Clinical predictive models that include race as a predictor have the potential to exacerbate disparities in healthcare. Such models can be respecified to exclude race or optimized to reduce racial bias. We investigated the impact of such respecifications in a predictive model - UTICalc - which was designed to reduce catheterizations in young children with suspected urinary tract infections. To reduce racial bias, race was removed from the UTICalc logistic regression model and replaced with two new features. We compared the two versions of UTICalc using fairness and predictive performance metrics to understand the effects on racial bias. In addition, we derived three new models for UTICalc to specifically improve racial fairness. Our results show that, as predicted by previously described impossibility results, fairness cannot be simultaneously improved on all fairness metrics, and model respecification may improve racial fairness but decrease overall predictive performance.
Download full-text PDF |
Source |
|---|---|
| http://www.ncbi.nlm.nih.gov/pmc/articles/PMC11141814 | PMC |
Publication Analysis
Top Keywords
Similar Publications
The challenges faced by early-career international medical graduates in general practice and the opportunities for support: a mixed methods study.
Educ Prim Care
January 2025
College of Medicine and Health, University of Exeter, Exeter, UK.
Background: Recruiting and retaining International Medical Graduates (IMGs) has been identified as a key component in addressing the ongoing workforce crisis in general practice. However, research shows IMG General Practitioners (GPs) face unique challenges compared to their UK-trained counterparts, impacting their welfare, retention in the workforce, and ability to provide patient care.
Aim: This study examined the challenges facing early-career IMG GPs, the help and support they access and want to access, and their perceptions and utilisation of the help and support available.
The Minority Tax: Stories from Family Physicians.
J Am Board Fam Med
January 2025
From the University of Texas Medical Branch, Galveston, TX (ZSD); University of Texas Medical Branch, Galveston, TX (NB); MPH University of Pittsburgh, Pittsburgh, PA (WTF); University of Texas Medical Branch, Galveston, TX (AC); Yuma Regional Medical Center, Yuma AZ (NOA); University of Arizona College of Medicine, Phoenix, AZ (NOA); University of Texas Medical Branch, Galveston, TX (KMC).
The minority tax has been defined as a set of disparities that those who are underrepresented in medicine face in addition to clinical care, education, and research responsibilities. These taxes include systemic racism, diversity efforts, clinical and promotion disparities, lack of faculty development, and isolation. Much has been added to the literature to better define and characterize the minority tax and propose suggestions for mitigations.
View Article and Find Full Text PDFPsychology and whiteness itself.
Am Psychol
January 2025
Department of Counseling and Clinical Psychology, Teachers College, Columbia University.
In the aftermath of its Apology to People of Color (American Psychological Association, 2021b), the American Psychological Association recently committed itself to a long-term process by which it aims to address racial equity within the field of psychology as well as society more broadly (Andoh, 2022). In service of these ends, what can psychology learn from an analysis of the discursive framework within which it conducts its racism-related work? This critical conceptual article begins with the premise that all professional discourse-the concepts, language, and logic structures by which a field creates and communicates knowledge-inevitably bears the markings of the society in which it was established. Examination of psychological discourse, therefore, can reveal information not only about social hierarchies but also about the field's potential reproduction of them (even when the field intends to do otherwise).
View Article and Find Full Text PDFDeveloping a Novel Integrative Health Equity and Anti-racism Tool (IHEART) for Pilot Application in a Multicenter Integrative Health Elective for Medical Students and Resident Physicians: A Study Protocol.
Glob Adv Integr Med Health
January 2025
Harvard Medical School, Boston, MA, USA.
Background: This study protocol introduces the Integrative Health Equity and Anti-Racism Tool (IHEART), an innovative instrument designed to infuse equity, diversity, and inclusion (EDI) into Integrative Health (IH) education. Recognizing the gaps in current IH training that fail to address social and systemic inequities adequately, the IHEART is intended to respond to the growing need for inclusivity in IH practices and educational materials. The tool is mainly focused on addressing issues such as accessibility of complementary and integrative health (CIH) therapies, cultural misappropriation, anti-racism, gender diversity, disability justice, trauma-informed care, weight inclusivity, and planetary health, which are currently inconsistently covered in IH training.
View Article and Find Full Text PDFWork-related grief and bereavement experiences of social and community service workers working with people experiencing social disadvantage: a rapid scoping review.
BMC Palliat Care
January 2025
Institute on Aging and Lifelong Health, University of Victoria, 3800 Finnerty Road, Victoria, BC, V8P 5C2, Canada.
Background: While an extensive body of research in palliative care exists on the experiences of grief and bereavement among family caregivers, much of this research is based on normative assumptions of who family caregivers are - housed, financially stable, and with extended family and/or friends to draw on for support. Research shows that in contexts of social disadvantage(e.g.
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!