Bio


Anesthesiologist and internist interested in artificial intelligence and large language models in medicine. Currently, my primary focus is on developing and evaluating applications of large language models to improve the "user experience" of patients (who spend too much time fighting the system that is tasked with healing them) and doctors (who spend too much time fighting the system that is supposed to help them be healers).

Interests:
- Benchmarking LLMs as clinical calculators
- Medical summarization by LLMs
- Agentic /tool-using language models
- GenerativeAI for Medical Education and Simulation
- Data equity in LLMs
- Novel benchmarks for clinical LLMs, including simulation
- Participatory research, open-source software

I'm a Clinical Scholar in the Dept of Anesthesiology and a Post-Doctoral Fellow in Anesthesiology / Biomedical Data Science in the lab of Nima Aghaeepour.

I completed medical school at the UC Berkeley - UCSF Joint Medical Program, followed by the Combined Internal Medicine/Anesthesiology Residency at the Stanford School of Medicine, and a fellowship in Anesthesia Informatics at the Stanford AIM Lab.

Clinical Focus


  • Anesthesia
  • Geriatric anesthesia

Academic Appointments


Professional Education


  • Certificate, Hasso Plattner Institute of Design, Stanford University, Design Thinking (Executive Education Bootcamp) (2023)
  • Residency in Anesthesiology, Anesthesiology Residency, Stanford University, Anesthesiology (2023)
  • DABIM, Internal Medicine Residency, Stanford Universirty, Internal Medicine (2023)
  • MD, University of California San Francisco, Medicine; Veteran's Affairs Clinical Track (2018)
  • MS, University of California Berkeley, Health and Medical Sciences; focus in epidemiology and health economics (2016)
  • BS, University of Oregon, Biology (2010)

Stanford Advisors


All Publications


  • Enhancing the Readability of Preoperative Patient Instructions Using Large Language Models. Anesthesiology Hong, H. J., Schmiesing, C. A., Goodell, A. J. 2024; 141 (3): 608-610

    View details for DOI 10.1097/ALN.0000000000005122

    View details for PubMedID 39136480

  • Prevalence, mutational spectrum and clinical implications of clonal hematopoiesis of indeterminate potential in plasma cell dyscrasias. Seminars in oncology Testa, S., Kumar, J., Goodell, A. J., Zehnder, J. L., Alexander, K. M., Sidana, S., Arai, S., Witteles, R. M., Liedtke, M. 2022

    Abstract

    Clonal hematopoiesis of indeterminate potential (CHIP) is common both in healthy individuals and patients with hematological cancers. Recent studies have showed worse prognosis for patients with multiple myeloma (MM) and non-Hodgkin lymphoma undergoing stem cell transplant, that have concomitant presence of CHIP. Data regarding the clinical and biological role of CHIP in plasma cell dyscrasias (PCDs) is rapidly increasing. However, the prevalence and prognostic implication of CHIP in patients with MM outside of the transplant setting, and in those with other more indolent PCDs remains elusive. Here we explored the prevalence and clinical implications of CHIP detected through next-generation sequencing in 209 patients with PCDs including MM, light chain (AL) amyloidosis (ALA), monoclonal gammopathy of undetermined significance (MGUS), and smoldering multiple myeloma (SMM). To avoid attributing the mutations to the plasma cell clone, CHIP was defined as the presence of DNMT3A, TET2, or ASXL1 mutations in the peripheral blood or bone marrow (DTA-CH). The prevalence of DTA-CH was 19% in patients with PCDs, with no difference between each PCD. TET2 (23%) and DNMT3A (22%), were the most frequently mutated genes. DTA-CH correlated with older age in MM (P = .001) and MGUS/SMM (P = 0.0007), as well as with coronary artery disease or congestive heart failure in MM (P = .03). DTA-CH did not predict worse OS or PFS in either MM or ALA, nor it predict higher risk of progression to MM in patients with MGUS/SMM. Our results overall further elucidate the prevalence and mutational spectrum of CHIP in PCDs, providing more information regarding the clinical relevance of CHIP in this patient population.

