Dr. Phadke is a dedicated clinician-educator and board-certified internal medicine physician. She divides her time between the clinical care of adult primary care patients, teaching, quality improvement implementation and evaluation, and health system leadership.

In her clinical care, she practices at Hoover Pavillion in Palo Alto. She enjoys forming deep relationships with patients. Her goal is to listen actively, provide expertise, and guide patients toward the best health outcomes.

Her teaching includes clinical teaching within the internal medicine continuity clinic and medical student ambulatory clerkship, and quality improvement coaching and teaching.

She hold several administrative roles including Associate Physician Improvement Leader for the Department of Medicine at Stanford, Quality Director for the Division of Primary Care and Population Health, Founding Director of the Primary Care Project Engagement Platform, and Director of Chronic Disease Management for the Stanford Healthcare Alliance insurance plan. Across these roles, she partners with physicians, quality improvement professionals, and care teams to improve clinical care within Stanford Medicine.

Her scholarly focus is primary care quality improvement evaluation. She has published and presented on a wide array of topics from team-based care in primary care to strategies to improve chronic disease management to the integration of emerging technologies. She enjoys working with trainees and students on scholarship.

Clinical Focus

  • Internal Medicine

Academic Appointments

Administrative Appointments

  • Medical Director Population Health, Division of of Primary Care and Population Health, Stanford University School of Medicine (2017 - Present)
  • Director of Quality, Division of Primary Care and Population Health, Stanford University School of Medicine (2018 - Present)
  • Founding Director, Primary Care Performance Enhancement Program, Stanford University School of Medicine (2018 - Present)
  • Medical Director, Chronic DIsease Management, Stanford Health Plans (2019 - Present)
  • Associate Physician Improvement Leader, Department of Medicine, Stanford University School of Medicine (2023 - Present)

Boards, Advisory Committees, Professional Organizations

  • Member, American College of Physicians (2014 - Present)
  • Member, Society for General Internal Medicine (2014 - Present)

Professional Education

  • Board Certification: American Board of Internal Medicine, Internal Medicine (2020)
  • Bachelors in Science, Yale University, Molecular, Cellular, and Developmental Biology (2007)
  • Medical Education: Pritzker School of Medicine University of Chicago Registrar (2011) IL
  • Internship: Stanford University Internal Medicine Residency (2012) CA
  • Residency: Stanford University Internal Medicine Residency (2014) CA
  • Fellowship: Palo Alto VA Healthcare System (2015) CA

All Publications

  • AI-Human Hybrid Workflow Enhances Teleophthalmology for the Detection of Diabetic Retinopathy. Ophthalmology science Dow, E. R., Khan, N. C., Chen, K. M., Mishra, K., Perera, C., Narala, R., Basina, M., Dang, J., Kim, M., Levine, M., Phadke, A., Tan, M., Weng, K., Do, D. V., Moshfeghi, D. M., Mahajan, V. B., Mruthyunjaya, P., Leng, T., Myung, D. 2023; 3 (4): 100330


