Version 5, 28.6.2004. Copyright KTL (National Public Health Institute, Finland). mtad 16. Aprta 2003 12:56 jtue 21. Novta 2003 15:46 48,24 760,32,1 48,29 1,1,1,1,1,1,0,0,0,0 1,40,8,866,552,17 Arial, 13 81,1,1,0,2,9,4744,6798,7 Contains data that comes from a referrable source. The reference must be mentioned in the Reference attribute. Colour 3R1B. 1 168,336,1 48,24 2,102,90,476,516 65535,52427,65534 Contains data that comes from a non-referrable source, i.e. some general knowledge or author judgement. Colour 4R2B. 1 168,392,1 48,24 52425,39321,65535 Contains information about general issues related to the structure and content of a model. Text is written to Description. Each addition is started with the date and the name of the user. The title of the node is Loki n or Log n (n=version number of the model). You should not write information related to a particular node, that should be written in the node itself so that the information will be inherited with the node. Colour 4L3B. 1 56,464,1 48,12 65535,54067,19661 Argument about a node, data, or relationship in a model; or a description of its importance. Colour: automatic (2L3B). 1 280,280,1 48,24 65535,31131,19661 This is the basic building block of an Analytica model. It is a variable that defines a (typically) measurable entity. Usually it is calculated based on data on and relationships about its causes. 1 168,280,1 48,24 Modules are used to create a hierarchical structure. Modules may contain nodes and other modules inside them. mtad 16. Aprta 2003 12:56 48,24 56,416,1 48,24 1,40,0,505,406,17 A conclusion is basically an argument. The colour is used to enhance the fact that the data for this argument originates from the results of the model. Colour 6L3B. 1 280,392,1 48,24 2,44,90,476,224 65535,65532,19661 Index related to the node beside it. Indexes should be as close as possible to the place where they are used. Otherwise there is the risk of a connection brake. Colour: automatic (5R2B). [0] 56,316,1 48,12 2,341,157,476,224 Colour description: xLyT describes the coordinates in the colour palette, xth cell from left and yth cell from top. Directions are L left, R right, T top, B bottom, e.g. 1R1B is the right bottom cell. 464,312,-1 112,56 Decision mode defines a decision under analysis. Other decisions (such as those decided by someone else) can be defined as uncertain variables instead of decisions. 0 56,280,1 48,24 Outcome of interest. The optimisation of this variable is often defined as the criteria for choosing between decision options. 0 56,360,1 48,24 An uncertain variable that is defined as a probability distribution. 0 168,448,1 48,24 This node describes the use of attributes. All attributes should be used as described in the manuals of Analytica. Exceptions are described here. This description is written for those who read as well as for those who write pyrkilo diagrams. Class: Shows the type of the node. Some of the variable classes have a special meaning in pyrkilo diagrams. These are presented in Help for pyrkilo diagrams module. Identifier: Name of the node that is used in definitions. This must be unique. When drafting a model, use automatic identifiers. Only when there is no expectation of many changes in node Titles, you can streamline all identifiers. Try to use short names and the same structure as in the Title. E.g. if title is Long-range transport, use identifier Lrt. Units: The unit of the value calculated in the node. Should always be defined when the node is a part of causal diagram. If the variable described by the node is dimensionless, '-' should be used. If an explicit unit is not (yet) defined, a description about how the variable could be measured should be used. E.g. 'mass', 'rate', 'age'. Title: The title that is shown on the node in diagram mode. First use a working name that roughly describes the contents. After your model is stabilised, you can rename nodes to better represent the final essence. Try to avoid intermediate nodes where it is difficult to understand what the outcome means; instead, combine consequent nodes so that the outcome is an understandable (measurable) variable. Description: A free-text area for describing the contents of the node. All nodes should be explicitly described and justified in the Description so that a user is able to get an idea of the node without looking at the Definition. If the definiton changes remarkably between versions, it is good to describe this and use dates and user's name for clarity (in a similar manner as in Log nodes). Definition: This defines what is calculated in this node. If the node is a non-numerical argument, it should contain a formula that refers to all nodes that are used to justify the claim; it this case, the numerical result will be nonsense. Pay attention to indexes: all important indexes must be included, but try to work with as few as possible. Use the ';' operator to chop the syntax into pieces. Check: Criteria to check the values entered. Reference: This contains a detailed information about the source of the data. The article or book should be found with this data. When building a model, there should be an accompanying RefMan file with the same name but a different extension. This file contains the full information about each reference cited in the model. The name and path of the RefMan file must be mentioned in the Reference of the model. If there is a relevant internet page or a network file, its path can be added here. Analytica 3.0 understands html-code and makes these links clickable. E.g. this is a link to the <a href=http://www.lumina.com>Lumina home page</a>. You can download Analytica Player from there to browse these models. 