    View details for DOI 10.1053/j.seminoncol.2022.11.001

    View details for PubMedID 36503855

  • Modeling the Impact of Recommendations for Primary Care-Based Screening for Latent Tuberculosis Infection in California. Public health reports (Washington, D.C. : 1974) Parriott, A., Kahn, J. G., Ashki, H., Readhead, A., Barry, P. M., Goodell, A. J., Flood, J., Shete, P. B. 2020; 135 (1_suppl): 172S-181S

    Abstract

    Targeted testing and treatment of persons with latent tuberculosis infection (LTBI) is a critical component of the US tuberculosis (TB) elimination strategy. In January 2016, the California Department of Public Health issued a tool and user guide for TB risk assessment (California tool) and guidance for LTBI testing, and in September 2016, the US Preventive Services Task Force (USPSTF) issued recommendations for LTBI testing in primary care settings. We estimated the epidemiologic effect of adherence to both recommendations in California.We used an individual-based Markov micro-simulation model to estimate the number of cases of TB disease expected through 2026 with baseline LTBI strategies compared with implementation of the USPSTF or California tool guidance. We estimated the risk of LTBI by age and country of origin, the probability of being in a targeted population, and the probability of presenting for primary care based on available data. We assumed 100% adherence to testing guidance but imperfect adherence to treatment.Implementation of USPSTF and California tool guidance would result in nearly identical numbers of tests administered and cases of TB disease prevented. Perfect adherence to either recommendation would result in approximately 7000 cases of TB disease averted (40% reduction compared with baseline) by 2026. Almost all of this decline would be driven by a reduction in the number of cases among non-US-born persons.By focusing on the non-US-born population, adherence to LTBI testing strategies recommended by the USPSTF and the California tool could substantially reduce the burden of TB disease in California in the next decade.

    View details for DOI 10.1177/0033354920927845

    View details for PubMedID 32735191

    View details for PubMedCentralID PMC7407051

  • Outlook for tuberculosis elimination in California: An individual-based stochastic model. PloS one Goodell, A. J., Shete, P. B., Vreman, R., McCabe, D., Porco, T. C., Barry, P. M., Flood, J., Marks, S. M., Hill, A., Cattamanchi, A., Kahn, J. G. 2019; 14 (4): e0214532

    Abstract

    RATIONALE: As part of the End TB Strategy, the World Health Organization calls for low-tuberculosis (TB) incidence settings to achieve pre-elimination (<10 cases per million) and elimination (<1 case per million) by 2035 and 2050, respectively. These targets require testing and treatment for latent tuberculosis infection (LTBI).OBJECTIVES: To estimate the ability and costs of testing and treatment for LTBI to reach pre-elimination and elimination targets in California.METHODS: We created an individual-based epidemic model of TB, calibrated to historical cases. We evaluated the effects of increased testing (QuantiFERON-TB Gold) and treatment (three months of isoniazid and rifapentine). We analyzed four test and treat targeting strategies: (1) individuals with medical risk factors (MRF), (2) non-USB, (3) both non-USB and MRF, and (4) all Californians. For each strategy, we estimated the effects of increasing test and treat by a factor of 2, 4, or 10 from the base case. We estimated the number of TB cases occurring and prevented, and net and incremental costs from 2017 to 2065 in 2015 U.S. dollars. Efficacy, costs, adverse events, and treatment dropout were estimated from published data. We estimated the cost per case averted and per quality-adjusted life year (QALY) gained.MEASUREMENTS AND MAIN RESULTS: In the base case, 106,000 TB cases are predicted to 2065. Pre-elimination was achieved by 2065 in three scenarios: a 10-fold increase in the non-USB and persons with MRF (by 2052), and 4- or 10-fold increase in all Californians (by 2058 and 2035, respectively). TB elimination was not achieved by any intervention scenario. The most aggressive strategy, 10-fold in all Californians, achieved a case rate of 8 (95% UI 4-16) per million by 2050. Of scenarios that reached pre-elimination, the incremental net cost was $20 billion (non-USB and MRF) to $48 billion. These had an incremental cost per QALY of $657,000 to $3.1 million. A more efficient but somewhat less effective single-lifetime test strategy reached as low as $80,000 per QALY.CONCLUSIONS: Substantial gains can be made in TB control in coming years by scaling-up current testing and treatment in non-USB and those with medical risks.

    View details for PubMedID 30964878

  • Health and economic benefits of reducing sugar intake in the USA, including effects via non-alcoholic fatty liver disease: a microsimulation model. BMJ open Vreman, R. A., Goodell, A. J., Rodriguez, L. A., Porco, T. C., Lustig, R. H., Kahn, J. G. 2017; 7 (8): e013543