    Detection of diabetic retinopathy (DR) outside of specialized eye care settings is an important means of access to vision-preserving health maintenance. Remote interpretation of fundus photographs acquired in a primary care or other nonophthalmic setting in a store-and-forward manner is a predominant paradigm of teleophthalmology screening programs. Artificial intelligence (AI)-based image interpretation offers an alternative means of DR detection. IDx-DR (Digital Diagnostics Inc) is a Food and Drug Administration-authorized autonomous testing device for DR. We evaluated the diagnostic performance of IDx-DR compared with human-based teleophthalmology over 2 and a half years. Additionally, we evaluated an AI-human hybrid workflow that combines AI-system evaluation with human expert-based assessment for referable cases.Prospective cohort study and retrospective analysis.Diabetic patients ≥ 18 years old without a prior DR diagnosis or DR examination in the past year presenting for routine DR screening in a primary care clinic.Macula-centered and optic nerve-centered fundus photographs were evaluated by an AI algorithm followed by consensus-based overreading by retina specialists at the Stanford Ophthalmic Reading Center. Detection of more-than-mild diabetic retinopathy (MTMDR) was compared with in-person examination by a retina specialist.Sensitivity, specificity, accuracy, positive predictive value, and gradability achieved by the AI algorithm and retina specialists.The AI algorithm had higher sensitivity (95.5% sensitivity; 95% confidence interval [CI], 86.7%-100%) but lower specificity (60.3% specificity; 95% CI, 47.7%-72.9%) for detection of MTMDR compared with remote image interpretation by retina specialists (69.5% sensitivity; 95% CI, 50.7%-88.3%; 96.9% specificity; 95% CI, 93.5%-100%). Gradability of encounters was also lower for the AI algorithm (62.5%) compared with retina specialists (93.1%). A 2-step AI-human hybrid workflow in which the AI algorithm initially rendered an assessment followed by overread by a retina specialist of MTMDR-positive encounters resulted in a sensitivity of 95.5% (95% CI, 86.7%-100%) and a specificity of 98.2% (95% CI, 94.6%-100%). Similarly, a 2-step overread by retina specialists of AI-ungradable encounters improved gradability from 63.5% to 95.6% of encounters.Implementation of an AI-human hybrid teleophthalmology workflow may both decrease reliance on human specialist effort and improve diagnostic accuracy.Proprietary or commercial disclosure may be found after the references.

    View details for DOI 10.1016/j.xops.2023.100330

    View details for PubMedID 37449051

    View details for PubMedCentralID PMC10336195

  • Catalyzing System Change: 100 Quality Improvement Projects in 1000 Days. Journal of general internal medicine Sattler, A., Phadke, A., Mickelsen, J., Seay-Morrison, T., Filipowicz, H., Chhoa, D., Srinivasan, M. 2023


    Health system change requires quality improvement (QI) infrastructure that supports frontline staff implementing sustainable innovations. We created an 8-week rapid-cycle QI training program, Stanford Primary Care-Project Engagement Platform (PC-PEP), open to patient-facing primary care clinicians and staff.Examine the feasibility and outcomes of a scalable QI program for busy practicing providers and staff in an academic medical center.Program evaluation.A total of 172 PCPH team members: providers (n = 55), staff (n = 99), and medical learners (n = 18) in the Stanford Division of Primary Care and Population Health (PCPH) clinics, 2018-2021.We categorized projects by the Institute for Healthcare Improvement's (IHI) Quintuple Aim (QA): better health, better patient experience, lower cost of care, better care team experience, and improved equity/inclusion. We assessed project progress with a modified version of The Ottawa Hospital Innovation Framework: step 1 (identified root causes), step 2 (designed/tested interventions), step 3 (assessed project outcome), step 4 (met project goal with target group), step 5A (intervention(s) spread within clinic), step 5B (intervention(s) spread to different setting). Participants rated post-participation QI self-efficacy.Within 1000 days, 172 unique participants completed 104 PC-PEP projects. Most projects aimed to improve patient health (55%) or care team experience (23%). Among projects, 9% reached step 1, 8% step 2, 16% step 3, 26% step 4, 21% step 5A, and 20% step 5B. Learner involvement increased likelihood of scholarly products (47% vs 10%). Forty-six of 47 (98%) survey respondents reported improved QI self-efficacy. Medical assistants, more so than physicians, reported feeling acknowledged by the health system for their QI efforts (100% vs 61%).With appropriate QI infrastructure, scalable QI training models like Stanford PC-PEP can empower frontline workers to create meaningful changes across the IHI QA.

