서브메뉴
검색
Statistical and Mixed-Methods Approaches for Analyzing Health Consequences of Medical Algorithms
Statistical and Mixed-Methods Approaches for Analyzing Health Consequences of Medical Algorithms
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202105621
- ISBN
- 9798265428561
- DDC
- 658
- 서명/저자
- Statistical and Mixed-Methods Approaches for Analyzing Health Consequences of Medical Algorithms
- 발행사항
- [Sl] : Stanford University, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 167 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 87-05, Section: B.
- 주기사항
- Advisor: Guestrin, Carlos;Rose, Sherri.
- 학위논문주기
- Thesis (Ph.D.)--Stanford University, 2025.
- 초록/해제
- 요약Inequitable equilibriums driving health disparities persist across the US. In order to ensure that medical algorithms do not make disparities worse, their role in driving health outcomes must be critically examined alongside other factors. In this dissertation, I explicitly consider downstream consequences of algorithms on health outcomes to develop meaningful evaluation metrics and generate data simulations that mitigate societal bias. I also look upstream, considering how social theories of health can be used to construct more equitable algorithms, and how community engaged research can elucidate relationships between factors driving health outcomes.First, I introduce a new measure of local calibration-the threshold calibration error-to evaluate a medical algorithm used to guide statin treatment decisions. Next, I propose a novel data transformation framework for attenuating societal biases in data using microsimulation models. With it, I simulate changes in disease trajectories likely to result from the removal of a race adjustment in the eGFR equation-an algorithm used for diagnosis and staging in chronic kidney disease. I also argue for incorporating area-level indices capturing social drivers of health into predictive models of health outcomes, and propose a set of considerations for implementation. Lastly, I describe a participatory research study I designed and carried out to understand specific ways in which social factors (e.g. housing, transportation, adverse experiences, access barriers, direct discrimination) impact chronic kidney disease patients' ability to manage their condition.This work, drawing from algorithmic fairness, biomedical informatics, decision science, social epidemiology, community engaged research, and systems dynamics, illustrates the importance of a multi-disciplinary approach for evaluating health and health equity consequences of medical algorithms.
- 일반주제명
- Decision making
- 일반주제명
- High density lipoprotein
- 일반주제명
- Physiology
- 기타저자
- Stanford University.
- 기본자료저록
- Dissertations Abstracts International. 87-05B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2025 us c eng d■001000017360802
■00520260202105621
■006m o d
■007cr#unu||||||||
■020 ▼a9798265428561
■035 ▼a(MiAaPQ)AAI32316510
■035 ▼a(MiAaPQ)Stanfordxj603cd1091
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a658
■1001 ▼aForyciarz, Agata Lidia.
■24510▼aStatistical and Mixed-Methods Approaches for Analyzing Health Consequences of Medical Algorithms
■260 ▼a[Sl]▼bStanford University▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a167 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 87-05, Section: B.
■500 ▼aAdvisor: Guestrin, Carlos;Rose, Sherri.
■5021 ▼aThesis (Ph.D.)--Stanford University, 2025.
■520 ▼aInequitable equilibriums driving health disparities persist across the US. In order to ensure that medical algorithms do not make disparities worse, their role in driving health outcomes must be critically examined alongside other factors. In this dissertation, I explicitly consider downstream consequences of algorithms on health outcomes to develop meaningful evaluation metrics and generate data simulations that mitigate societal bias. I also look upstream, considering how social theories of health can be used to construct more equitable algorithms, and how community engaged research can elucidate relationships between factors driving health outcomes.First, I introduce a new measure of local calibration-the threshold calibration error-to evaluate a medical algorithm used to guide statin treatment decisions. Next, I propose a novel data transformation framework for attenuating societal biases in data using microsimulation models. With it, I simulate changes in disease trajectories likely to result from the removal of a race adjustment in the eGFR equation-an algorithm used for diagnosis and staging in chronic kidney disease. I also argue for incorporating area-level indices capturing social drivers of health into predictive models of health outcomes, and propose a set of considerations for implementation. Lastly, I describe a participatory research study I designed and carried out to understand specific ways in which social factors (e.g. housing, transportation, adverse experiences, access barriers, direct discrimination) impact chronic kidney disease patients' ability to manage their condition.This work, drawing from algorithmic fairness, biomedical informatics, decision science, social epidemiology, community engaged research, and systems dynamics, illustrates the importance of a multi-disciplinary approach for evaluating health and health equity consequences of medical algorithms.
■590 ▼aSchool code: 0212.
■650 4▼aDecision making
■650 4▼aHigh density lipoprotein
■650 4▼aPhysiology
■690 ▼a0719
■71020▼aStanford University.
■7730 ▼tDissertations Abstracts International▼g87-05B.
■790 ▼a0212
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17360802▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
