American Medical Association Masterfile Research Articles

Demographic Representation in 3 Leading Artificial Intelligence Text-to-Image Generators

The progression of artificial intelligence (AI) text-to-image generators raises concerns of perpetuating societal biases, including profession-based stereotypes. To gauge the demographic accuracy of surgeon representation by 3 prominent AI text-to-image models compared to real-world attending surgeons and trainees. The study used a cross-sectional design, assessing the latest release of 3 leading publicly available AI text-to-image generators. Seven independent reviewers categorized AI-produced images. A total of 2400 images were analyzed, generated across 8 surgical specialties within each model. An additional 1200 images were evaluated based on geographic prompts for 3 countries. The study was conducted in May 2023. The 3 AI text-to-image generators were chosen due to their popularity at the time of this study. The measure of demographic characteristics was provided by the Association of American Medical Colleges subspecialty report, which references the American Medical Association master file for physician demographic characteristics across 50 states. Given changing demographic characteristics in trainees compared to attending surgeons, the decision was made to look into both groups separately. Race (non-White, defined as any race other than non-Hispanic White, and White) and gender (female and male) were assessed to evaluate known societal biases. Images were generated using a prompt template, "a photo of the face of a [blank]", with the blank replaced by a surgical specialty. Geographic-based prompting was evaluated by specifying the most populous countries on 3 continents (the US, Nigeria, and China). The study compared representation of female and non-White surgeons in each model with real demographic data using χ2, Fisher exact, and proportion tests. There was a significantly higher mean representation of female (35.8% vs 14.7%; P < .001) and non-White (37.4% vs 22.8%; P < .001) surgeons among trainees than attending surgeons. DALL-E 2 reflected attending surgeons' true demographic data for female surgeons (15.9% vs 14.7%; P = .39) and non-White surgeons (22.6% vs 22.8%; P = .92) but underestimated trainees' representation for both female (15.9% vs 35.8%; P < .001) and non-White (22.6% vs 37.4%; P < .001) surgeons. In contrast, Midjourney and Stable Diffusion had significantly lower representation of images of female (0% and 1.8%, respectively; P < .001) and non-White (0.5% and 0.6%, respectively; P < .001) surgeons than DALL-E 2 or true demographic data. Geographic-based prompting increased non-White surgeon representation but did not alter female representation for all models in prompts specifying Nigeria and China. In this study, 2 leading publicly available text-to-image generators amplified societal biases, depicting over 98% surgeons as White and male. While 1 of the models depicted comparable demographic characteristics to real attending surgeons, all 3 models underestimated trainee representation. The study suggests the need for guardrails and robust feedback systems to minimize AI text-to-image generators magnifying stereotypes in professions such as surgery.

Do clinical and communication skills scores on credentialing exams predict potentially inappropriate antibiotic prescribing?

There is considerable variation among physicians in inappropriate antibiotic prescribing, which is hypothesized to be attributable to diagnostic uncertainty and ineffective communication. The objective of this study was to evaluate whether clinical and communication skills are associated with antibiotic prescribing for upper respiratory infections and sinusitis. A cohort study of 2,526 international medical graduates and 48,394 U.S. Medicare patients diagnosed by study physicians with an upper respiratory infection or sinusitis between July 2014 and November 2015 was conducted. Clinical and communication skills were measured by scores achieved on the Clinical Skills Assessment examination administered by the Educational Commission for Foreign Medical Graduates (ECFMG) as a requirement for entry into U.S residency programs. Medicare Part D data were used to determine whether patients were dispensed an antibiotic following an outpatient evaluation and management visit with the study physician. Physician age, sex, specialty and practice region were retrieved from the ECFMG databased and American Medical Association (AMA) Masterfile. Multivariate GEE logistic regression was used to evaluate the association between clinical and communication skills and antibiotic prescribing, adjusting for other physician and patient characteristics. Physicians prescribed an antibiotic in 71.1% of encounters in which a patient was diagnosed with sinusitis, and 50.5% of encounters for upper respiratory infections. Better interpersonal skills scores were associated with a significant reduction in the odds of antibiotic prescribing (OR per score decile 0.93, 95% CI 0.87-0.99), while greater proficiency in clinical skills and English proficiency were not. Female physicians, those practicing internal medicine compared to family medicine, those with citizenship from the US compared to all other countries, and those practicing in southern of the US were also more likely to prescribe potentially unnecessary antibiotics. Based on this study, physicians with better interpersonal skills are less likely to prescribe antibiotics for acute sinusitis and upper respiratory infections. Future research should examine whether tailored interpersonal skills training to help physicians manage patient expectations for antibiotics could reduce unnecessary antibiotic prescribing.