    View details for DOI 10.1007/s11606-023-08431-w

    View details for PubMedID 37985609

    View details for PubMedCentralID 9341176

  • Artificial Intelligence Improves Patient Follow-Up in a Diabetic Retinopathy Screening Program. Clinical ophthalmology (Auckland, N.Z.) Dow, E. R., Chen, K. M., Zhao, C. S., Knapp, A. N., Phadke, A., Weng, K., Do, D. V., Mahajan, V. B., Mruthyunjaya, P., Leng, T., Myung, D. 2023; 17: 3323-3330


    We examine the rate of and reasons for follow-up in an Artificial Intelligence (AI)-based workflow for diabetic retinopathy (DR) screening relative to two human-based workflows.A DR screening program initiated September 2019 between one institution and its affiliated primary care and endocrinology clinics screened 2243 adult patients with type 1 or 2 diabetes without a diagnosis of DR in the previous year in the San Francisco Bay Area. For patients who screened positive for more-than-mild-DR (MTMDR), rates of follow-up were calculated under a store-and-forward human-based DR workflow ("Human Workflow"), an AI-based workflow involving IDx-DR ("AI Workflow"), and a two-step hybrid workflow ("AI-Human Hybrid Workflow"). The AI Workflow provided results within 48 hours, whereas the other workflows took up to 7 days. Patients were surveyed by phone about follow-up decisions.Under the AI Workflow, 279 patients screened positive for MTMDR. Of these, 69.2% followed up with an ophthalmologist within 90 days. Altogether 70.5% (N=48) of patients who followed up chose their location based on primary care referral. Among the subset of patients that were seen in person at the university eye institute under the Human Workflow and AI-Human Hybrid Workflow, 12.0% (N=14/117) and 11.7% (N=12/103) of patients with a referrable screening result followed up compared to 35.5% of patients under the AI Workflow (N=99/279; χ2df=2 = 36.70, p < 0.00000001).Ophthalmology follow-up after a positive DR screening result is approximately three-fold higher under the AI Workflow than either the Human Workflow or AI-Human Hybrid Workflow. Improved follow-up behavior may be due to the decreased time to screening result.

    View details for DOI 10.2147/OPTH.S422513

    View details for PubMedID 38026608

    View details for PubMedCentralID PMC10665027

  • How Many Lives Will You Save? A Mixed Methods Evaluation of a Novel, Online Game for Patient Safety and Quality Improvement Education. American journal of medical quality : the official journal of the American College of Medical Quality Ruiz Colon, G., Evans, K., Kanzawa, M., Phadke, A., Katznelson, L., Shieh, L. 2023


    Medical trainees have limited knowledge of quality improvement and patient safety concepts. The authors developed a free quality improvement/patient safety educational game entitled Safety Quest (SQ). However, 1803 undergraduate medical trainees, graduate medical trainees, and continuing medical education learners globally completed at least 1 level of SQ. Pre- and post-SQ knowledge and satisfaction were assessed among continuing medical education learners. Thematic analysis of feedback given by trainees was conducted. Among graduate medical trainees, SQ outranked other learning modalities. Three content areas emerged from feedback: engagement, ease of use, and effectiveness; 87% of comments addressing engagement were positive. After completing SQ, 98.6% of learners passed the post-test, versus 59.2% for the pretest (P < 0.0001). Ninety-three percent of learners agreed that SQ was engaging and interactive, and 92% believed it contributed to their professional growth. With an increased need for educational curricula to be delivered virtually, gamification emerges as a unique strategy that learners praise as engaging and effective.

    View details for DOI 10.1097/JMQ.0000000000000153

    View details for PubMedID 37882817

  • Targeted Electronic Patient Portal Messaging Increases Hepatitis C Virus Screening in Primary Care: a Randomized Study. Journal of general internal medicine Halket, D., Dang, J., Phadke, A., Jayasekera, C., Kim, W. R., Kwo, P., Downing, L., Goel, A. 2022


