{"id":2632,"date":"2023-02-14T12:45:24","date_gmt":"2023-02-14T11:45:24","guid":{"rendered":"https:\/\/www.gmds2023.de\/program\/tutorials\/tutorial-02\/"},"modified":"2023-02-15T15:40:38","modified_gmt":"2023-02-15T14:40:38","slug":"tutorial-02","status":"publish","type":"page","link":"https:\/\/www.gmds2023.de\/en\/program\/tutorials\/tutorial-02\/","title":{"rendered":"Tuto\u00adr\u00adi\u00adal 02"},"content":{"rendered":"\n

T02 \u2014 Causal Infer\u00adence in Biosta\u00adtis\u00adtics and Epidemi\u00adol\u00ado\u00adgy: DAGs, g\u2011Methods and Target Trial Emula\u00adtion \u2013 A Tuto\u00adr\u00adi\u00adal for Researchers and Educators<\/h2>\n\n
<\/div>\n\n

Day, Time, Duration<\/mark><\/h3>\n\n

Sunday morn\u00ading, 9 am\u20111 pm, 4h<\/p>\n\n

Language offered<\/mark><\/h3>\n\n

Englisch<\/p>\n\n

Content Summa\u00adry<\/mark><\/h3>\n\n

One of the most impor\u00adtant tasks of deci\u00adsion makers is to derive causal inter\u00adpre\u00adta\u00adtions from statis\u00adti\u00adcal and epidemi\u00ado\u00adlog\u00adi\u00adcal analy\u00adses of obser\u00adva\u00adtion\u00adal and real-world datasets. Often an inter\u00adven\u00adtion, action or risk factor is inter\u00adpret\u00aded to have a \u201ccausal effect\u201d on one or more outcomes (e.g., prob\u00ada\u00adbil\u00adi\u00adty, rate, or mean of endpoint). There\u00adfore, both the biosta\u00adtis\u00adti\u00adcian and the epidemi\u00adol\u00ado\u00adgist need tools to under\u00adstand: (1) when effect esti\u00admates have a causal inter\u00adpre\u00adta\u00adtion and when they do not (using direct\u00aded acyclic graphs, DAGs); and (2) the causal frame\u00adwork for the hypo\u00adthet\u00adi\u00adcal target exper\u00adi\u00adment (e.g., causal research ques\u00adtion, target trial emula\u00adtion design), (3) the appro\u00adpri\u00adate meth\u00adods to derive causal effects instead of mere\u00adly statis\u00adti\u00adcal asso\u00adci\u00ada\u00adtions (e.g., tradi\u00adtion\u00adal multi\u00advari\u00adate regres\u00adsion analy\u00adsis or propen\u00adsi\u00adty score for time-independent confound\u00ading and causal g\u2011methods for time-dependent confound\u00ading). Over the last years, it has become evident, that bias in obser\u00adva\u00adtion\u00adal stud\u00adies is often not main\u00adly caused by uncon\u00adtrolled confound\u00ading but by self-inflicted bias\u00ades relat\u00aded to incor\u00adrect analyt\u00adic designs lead\u00ading to incor\u00adrect exposure\/treatment assign\u00adment and time-related bias\u00ades (e.g., immor\u00adtal time bias). This tuto\u00adr\u00adi\u00adal intends to provide basic knowl\u00adedge on causal think\u00ading and visu\u00adal, struc\u00adtur\u00adal, and statis\u00adti\u00adcal tools to guide valid causal analy\u00adsis and to know which esti\u00admates are suit\u00adable for causal inter\u00adpre\u00adta\u00adtion. The tuto\u00adr\u00adi\u00adal covers inno\u00adv\u00ada\u00adtive causal designs as well as causal infer\u00adence concepts and meth\u00adods that are need\u00aded for the design and analy\u00adsis of obser\u00adva\u00adtion\u00adal data and prag\u00admat\u00adic trials with time-varying expo\u00adsures or treat\u00adments. The tuto\u00adr\u00adi\u00adal is struc\u00adtured into the follow\u00ading sections: <\/p>\n\n

