PDPooja Desai

Pooja Desai

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Pooja Desai has a total of 22 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Impact of clinical trial virtualization on recruitment in underserved communities for a type 2 diabetes mHealth intervention

    by Elizabeth Campbell, Andrea Cassells, T.J. Lin, Jacqeline Cortez Lainez, Arlene Smaldone, Pooja Desai, Haomiao Jia, George Hripcsak, Jonathan Tobin, Lena Mamykina

  • Open Access

    A qualitative Interview Study Investigating Patient, Health Professional, and Developer Perspectives on Real-World Implementation of Patient-Centered AI Systems

    by Natalie Benda, Pooja Desai, Zayan Reza, Victoria Winogora, Uday Suresh, Yiye Zhang, Alison Hermann, Rochelle Joly, Jyotishman Pathak, Meghan Reading Turchioe

    Our objective was to triangulate patient, health professional, and developer perspectives for implementing patient-centered artificial intelligence (AI) systems. We conducted semi-structured interviews with patients (N = 18), health professionals (N = 8), and AI developers (N = 8). We created interview guides informed by frameworks in bioethics and health information informatics. We utilized a predictive algorithm for determining risk for postpartum depression as a use case to concretize our discussions. Our team analyzed transcripts from interview recordings using thematic, directed content analysis and the constant comparative process. Participants found mitigating potential harms caused by AI (e.g., bias, stigma, or patient anxiety) greatly important. They also believed that AI must provide clinical benefits by allowing health professionals and patients to easily take actions based on AI output. To take safe action, end users needed transparency to understand the AI’s accuracy and predictors driving risk. Patient participants wanted health professionals to interpret AI output, but health professionals did not always feel they had the time or training to do so. Participants also raised concerns regarding how data quality may affect AI accuracy, who may be responsible for inappropriate actions taken based on AI, and issues regarding data security, privacy, and accessibility. Our results support real-world implementation of more patient-centered AI tools by: providing health professionals with competencies for discussing AI-based risks; engaging patients and health professionals throughout the development process; inclusively communicating AI output to health professionals and patients; and implementing multi-layer systems of AI governance.

  • Open Access

    Psycho-social and behavioral trends in Type 2 Diabetes self-management amongst medically underserved patients during the COVID-19 pandemic in an mHealth intervention

    by Elizabeth Campbell, Haomiao Jia, Andrea Cassells, Tj Lin, Jacqueline Cortez Lainez, Jonathan Tobin, Arlene Smaldone, Pooja Desai, George Hripcsak, Lena Mamykina

    Objective To examine behavioral and psychosocial aspects of Type 2 Diabetes Mellitus (T2DM) self-management among medically underserved adults during the COVID-19 pandemic. Materials and Methods Baseline data were analyzed from T2 Coach, a randomized controlled trial testing a behavioral mHealth intervention using a conversational agent to support T2DM self-management. Participants (n=279) were recruited from Federally Qualified Health Centers in underserved areas of the New York City metropolitan area (2020–2023). Measures included demographics, HbA1c, BMI, mobile device proficiency, and validated surveys assessing self-care adherence (SCI-R), self-efficacy, diabetes distress (PAID), and pandemic-related experiences (e.g., food insecurity, dietary and exercise changes). Multiple linear regression models examined associations between demographic, clinical, and COVID-19-related variables with SCI-R, self-efficacy, PAID, and HbA1c outcomes. Results Participants were primarily female (60.6%), Hispanic (63.4%), and foreign-born (65.9%), with substantial socioeconomic disadvantage (43.7% Medicaid enrollment; 42.7% food insecurity). Mean HbA1c was 10.1 (SD 1.7). Pandemic-related disruptions were common: 18.7% ate more, 34.5% exercised less, and ∼40% changed food purchasing habits. Regression models indicated that poorer self-reported health and dietary changes were linked to higher distress and lower self-efficacy. Increased exercise predicted better adherence (higher SCI-R), and women demonstrated stronger adherence than men. Older age was associated with lower HbA1c. Digital literacy was not associated with self-management outcomes. Discussion COVID-19 exacerbated challenges to diabetes self-management among underserved populations, particularly through lifestyle disruptions and psychosocial stressors. Conclusion Findings underscore the need for culturally responsive, technology-enabled interventions that address diet, exercise, and emotional well-being to improve diabetes outcomes during public health crises.

  • Open Access

    Impact of Clinical Trial Virtualization on Recruitment in Underserved Communities for a Type 2 Diabetes mHealth Intervention

    by Elizabeth Campbell, Andrea Cassells, Tj Lin, Jacqeline Cortez Lainez, Arlene Smaldone, Pooja Desai, Haomiao Jia, George Hripcsak, Jonathan Tobin, Lena Mamykina

    Objective Approaches to randomized clinical trial (RCT) implementation greatly changed during the COVID-19 pandemic as clinical trials transitioned to largely virtual implementation; virtual RCT implementation may continue in the future. Without careful consideration, virtual RCTs may exacerbate the historical under-representation of women, the elderly, and racial and ethnic minorities that has occurred in past clinical trials. The following study presents an approach to virtualizing an RCT for an mHealth intervention. Materials and Methods Participants for the trial were recruited from Federally Qualified Health Centers in medically underserved areas of the New York City metropolitan area. Recruitment began in January 2020 but was paused in March 2020 due to the COVID-19 pandemic’s onset. The research team developed a virtual protocol for recruitment, onboarding, and study implementation. Results Study participants were predominantly from immigrant, low-income, and racial and ethnic minority groups. Our virtualization approach included an easier-to-understand consent form, expanded virtual training materials, and greater one-on-one attention during training. Our approach did not lead to significant changes in the population we recruited into the study or introduce additional biases and restriction. Discussion Our study represents an early investigation into how the change from in-person clinical trial recruitment and study implementation to virtual during the COVID-19 pandemic may impact recruitment of participants from medically underserved communities into RCTs. Conclusion This approach may be used in future trials testing mHealth and other technological interventions without exacerbating the under-representation of medically underserved populations. Highlights The COVID-19 pandemic accelerated the shift to virtual randomized clinical trials (RCTs), raising concerns about exacerbating the under-representation of marginalized groups. This study implemented a fully virtual RCT for a mHealth intervention with consideration of inclusivity. Participants were recruited from Federally Qualified Health Centers in medically underserved areas of New York City, with most coming from immigrant, low-income, and minority communities. Adaptations such as simplified consent, expanded virtual training, and individualized support ensured accessibility without altering participant demographics or introducing bias, providing a replicable model for equitable virtual trials.