    IMPORTANCE: Electronic health record (EHR) tools such as direct-to-patient messaging and automated lab orders are effective at improving uptake of preventive health measures. It is unknown if patient engagement in primary care impacts efficacy of such messaging.OBJECTIVE: To determine whether more engaged patients, defined as those who have an upcoming visit scheduled, are more likely to respond to a direct-to-patient message with an automated lab order for hepatitis C virus (HCV) screening.DESIGN: Randomized trial PARTICIPANTS: One thousand six hundred randomly selected Stanford Primary Care patients, 800 with an upcoming visit within 6 months and 800 without, born between 1945 and 1965 who were due for HCV screening. Each group was randomly divided into cohorts of 400 subjects each. Subjects were followed for 1 year.INTERVENTION: One 400 subject cohort in each group received a direct-to-patient message through the EHR portal with HCV antibody lab order.MAIN OUTCOME AND MEASURE: The EHR was queried on a monthly basis for 6 months after the intervention to monitor which subjects completed HCV screening. For any subjects screened positive for HCV, follow-up through the cascade of HCV care was monitored, and if needed, scheduled by the study team.KEY RESULTS: Of 1600 subjects, 538 (34%) completed HCV screening. In the stratum without an upcoming appointment, 18% in the control group completed screening compared to 26% in intervention group (p<0.01). Similarly, in the stratum with an upcoming appointment, 34% in the control group completed screening compared to 58% in the intervention group (p<0.01).CONCLUSION: Direct-to-patient messaging coupled with automated lab orders improved HCV screening rates compared to standard of care, particularly in more engaged patients. Including this intervention in primary care can maximize screening with each visit, which is particularly valuable in times when physical throughput in the healthcare system may be low.

    View details for DOI 10.1007/s11606-022-07460-1

    View details for PubMedID 35230622

  • Transforming Interprofessional Roles During Virtual Health Care: The Evolving Role of the Medical Assistant, in Relationship to National Health Profession Competency Standards. Journal of primary care & community health Rokicki-Parashar, J., Phadke, A., Brown-Johnson, C., Jee, O., Sattler, A., Torres, E., Srinivasan, M. 2021; 12: 21501327211004285


    INTRODUCTION: Medical assistants (MAs) were once limited to obtaining vital signs and office work. Now, MAs are foundational to team-based care, interacting with patients, systems, and teams in many ways. The transition to Virtual Health during the COVID-19 pandemic resulted in a further rapid and unique shift of MA roles and responsibilities. We sought to understand the impact of this shift and to place their new roles in the context of national professional competency standards.METHODS: In this qualitative, grounded theory study we conducted semi-structured interviews with 24 MAs at 10 primary care sites at a major academic medical center on their experiences during the shift from in-person to virtual care. MAs were selected by convenience sample. Coding was done in Dedoose version 8.335. Consensus-based inductive and deductive approaches were used for interview analysis. Identified MA roles were compared to national MA, Institute of Medicine, physician, and nursing professional competency domains.RESULTS: Three main themes emerged: Role Apprehension, Role Expansion, and Adaptability/Professionalism. Nine key roles emerged in the context of virtual visits: direct patient care (pre-visit and physical care), panel management, health systems ambassador, care coordination, patient flow coordination, scribing, quality improvement, and technology support. While some prior MA roles were limited by the virtual care shift, the majority translated directly or expanded in virtual care. Identified roles aligned better with Institute of Medicine, physician, and nursing professional competencies, than current national MA curricula.CONCLUSIONS: The transition to Virtual Health decreased MA's direct clinical work and expanded other roles within interprofessional care, notably quality improvement and technology support. Comparison of the current MA roles with national training program competencies identified new leadership and teamwork competencies which could be expanded during MA training to better support MA roles on inter-professional teams.