Training in Residency and Provision of Reproductive Health Services Among Family Medicine Physicians

Contraception and abortion services are essential health care, and family medicine (FM) physicians are an important part of the workforce providing this care. Residency could inform the reproductive health services FM physicians provide. To determine which residency training factors are associated with FM physicians' provision of reproductive health services to Medicaid beneficiaries. This cross-sectional, population-based observational study of inpatient and outpatient FM physicians who completed residency between 2008 and 2018 and treated at least 1 Medicaid beneficiary in 2019 was conducted from November 2022 to March 2023. The study used 2019 American Medical Association Masterfile and Historical Residency file, as well as the 2019 Transformed Medicaid Statistical Information System claims. Residency training in community-based or reproductive health-focused programs. The outcomes were providing the following to at least 1 Medicaid beneficiary in 2019: prescription contraception (pill, patch, and/or ring), intrauterine device (IUD) and/or contraceptive implant, and dilation and curettage (D&C). Odds of providing each outcome were measured using correlated random-effects regression models adjusted for physician, residency program, and county characteristics. In the sample of 21 904 FM physician graduates from 410 FM residency programs, 12 307 were female (56.3%). More than half prescribed contraception to Medicaid beneficiaries (13 373 physicians [61.1%]), with lower proportions providing IUD or implant (4059 physicians [18.5%]) and D&C (152 physicians [.7%]). FM physicians who graduated from a Reproductive Health Education in Family Medicine program, which fully integrates family planning into residency training, had significantly greater odds of providing prescription contraception (odds ratio [OR], 1.23; 95% CI, 1.07-1.42), IUD or implant (OR, 1.79; 95% CI, 1.28-2.48), and D&C (OR, 3.61; 95% CI, 2.02-6.44). Physicians who completed residency at a Teaching Health Center, which emphasizes community-based care, had higher odds of providing an IUD or implant (OR, 1.51; 95% CI, 1.19-1.91). In this cross-sectional study of FM physicians providing Medicaid service, characteristics of residency training including community-based care and integration of family planning training are associated with greater odds of providing reproductive health services. With growing reproductive health policy restrictions, providing adequate training in reproductive health is critical to maintaining access to care, especially for underserved populations.

Burnout in US hematologists and oncologists: impact of compensation models and advanced practice provider support.

Burnout is prevalent throughout medicine. Few large-scale studies have examined the impact of physician compensation or clinical support staff on burnout among hematologists and oncologists. In 2019, the American Society of Hematology conducted a practice survey of hematologists and oncologists in the AMA (American Medical Association) Masterfile; burnout was measured using a validated, single-item burnout instrument from the Physician Work-Life Study, while satisfaction was assessed in several domains using a 5-point Likert scale. The overall survey response rate was 25.2% (n= 631). Of 411 respondents with complete responses in the final analysis, 36.7% (n= 151) were from academic practices and 63.3% (n= 260) from community practices; 29.0% (n= 119) were female. Over one-third (36.5%; n= 150) reported burnout, while 12.0% (n= 50) had a high level of burnout. In weighted multivariate logistic regression models incorporating numerous variables, compensation plans based entirely on relative value unit (RVU) generation were significantly associated with high burnout among academic and community physicians, while the combination of RVU+ salary compensation showed no significant association. Female gender was associated with high burnout among academic physicians. High advanced practice provider utilization was inversely associated with high burnout among community physicians. Distinct patterns of career dissatisfaction were observed between academic and community physicians. We propose that the implementation of compensation models not based entirely on clinical productivity increased support for women in academic medicine, and expansion of advanced practice provider support in community practices may address burnout among hematologists and oncologists.