0 736,272,1 48,24 2,236,80,581,552 mtad 30. Aprta 2003 10:22 jtue 7. Mayta 2003 10:26 48,24 736,416,1 48,24 1,1,1,1,1,1,0,0,0,0 1,40,15,934,679,17 Arial, 13 79,1,1,0,2,9,4744,6798,7 0 56,448,1 48,24 1,1,1,1,1,1,0,,0, 65535,52427,65534 0 160,576,1 48,24 1,1,1,1,1,1,0,,0, 65535,52427,65534 0 160,392,1 48,24 1,1,1,1,1,1,0,,0, 65535,52427,65534 D_+R_+W_ 160,448,1 48,24 2,102,90,476,507 B_ 160,512,1 48,24 1,1,1,1,1,1,0,,0, 2,102,90,476,224 65535,52427,65534 Q_+Model_variable 264,448,1 48,24 0 552,424,1 64,40 65535,52427,65534 0 680,576,1 64,38 65535,52427,65534 0 680,352,1 48,28 49151,49151,49151 D__hoek_et_al_2002__+R__if_the_effect_was+W__design_was_approp 680,424,1 48,24 19661,48336,65535 B__greenland__cohort 680,488,1 52,24 65535,31131,19661 Hoek et al (Lancet 2002) found out in a cohort study that a higway near home was a risk factor for overall and cardiopulmonary mortality. The study was well designed and performed, and the results are convincing, althought the risk estimates were large. However, the exposure estimation was based on the vicinity of a major road and black smoke and nitric dioxide ambient concentrations. The current knowledge tends to associate especially fine particles (PM) with cardiopulmonary mortality, and in this study the exposure measure clearly was an indicator of traffic-related PM. However, other explanations exist as well, and e.g. noise was not ruled out as a confounder. Taken together, the study increases the plausibility of PM being causally linked to cardiopulmonary mortality. Q__therefore_probabl+Pm_plausibility 792,424,1 48,40 2,412,106,476,224 65535,31131,19661 0 576,136,1 48,28 52425,39321,65535 0 688,232,1 48,55 65535,52427,65534 0 688,56,1 48,38 49151,49151,49151 D__harry_is_born_in_+R__unless_harry_was_+W__because_b__those_ 688,136,1 48,29 1,0,-23 19661,48336,65535 Q__therefore_certain 800,136,1 48,24 65535,31131,19661 probability 0.9 904,424,1 48,24 2,102,90,476,224 52425,39321,65535 0 376,448,1 48,24 19661,48336,65535 Here we show how to describe a complicated reasoning starting with different kinds of data and ending up with a claim. This follows the ideas of Stephen Toulmin (Uses of Argument, Cambridge University Press, 1958). In a simple case or drafting phase it is enough to create an argument by joining a piece of data to the claim. A claim node is often called an Argument node, but it should be remembered, that a full argument includes the reasoning in addition to the outcome. Instead of using several nodes to describe the reasoning, as is done here, it is often convenient to write it down inside the argument node, in description attribute (see 'C: Traffic PM increases cardiopulm mortality'). Note that the arrows usually point from the model to the claim. The logic of this is that the model is seen as a description of reality. And the reality affects the claim, not vice versa. - Data (D) or premises describes the information that supports the claim. - Warrant (W) makes the argument from data to claim a legitimate one. Often it is not explicitly mentioned in discussion, but it should be described in a pyrkilo unless it is obvious. - Backing (B) contains some background information needed for warrant, and often it is practical to combine these two. - Qualifier (Q) describes the strength of the argument. - Rebuttal (R) defines the scope when the argument is valid. 256,180,-1 248,172 jtue 28. Junta 2004 18:03 48,24 736,480,1 48,29 1,257,40,510,515,17 The set module is used as a set containing items, such as a set of emission sources summing up to the total emission. However, it is not possible to refer to a module. Therefore a set node is used instead of the module for refering in the model. If there are items that belong to several sets, aliases are created and placed in each set module. A set node is located in the respective set module. Colour 9L3T. jtue 30. Octta 2003 10:34 48,24 56,144,1 48,24 1,60,78,398,477,17 1,52427,26212 This node is like an alias to a set module. The set module is used as a set containing items, but it is not possible to refer to a module. Therefore a set node is created, and it is used instead of the module for refering in the model. If there are items that belong to several sets, aliases are created and placed in each set. Set node is located in the respective set module. Colour 9L3T. Item1+Item2+Item3+Item4+Item5 48,24,1 48,24 1,724,97 1,52427,26212 This is one item belonging to the set 'Set'. 0 168,24,1 48,24 This is one item belonging to the set 'Set'. 0 168,248,1 48,24 This is one item belonging to the set 'Set'. 0 168,192,1 48,24 This is one item belonging to the set 'Set'. 0 168,136,1 48,24 This is one item belonging to the set 'Set'. 0 168,80,1 48,24 Shows that there is a need for a model, but it is not yet defined explicitly. Colour 7L2B. Model prototype has the same meaning, either can be used depending on whether a node or a module is more convenient. Colour 7L2B. 1 56,88,1 48,24 52429,65535,39321 Shows that there is a need for a model, but it is not yet defined explicitly. Colour 7L2B. Placeholder for a model has the same meaning, either can be used depending on whether a node or a module is more convenient. Colour 7L2B. mtad 16. Aprta 2003 12:56 48,24 56,32,1 48,24 52429,65535,39321 A question or a lack of knowledge. An item that would be important in a model, but there is no information about its value and it cannot therefore be used in calculations. Colour: 5R1T. 0 56,272,1 52,28 49151,49151,49151 Note: colour depends on the source of information and can be that of original data, author judgement or causal node. Do NOT use the automatic colour (L2B3), because it is reserved for the argument. 0 168,256,1 48,24 65535,52427,65534 (param1) 0 168,32,1 48,24 param1 Actually I don't know what can be done with buttons (Jouni Tuomisto 29.10.2003). 