    Abstract

    Excessive consumption of added sugars in the human diet has been associated with obesity, type 2 diabetes (T2D), coronary heart disease (CHD) and other elements of the metabolic syndrome. Recent studies have shown that non-alcoholic fatty liver disease (NAFLD) is a critical pathway to metabolic syndrome. This model assesses the health and economic benefits of interventions aimed at reducing intake of added sugars.Using data from US National Health Surveys and current literature, we simulated an open cohort, for the period 2015-2035. We constructed a microsimulation model with Markov chains for NAFLD (including steatosis, non-alcoholic steatohepatitis (NASH), cirrhosis and hepatocellular carcinoma (HCC)), body mass index, T2D and CHD. We assessed reductions in population disease prevalence, disease-attributable disability-adjusted life years (DALYs) and costs, with interventions that reduce added sugars consumption by either 20% or 50%.The model estimated that a 20% reduction in added sugars intake will reduce prevalence of hepatic steatosis, NASH, cirrhosis, HCC, obesity, T2D and CHD. Incidence of T2D and CHD would be expected to decrease by 19.9 (95% CI 12.8 to 27.0) and 9.4 (95% CI 3.1 to 15.8) cases per 100 000 people after 20 years, respectively. A 20% reduction in consumption is also projected to annually avert 0.767 million (M) DALYs (95% CI 0.757M to 0.777M) and a total of US$10.3 billion (B) (95% CI 10.2B to 10.4B) in discounted direct medical costs by 2035. These effects increased proportionally when added sugars intake were reduced by 50%.The decrease in incidence and prevalence of disease is similar to results in other models, but averted costs and DALYs were higher, mainly due to inclusion of NAFLD and CHD. The model suggests that efforts to reduce consumption of added sugars may result in significant public health and economic benefits.

    View details for DOI 10.1136/bmjopen-2016-013543

    View details for PubMedID 28775179

    View details for PubMedCentralID PMC5577881

  • Where teachers are few: documenting available faculty in five Tanzanian medical schools. Global health action Mkony, C. A., Kaaya, E. E., Goodell, A. J., Macfarlane, S. B. 2016; 9: 32717

    Abstract

    Faced with one of the lowest physician-to-population ratios in the world, the Government of Tanzania is urging its medical schools to train more physicians. The annual number of medical students admitted across the country rose from 55 in the 1990s to 1,680 approved places for the 2015/16 academic year. These escalating numbers strain existing faculty.To describe the availability of faculty in medical schools in Tanzania.We identified faculty lists published on the Internet by five Tanzanian medical schools for the 2011/12 academic year and analyzed the appointment status, rank, discipline, and qualifications of faculty members.The five schools reported 366 appointed faculty members (excluding visiting, part-time, or honorary appointments) for an estimated total enrolled student capacity of 3,275. Thirty-eight percent of these faculty were senior lecturers or higher. Twenty-seven percent of the appointments were in basic science, 51% in clinical science, and 21% in public health departments. The most populated disciplines (more than 20 faculty members across the five institutions) were biochemistry and molecular biology, medicine, obstetrics and gynecology, pediatrics, and surgery; the least populated disciplines (less than 10 faculty members) were anesthesiology, behavioral sciences, dermatology, dental surgery, emergency medicine, hematology, ophthalmology, orthopedics, otorhinolaryngology, oncology and radiology, psychiatry. These figures are only indicative of faculty numbers because of differences in the way the schools published their faculty lists.Universities are not recruiting faculty at the same rate that they are admitting students, and there is an imbalance in the distribution of faculty across disciplines. Although there are differences among the universities, all are struggling to recruit and retain staff. If Tanzanian universities, the government, donors, and international partners commit resources to develop, recruit, and retain new faculty, Tanzania could build faculty numbers to permit a quality educational experience for its doctors of tomorrow.

    View details for DOI 10.3402/gha.v9.32717

    View details for PubMedID 27741957

    View details for PubMedCentralID PMC5065695

  • Modeling solutions to Tanzania's physician workforce challenge. Global health action Goodell, A. J., Kahn, J. G., Ndeki, S. S., Kaale, E., Kaaya, E. E., Macfarlane, S. B. 2016; 9: 31597

    Abstract

    There is a great need for physicians in Tanzania. In 2012, there were approximately 0.31 physicians per 10,000 individuals nationwide, with a lower ratio in the rural areas, where the majority of the population resides. In response, universities across Tanzania have greatly increased the enrollment of medical students. Yet evidence suggests high attrition of medical graduates to other professions and emigration from rural areas where they are most needed.To estimate the future number of physicians practicing in Tanzania and the potential impact of interventions to improve retention, we built a model that tracks medical students from enrollment through clinical practice, from 1990 to 2025.We designed a Markov process with 92 potential states capturing the movement of 25,000 medical students and physicians from medical training through employment. Work possibilities included clinical practice (divided into rural or urban, public or private), non-clinical work, and emigration. We populated and calibrated the model using a national 2005/2006 physician mapping survey, as well as graduation records, graduate tracking surveys, and other available data.The model projects massive losses to clinical practice between 2016 and 2025, especially in rural areas. Approximately 56% of all medical school students enrolled between 2011 and 2020 will not be practicing medicine in Tanzania in 2025. Even with these losses, the model forecasts an increase in the physician-to-population ratio to 1.4 per 10,000 by 2025. Increasing the absorption of recent graduates into the public sector and/or developing a rural training track would ameliorate physician attrition in the most underserved areas.Tanzania is making significant investments in the training of physicians. Without linking these doctors to employment and ensuring their retention, the majority of this investment in medical education will be jeopardized.