Advancing federal resource allocation: HPSA and the GME impact factor

Health Professional Shortage Areas (HPSA) is a new Graduate Medical Education (GME) resource allocation approach by the US Health and Human Services aimed at producing physicians who practice in these areas. We investigated evidence whether HPSA—an ecological index—is associated with closing physician shortage gaps and the added value of a performance-based index. Using linked administrative data from 1) the American Medical Association's Masterfile (2016 and 2021); 2) the Accreditation Council for Graduate Medical Education sponsoring institutions roster (2021–2022), and the primary care geographic HPSA designation (2021), we assembled a prospective cohort and collapsed these data to the organizational level (N = 622). The diagnostic test is the GME Impact Factor, a performance-based index quantifying the proportion of physicians who practice in a HPSA following medical residency. The reference test is HPSA designation of the sponsoring institution. We assessed sensitivity, specificity, and likelihood ratio test of a positive test (LRT) to appraise HPSA utility on outcome, and the Interclass Coefficient (ICC) to assess similarity, where <0.5 indicates considered poor reliability. Sensitivity is 86.2% (95% CI: 81.9–89.7), specificity is 45.1 (95% CI: 39.7–50.5). LRT is 1.57 (95% CI: 1.55–1.59) and ICC is 0.45 (95% CI: 0.37–0.51). Partial evidence that training sites located within HPSAs are associated with producing physicians who will practice in health professional shortage area following residency training. A performance-based approach — the GME Impact Factor—which utilizes historical production, holds added value in distributing fiscal and personnel resources that yields population-level results. ( Fig. 1 ).

Physician-level variation in axillary surgery in older adults with T1N0 hormone receptor–positive breast cancer.

284 Background: Randomized controlled trial data support safe omission of axillary surgery in patients &gt;70 years old with T1N0 hormone receptor-positive (HR+) breast cancer, but utilization of axillary surgery in the U.S. remains &gt; 80%. We sought to assess the proportion of variance in axillary surgery receipt that is attributable to surgeon-level differences vs. patient level differences, and assess the patient- and surgeon-level factors that affect axillary surgery receipt in this population. Methods: Women &gt;70 years old with T1N0 HR+/HER2-negative breast cancer diagnosed from 2013-2015 in SEER-Medicare were linked to surgeon-level data from the American Medical Association Masterfile. Patients of low-volume surgeons (surgeons treating &lt; 5 patients in this population over the study period) were excluded. To assess patient-and surgeon-level factors associated with axillary surgery receipt, we estimated hierarchical logistic regression models regressing receipt of axillary surgery on surgeon-specific random intercepts (null model). We then re-estimated this model including patient factors (age, Charlson Comorbidity Index (CCI), and a validated claims-based frailty index). Our final model included other surgeon-, patient-, and disease-level factors. The intracluster correlation (ICC) was used to estimate the proportion of total variance in the outcome attributable to the surgeon level. Results: Of 4,410 women included, 58.3% were &gt;75 years old, 20.7% were frail, and 13.9% had a CCI&gt;2; 86.1% of women underwent axillary surgery. Of 432 surgeons represented, 52.6% were female and 9.5% identified as surgical oncologists. In the null model, 10.5% of the variance in the axillary evaluation was attributable to the surgeon-level, with 89.5% attributable to patient-level differences (Table). Adjusting for key patient-level variables (age, CCI, and frailty status) reduced unexplained patient-level variance, resulting in a greater proportion of total unexplained variance attributable to the surgeon level (13%). In our final model, physician subspecialty and years in practice were non-significant, but patients of female surgeons were less likely to undergo axillary surgery (OR 0.69, 95% CI [0.51-0.95]). Conclusions: In older women with T1N0 HR+ disease, patient-level differences reassuringly account for most of the variation in receipt of axillary surgery. Drivers of residual patient-level variation may include unmeasured factors such as differences in surgeon-patient communication and patient preferences.[Table: see text]

Assessing the effectiveness of a LGBT cultural competency training for oncologists: study protocol for a randomized pragmatic trial

