About

Jason Karlawish, M.D., is a Professor of Medicine (Geriatrics) at the University of Pennsylvania's Perelman School of Medicine. He is a Fellow at the Institute on Aging and serves as the Director of the Outreach: Recruitment Core at the Alzheimer's Disease Center. Dr. Karlawish is also an Attending Physician at the Penn Memory Center and a Senior Fellow at the Leonard Davis Institute. He holds multiple roles within the university, including faculty positions at the Penn Center for Neuroscience and Society, the Penn Neurodegenerative Disease Ethics and Policy Program, and the Penn Healthy Brain Research Center. His work focuses on Alzheimer's disease, neurodegenerative diseases, and aging, with a particular emphasis on ethics, policy, and improving diagnosis, treatment, and care for patients. Dr. Karlawish has contributed to advancing understanding in these areas through research, leadership, and public engagement.

Research topics

• Medicine
• Psychology
• Internal medicine
• Pathology
• Biology
• Neuroscience
• Computer Science
• Machine Learning
• Psychiatry
• Oncology
• Artificial Intelligence
• Family medicine
• Audiology
• Mathematics
• Cognitive science
• Nuclear medicine
• Genetics
• Data science
• Epistemology
• Econometrics
• Physical therapy
• Philosophy
• Theoretical computer science

Selected publications

• Expert consensus on communicating tau PET results to persons living with MCI or dementia: Findings from a modified Delphi study

  Alzheimer s & Dementia · 2026-04-01

  articleOpen accessSenior authorCorresponding

  INTRODUCTION: There is interest in incorporating tau imaging into clinical care because it provides unique diagnostic and prognostic information. Yet, clinicians lack guidance on communicating results. METHODS: We conducted a modified Delphi process with practicing US-based expert clinicians in human tau imaging. Expert clinicians were interviewed to elicit input on best practices for communicating tau positron emission tomography (PET) results to cognitively impaired patients. These data were used to develop candidate practices and statements, which expert clinicians rated across two online survey rounds. RESULTS: Eighteen expert clinicians completed interviews and both surveys. They reached consensus on 12 practices - like showing individuals their tau PET scan images - and eight statements for communicating tau PET results - one for tau alone, three for concordant amyloid and tau results, and four for discordant amyloid and tau results - to cognitively impaired patients. DISCUSSION: This study provides novel consensus recommendations for communicating tau PET results to cognitively impaired patients.

  PublisherDOI

• Study of the utility and impact of a plasma p‐tau181 Alzheimer's biomarker (SUITABLE)

  Alzheimer s & Dementia Translational Research & Clinical Interventions · 2026-01-01

  articleOpen accessSenior author

  Abstract INTRODUCTION Blood‐based Alzheimer's disease (AD) biomarkers are increasingly used in clinical care, but there are few real‐world studies of their clinical utility. We evaluated the impact of plasma phosphorylated tau at threonine 181 (p‐tau181) on the diagnosis and management of patients with cognitive impairment in a memory clinic. METHODS Dementia specialists referred patients for plasma p‐tau181 testing during clinical care. We used surveys and chart review to assess the impact of plasma p‐tau181 testing on clinical care and patient and care partner wellbeing. We assessed co‐primary outcomes of (1) change in diagnostic confidence before vs after testing, (2) change in diagnosis, and (3) change in management. A diagnosis or management change of greater than 30% was considered to be clinically significant. Secondary and exploratory outcomes included impact of plasma p‐tau181 testing on the use of additional diagnostics and impact on patient and care partner psychological wellbeing, health behaviors, and future planning. RESULTS One hundred patients completed plasma p‐tau181 testing; 81% had an elevated result, and 19% had a non‐elevated result. Clinician diagnostic confidence averaged 1.43 points higher after p‐tau181 testing (95% confidence interval [CI] 1.02 to 1.84, p < 0.0001). After testing, 27.7% (95% CI 18.9% to 36.4%, p = 0.31) of patients had a change in management and 17.2% of patients had a change in diagnosis – significantly lower than the 30% threshold (95% CI 9.5% to 24.9%, p = 0.004). Patients with elevated results were less likely to have a diagnosis changed after testing than those with non‐elevated results (OR 0.08 [95% CI 0.01 to 0.47], p = 0.005). A non‐elevated result was associated with decreased use of amyloid positron emission tomography (PET) (odds ratio [OR] 0.25 [95% CI 0.07 to 0.86], p = 0.028). Patients with an elevated result had higher anxiety at 2 to 3 months after disclosure ( p = 0.01). DISCUSSION Non‐elevated plasma p‐tau181 results impacted the diagnosis and use of amyloid PET imaging in a memory clinic. There was increased anxiety in patients after testing, affirming the need for post‐disclosure psychological support.

