Health care provider movement increased through COVID-19.

COVID-19 placed unprecedented strain on the health workforce, raising concerns of increasing worker turnover and attrition. This study explores the use of 2 publicly available Medicare datasets-Provider Enrollment, Chain, and Ownership System (PECOS) and Doctors and Clinicians-to track provider movement across states and organizations from 2017 to 2023. We found an increase in state-to-state movement of providers post-COVID-19, with an initial spike in physician movement in the first year (April 2020 to March 2021). Movement varied across specialties and professions. Between organizations, we saw an initial increase in movement for family physicians but not internal medicine physicians. Overall, provider movement was generally to larger organizations. Our study finds increasing movement of providers in the post-COVID-19 period through the novel use of 2 publicly available Medicare datasets. Tracking health care workforce movement closer to real time is important to understand a changing workforce-with differences across communities-and to guide policies to ensure sufficient workforce and prevent worsening disparities over time.

Deciphering the Neighborhood Atlas Area Deprivation Index: the consequences of not standardizing.

The Area Deprivation Index (ADI) is a widely used measure recently selected for several federal payment models that adjusts payments based on where beneficiaries live. A recent debate in Health Affairs focuses on seemingly implausible ADI rankings in major cities and across New York. At the root of the issue is the importance of standardization of measures prior to calculating index scores. Neighborhood Atlas researchers are implicitly arguing that their choice to not standardize is of little consequence. Using the same data and methods as the Neighborhood Atlas, this paper focuses on this choice by calculating and comparing standardized and unstandardized ADI scores. The calculated unstandardized ADI nearly perfectly matches the Neighborhood Atlas ADI (r > 0.9999), whereas the correlation with a standardized version is much lower (r = 0.7245). The main finding is that, without standardization, the ADI is reducible to a weighted average of just 2 measures-income and home values-certainly not the advertised multidimensional measure. Federal programs that have incorporated the ADI risk poorly allocating scarce resources meant to reduce health inequities.

Family Medicine Residencies: How Rural Training Exposure in GME Is Associated With Subsequent Rural Practice.

Background: Rural US populations face a chronic shortage of physicians and an increasing gap in life expectancy compared to urban US populations, creating a need to understand how to increase residency graduates' desire to practice in such areas. Objective: This study quantifies associations between the amount of rural training during family medicine (FM) residencies and subsequent rural work. Methods: American Medical Association (AMA) Masterfile, AMA graduate medical education (GME) supplement, American Board of Family Medicine certification, Accreditation Council for Graduate Medical Education (ACGME), and Centers for Medicare and Medicaid Services hospital costs data were merged and analyzed. Multiple logistic regression measured associations between rural training and rural or urban practice in 2018 by all 12 162 clinically active physicians who completed a US FM residency accredited by the ACGME between 2008 and 2012. Analyses adjusted for key potential confounders (age, sex, program size, region, and medical school location and type) and clustering by resident program. Results: Most (91%, 11 011 of 12 162) residents had no rural training. A minority (14%, 1721 of 12 162) practiced in a rural location in 2018. Residents with no rural training comprised 80% (1373 of 1721) of those in rural practice in 2018. Spending more than half of residency training months in rural areas was associated with substantially increased odds of rural practice (OR 5.3-6.3). Only 4% (424 of 12 162) of residents spent more than half their training in rural locations, and only 5% (26 of 436) of FM training programs had residents training mostly in rural settings or community-based clinics. Conclusions: There is a linear gradient between increasing levels of rural exposure in FM GME and subsequent rural work.

Patient-Physician Racial Concordance Associated with Improved Healthcare Use and Lower Healthcare Expenditures in Minority Populations.