BackgroundLGBT patients may have unique psychosocial cancer care needs, and healthcare providers should have knowledge and understanding of these unique needs to effectively address disparities through the delivery of personalized healthcare. As such, our group developed and piloted a web-based LGBT cultural competency training designed specifically for oncologists called the Curriculum for Oncologists on LGBT populations to Optimize Relevance and Skills (COLORS). We designed a randomized pragmatic trial for oncologists to compare the effectiveness of the COLORS training versus a general online LGBT cultural competency training in improving LGBT-related knowledge, attitudes, and clinical practices.Methods/designStudy procedures include an 8-step approach for recruitment, randomization, retention, and completion of the interventions. Oncologists of any subspecialty who are currently practicing physicians will be identified from the American Medical Association Masterfile. Approximately 5000 oncologists will be sent a FedEx envelope with an invitation letter and study timeline. Electronic consent is obtained using a secure REDCap (Research Electronic Data Capture) portal hosted at the Moffitt Cancer Center (Tampa, FL) where the oncologists will complete the eligibility questionnaire, pre-training assessments, and then will be randomized to complete the COLORS training or an online general healthcare training offered by the National LGBT Health Education Center. Effectiveness of both trainings will be assessed utilizing self-reported measures of LGBT-related knowledge, attitudes, and affirming clinical practices. The measures will be collected before and directly after training completion, as well as 3-month post-training completion. The primary outcomes are changes in knowledge, attitudes, and practice behaviors regarding LGBT cancer patients from pre-test to post-test training in the COLORS training vs. the comparison training.DiscussionThe overarching premise of this trial is to assess the effectiveness of the COLORS cultural competency training program. If successful, among oncologists who completed the COLORS training should yield statistically significantly improvements in knowledge, attitudes, and affirming practice.

Health Care Expenditures Attributable to Primary Care Physician Overall and Burnout-Related Turnover: A Cross-sectional Analysis

ObjectiveTo estimate the excess health care expenditures due to US primary care physician (PCP) turnover, both overall and specific to burnout. MethodsWe estimated the excess health care expenditures attributable to PCP turnover using published data for Medicare patients, calculated estimates for non-Medicare patients, and the American Medical Association Masterfile. We used published data from a cross-sectional survey of US physicians conducted between October 12, 2017, and March 15, 2018, of burnout and intention to leave one’s current practice within 2 years by primary care specialty to estimate excess expenditures attributable to PCP turnover due to burnout. A conservative estimate from the literature was used for actual turnover based on intention to leave. Additional publicly available data were used to estimate the average PCP panel size and the composition of Medicare and non-Medicare patients within a PCP’s panel. ResultsTurnover of PCPs results in approximately $979 million in excess health care expenditures for public and private payers annually, with $260 million attributable to PCP burnout-related turnover. ConclusionTurnover of PCPs, including that due to burnout, is costly to public and private payers. Efforts to reduce physician burnout may be considered as one approach to decrease US health care expenditures.

A cross-sectional study of the practice types of US adult primary care physician specialists.

Many physicians listed as primary care in databases such as the American Medical Association (AMA) Masterfile do not provide traditional ambulatory primary care. To compare physicians listed in the AMA Masterfile as primary care physician (PCPs) specialists for adult patients with their actual practice type. We conducted a cross-sectional study of the AMA Masterfile report for PCPs who care for adults (listed as family medicine, internal medicine, medicine-paediatrics, and geriatrics) in the summer and fall of 2018 (spring of 2019 for Hartford, CT) in the primary counties of 8 metropolitan areas across the United States. We searched multiple websites to determine the actual practice type of each physician in the study counties. We correlated the 2 datasets: the AMA Masterfile list vs the results of our searches. Family physicians were more likely to function as traditional ambulatory PCPs than internists [1,738/2,101 (82.7%) vs 1,241/2,025 (60.9%), P < 0.001], and less likely to be hospitalists [83/2,101 (4.0%) vs 631/2,025 (31.0%), P < 0.001]. Other practice types included urgent care [105 (5.0%) family physicians, 16 (0.8%) internists] and emergency medicine [49 (2.3%) family physicians, 20 (1.0%) internists]. The AMA Masterfile identified 4,892 practicing PCPs for adult patients in the study counties, of which 3,084 (63.0%) matched by location and ambulatory PCP practice type [3,695 (75.5%) for ambulatory PCP practice type only]. We provide an updated estimate using a unique methodology to estimate how to correct the AMA Masterfile for PCPs who actually provide traditional ambulatory primary care to adult patients.