  PublisherDOI

• Amyloid Imaging and <i>APOE</i> Genotype Disclosure and Short-Term Psychological Distress

  JAMA Network Open · 2026-03-30

  articleOpen accessSenior author

  Importance: Alzheimer disease (AD) biomarker and genetic testing results are increasingly disclosed to cognitively unimpaired adults in research and could in the future inform clinical treatment decisions in this population. Objectives: To assess psychological outcomes after returning 3 categories of amyloid biomarker results as well as apolipoprotein E (APOE) genotypes. Design, Setting, and Participants: This cohort study was a secondary analysis of data collected as part of screening for the multisite AHEAD preclinical AD trial. Participants were individuals aged 55 to 80 years undergoing screening from July 14, 2020, to October 15, 2024. Exposure: Participants were informed whether they had not-detected, intermediate, or elevated amyloid positron emission tomography levels, as well as their APOE genotype, which were categorized as noncarrier, ε4 heterozygote, or ε4 homozygote. Main outcomes and measures: Impact of Events Scale (IES; 15 items to assess intrusive thoughts and avoidance; each item is scored as not at all [0], rarely [1], sometimes [3], or often [5]; total range, 0-75), collected 24 to 72 hours after disclosure, and change in a scale measuring concerns about AD dementia (adapted scale using 6 items in which participants indicated their level of agreement with statements related to their perceived probability of developing AD dementia; items scored as strongly disagree [1] through strongly agree [5]; total range, 6-30), calculated by subtracting the score collected before biomarker testing from 1 collected after biomarker and genetic test results disclosure. Results: Among 3414 included individuals, the mean (SD) age was 68.8 (6.0) years and 2116 (62%) were female. Group mean IES scores were below clinically significant thresholds. Nevertheless, across genetic groups, learning an elevated amyloid result (1184 participants) was associated with higher IES (mean [SD], 10.5 [10.9]) than intermediate amyloid (482 participants; mean [SD] IES, 8.8 [9.8]), and intermediate amyloid was associated with higher scores than not-detected amyloid (1748 participants; mean [SD] IES, 6.5 [8.4]). Across amyloid groups, learning APOE ε4 homozygosity (337 participants) was associated with higher mean (SD) IES (12.7 [11.6]) than heterozygosity (1609 participants; 9.1 [10.2]), and heterozygosity was associated with higher IES than noncarrier status (1468 participants; mean [SD] IES, 6.2 [8.1]). Both types of information were significant in an analysis of covariance model; no interaction effect was observed. In contrast, only biomarker disclosure was associated with differential change in concerns about AD dementia. Those with elevated amyloid showed a mean (SD) increase in concern (0.8 [3.5]), those with intermediate amyloid showed a smaller increase (0.4 [3.7]), and those with not-detected amyloid showed decreased concerns (-1.1 [4.2]). Conclusions and Relevance: In this cohort study of cognitively unimpaired adults, associations with intrusive thoughts were observed to differ among genetic and biomarker subgroups; such associations were limited to biomarker subgroups for measures of perceived dementia risk.

  PublisherDOI

• Characterizing the research participant recruitment funnel for Alzheimer's secondary prevention trials: Results from a survey of U.S. adults

  Alzheimer s & Dementia Behavior & Socioeconomics of Aging · 2025-08-28

  articleOpen access

  Introduction: The steps of participant recruitment are described as a "funnel." In AD secondary prevention trials, the funnel typically yields an unrepresentative study population. Methods: A national survey of U.S. adults aged 65 and over assessed intention to enroll in a hypothetical secondary prevention trial. Among those intending to enroll, intention to ask someone to serve as their study partner was also measured. Results: In the full sample (n=603), intention to enroll - measured on a scale of 1 (very unlikely) to 5 (very likely) - was low (M: 2.37, SD: 1.31). Among those somewhat or very likely to enroll (24%, n=168), mean intention to ask a study partner was 3.45 out of 5 (SD 1.44). There were no significant differences between Black and White respondents. Discussion: Our results suggest underrepresentation of Black older adults in secondary prevention trials is not attributable to lower intention to participate or ask a study partner.