168,88,1 48,24 2,356,138,476,224 Shows the strength of an argument (or more precisely, the strength of the data, warrant, and backing to support the argument). There may be several pieces of data connected by several qualifiers to one argument (claim). Colour: automatic (8R3B). 0 168,144,1 48,24 This is used to emphasize that there is a lack of knowledge about this variable, but the variable is still important for understanding causality. In a well-built model, the nuisance parameters are cancelled out before the outcome, so that they do not affect the result given the data used. E.g., we may want to model the human exposure to traffic emissions. We have an estimate about the relative contributions of the most important sources to the exposure, but we don't have absolute values from fate and dispersion models. Therefore, when we want to express causal connections, we must 0 56,208,1 48,29 58981,65535,52427 Non-causal nodes are used when the relation between nodes is not causal. Ice cream consumption and drowning accidents are correlated (at least in Finland). One is not a cause of the other, but both are affected by the outside temperature during summer. Non-causal connections can be used in models for deducting values for one node when the other is known. A typical example of this is a top-down model, where we may have information on 1) total concentration of a pollutant and 2) source contribution of certain emission sources (as a fraction of total concentration). The source contribution is not a causal variable, it is merely an index of all (unknown) causal variables, but it can be used to estimate the concentration caused by a particular emission source. 0 168,200,1 48,24 2,587,271,476,412 39321,55707,65535 29.10.2003 Jouni Tuomisto This is the third version of the Help model. Originally it was planned that it would habe been included in each model as a linked module, but soon it was found that the link caused much more trouble than benefit. It is therefore used as a stand alone model. However, it may be a good idea to embed it into each model using Copy (not Link). N:\huippuyksikko\tutkimus\mallit\TainioTheUsesOfArgument.ana (7.5.2003) is closed and connected to this file in the module 'Argument strucure'. It is edited to reflect the current thinking about the argument structure. 1 736,368,1 48,12 1,1,1,1,1,1,1,,0, 2,433,83,476,398 65535,54067,19661 Arial, 13 A value or preference. Colour 8L4B. 0 280,448,1 48,24 5,65535,1 Pyrkilo diagram method (or structured deliberation as it is sometimes called) has been developed to facilitate the Science-Policy Interface. There is a need for methods facilitating the flow of information and understanding between science and policy. The principle is to describe a risk situation in a formal manner. Pyrkilo is an enhanced causal diagram that contains items along a causal pathway (or network) from e.g. abatement strategies to emissions to dispersion to exposure to effects. It has been designed to describe also other than causal connections such as non-causal reasoning, values, preferences, and arguments. These diagrams use Analytica(TM) platform, a graphical Monte Carlo simulation program. It is based on nodes (or variables or objects). They are used to describe and define all the pieces needed for a description of the situation under scrutiny. Many nodes are used as described in Analytica manuals. However, there are also special colours and shapes representing features that are important for pyrkilo diagrams. See Description of each node for more details. You can see the definitions and descriptions by clicking or double-clicking the nodes. 408,124,-1 400,116 2,402,92,530,558 Understanding of a particular risk develops simultaneously at various levels and using differently structured methods. At one end there is a "discussion layer": political and public discussions about risks of a hazard with a wide interest on e.g. economical consequences of available decision options, public health, and social justification equity. This discussion is sometimes poorly structured and does not give scientists or risk assessors clear questions that could be answered by scientific methods. At the other end there is a "modelling layer": models dealing with specific questions such as air concentrations of a pollutant from a specific source or risk-benefit analyses of particular actions. The models are complex and use several assumptions that are not abvious to outsiders. There is a need for methods facilitating the flow of information and understanding between the existing layers. Pyrkilo (from the Finnish word pyrkiŠ, to aim at) diagrams aim at offering an interface between the two disciplines. The principle is to describe a risk situation in a formal manner. It is an enhanced causal diagram that contains items along the causal pathway (or network) from e.g. emissions to dispersion to exposure to effects. It has been designed to describe also other than causal connections such as values, preferences, and arguments. The pyrkilo method has several objecitves. The structure is relatively easily understandable and readable. It has a basic structure similar to but simpler than mathematical causal models. Translation into a natural language or mathematical model is relatively easy, as is required from an interface. The diagram is fast to build with little data, and the critical parts of the diagram can subsequently be developed into a full-scale model. It is easy to expand into new areas, as the political discussion proceeds. Because of its structured and formal nature, it requires that many assumptions are made explicit unlike in political rhetoric, and therefore it is easier to identify possibly illogical or conflicting issues. It can be used to explore the validity or importance of an aspect before it is brought into the other discipline of heavy modelling or political discourse. 