Association of clinical competence, specialty and physician country of origin with opioid prescribing for chronic pain: a cohort study

BackgroundAlthough little is known about why opioid prescribing practices differ between physicians, clinical competence, specialty training and country of origin may play a role. We hypothesised that physicians with stronger...

Overall survival based on oncologist density in the United States: A retrospective cohort study.

Medically underserved areas (MUA) or health professional shortage areas (HPSA) designations are based on primary care health services availability. These designations are used in recruiting international medical graduates (IMGs) trained in primary care or subspecialty (e.g., oncology) to areas of need. Whether the MUA/HPSA designation correlates with Oncologist Density (OD) and supports IMG oncologists’ recruitment to areas of need is unknown. We evaluated the concordance of OD with the designation of MUAs/HPSAs and evaluated the impact of OD and MUA/HPSA status on overall survival. We conducted a retrospective cohort study of patients diagnosed with hematological malignancies or metastatic solid tumors in 2011 from the Surveillance Epidemiology and End Results (SEER) database. SEER was linked to the American Medical Association Masterfile to calculate OD, defined as the number of oncologists per 100,000 population at the county level. We calculated the proportion of counties with MUA or HPSA designation for each OD category. Overall survival was estimated using the Kaplan-Meier method and compared between the OD category using a log-rank test. We identified 68,699 adult patients with hematologic malignancies or metastatic solid cancers in 609 counties. The proportion of MUA/HPSA designation was similar across counties categorized by OD (93.2%, 95.4%, 90.3%, and 91.7% in counties with <2.9, 2.9–6.5, 6.5–8.4 and >8.4 oncologists per 100K population, p = 0.7). Patients’ median survival in counties with the lowest OD was significantly lower compared to counties with the highest OD (8 vs. 11 months, p<0.0001). The difference remained statistically significant in multivariate and subgroup analysis. MUA/HPSA status was not associated with survival (HR 1.03, 95%CI 0.97–1.09, p = 0.3). MUA/HPSA designation based on primary care services is not concordant with OD. Patients in counties with lower OD correlated with inferior survival. Federal programs designed to recruit physicians in high-need areas should consider the availability of health care services beyond primary care.

The Supply and Distribution of the Preventive Medicine Physician Workforce.

Preventive medicine physicians work at the intersection of clinical medicine and public health. A previous report on the state of the preventive medicine workforce in 2000 revealed an ongoing decline in preventive medicine physicians and residents, but there have been few updates since. The purpose of this study was to describe trends in both the number of board-certified preventive medicine physicians and those physicians who self-designate preventive medicine as a primary or secondary specialty and examine the age, gender distribution, and geographic distribution of this workforce. Analysis of the supply of preventive medicine physicians using data derived from board certification files of the American Board of Preventive Medicine and self-designation data from the American Medical Association Masterfile. The 50 US states and District of Columbia. Board-certified and self-designated preventive medicine physicians in the United States. Number, demographics, and location of preventive medicine physicians in United States. From 1999 to 2018, the total number of physicians board certified in preventive medicine increased from 6091 to 9270; the number of self-identified preventive medicine physicians has generally decreased since 2000, with a leveling off in the past 4 years matching the trend of preventive medicine physicians per 100 000 population; there is a recent increase in women in the specialty; the practice locations of preventive medicine physicians do not match the US population in rural or micropolitan areas; and the average age of preventive medicine physicians is increasing. The number of preventive medicine physicians is not likely to match population needs in the United States in the near term and beyond. Assessing the preventive medicine physician workforce in the United States is complicated by difficulties in defining the specialty and because less than half of self-designated preventive medicine physicians hold a board certification in the specialty.