  PublisherOA PDFDOI

• Evaluation of the Revised Criteria for Biological and Clinical Staging of Alzheimer Disease

  JAMA Neurology · 2025-05-19 · 34 citations

  articleOpen access

  Importance: While clinical disease stages remained largely unchanged in the 2024 update of the Alzheimer disease (AD) criteria, tau-positron emission tomography (PET) was introduced as a core biomarker and its spatial extent was incorporated into the revised biological stages of the disease. It is important to consider both the clinical and the biological stages and understand their discrepancies. Objective: To compare individuals who have discrepant biological and clinical stages with those who have congruent stages in terms of copathologies, comorbidities, and demographics. Design, Setting, and Participants: Participants were from the Swedish BioFINDER-2 (inclusion from 2017 through 2023) and the Alzheimer's Disease Neuroimaging Initiative (ADNI) (inclusion from 2015 through 2024). BioFINDER-2 included a prospective population-based (cognitively normal [CN] older adults) and memory clinic-based cohort (participants with subjective cognitive impairment [SCD], mild cognitive impairment [MCI], and dementia). ADNI included a volunteer-based sample. All participants who were amyloid-β positive and had undergone tau-PET were included. In BioFINDER-2, 838 participants of a total of 1979 were included, and of 927 with tau-PET in ADNI, 380 were included. Exposures: The clinical (CN to dementia) and biological (based on PET; initial [amyloid-β-positive only] to advanced [amyloid-β-positive, elevated, and widespread tau]) stages from the revised AD criteria. Main Outcomes and Measures: Cross-sectional measures of neurodegeneration (cortical thickness, TAR DNA-binding protein 43 [TDP-43] imaging signature, neurofilament light [NfL]), α-synuclein cerebrospinal fluid status, plasma glial fibrillary acidic protein, white matter lesions, infarcts, microbleeds, comorbidities, and demographics. Results: There were 838 BioFINDER-2 participants (mean age, 73.9 [SD, 7.3] years; 431 women [51%]; 407 men [49%]) and 380 ADNI participants (average age, 72.9 [SD, 7.0] years; 194 women [51%]; 186 mean [49%]) included. In BioFINDER-2, 37.7% of the sample had congruent biological and clinical stages (reference group), 51.3% had more advanced clinical impairment compared with their clinical stage (clinical > biological) and 11.0% had the opposite (biological > clinical). The main differences were between the reference group and the clinical > biological group: the latter participants were more often positive for α-synuclein pathology, had higher NfL levels, greater TDP-43-like atrophy, and higher burden of cerebral small vessel disease lesions (all false discovery rate P < .05). The only difference between the biological > clinical and the reference group was that the former had less neurodegeneration (thicker cortex; all false discovery rate P < .001). The main results were replicated in the independent ADNI cohort, where congruent 56.1% of participants had biological and clinical stages; 36.1% were in the category clinical > biological, and 7.9% in biological > clinical. Conclusions and Relevance: Copathologies play an important role in symptom severity in individuals who harbor less tau-tangle pathology than expected for their clinical impairment. These results highlight the importance of measuring non-AD biomarkers in patients with AD with worse cognitive impairment than expected based on their biological stage, which could impact the clinical diagnosis and prognosis.

  PublisherOA PDFDOI

• Ethics From the Outset: Incorporating Ethical Considerations into the Artificial Intelligence and Technology Collaboratories for Aging Research Pilot Projects

  The Journals of Gerontology Series A · 2025-03-28 · 2 citations

  articleOpen access

  There is an urgent need to develop tools to enable older adults to live healthy, independent lives for as long as possible. To address this need, the National Institute on Aging (NIA) Artificial Intelligence and Technology Collaboratories (AITCs) for Aging Research were created to identify, develop, evaluate, commercialize, and disseminate innovative technologies and artificial intelligence (AI) methods to promote healthy aging and to support persons with Alzheimer's disease and Alzheimer's disease-related dementias (AD/ADRD). In 2023, AITC pilot grant applicants were required to answer questions about how, if at all, they would safeguard older adults' data privacy and confidentiality, advance health equity, address bias, and protect vulnerable participants. Our team analyzed applicants' answers to these ethics-focused questions using a constructivist grounded theory approach. In this article, we present what we learned and discuss modifications to our approach moving forward.