0 736,328,1 48,24 2,54,111,476,224 A scope node is basically an argument. The bevel is used to enhance the fact that the argument is about the scope of the model, (i.e. about the existence of a node or module). Colour: automatic (2L3B). 1 280,336,1 48,24 1,1,1,1,1,1,0,,1, 65535,31131,19661 Pieta 16. tamta 2004 1:32 48,24 136,328,1 48,46 1,141,229,418,300,17 (mg/kg/d)^Æ1 Potency of pollutants. Cancer slope factors (CSF) are used for cancer and Reference Dose (RfD) values are used for non-cancer endpoints. The data comes from the U.S.EPA. Table(Pollutant1,Self)( 16,50u, 1.1,250u, 156K,0, 2,20u ) ['Ca (CSF)','Non-Ca (RfD)'] 64,40,1 48,24 2,249,11,476,457 2,480,276,416,303,0,MIDM 2,103,144,416,303,0,MIDM [Alias Potency1] 65535,52427,65534 [Self,Pollutant1] [Self,Pollutant1] U.S.EPA. Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisory. Volume 2: Risk Assessment and Fish Consumption Limits, 3rd Edition. 2000. Table 3-1. <a href="http://www.epa.gov/waterscience/fish/guidance.html">Open access Internet file</a> <a href="http://ytoswww/yhteiset/Huippuyksikko/Kirjallisuus/Fish_and_health/EPAFishAdvisory/">Intranet file</a> (mg/kg/d)-1 The response assessment is restricted to cancer endpoints, because it is the more sensitive endpoint. potency[Potency='Ca (CSF)'] 64,128,1 48,38 2,102,90,476,279 2,136,146,416,303,0,MIDM Risk_slope 232,128,1 72,72 2,436,19,476,224 jtue 16. Janta 2004 12:54 48,24 320,200,1 48,24 1,81,109,605,297,17 ÒWeÕre telling people that if they want to reduce their risk of cancer, they should not eat more than one meal of farmed salmon a month,Ó Carpenter said. He added that the cancer risk from the toxins effectively cancels out the benefits of omega-3 fatty acids found in farmed salmon, which have not been proven to prevent or reduce the risk of cancer. ÒThere are other places to get omega-3s - wild salmon, other fish, canola oil, flaxseed oil,Ó Carpenter said. F 120,56,1 60,44 2,102,90,609,347 <a href="http://www.intrafish.com/articlea.php?articleID=41061&s=1">Intrafish.com press release 9 Jan 2004</a> g/d Data comes from EPIC study looking at fish consumption in 10 European countries by gender. We take the minimum, the unweighed average and the maximum of these values in the distribution to represent uncertainty in population average fatty fish intake. All fatty fish is assumed to be salmon. Triangular( 7.5, 15.3, 31 ) 352,48,1 48,38 2,376,70,476,496 2,0,0,793,492,1,PDFP [Variable Welch_et_al_2002] Welch AA, Lund E, Amiano P, Dorronsoro M. Variability in fish consumption in 10 European countries. In: Riboli E, Lambert R, editors. Nutrition and lifestyle: opportunities for cancer prevention. Lyon: International Agency for Research on Cancer, 2002: 221-222. <a href="http://ytoswww/yhteiset/Huippuyksikko/Kirjallisuus/Fish_and_health/RiboliNutritionLifestyle_IARC156_2002.pdf">PDF of article</a> <a href="http://ytoswww/yhteiset/Huippuyksikko/Tutkimus/Viljelylohi/Materiaali/ConsumptionOfFish.xls">Data in Excel</a> fraction Fraction of farmed salmon of total salmon consumption in Western Europe. The current estimate is based on author judgement after discussions with people from the Finnish Game and Fisheries Research Institute. Uniform( .8, 1 ) 248,48,1 52,44 2,102,90,476,367 g/d Intake of farmed and wild salmon after the two decisions (regulate fish feed pollutants / recommend restrictions for farmed salmon use) has been made. Although market salmon exists in the index, it is not used in this version of the model. Wild salmon use after restricting farmed salmon use has a triangular probability distribution. Min assumes the same relative decrease as in farmed salmon; mode assumes no change; max assumes that wild salmon intake increases so much that it totally compensates the decrease in farmed salmon use. Estimates are based on author judgement. The wild salmon production capacity is probably much than the max used for the variable. This overestimation causes bias towards smaller costs due to salmon use restrictions. Table(Salmon,Reg_poll,Recommendation)( F,(A*F), (F+Pollutant_scare),((A*F)+Pollutant_scare), W,Triangular((A*W),W,(W+((1-A)*F))), W,Triangular((A*W),W,(W+((1-A)*F))), 0,0, 0,0 ) 248,216,1 48,24 2,516,77,476,224 2,58,276,500,232,0,MIDM 2,248,258,443,303,1,PDFP [Recommendation,Salmon] [Index Salmon] g/d Average farmed salmon consumption in Western Europe in the base case. Fraction_farmed*crude_salmon 248,144,1 48,29 2,469,178,476,300 g/d Average wild salmon consumption in Western Europe in the base case. (1-Fraction_farmed)*crude_salmon 352,144,1 48,24 fraction Farmed salmon use per baseline after a recommendation to restrict farmed salmon use. A uniform distribution between 1 American meal/month (227 g) and no change to baseline. If baseline is lower than the recommendation, no change occurs. var a:= 1*227/(365.25/12); var b:= min([a/f,1]); uniform(b,1) 128,144,1 56,36 2,264,94,476,252 2,552,65,424,320,0,MIDM g/d Change in farmed salmon use when fish feed is more strictly regulated. Consumer may consume more salmon, when pollutant problems are handled. However, there is a possibility of bad reputation ('There is a big problem, because authorities have to intervene'). The range overlaps zero to reflect this uncertainty. The expectation is slightly positive. The estimate is based on author judgement. Triangular( -1, 0.5, 1 ) 128,216,1 52,32 2,151,377,416,303,1,PDFP 0 472,48,1 48,24 2,102,90,476,379 65535,52427,65534 [Chance Crude_salmon] Welch AA, Lund E, Amiano P, Dorronsoro M. Variability in fish consumption in 10 European countries. In: Riboli E, Lambert R, editors. Nutrition and lifestyle: opportunities for cancer prevention. Lyon: International Agency for Research on Cancer, 2002: 221-222. <a