Disparities in the Geographic Distribution of Neurosurgeons in the United States: A Geospatial Analysis

Large disparities in access to neurosurgical care are known, but there are limited data on whether geographic distribution of the neurosurgery workforce potentially plays a role in these disparities. The goal of this study was to identify the geographic distribution of neurosurgeons in the United States and to study the association of the per capita workforce distribution with socioeconomic characteristics of the population. The number of practicing neurosurgeons in the United States in 2016 was obtained from the 2017-2018 American Medical Association Masterfile contained within the Area Health Resource File. The association of the number of neurosurgeons per 100,000 population with socioeconomic characteristics was assessed through linear regression analysis at Hospital Referral Region (HRR) level. The median number of neurosurgeons per capita across all HRRs was 1.47 neurosurgeons per 100,000 population (interquartile range, 1.02-2.27). Bivariable analysis showed that greater supply of neurosurgeons was positively associated with regional levels of college education, median income, and median age. The number of neurosurgeons per capita at the HRR level was negatively associated with unemployment, poverty, and percent uninsured. Regions characterized by low socioeconomic status have fewer neurosurgeons per capita in the United States. Low income, low number of college graduates, and high unemployment rate are associated with fewer numbers of neurosurgeons per capita. Further research is needed to determine if these geographic workforce disparities contribute to poor access to quality neurosurgical care.

Physician network connections to specialists and HIV quality of care.

To estimate novel measures of generalist physicians' network connectedness to HIV specialists and their associations with two dimensions of HIV quality of care. Medicare and Medicaid claims and the American Medical Association Masterfile data on people living with HIV (PLWH) and the physicians providing their HIV care in California between 2007 and 2010. I construct regional patient-sharing physician networks from the shared treatment of PLWH and calculate (a) measures of network connectedness to all physician types and (b) specialty-weighted measures to describe connectedness to HIV specialists. Two HIV quality of care outcomes are then evaluated: medication quality (prescribing antiretroviral drugs from at least two drug classes) and monitoring quality (at least two annual HIV virus monitoring scans). Linear probability models estimate the associations between network statistics and the two dimensions of HIV quality of care, and a policy simulation demonstrates the importance of these statistical relationships. These analyses include 16124 PLWH, 3240 generalists, and 1031 HIV specialists. PLWH are identified from claims for patients with any indication of HIV using an existing algorithm from the literature. Generalists' network connectedness to HIV specialists is positively related with their own HIV medication quality; one additional HIV specialist connection is associated with a 1.46 percentage point (SE 0.42, P<.01) increase in generalist's medication quality. Based on the estimated associations, a simulated policy that increases connectedness between generalists and HIV specialists reduces the annual rate of HIV infections by up to 6%, roughly 290 fewer infections per year. Only network connectedness to all physician types is associated with improved monitoring quality. Network connectedness to HIV specialists is positively associated with generalists' HIV medication quality, which suggests that specialists provide clinical support through patient-sharing for complex treatment protocol.

Clarifying the Role of the Modern General Surgeon: The Effect of Specialization on Inpatient Practice Patterns in General Surgery

The growing concentration of fellowship-trained and integrated residency-trained subspecialty surgeons has encroached on the breadth and volume of a so-called "true" general surgery practice, leaving the role of new general surgeons in flux. We aimed to describe the surgical practice of new general surgeons with and without subspecialty fellowship training. In this retrospective cohort study, state discharge data was linked to American Medical Association Masterfile and American Hospital Association annual survey data. New-to-practice general surgeons with and without subspecialty board-certification in colorectal surgery (CRS) or cardiothoracic surgery (CTS) were identified in 2008, and followed over 10 years. Surgeon overall inpatient case volume, colorectal resection case volume, and thoracic lobectomy case volume were compared between surgeons with and without related subspecialty training. NY and FL (2008-2017). The study population included 276 new-to-practice surgeons with mean age of 36.9 years. New-to-practice surgeons were defined as those with zero to three years of experience in 2008. Of all surgeons, 11.2% were subspecialty board-certified in CRS and 11.6% were subspecialty board-certified in CTS. Board-certified CRS surgeons performed more colorectal resections than the non-CRS general surgeons each year (p-value <0.001 for all). Overall, non-CRS general surgeons performed 60.7% of all colorectal resections. Board-certified CTS surgeons performed more thoracic lobectomies than non-CTS surgeons each year. Non-CTS surgeons performed 1.1% of all thoracic lobectomies. On average, new subspecialty surgeons perform significantly more specialty operations than non-subspecialty new general surgeons. However, as a group, new non-colorectal general surgeons perform the majority of colorectal resections. In contrast, new non-cardiothoracic general surgeons perform less than two percent of the thoracic lobectomies. This may have implications for a shift in the training paradigm going forward.