BACKGROUND: Racial concordance between patients and clinician has been linked to improved satisfaction and patient outcomes. OBJECTIVES: (1) To examine the likelihood of clinician-patient racial concordance in non-Hispanic White, non-Hispanic Black, Asian, and Hispanic patients and (2) to evaluate the impact of patient-clinician race concordance on healthcare use and expenditures within each racial ethnic group. METHODS: We analyzed data from the 2010-2016 Medical Expenditure Panel Survey (MEPS). We used bivariate and multivariate models to assess the association between patient-clinician race concordance and emergency department (ED) use, hospitalizations, and total healthcare expenses, controlling for patient socio-demographic factors, insurance coverage, health status, and survey year fixed effects. RESULTS: Of the 50,626 adults in the analysis sample, 32,350 had racial concordance with their clinician. Among Asian and Hispanic patients, low income, less education, and non-private insurance were associated with an increased likelihood of patient-clinician racial concordance. Emergency department use was lower among Whites and Hispanics with concordant clinicians compared to those without a discordant clinician (15.6% vs. 17.3%, p = 0.02 and 12.9% vs. 16.2%, p = 0.01 respectively). Total healthcare expenditures were lower among Black, Asian, and Hispanic patients with race-concordant clinicians than those with discordant clinicians (14%, 34%, and 20%, p < 0.001 respectively). CONCLUSIONS: These results add to the body of evidence supporting the hypothesis that racial concordance contributes to a more effective therapeutic relationship and improved healthcare. These results emphasize the need for medical education surrounding cultural humility and the importance of diversifying the healthcare workforce.

Family Practices in Transforming Clinical Practice Initiative Showed No Changes in Medicare Costs or Utilization.

BACKGROUND: The Centers for Medicare and Medicaid Services proposed that the Transforming Clinical Practice Initiative (TCPI) would improve health outcomes for patients, reduce utilization of institutional services, and generate significant savings for payers by the end of September 2019. OBJECTIVE: The objective of this study was to investigate whether participation in TCPI's Practice Transformation Networks (PTNs) was associated with improved cost and utilization outcomes for Medicare patients of family medicine-based practices in the first 2 years, that is, 2016-2017, of the Initiative. STUDY DESIGN: A quasi-experimental design with a longitudinal cohort of family medicine-based practices and a propensity-matched comparison sample. SUBJECTS: A total of 761 PTN practices and 3451 non-PTN practices. MEASURES: To measure practice-level patient outcomes, we attributed patients to practice based on the plurality of office visits. We obtained Medicare claims from 2011 to 2017 to assess PTN participation effects for Medicare Part A and B costs, hospital admission, and emergency department visit rates using a Difference-in-Differences design, adjusting for baseline characteristics. RESULTS: The differences in Medicare Part A and B costs (-1.71%, P=0.25), annual rates of hospitalization (-0.59%, P=0.12) and emergency department visit (-0.29%, P=0.46) were not significantly lower among PTN practices (N=761) than among propensity score-matched non-PTN practices (N=3541). CONCLUSIONS: TCPI's transforming efforts, such as the outcomes examined in the study, might need a longer time frame to manifest and require evaluation after the full 4-year participation period. The indistinguishable effect of PTN participation may also be attributed to the fact that non-PTN practices might have participated in other initiatives that changed their care and curbed health care utilization and costs consequently.

United States Family Medicine research collaborations associated with higher citation and funding rates.

INTRODUCTION Among academic medical disciplines, Family Medicine (FM) research is notable for its breadth of health-care content areas, making it particularly susceptible to interdisciplinary collaboration. AIM This study characterises the degree and typology of such collaborations, and determines whether collaboration patterns are associated with citation frequency and funding. METHODS This cross-sectional study describes collaboration patterns for publications from 2015 indexed in Web of Science and authored by faculty from United States (US) departments of family medicine (DFMs). We determined mean number of total and FM authors per publication, and percentage of publications with FM first or last authors. Publications were categorised by inclusion of non-FM faculty author(s) and number of DFMs represented. RESULTS Overall, 919 FM faculty from 109 DFMs authored a total of 1872 unique publications in 2015. There was an average of 6.8 authors per publication with 1.4 authors being FM faculty. FM faculty were first author on 26.2% and last author on 29.2% of publications. Of all publications, 0.9% were single FM Author; 1.0% were same DFM; 0.3% were multiple DFMs; 72.4% were single FM Author+non-FM; 19.3% were same DFM+non-FM; 6.0% were multiple DFMs+non-FM. FM publications with non-FM faculty authors showed higher citation rates, higher rates of funding, and lower rates of having no funding source. DISCUSSION Most FM publications involved non-FM faculty authors. Collaborations involving non-FM authors were correlated with higher impact publications and projects that were more likely to have been funded.