href="http://ytoswww/yhteiset/Huippuyksikko/Kirjallisuus/Fish_and_health/RiboliNutritionLifestyle_IARC156_2002.pdf">PDF of article</a> <a href="http://ytoswww/yhteiset/Huippuyksikko/Tutkimus/Viljelylohi/Materiaali/ConsumptionOfFish.xls">Data in Excel</a> A decision about whether a general recommendation should be given to avoid the consumption of European farmed salmon to one meal (227 g) per month or not (business as usual, BAU). ['BAU','Restrict farmed salmon use'] 424,136,1 76,28 2,102,90,476,354 g/g Omega-3 fatty acid content in salmon. Estimate is from Din 2004 Table 2. In a previous version, there were other fish types as well: min is the lowest value, max is the highest value and mode is the average of the two. Uniform(0.0128,0.0215) 464,200,1 48,29 1,1,1,1,1,1,0,0,0,0 2,102,90,476,445 2,106,70,416,303,0,MIDM 52425,39321,65535 Din JN, Newby DE, Flapan AD. Science, medicine, and the future - Omega 3 fatty acids and cardiovascular disease - fishing for a natural treatment. British Medical Journal 2004; 328(7430):30-35. <a href="http://ytoswww/yhteiset/Huippuyksikko/Kirjallisuus/Fish_and_health/Din_Omega3andCVD_BMJ2004.pdf>Intranet file</a> g/d Omega-3 fatty acid intake from salmon. salmon_intake*N3_content 384,264,1 48,24 2,57,94,723,303,0,MIDM jtok 16. tamta 2004 22:14 48,24 168,200,1 48,24 1,60,117,556,308,17 µg/kg Pollutant concentration data from Hites et al, Fig 2. Table(Location1,Pollutant1)( 5.607,160.008,2.94m,50.904, 5.607,190.0095,2.5725m,48.177, 6.8085,136.6735,2.52m,47.268, 3.738,126.673,2.31m,49.995, 4.539,113.339,2.31m,41.814, 4.9395,150.0075,1.995m,37.269, 3.6045,93.338,2.52m,46.359, 4.4055,153.341,2.1m,34.542, 3.204,66.67,1.7325m,39.087, 2.5365,43.3355,1.89m,34.1784, 3.3375,70.0035,1.4175m,29.997, 2.8035,43.3355,1.449m,34.542, 2.5365,36.6685,1.5225m,34.542, 2.403,36.6685,1.3545m,32.724, 2.0025,30.0015,1.554m,26.361, 1.7355,32.0016,1.1025m,21.816, 2.0025,58.6696,840u,12.726, 1.2015,23.3345,1.344m,22.725, 1.2015,23.3345,945u,18.18, 1.2015,26.668,787.5u,15.453, 0.9345,14.6674,1.344m,21.816, 1.068,18.6676,892.5u,21.816, 1.2015,53.336,1.575m,7.272, 0.801,18.6676,1.155m,14.544, 0.534,20.6677,430.5u,15.453, 0.6675,23.3345,315u,7.272, 0.534,26.668,283.5u,8.181, 0.4806,20.001,315u,6.363, 0.4806,13.334,735u,9.09, 0.4005,16.6675,178.5u,4.545, 0.267,13.334,210u,3.636, 0.4005,14.0007,105u,3.636, 0.4005,15.3341,105u,3.636, 0.4005,12.0006,126u,3.636, 0.4005,10.0005,105u,2.727, 0.4005,10.0005,105u,1.818, 0.267,6.667,105u,1.818, 0.267,6.667,105u,1.818, 0.267,6.667,105u,1.818 ) 56,48,1 48,29 2,354,132,476,475 2,17,12,416,638,0,MIDM [Alias Pollutants_in_salmo2] 65535,52427,65534 [Pollutant1,Location1] [Pollutant1,Location1] , , , Hites RA, Foran JA, Carpenter DO, Hamilton MC, Knuth BA, Schwager SJ. Global assessment of organic contaminants in farmed salmon. Science 2004; 303(5655):226-229. <a href="http://ytoswww/yhteiset/Huippuyksikko/Kirjallisuus/Fish_and_health/HitesRA%26al_Science2004.pdf>Intranet file</a> µg/kg Pollutant concentrations in salmon Poll_i_types*Feed_poll 312,112,1 48,24 [Salmon,Pollutant1] Data from Hites classified based on salmon type (farmed, wild, market) Table(Location1,Self)( 'Farmed salmon','Europe', 'Farmed salmon','Europe', 'Market salmon','Europe', 'Market salmon','Europe', 'Farmed salmon','Europe', 'Market salmon','Europe', 'Market salmon','Europe', 'Market salmon','Europe', 'Farmed salmon','North America', 'Market salmon','North America', 'Farmed salmon','North America', 'Market salmon','North America', 'Farmed salmon','North America', 'Market salmon','North America', 'Market salmon','North America', 'Market salmon','North America', 'Market salmon','North America', 'Market salmon','North America', 'Market salmon','North America', 'Market salmon','North America', 'Farmed salmon','North America', 'Farmed salmon','South America', 'Wild salmon','North America', 'Market salmon','North America', 'Wild salmon','North America', 'Wild salmon','North America', 'Wild salmon','North America', 'Wild salmon','North America', 'Wild salmon','North America', 'Wild salmon','North America', 'Wild salmon','North America', 'Wild salmon','North America', 'Wild salmon','North America', 'Wild salmon','North America', 'Wild salmon','North America', 'Wild salmon','North America', 'Wild salmon','North America', 'Wild salmon','North America', 'Wild salmon','North America' ) ['Type','Region'] 56,112,1 48,24 2,102,90,476,395 2,72,82,416,614,0,MIDM 52425,39321,65535 [Self,Location1] [Self,Location1] µg/kg Dieldrin, toxaphene, PCB, and dioxin concentrations in farmed, wild, and market salmon. Triangular probability distribution is used for each pollutant and salmon type. Estimates are based on data from Hites 2004. Parameters min, mode, and max are the minimum, average, and maximum of Hites' data, respectively. var typ:= salmon_type[salmon_type='Type']; var a:= (if salmon=typ then poll_salmon_hites else 0); var b:= (if salmon=typ then 1 else 0); var c:= max(a,location1); var d:= sum(a,location1)/sum(b,location1); var e:= (if salmon=typ then poll_salmon_hites else 1M); var f:= min(e,location1); triangular(f,d,c) 192,113,1 48,38 2,36,56,476,309 2,489,195,416,303,1,PDFP jtue 28. Junta 2004 18:03 48,24 488,40,1 48,24 1,0,1,1,1,1,0,,0, 1,40,0,517,300,17 µg/kg Triangular probability distribution for concentrations indexed by salmon type AND THE THREE REGIONS based on data from Hites 2004. Min is the minimum of Hites, Max is the maximum of Hites, and Mode is the average of Hites. Currently not used, but probably we should look at European values, because the consumption and population data comes from Europe. var typ:= salmon_type[salmon_type='Type']; var reg:= salmon_type[salmon_type='Region']; var a:= (if typ=salmon and reg=Region then poll_salmon_hites else 0); var b:= sum((if typ=salmon and reg=Region