Disruptions in Medical Education Catalysed by the Pandemic

On March 17, 2020, the AAMC announced changes in the medical education system in view of the SARS-CoV-2 pandemic declared by the World Health Organisation (WHO). Since then, medical education has continued to experience a paradigm shift in its manner of delivery, methods of teaching, technology used, definition of classroom and clinical setting, and in how best to meet the needs of the nation’s rising students and junior doctors. The effects of the novel coronavirus have been far reaching and are proving to strain every facet of our system, including its effects on medical students and graduates and their resultant pathway into U.S. residency programs. In March 2020, students across the world experienced an unprecedented scenario of having all of their in person classes transition to an electronic platform. This measure was implemented in concordance with what the world had learned previously from the influenza epidemic: social distancing, closing down schools, and ultimately reducing the peak number of cases by reducing the transmission of the virus. [1] As students slowly transition back into the hospital setting after months of backlog or temporarily suspended examinations and exclusively receiving instruction via online and simulated learning, we are forced to wonder how are these changes affecting the inculcation of the six core competencies in medical students and residents? As per the American Medical Association Masterfile in 2020, International Medical Graduates constitute 24% of the total practicing physicians in the United States [1], and more than 6,000 participate each year in the NRMP Match.How are these changes going to affect the patient care delivered by these future doctors? Furthermore, what can be done to mitigate the effects of these changes and make an improvement in patient care?

Adoption of Active Surveillance for Very Low-Risk Differentiated Thyroid Cancer in the United States: A National Survey.

Active surveillance (AS) of thyroid cancer with serial ultrasounds is a newer management option in the United States. This work aimed to understand factors associated with the adoption of AS. We surveyed endocrinologists and surgeons in the American Medical Association Masterfile. To estimate adoption, respondents recommended treatment for 2 hypothetical cases appropriate for AS. Established models of guideline implementation guided questionnaire development. Outcome measures included adoption of AS (nonadopters vs adopters, who respectively did not recommend or recommended AS at least once; and partial vs full adopters, who respectively recommended AS for one or both cases). The 464 respondents (33.3% response) demographically represented specialties that treat thyroid cancer. Nonadopters (45.7%) were significantly (P < .001) less likely than adopters to practice in academic settings, see more than 25 thyroid cancer patients/year, be aware of AS, use applicable guidelines (P = .04), know how to determine whether a patient is appropriate for AS, have resources to perform AS, or be motivated to use AS. Nonadopters were also significantly more likely to be anxious or have reservations about AS, be concerned about poor outcomes, or believe AS places a psychological burden on patients. Among adopters, partial and full adopters were similar except partial adopters were less likely to discuss AS with patients (P = .03) and more likely to be anxious (P = .04), have reservations (P = .03), and have concerns about the psychological burden (P = .009) of AS. Few respondents (3.2%) believed patients were aware of AS. Widespread adoption of AS will require increased patient and physician awareness, interest, and evaluation of outcomes.

(Re)thinking the Residential in Residency: Modern Surgical Practice Continues to Move Away From the Inpatient Setting

Despite the overall shift in care delivery to an ambulatory setting, the majority of general surgical education still relies on the experience of caring for inpatients. We aimed to investigate how the inpatient practice patterns of newly minted general surgeons (GS) have changed since 2008, in order to better inform education policies regarding both training approach and setting for modern surgical trainees. State discharge data from NY and FL (2008-2017) were linked to data on GS from the American Medical Association Masterfile, and to hospital data from the American Hospital Association annual survey. Mean annual inpatient case volume (CV) and case type breadth (CB) were compared between surgeons who were new-to-practice (0-3 years of experience) in 2008 and in 2013. Each new surgeon cohort was followed for 5 years. Case type was classified by organ system. The 2008 cohort included 328 GS with a mean age of 37.1, 79.6% male and 94.2% board-certified. The 2013 cohort included 359 GS with a mean age of 36.2, 73.0% male and 93.9% board-certified. CV was higher among the 2008 cohort than the 2013 cohort for each year of practice in the study period. CB included at least 4 organ system types for all new GS with greater breadth among the 2008 cohort for each year in the study period. Declining rates of inpatient surgery affect general surgeons who were new-to-practice in 2013 significantly more than those entering practice only 5 years ahead of them. New surgeons continue to start their practices broadly, suggesting a need to continue broad training while expanding formal educational policies to include the full spectrum of ambulatory surgery.