  PublisherOA PDFDOI

• Disclosure of elevated amyloid status is not associated with long‐term suicidality in a preclinical AD trial

  Alzheimer s & Dementia · 2025-02-01 · 5 citations

  articleOpen access

  INTRODUCTION: The long-term implications of disclosing Alzheimer's disease (AD) biomarker information to cognitively unimpaired individuals are unknown. METHODS: We compared participants who disclosed their elevated amyloid imaging result in a preclinical AD trial to those who disclosed a not elevated result and enrolled in an observational cohort that underwent parallel assessments. Our primary outcome was a score > 0 on the Columbia Suicidality Severity Rating Scale (CSSRS) at any visit; we also considered suicidal behaviors (CSSRS > 5). RESULTS: Among 1707 total participants (68% elevated amyloid, mean [standard deviation] age 71.5 [4.7], 60% female, 90% non-Hispanic White), followed for a mean 218 (74.1) weeks, there were no suicides and few indications of suicidal thoughts (n = 124 [7%]) or behaviors (n = 13 [<1%]). In a generalized estimating equation model controlling for covariates, we observed no effect of amyloid status on the primary outcome of CSSRS > 0 (odds ratio = 1.6, 95% confidence interval = 0.76, 3.37). DISCUSSION: With a structured approach, brain amyloid results can be returned safely. HIGHLIGHTS: The Anti-Amyloid Treatment in Asymptomatic Alzheimer's study was among the first and largest studies to include biomarker disclosure in a population without cognitive impairment. Routine psychological assessment provided a novel assessment of the impact of disclosure in this sample. Learning an elevated brain amyloid result through a protocolized approach was not associated with suicidal thoughts or behaviors compared to a matched cohort who learned they did not have elevated brain amyloid. Future research will be needed to ensure similar safety in more real-world settings.

  PublisherOA PDFDOI

• Task Order Moderates the Effects of APOE Genotype Awareness on Digit Span Performance in Cognitively Unimpaired Older Adults

  Alzheimer s & Dementia · 2025-12-01

  articleOpen access

  BACKGROUND: As the prevalence of APOE testing rises, it is crucial to understand how APOE genotype and genotype awareness impact cognitive performance. This study investigates the effects of APOE genotype, genotype awareness, and task order on digit span performance in cognitively unimpaired older adults. METHOD: = 71.14, range = 63-79), 69.4% were APOE ε4 carriers. A total of 69.1% of ε4 carriers and 53.3% of ε4 non-carriers were aware of their genotype. Participants were told the study aimed to examine the effects of APOE genotype on cognitive performance. After receiving this instruction participants completed a randomized battery of cognitive tests, including Forward and Backward Digit Span (FDS, BDS). The primary analysis was a MANOVA examining the effects of APOE genotype, and genotype awareness, on FDS and BDS performance. An exploratory analysis assessed whether the results varied based on task order. RESULT: In the primary analysis there were no significant results (ps > .174). However, adding task order as a factor resulted in a significant interaction between genotype, genotype awareness, and task order for BDS (p = .026). Follow-up 2(genotype awareness) X 2(task order) ANOVAs on BDS revealed a significant interaction between these factors for ε4 carriers (p = .002) but not for ε4 non-carriers (p = .552). Among the ε4 carriers, knowledge of genotype was associated with numerically poorer performance when digit span tasks were completed first (Aware: M = 8.69, Unaware: M = 10.69; p = .052) but significantly better performance when digit span tasks were completed last (Aware: M = 10.35, Unaware: M = 7.56; p = .016). CONCLUSION: Among APOE ε4 carriers the effects of genotype awareness on BDS performance varied based upon whether the digit span tasks were completed first or last within the cognitive battery. Among ε4 carriers who performed the digit span tasks first, genotype awareness was linked to lower BDS performance, possibly due to stereotype threat. This effect reversed for among ε4 carriers who completed the digit span tasks last. These findings align with research suggesting stronger threat effects when the dependent variable is measured soon after the manipulation (Lamont et al., 2015).