then 1 else 0),location1); var c:= max(a,location1); var d:= if b>0 then sum(a,location1)/b else 0; var e:= (if salmon=salmon_type then poll_salmon_hites else 1M); var f:= min(e,location1); d 56,32,1 48,29 2,102,90,476,402 2,72,82,451,390,0,MIDM [Salmon,Pollutant1] The three regions considered in Hites et al 2004. ['Europe','North America','South America'] 56,72,1 48,12 µg/kg Triangular probability distribution for concentrations indexed by salmon type based on data from Hites et al, 2004. Min is the minimum, Mode is the average, and Max is the maximum calculated for each salmon type (farmed, wild, and market) separately. var typ:= salmon_type[salmon_type='Type']; var a:= (if salmon=typ then poll_salmon_hites else 0); var b:= (if salmon=typ then 1 else 0); var c:= max(a,location1); var d:= sum(a,location1)/sum(b,location1); var e:= (if salmon=typ then poll_salmon_hites else 1M); var f:= min(e,location1); index x:=['Min','Mode','Max']; array(x,[f,d,c]) 56,136,1 48,38 2,102,90,476,365 2,561,214,416,303,0,MIDM 1,D,4,2,0,0 jtuomist Tue, Mar 27, 2001 11:26 48,24 760,96,1 48,29 1,0,0,1,1,1,0,,0, 1,40,84,859,492,17 100,1,1,1,2,9,4744,6798,7 - Index for making classes number of bins for classification. sequence(1/9,1,1/9) 760,64,1 48,12 (input,output) vary A function that classifies the values of output variable into classes number of different bins (defined by centiles) based on the values of the input variable. (if (input >= getfract(input,Frac-1/9) and input<getfract(input,Frac) ) then output else 0) + (if getfract(input,1)=input and Frac=1 then output else 0) 760,32,1 48,24 input,output avoided deaths/a Total VOI (EVPI) mean(max(outcome,recommendation))-Nbuu 320,56,1 48,24 2,537,64,416,303,0,MIDM [Reg_poll,Salmon] 1,I,4,2,0,0 cases/a Expected net benefits under uncertainty max(mean(outcome),recommendation) 208,56,1 48,24 1,114,51,409,331,0,MIDM cases/a classify(outcome_inputs,outcome) 208,112,1 48,24 2,155,110,843,183,0,MEAN Graphtool:0 Distresol:10 Diststeps:1 Cdfresol:5 Cdfsteps:1 Symbolsize:6 Baroverlap:0 Linestyle:1 Frame:1 Grid:1 Ticks:1 Mesh:1 Scales:1 Rotation:45 Tilt:0 Depth:70 Frameauto:1 Showkey:1 Xminimum:0 Xmaximum:1 Yminimum:0 Ymaximum:1 Zminimum:0 Zmaximum:1 Xintervals:0 Yintervals:0 Includexzero:0 Includeyzero:0 Includezzero:0 Statsselect:[1,1,1,1,1,0,0,0] Probindex:[0.05,0.25,0.5,0.75,0.95] [Frac,Recommendation] avoided deaths/a Collects all VOI to one table (EVPI and EVPPI). var a:= Max(mean(Nb),recommendation); sum(a,Frac)-Nbuu 320,112,1 48,24 1,533,191,416,303,0,MIDM 2,165,255,459,391,0,MIDM Graphtool:0 Distresol:10 Diststeps:1 Cdfresol:5 Cdfsteps:1 Symbolsize:6 Baroverlap:0 Linestyle:10 Frame:1 Grid:1 Ticks:1 Mesh:1 Scales:1 Rotation:45 Tilt:0 Depth:70 Frameauto:0 Showkey:1 Xminimum:0 Xmaximum:1 Yminimum:0 Ymaximum:13K Zminimum:1 Zmaximum:8 Xintervals:0 Yintervals:0 Includexzero:0 Includeyzero:0 Includezzero:0 Statsselect:[1,1,1,1,1,0,0,0] Probindex:[0.05,0.25,0.5,0.75,0.95] Arial, 2 1,I,6,2,0,0 1 56,56,1 48,46 Outcome Table(outcome_inputs1)( Impr_in_feed,Pollutant_scare,All_or_chd,Erf_hcrude,Benefit_limit,Crude_salmon,Fraction_farmed,N3_content,Poll_or_net,Poll_i_types[Pollutant1='Dieldrin', Salmon='Farmed salmon'],Poll_i_types[Pollutant1='Toxaphene', Salmon='Farmed salmon'],Poll_i_types[Pollutant1='PCB', Salmon='Farmed salmon'],A,Regulate_pollutants_,Recommend) ['Pollutant levels in fish feed after lower limits (S+P)','Salmon consumption after feed limits (S+P)','Does omega-3 help CHD patients or everyone? (S)','Dose-response of health benefit (S)','Highest omega-3 dose with health benefit (S)','Current average consumption of salmon (S)','Fraction of farmed from total salmon use (S)','Omega3 content in salmon (S)','Consider pollutant or net health effect? (P)','Dieldrin concentration in farmed salmon (S)','Toxaphene concentration in farmed salmon (S)','PCB concentration in farmed salmon (S)','Farmed salmon use after recommendation (S)','Lower limits for pollutants in fish feed? (P)','Recommend restricted farmed salmon consumption? (P)'] 752,112,1 48,24 1,1,1,1,1,1,0,0,0,0 2,377,196,476,275 2,443,237,688,342,0,MIDM 2,168,178,582,361,0,MIDM [Self,Self] rank correlation Abs( RankCorrel( Outcome_inputs, Outcome ) ) 208,168,1 48,24 1,1,1,1,1,1,0,0,0,0 2,212,345,590,329,0,MIDM [Recommendation,Outcome_inputs] 1 56,240,1 48,46 Outcome2 avoided deaths/a Total VOI (EVPI) mean(max(outcome2,reg_poll))-Nbuu1 320,240,1 48,24 2,559,23,416,303,0,MIDM [Reg_poll,Salmon] cases/a Expected net benefits under uncertainty max(mean(outcome2),reg_poll) 208,240,1 48,24 1,114,51,409,331,0,MIDM cases/a classify(outcome_inputs,outcome2) 208,296,1 48,24 2,45,47,932,338,0,SAMP Graphtool:0 Distresol:10 Diststeps:1 Cdfresol:5 Cdfsteps:1 Symbolsize:6 Baroverlap:0 Linestyle:1 Frame:1 Grid:1 Ticks:1 Mesh:1 Scales:1 Rotation:45 Tilt:0 Depth:70 Frameauto:1 Showkey:1 Xminimum:0 Xmaximum:1 Yminimum:0 Ymaximum:1 Zminimum:0 Zmaximum:1 Xintervals:0 Yintervals:0 Includexzero:0 Includeyzero:0 Includezzero:0 Statsselect:[1,1,1,1,1,0,0,0] Probindex:[0.05,0.25,0.5,0.75,0.95] [Frac,Reg_poll] avoided deaths/a Collects all VOI to one table (EVPI and EVPPI). var a:= Max(mean(Nb1),reg_poll); sum(a,Frac)-Nbuu1 320,296,1 48,24 1,533,191,416,303,0,MIDM 2,61,82,455,383,0,MIDM Graphtool:0 Distresol:10 Diststeps:1 Cdfresol:5 Cdfsteps:1 Symbolsize:6 Baroverlap:0 Linestyle:10 Frame:1 Grid:1 Ticks:1 Mesh:1 Scales:1 Rotation:45 Tilt:0 Depth:70 Frameauto:0 Showkey:1 Xminimum:0 Xmaximum:1 Yminimum:0 Ymaximum:13K Zminimum:1 Zmaximum:8 Xintervals:0 Yintervals:0 Includexzero:0 Includeyzero:0 Includezzero:0 Statsselect:[1,1,1,1,1,0,0,0] Probindex:[0.05,0.25,0.5,0.75,0.95] Arial, 2 [Pollutant1,Outcome_inputs] 1,I,6,2,0,0 rank correlation Abs( RankCorrel( Outcome_inputs, Outcome2 ) ) 208,352,1 48,24 1,1,1,1,1,1,0,0,0,0 2,212,345,590,329,0,MIDM [Recommendation,Outcome_inputs] var b:= nb1; var c:=0; var x:=1; while x<=samplesize do (c:=c+b[run=x]; x:=x+1); var d:= c/samplesize; var a:= Max(d,reg_poll); sum(a,Frac)-Nbuu1; x 320,352,1 48,24 2,446,377,658,349,0,MIDM [Frac,Outcome_inputs] 1,I,4,2,0,0 index a:= ['Total VOI']; index b1:= concat(a,outcome_inputs1); index b2:= concat(a,outcome_inputs1); index d:= concat(b1,b2); var c:= (if a='Total VOI' then total_voi1 else 0); var e:= concat(c,v3,a,outcome_inputs1,b1); var f:= (if a='Total VOI' then total_voi2 else 0); var g:= concat(f,v1,a,outcome_inputs1,b2); var h:= concat(e,g,b1,b2,d); h 432,184,1 48,24 2,217,23,507,382,0,MIDM Graphtool:0 Distresol:10 Diststeps:1 Cdfresol:5 Cdfsteps:1 Symbolsize:6 Baroverlap:0 Linestyle:1 Frame:1 Grid:1 Ticks:1 Mesh:1 Scales:1 Rotation:45 Tilt:0 Depth:70 Frameauto:1 Showkey:1 Xminimum:0 Xmaximum:1 Yminimum:0 Ymaximum:1 Zminimum:0 Zmaximum:1 Xintervals:0 Yintervals:0 Includexzero:0 Includeyzero:0 Includezzero:0 Statsselect:[1,1,1,1,1,0,0,0] Probindex:[0.05,0.25,0.5,0.75,0.95] ['Pollutant levels in fish feed after lower limits (S+P)','Salmon consumption after feed limits (S+P)','Does omega-3 help CHD patients or everyone? (S)','Dose-response of health benefit (S)','Highest omega-3 dose with health benefit (S)','Current average consumption of salmon (S)','Fraction of farmed from total salmon use (S)','Omega3 content in salmon (S)','Consider pollutant or net health effect? (P)','Dieldrin concentration in farmed salmon (S)','Toxaphene concentration in farmed salmon (S)','PCB concentration in farmed salmon (S)','Farmed salmon use after recommendation (S)','Lower limits for pollutants in fish feed? (P)','Recommend restricted farmed salmon consumption? (P)'] 752,168,1 48,24 1,1,1,1,1,1,0,0,0,0 2,102,90,476,469 2,351,356,688,342,0,MIDM 2,168,178,582,361,0,MIDM [Self,Self] ['Total VOI','Pollutant levels in fish feed after lower limits (S+P)','Salmon consumption after feed limits (S+P)','Does omega-3 help CHD patients or everyone? (S)','Dose-response of health benefit (S)','Highest omega-3 dose with health benefit (S)','Current average consumption of salmon (S)','Fraction of farmed from total salmon use (S)','Omega3 content in salmon (S)','Consider pollutant or net health effect? (P)','Dieldrin concentration in farmed salmon (S)','Toxaphene concentration in farmed salmon (S)','PCB concentration in farmed salmon (S)','Farmed salmon use after recommendation (S)','Lower limits for pollutants in fish feed? (P)','Recommend restricted farmed salmon consumption? (P)','Total VOI','Pollutant levels in fish feed after lower limits (S+P)','Salmon consumption after feed limits (S+P)','Does omega-3 help CHD patients or everyone? (S)','Dose-response of health benefit (S)','Highest omega-3 dose with health benefit (S)','Current average consumption of salmon (S)','Fraction of farmed from total salmon use (S)','Omega3 content in salmon (S)','Consider pollutant or net health effect? (P)','Dieldrin concentration in farmed salmon (S)','Toxaphene concentration in farmed salmon (S)','PCB concentration in farmed salmon (S)','Farmed salmon use after recommendation (S)','Lower limits for pollutants in fish feed? (P)','Recommend restricted farmed salmon consumption? (P)'] 536,216,1 48,12 28.1.2004 Jouni Tuomisto Tulos on laskettu 28.1.2004 malliversiolla TuomistoViljelylohi2004_4.ana (28.1.2004 14:42). Iteraatioita on ollut 2000 kussakin ajossa, ja siemenlukua on vaihdettu 13, 14...22. Table(Dee,Self)( 45.99,54.68,52.26,36.1612390137561,54.3154602065333,34.9994540997468,48.2712149729705,43.4418171126272,36.5685772254328,48.5061584782452, -7.28p,3.64p,29.1p,-40.0177668780088p,-18.1898940354586p,-10.9139364212751p,-3.63797880709171p,32.7418092638254p,-25.465851649642p,7.27595761418343p, -10.9p,3.64p,29.1p,-43.6557456851006p,-18.1898940354586p,-7.27595761418343p,-3.63797880709171p,32.7418092638254p,-25.465851649642p,7.27595761418343p, 0,-10.9p,40p,-25.465851649642p,-3.63797880709171p,-3.63797880709171p,-18.1898940354586p,10.9139364212751p,-40.0177668780088p,3.63797880709171p, -10.9p,0,21.8p,-40.0177668780088p,-14.5519152283669p,-7.27595761418343p,-3.63797880709171p,32.7418092638254p,-29.1038304567337p,3.63797880709171p, -3.64p,7.28p,25.5p,-43.6557456851006p,-10.9139364212751p,-7.27595761418343p,-7.27595761418343p,29.1038304567337p,-32.7418092638254p,3.63797880709171p, -7.28p,3.64p,25.5p,-43.6557456851006p,-14.5519152283669p,-7.27595761418343p,-7.27595761418343p,32.7418092638254p,-21.8278728425503p,3.63797880709171p, -7.28p,3.64p,29.1p,-43.6557456851006p,-14.5519152283669p,-7.27595761418343p,-3.63797880709171p,29.1038304567337p,-25.465851649642p,7.27595761418343p, -7.28p,3.64p,25.5p,-40.0177668780088p,-14.5519152283669p,-7.27595761418343p,-7.27595761418343p,29.1038304567337p,-25.465851649642p,3.63797880709171p, 19.73,19.19,19.44,19.4379094993274,20.0175573351808,20.3445513117949,20.355546642757,19.7440144359389,20.7034870828174,19.9763413781584, -7.28p,3.64p,29.1p,-43.6557456851006p,-10.9139364212751p,-7.27595761418343p,-3.63797880709171p,32.7418092638254p,-25.465851649642p,7.27595761418343p, -7.28p,0,29.1p,-43.6557456851006p,-14.5519152283669p,-7.27595761418343p,-7.27595761418343p,29.1038304567337p,-29.1038304567337p,3.63797880709171p, -7.28p,3.64p,29.1p,-40.0177668780088p,-18.1898940354586p,-7.27595761418343p,-3.63797880709171p,32.7418092638254p,-25.465851649642p,3.63797880709171p, -7.28p,0,29.1p,-43.6557456851006p,-14.5519152283669p,-10.9139364212751p,-3.63797880709171p,29.1038304567337p,-29.1038304567337p,7.27595761418343p, 14.6p,7.28p,29.1p,-36.3797880709171p,-14.5519152283669p,-18.1898940354586p,-7.27595761418343p,43.6557456851006p,-40.0177668780088p,7.27595761418343p, 3.64p,-3.64p,32.7p,-29.1038304567337p,-29.1038304567337p,-29.1038304567337p,-10.9139364212751p,43.6557456851006p,-32.7418092638254p,0, 