Distribution of Physician Specialties by Rurality.

Physicians of all specialties are more likely to live and work in urban areas than in rural areas. Physician availability affects the health and economy of rural communities. This study aimed to measure and update the availability of physician specialties in rural counties. This analysis included all counties with a Rural-Urban Continuum Code (RUCC) between 4 and 9. Geographically identified physician data from the 2019 American Medical Association Masterfile was merged with 2019 County Health Rankings, the Census Bureau's 2010 county-level population data, and 2010 Topologically Integrated Geographic Encoding and Referencing shapefiles. Multivariate logistic regression was performed to assess the availability of physicians by specialty in rural counties. Of the 1,947 rural counties in our sample, 1,825 had at least 1 physician. Specialties including emergency medicine, cardiology, psychiatry, diagnostic radiology, general surgery, anesthesiology, and OB/GYN were less available than primary care physicians (PCPs) in all rural counties. The probability of a rural county having a PCP was the highest in RUCC 4 (1.0) and lowest in RUCC 8 (0.93). Of all primary care specialties, family medicine was the most evenly distributed across the rural continuum, with a probability of 1.0 in RUCC 4 and 0.88 in RUCC 9. Family medicine is the physician specialty most likely to be present in rural counties. Policy efforts should focus on maintaining the training and scope of practice of family physicians to serve the health care needs of rural communities where other specialties are less likely to practice.

Abstract PO-210: Distribution of genomic testing resources by oncology practice setting and rurality: A nationally representative analysis

Abstract Purpose: This study seeks to understand how the availability of genomic testing resources for cancer treatment vary by oncology practice setting and rurality. Background: Pervasive inequities in access to cancer care resources exist at the neighborhood and health system level. In particular, rurality and practice type have been linked to diminished access to care and poorer health outcomes. While these inequities persist, there have been many advancements in cancer care technology and oncologists are increasingly using genomic testing and next-generation sequencing technology to inform treatment decisions. Little is known about the availability of genomic testing resources in different oncology practice settings and at differing levels of rurality. Methods: This study used data from the National Survey of Precision Medicine in Cancer Treatment, the first nationally representative sample of oncologists (N=1,281) practicing in the United States. Oncologists were identified from the American Medical Association Masterfile for 2017 and were sampled by specialty, census region, metropolitan statistical area, sex, and age. The outcome, availability of genomic testing resources, was measured using one survey question asking oncologists if their primary practice has the following genomic testing resources: on-site pathology, contracts with outside labs, on-site genetic counselors, internal protocols for using genomic testing, electronic medical records (EMR) that alert providers when to order a genomic test, and molecular tumor boards. The predictor variables were rurality and practice type. Rurality was determined using Rural Urban Continuum Codes (RUCC) from 2013. Practice type was measured using the following six categories: solo practice academic, solo practice non-academic, single specialty group academic, single specialty group non-academic, multispecialty group academic, and multispecialty group non-academic. Descriptive statistics were presented, and chi-square tests were used to assess statistically significant difference. The weighted sampling design of the survey was accounted for using SAS statistical package version 9.4 (SAS, Cary NC). Results: Higher proportions of academic practices had genomic testing resources than non-academic practices. Higher proportions of multispecialty groups had genomic testing resources compared to single specialty group and solo practices. Compared to urban practices, lower proportions of rural practices had genomic testing resources for all except contracts with outside labs and EMR alerts. Conclusion: These data highlight the unequal distribution of genomic testing resources for cancer treatment by practice type and rurality. Future studies and interventions should strive to further assess factors affecting access to cancer care resources and explore health system level approaches to improve health equity. Citation Format: Brittany D. Gardner, Janeth Sanchez, Michelle Doose, Sallie J. Weaver, Shobha Srinivasan, Andrew Freedman, Janet S. de Moor. Distribution of genomic testing resources by oncology practice setting and rurality: A nationally representative analysis [abstract]. In: Proceedings of the AACR Virtual Conference: Thirteenth AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2020 Oct 2-4. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(12 Suppl):Abstract nr PO-210.