  PublisherOA PDFDOI

• A Short Story Changed How I Care for Persons Living With Dementia

  Journal of the American Geriatrics Society · 2025-04-01

  articleOpen access1st authorCorresponding

  “A woman goes to her doctor to have a prescription renewed. But the doctor is not there. It's her day off. In fact, the woman has got the day wrong, she has mixed up Monday with Tuesday. This is the very thing she wants to talk to the doctor about.” [1] These opening lines of Alice Munro's “In Sight of the Lake”—a short story about Nancy, an elderly woman worried her mind is “slipping a bit”—pull me out of my experience as a dementia specialist and into the experience of a person living with dementia. The practice of medicine relies on facts as best as we can gather and assess them. Narratives, however, can have a role in shaping medical practice as well. This story, (Because, even after publication, Munro often revised her stories, several versions are in print; this essay is based on the publication in the collection Dear Life1) a fiction none-the-less, is one such example. It has had a great influence on how I think about the care of persons living with dementia. Nancy's visit to the doctor's office launches her on a journey. It begins the day after the visit to renew the prescription, when an unnamed “office girl”—Munro's words—calls Nancy. The prescription is ready to pick up and an appointment has been made with a specialist to look into her “mind problem.” Nancy rejects the term. “It isn't mind. It's just memory.” The reader can hear the girl's dismissive shrug as she replies, “Whatever. The specialist deals with elderly patients.” Nancy parries with a quip: “Elderly patients who are off their nuts.” Still, she accepts the referral to the specialist, located, the girl explains, in the village whose name Nancy hears as “hymen.” One more quip. “Oh dear,” Nancy remarks, “a marriage specialist.” Afraid she might struggle to find the doctor's office or, arrive late and flustered, and so create a bad impression, Nancy makes a plan: On the eve of her appointment, she sets off in her car to become comfortable with the route from her home to the Elderly Specialist's office in the village of Highman. There, she sets out to find his office, but in her walk about the village, she struggles to locate the office of the doctor whose name she has forgotten. She also struggles over whether to ask for help. Finally, she relents and asks a man for directions. With these, she resumes her journey. Right from the opening—Nancy's mix up of the dates, the very thing she wants to talk to the doctor about—we're drawn into Nancy's conscious experiences. Consciousness is more than simply being awake and alert. The term describes our capacity to experience the world. It is what happens when cognition, perceptions and feelings assemble into a subjective experience. Consciousness ought to matter to doctors who care for persons living with dementia, especially geriatricians. A conscious creature has a mind, and dementia is widely understood as a condition where mind is incrementally transformed, even lost. No wonder mind is one of the “four M's” of geriatric medicine. “Crazy doctor,” “elderly patients off their nuts,” “oh dear, a marriage specialist,” “It isn't mind. It's just memory”—Nancy's wit, word play and quibbles with terms might be interpreted as her minimizing or trying to deflect her problems. They might, but they also reflect her experiences struggling to live with dementia. She is intensely aware of her problems and fears they will cause stigma and a loss of connection with people. I think of her pre-visit trip to find the doctor's office as yet another example of the supports a patient uses to minimize the impact of their memory loss. Other cognitive supports are notes and white boards (I once found a patient's handwritten note “Aricept—for memory lapses.”). As long as these work, they're why the common answer to my question “How are you are doing with daily tasks?” is “I'm doing fine.” For years, I made little effort to probe this answer. It was proof enough: The patient lacked awareness of their problems with day-to-day life. But then I read this story. It got me thinking. What is mind, and how do we assess it? Here's what I discovered. “I'm doing fine” may in fact signify a lack of awareness, but it may not. How do I discover whether the patient is like Nancy, aware of the changes in how their mind functions? To answer this, I assess the patient's conscious experiences. I routinely ask a question taken from philosophical inquiries into consciousness: What is it like to be you? [2] Something we conclude has a ‘feeling of what it is like’ is a conscious entity. I think my dog Moby has a feeling of what it is like to be. I don't think that ChatGPT does. In answering this question, many patients recount vivid experiences like Nancy's confusion of time and her struggles to find her way around Highman. Following up with How do you cope with those experiences? often reveals an important finding. They don't know how they'd mange without their notepad, smartphone, geotracker, pill box and especially without the person we call a caregiver. With these mind supports in place, they're in fact “doing fine.” This is an entre for an important, albeit difficult conversation. What kinds of supports might they need when the supports they have in place are no longer working. A critical reader should rightly ask, there are many fictional accounts of persons living with dementia. What's special about this story? What's special are the events that ensue after Nancy resumes her journey. They show how mind isn't simply in the brain and the experience of dementia isn't simply the consequence of damage to that brain. Mind emerges from an interaction between the brain and the world around it. For persons living with dementia, this interaction is of great ethical importance. We the caregivers have at least some control over their world. Nancy's journey to find the “Crazy Doctor” (she's changed the description) leads her to the Lakeview Rest Home. Initial feelings of pleasure stirred by the sight of the structure's lattice fences soon transform into dread and fear. The silvery tiled floor looks slippery. The panes of the glass doors are wavy and distorted. The doors have knobs, but they don't open. This built environment is threatening. “There is nothing to do but get out of this place and go home,” she decides. But she cannot leave. We discover this has all been a dream. Nancy is a resident of Lakeview. We meet Sandy, an aide at Lakeview, helping Nancy get ready for bed. “What are we going to do with you,” Sandy says. “All we want is to get you into your nightie. And you go and carry on like a chicken that's scared of being et for dinner.” In the diction of long-term care, Nancy needs assistance with basic activities of daily living but she's resisting this assistance. She's agitated. The seemingly ordinary dialogue that follows is profoundly disturbing. Nancy says she was dreaming of a time when her husband was alive, and she was still driving a car. Sandy's reply is seemingly ordinary—“You have a nice car?”—but it is her choice of verb tense—“have” not “had”—and the topic of her question that are, together, extraordinarily absurd. Nancy was talking about her past (“When I was still driving a car”) and yet Sandy—and you can hear the weird, sing-song elder speak—asks about the car in the present tense, and her question neither substantively nor emotionally connects with what Nancy told her. Nancy was dreaming about two significant losses—a vehicle that gave her independence and her husband, but Sandy blandly replies with a question about whether the car was “nice,” whatever that means. But Nancy doesn't argue. Instead, she yet again acquiesces to absurdity. I write “yet again,” because Munro explains Nancy “knows that she has to behave differently, and more than that has to believe differently.” She feels the absurdity of how she is treated. Humans designed and named the frightening place called Lakeview Rest Home. Sandy chose her absurd words—“you have a nice car?”—that unmoored Nancy from time and denied her emotional experiences. She described Nancy not as a human in distress, but an agricultural animal off to the slaughter. Sandy extinguished a lucid moment and so an opportunity to connect with Nancy. The message is to us, the people who care for persons like Nancy. When we talk like Sandy and build structures like Lakeview, we're harming the person's conscious experiences. We are harming the patient's mind. We are contributing to dementia. Jason Karlawish drafted, revised and approves of the final manuscript. The author has nothing to report. The sponsor had no role in drafting, revising or approval of the final manuscript. In the last 12 months, Jason Karlawish was on the scientific advisory board of Linus Health and provided consultation to Biogen.