    View details for DOI 10.3402/gha.v9.31597

    View details for PubMedID 27357075

    View details for PubMedCentralID PMC4926102

  • Babesia screening: the importance of reporting and calibration in cost-effectiveness models. Transfusion Goodell, A. J., Bloch, E., Simon, M. S., Shaz, B., Custer, B. 2016; 56 (3): 774-5

    View details for DOI 10.1111/trf.13436

    View details for PubMedID 26954454

  • Costs, consequences, and cost-effectiveness of strategies for Babesia microti donor screening of the US blood supply. Transfusion Goodell, A. J., Bloch, E. M., Krause, P. J., Custer, B. 2014; 54 (9): 2245-57

    Abstract

    Babesia microti is regarded as the foremost infectious risk to the US blood supply for which a regulatory-approved screening test is unavailable. More than 160 cases of transfusion-transmitted Babesia microti (TTB) have been reported to date, yet there is little consensus regarding a mitigation strategy.This study sought to assess the cost-utility of donation screening by mode of testing (immunofluorescence assay, enzyme-linked immunosorbent assay [ELISA], polymerase chain reaction [PCR], and combinations thereof) as well as extent of geographic inclusion (4-state, 7-state, 20-state, or national screening). A discrete-time Markov cohort model to simulate the outcomes of B. microti infection and survival of the transfused population was developed. Seroprevalence was estimated by extrapolating babesiosis claims from the Centers for Medicaid and Medicare Services and reports to the Centers for Disease Control and Prevention. Test performance was estimated from clinical diagnostics and limited donor screening studies, while transmissibility was estimated as a weighted average of three studies. Results are reported as the cost per quality-adjusted life-year (QALY) for each strategy compared to no screening.Given model inputs, 4-state and 7-state ELISA in combination with PCR would cost $5.2 million and $6.6 million/QALY, respectively. Cost-effectiveness for 20-state and national screening strategies were less favorable.Targeted screening in states with the highest seroprevalence of infection is likely to exceed an implicit threshold of $1 million/QALY often used in blood safety. However, the proportion of donor-seronegative parasitemia, transmissibility, and clinical outcomes resulting from TTB are uncertain.

    View details for DOI 10.1111/trf.12805

    View details for PubMedID 25109338

  • An institutional research agenda: focusing university expertise in Tanzania on national health priorities. Journal of public health policy Masalu, J. R., Aboud, M., Moshi, M. J., Mugusi, F., Kamuhabwa, A., Mgimwa, N., Freeman, P., Goodell, A. J., Kaaya, E. E., Macfarlane, S. B. 2012; 33 Suppl 1: S186-201

    Abstract

    A well-articulated institutional health research agenda can assist essential contributors and intended beneficiaries to visualize the link between research and community health needs, systems outcomes, and national development. In 2011, Tanzania's Muhimbili University of Health and Allied Sciences (MUHAS) published a university-wide research agenda. In developing the agenda, MUHAS leadership drew on research expertise in its five health professional schools and two institutes, its own research relevant documents, national development priorities, and published literature. We describe the process the university underwent to form the agenda and present its content. We assess MUHAS's research strengths and targets for new development by analyzing faculty publications over a five-year period before setting the agenda. We discuss implementation challenges and lessons for improving the process when updating the agenda. We intend that our description of this agenda-setting process will be useful to other institutions embarking on similar efforts to align research activities and funding with national priorities to improve health and development.

    View details for DOI 10.1057/jphp.2012.50

    View details for PubMedID 23254843

  • Tracking university graduates in the workforce: information to improve education and health systems in Tanzania. Journal of public health policy Pemba, S., Macfarlane, S. B., Mpembeni, R., Goodell, A. J., Kaaya, E. E. 2012; 33 Suppl 1: S202-15

    Abstract

    With a severe shortage of highly trained health professionals, Tanzania must make the best possible use of available human resources and support training institutions to educate more graduates. We highlight the overlooked but significant role of universities in collecting, managing, and using human resources data in Tanzania and in other countries struggling to build their health workforces. Although universities, professional councils, ministries of health, education, and finance, and non-governmental organizations in Tanzania all maintain databases that include details of health professionals' education, registration, and employment, they do not make the information easily accessible to one another. Using as an example Muhimbili University of Health and Allied Sciences - the leading public institution for health professions education in Tanzania - we explore how training institutions can gather and use data to target and improve the quality of education for increasing numbers of graduates. We specifically examine the substantial challenge universities face in locating more members of each graduating class and conclude with recommendations about how the situation can be improved.

    View details for DOI 10.1057/jphp.2012.48

    View details for PubMedID 23254844