1. Intro\u00adduc\u00adtion to the prin\u00adci\u00adples of causa\u00adtion in biosta\u00adtis\u00adtics and epidemi\u00adol\u00ado\u00adgy
2. Use of causal diagrams (direct\u00aded acyclic graphs, DAGs)
3. Demon\u00adstra\u00adtion of the target trial exper\u00adi\u00adment and the target trial emula\u00adtion approach
4. Brief intu\u00aditive illus\u00adtra\u00adtion of the prin\u00adci\u00adples of g\u2011methods: a) g\u2011formula, b) margin\u00adal struc\u00adtur\u00adal models with inverse prob\u00ada\u00adbil\u00adi\u00adty of treat\u00adment weight\u00ading, and c) struc\u00adtur\u00adal nest\u00aded models with g\u2011estimation
5. Intro\u00adduc\u00adtion of the target exper\u00adi\u00adment prin\u00adci\u00adple and appli\u00adca\u00adtion of the target trial emula\u00adtion approach combined with a coun\u00adter\u00adfac\u00adtu\u00adal approach using the cloning-censoring-weighting (CCW) approach for dynam\u00adic treat\u00adment strate\u00adgies
6. Appli\u00adca\u00adtion of g\u2011methods in obser\u00adva\u00adtion\u00adal stud\u00adies and prag\u00admat\u00adic trials with post-randomization confound\u00ading and selec\u00adtion bias (treat\u00adment switching\/non-adherence\/2nd-line-treatment etc.)<\/p>\n\n

Orga\u00adniz\u00ader<\/mark><\/h3>\n\n

Felic\u00adi\u00adtas K\u00fchne<\/p>\n\n

Insti\u00adtu\u00adtion<\/mark><\/h3>\n\n

UMIT TIROL \u2014 Univer\u00adsi\u00adty for Health Sciences and Technology<\/p>\n\n

Contact<\/mark><\/h3>\n\n

felicitas.kuehne [at] umit-tirol.at<\/p>\n\n

Addi\u00adtion\u00adal Speaker<\/mark><\/h3>\n\n

Uwe Siebert<\/p>\n\n

Insti\u00adtu\u00adtion<\/mark><\/h3>\n\n

UMIT TIROL \u2014 Univer\u00adsi\u00adty for Health Sciences and Technology<\/p>\n\n

<\/div>\n\n

<\/p>

<\/div>\n","protected":false},"excerpt":{"rendered":"

T02 \u2014 Causal Infer\u00adence in Biosta\u00adtis\u00adtics and Epidemi\u00adol\u00ado\u00adgy: DAGs, g\u2011Methods and Target Trial Emula\u00adtion \u2013 A Tuto\u00adr\u00adi\u00adal for Researchers and Educa\u00adtors Day, Time, Dura\u00adtion Sunday morn\u00ading, 9 am\u20111 pm, 4h Language offered Englisch Content Summa\u00adry One of the most impor\u00adtant tasks of deci\u00adsion makers is to derive causal inter\u00adpre\u00adta\u00adtions from statis\u00adti\u00adcal and epidemi\u00ado\u00adlog\u00adi\u00adcal analy\u00adses of observational<\/p>\n

Weiterlesen<\/a><\/p>\n","protected":false},"author":18,"featured_media":0,"parent":1348,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"wp_typography_post_enhancements_disabled":false,"footnotes":""},"_links":{"self":[{"href":"https:\/\/www.gmds2023.de\/en\/wp-json\/wp\/v2\/pages\/2632"}],"collection":[{"href":"https:\/\/www.gmds2023.de\/en\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.gmds2023.de\/en\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.gmds2023.de\/en\/wp-json\/wp\/v2\/users\/18"}],"replies":[{"embeddable":true,"href":"https:\/\/www.gmds2023.de\/en\/wp-json\/wp\/v2\/comments?post=2632"}],"version-history":[{"count":13,"href":"https:\/\/www.gmds2023.de\/en\/wp-json\/wp\/v2\/pages\/2632\/revisions"}],"predecessor-version":[{"id":2819,"href":"https:\/\/www.gmds2023.de\/en\/wp-json\/wp\/v2\/pages\/2632\/revisions\/2819"}],"up":[{"embeddable":true,"href":"https:\/\/www.gmds2023.de\/en\/wp-json\/wp\/v2\/pages\/1348"}],"wp:attachment":[{"href":"https:\/\/www.gmds2023.de\/en\/wp-json\/wp\/v2\/media?parent=2632"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}