320.6,318.5,264.9,335.708549076064,337.442817728655,275.665732995609,290.791226693727,334.370728607537,273.164494316921,276.459606918472, 10.57,14.86,41.8p,5.74955088028401,24.7671730730817,5.45696821063757p,-14.5519152283669p,4.67620054120562,-40.0177668780088p,-1.81898940354586p, 123.2,86.48,60.06,90.5814652189947,101.172076686928,59.0112207949551,50.290428820892,122.376689417411,54.4520162740264,81.5371850319334, -18.2p,18.2p,34.6p,9.09494701772928p,-3.63797880709171p,20.0088834390044p,-9.09494701772928p,18.1898940354586p,-36.3797880709171p,-16.3709046319127p, 2.138,1.82p,38.2p,-5.45696821063757p,11.8073357625708,7.27595761418343p,-12.732925824821p,9.96986357776404,-43.6557456851006p,1.81898940354586p, 6.398,3.079,43.7p,-9.09494701772928p,21.7569893201817,3.63797880709171p,-14.5519152283669p,18.4151329940305,-41.8367562815547p,-1.81898940354586p, 11.02,3.64p,41.8p,11.9510390913656,19.2005104308973,3.63797880709171p,-12.732925824821p,13.7459735344328,-40.0177668780088p,1.81898940354586p, 13.24,3.928,41.8p,-10.9139364212751p,6.94312189276934,3.84243531387074,-12.732925824821p,3.79393370039361,-41.8367562815547p,-1.81898940354586p, -20p,3.64p,13.18,2.94388484912088,6.21058777152939,5.45696821063757p,-12.732925824821p,3.40300298175498,-43.6557456851006p,6.49686528052371, -18.2p,3.64p,30.9p,-16.3709046319127p,10.9139364212751p,5.45696821063757p,-5.45696821063757p,9.09494701772928p,-12.732925824821p,1.81898940354586p, -20p,9.1p,41.8p,9.43435100685565,1.66582958040635,7.27595761418343p,-12.732925824821p,29.1038304567337p,-41.8367562815547p,-1.81898940354586p, 10.8,7.28p,41.8p,6.27069051176841,5.9436259337399,2.82376602475961,-14.5519152283669p,21.799220341687,-43.6557456851006p,-1.81898940354586p, 32.63,1.013,43.7p,-9.09494701772928p,14.1657807083811,0.412992603269231,-16.3709046319127p,29.1038304567337p,-41.8367562815547p,-1.81898940354586p, 7.526,5.46p,41.8p,16.9614523061118,33.5381902710542,7.27595761418343p,-12.732925824821p,29.0075866389925,-41.8367562815547p,1.81898940354586p, -10.9p,3.64p,34.6p,9.09494701772928p,-12.732925824821p,12.732925824821p,-20.0088834390044p,36.3797880709171p,-43.6557456851006p,-9.09494701772928p, -10.9p,9.1p,34.6p,5.45696821063757p,-9.09494701772928p,10.9139364212751p,-20.0088834390044p,32.7418092638254p,-40.0177668780088p,-10.9139364212751p ) ['Seed 13','Seed 14','Seed 15','Seed 16','Seed 17','Seed 18','Seed 19','Seed 20','Seed 21','Seed 22'] 536,184,1 48,24 2,130,115,985,303,0,MIDM 2,136,146,921,371,0,MIDM 65535,52427,65534 Graphtool:0 Distresol:10 Diststeps:1 Cdfresol:5 Cdfsteps:1 Symbolsize:6 Baroverlap:0 Linestyle:1 Frame:1 Grid:1 Ticks:1 Mesh:1 Scales:1 Rotation:45 Tilt:0 Depth:70 Frameauto:1 Showkey:1 Xminimum:0 Xmaximum:1 Yminimum:0 Ymaximum:1 Zminimum:0 Zmaximum:1 Xintervals:0 Yintervals:0 Includexzero:0 Includeyzero:0 Includezzero:0 Statsselect:[1,1,1,1,1,0,0,0] Probindex:[0.05,0.25,0.5,0.75,0.95] [Self,Dee] [Self,Dee] 28.1.2004 Jouni Tuomisto The result has been calculated with the model version TuomistoViljelylohi2004_4.ana (28.1.2004 14:42). We used 2000 iterations in each ten runs, and the random seed number was 13, 14...22. average(voi_result,voi_result) 640,184,1 48,24 2,120,59,514,669,0,MIDM cases/a Total mortality in European Economic Area countries (386.63 million inhabitants) 3.8664M 136,408,1 48,29 2,102,90,476,330 65535,52427,65534 <a href="http://www.who.int">WHO data</a> cases/a Coronary heart disease mortality in European Economic Area countries (386.63 million inhabitants). The estimate consists of acute myocardial infarction and other ischaemic heart diseases (ICD 10: 270, 279). 615.3k 504,408,1 48,29 65535,52427,65534 <a href="http://www.who.int">WHO data</a> 1 568,136,1 48,55 Should_we_change_fis 1 760,208,1 48,12 65535,54067,19661 Loki_v4 Eff_net+Eff_poll 256,480,1 60,44 65535,65532,19661 Recommendation+V3 424,48,1 64,44 65535,65532,19661 Reg_poll+V1 280,48,1 72,40 65535,65532,19661 DC-attribute with refinement Scheme (if any) Value Title Risk benefit analysis of eating farmed salmon Creator.personalName Tuomisto, Jouni T Creator.personalName Tuomisto, Jouko Creator.personalName Tainio, Marko Creator.personalName Niittynen, Marjo Creator.personalName Verkasalo, Pia Creator.personalName Vartiainen, Terttu Creator.personalName Kiviranta, Hannu Creator.personalName Pekkanen, Juha Subject risk benefit analysis Subject persistent organic pollutants Subject omega-3 fatty acids Subject MeSH polychlorinated biphenyls Subject MeSH salmon Subject MeSH risk assessment Subject MeSH fatty acids, omega-3 Subject UDC 614 Public health and hygiene. Description.abstract In their Report ÒGlobal assessment of organic contaminants in farmed salmon,Ó R. A. Hites and co-workers analyzed wild and farmed salmon samples from North and South America and Europe for organic pollutants (9 Jan. 2004, p. 226). The authors conclude that, because of chemical contaminants, farmed salmon should not be eaten more often than 0.25 to 1 times per month. However, the model used does not take into account any beneficial effects of eating fish. We analyzed both risks and benefits. We also performed a value-of-information analysis to see which uncertainties were relevant for decision-making. This is the version 4 of the model calculating risks and benefits of farmed salmon. (c) Copyright Kansanterveyslaitos (KTL; National Public Health Institute, Finland). Publisher Kansanterveyslaitos Date.issued W3C-DTF 2004-07-23 Type DCMIType Software Format IMT text/xml Format.medium computerFile Format 115 kB Identifier http://www.ktl.fi/risk Identifier URN URN:NBN:fi-fe20042774 Language ISO639-2 en Relation.hasPart URL http://www.sciencemag.org/cgi/content/full/305/5683/476 Rights Copyright Kansanterveyslaitos, 2004 0 744,240,1 80,12 65535,54067,19661