    View details for DOI 10.1177/21501327211004285

    View details for PubMedID 33764223

  • Qualitative Assessment of Rapid System Transformation to Primary Care Video Visits at an Academic Medical Center. Annals of internal medicine Srinivasan, M. n., Asch, S. n., Vilendrer, S. n., Thomas, S. C., Bajra, R. n., Barman, L. n., Edwards, L. M., Filipowicz, H. n., Giang, L. n., Jee, O. n., Mahoney, M. n., Nelligan, I. n., Phadke, A. J., Torres, E. n., Artandi, M. n. 2020


    The coronavirus disease 2019 pandemic spurred health systems across the world to quickly shift from in-person visits to safer video visits.To seek stakeholder perspectives on video visits' acceptability and effect 3 weeks after near-total transition to video visits.Semistructured qualitative interviews.6 Stanford general primary care and express care clinics at 6 northern California sites, with 81 providers, 123 staff, and 97 614 patient visits in 2019.Fifty-three program participants (overlapping roles as medical providers [n = 20], medical assistants [n = 16], nurses [n = 4], technologists [n = 4], and administrators [n = 13]) were interviewed about video visit transition and challenges.In 3 weeks, express care and primary care video visits increased from less than 10% to greater than 80% and from less than 10% to greater than 75%, respectively. New video visit providers received video visit training and care quality feedback. New system workflows were created to accommodate the new visit method.Nine faculty, trained in qualitative research methods, conducted 53 stakeholder interviews in 4 days using purposeful (administrators and technologists) and convenience (medical assistant, nurses, and providers) sampling. A rapid qualitative analytic approach for thematic analysis was used.The analysis revealed 12 themes, including Pandemic as Catalyst; Joy in Medicine; Safety in Medicine; Slipping Through the Cracks; My Role, Redefined; and The New Normal. Themes were analyzed using the RE-AIM (reach, effectiveness, adoption, implementation, and maintenance) framework to identify critical issues for continued program utilization.Evaluation was done immediately after deployment. Although viewpoints may have evolved later, immediate evaluation allowed for prompt program changes and identified broader issues to address for program sustainability.After pandemic-related systems transformation at Stanford, critical issues to sustain video visit long-term viability were identified. Specifically, technology ease of use must improve and support multiparty videoconferencing. Providers should be able to care for their patients, regardless of geography. Providers need decision-making support with virtual examination training and home-based patient diagnostics. Finally, ongoing video visit reimbursement should be commensurate with value to the patients' health and well-being.Stanford Department of Medicine and Stanford Health Care.

    View details for DOI 10.7326/M20-1814

    View details for PubMedID 32628536

  • Enhancing patient engagement during virtual care: A conceptual model and rapid implementation at an academic medical center NEJM: Catalyst Innovations in Care Srinivasan, M., Phadke, A., Zulman, D., Thadaney, S., Madill, E., Savage, T., Downing, N., Nelligan, I., Artandi, M., Sharp, C. 2020

    View details for DOI 10.1056/CAT.20.0262

  • Differences and Trends in DNR Among California Inpatients With Heart Failure. Journal of cardiac failure Phadke, A., Heidenreich, P. A. 2016; 22 (4): 312-315


    Do-not-resuscitate (DNR) orders reflect an important means of respecting patient autonomy while minimizing the risk of nonbeneficial interventions. We sought to clarify trends and differences in rates of DNR orders for patients hospitalized with heart failure.We used statewide data from California's Healthcare Cost and Utilization dataset (2007-2010) to determine trends in DNR orders within 24 hours of admission for patients with a primary discharge diagnosis of heart failure.Among 347,541 hospitalizations for heart failure, the rate of DNR order within 24 hours increased from 10.4% in 2007 to 11.3% in 2010 (P < .0001). After adjustment, DNR status correlated with older age, female gender, white race, frequent comorbidities (Charlson Score), and residence in higher income area (P < .0001). DNR use was more likely in hospitals with public or nonprofit financing or medical school affiliation, but not being a member of the Council on Teaching Hospitals (all P < .001).DNR order use among inpatients with heart failure is low but increasing slowly and varies by patient demographics and hospital characteristics.

    View details for DOI 10.1016/j.cardfail.2015.12.005

    View details for PubMedID 26700659