  PublisherOA PDFDOI

• Perspectives of Patients and Care Partners on Benefits or Burdens of Delaying Dementia Progression

  JAMA Network Open · 2025-06-23 · 1 citations

  articleOpen access

  This qualitative study examines patient and care partner perspectives about the potential of disease-modifying therapies to delay dementia progression.

  PublisherOA PDFDOI

Recent grants

• Ethics and Regulation Core (C)

  NIH · $68.6M · 2019–2030

• NIH Grant K01AG000931

  NIH · $538k · 2004

• NIH Grant R01MH071643

  NIH · $1.2M · 2010

• NIH Grant R01AG020627

  NIH · $828k · 2008

• NIH Grant RF1AG047866

  NIH · $6.0M · 2020

Frequent coauthors

• Joshua D. Grill

  University of California, Irvine

  252 shared

• Michael Donohue

  Janssen (United States)

  231 shared

• Reisa A. Sperling

  Harvard University

  223 shared

• Paul Aisen

  University of Southern California

  211 shared

• Judith L. Heidebrink

  206 shared

• Steven E. Arnold

  Harvard University

  200 shared

• Gregory A. Jicha

  199 shared

• John C. Morris

  Washington University in St. Louis

  194 shared

Education

• BS

  Northwestern University

• MD

  Northwestern University Feinberg School of Medicine