Assessing Primary Care Contributions to Behavioral Health: A Cross-sectional Study Using Medical Expenditure Panel Survey.

OBJECTIVES: To assess primary care contributions to behavioral health in addressing unmet mental healthcare needs due to the COVID-19 pandemic. METHODS: Secondary data analysis of 2016 to 2018 Medical Expenditure Panel Survey of non-institutionalized US adults. We performed bivariate analysis to estimate the number and percentage of office-based visits and prescription medications for depression and anxiety disorders, any mental illness (AMI), and severe mental illness (AMI) by physician specialty (primary care, psychiatry, and subspecialty) and medical complexity. We ran summary statistics to compare the differences in sociodemographic factors between patients with AMI by seeing a primary care physician versus those seeing a psychiatrist. Binary logistic regression models were estimated to examine the likelihood of having a primary care visit versus psychiatrist visit for a given mental illness. RESULTS: There were 394 023 office-based visits in the analysis sample. AMI patients seeing primary care physician were thrice as likely to report 1 or more chronic conditions compared to those seeing psychiatrist. Among patients with a diagnosis of depression or anxiety and AMI the proportion of primary care visits ([38% vs 32%, P < .001], [39% vs 34%, P < .001] respectively), and prescriptions ([50% vs 40%, P < .001], [47% vs 44%, P < .05] respectively) were higher compared to those for psychiatric care. Patients diagnosed with SMI had a more significant percentage of prescriptions and visits to a psychiatrist than primary care physicians. CONCLUSION: Primary care physicians provided most of the care for depression, anxiety, and AMI. Almost a third of the care for SMI and a quarter of the SMI prescriptions occurred in primary care settings. Our study underscores the importance of supporting access to primary care given primary care physicians' critical role in combating the COVID-19 related rise in mental health burden.

Proportion of Family Physicians in Solo and Small Practices is on the Decline.

Although solo and small practices are a vital part of primary care, the proportion of family physicians reporting working in practices with 5 or fewer providers declined from 15% to 11% for solo and 37% to 34% for small (2 to 5 providers) practices from 2014 to 2018. These decreasing trends are concerning, mainly when a low proportion of family physicians have solo practices in rural locations given the access to care challenges in these underserved populations.

Primary Care's Historic Role in Vaccination and Potential Role in COVID-19 Immunization Programs.

PURPOSE: Coronavirus disease 2019 (COVID-19) pandemic recovery will require a broad and coordinated effort for infection testing, immunity determination, and vaccination. With the advent of several COVID-19 vaccines, the dissemination and delivery of COVID-19 immunization across the nation is of concern. Previous immunization delivery patterns may reveal important components of a comprehensive and sustainable effort to immunize everyone in the nation. METHODS: The delivery of vaccinations were enumerated by provider type using 2017 Medicare Part B Fee-For-Service data and the 2013-2017 Medical Expenditure Panel Survey. The delivery of these services was examined at the service, physician, and visit level. RESULTS: In 2017 Medicare Part B Fee-For-Service, primary care physicians provided the largest share of services for vaccinations (46%), followed closely by mass immunizers (45%), then nurse practitioners/physician assistants (NP/PAs) (5%). The Medical Expenditure Panel Survey showed that primary care physicians provided most clinical visits for vaccination (54% of all visits). CONCLUSIONS: Primary care physicians have played a crucial role in delivery of vaccinations to the US population, including the elderly, between 2012-2017. These findings indicate primary care practices may be a crucial element of vaccine counseling and delivery in the upcoming COVID-19 recovery and immunization efforts in the United States.

Site of Care for COVID-19-Like Respiratory Illnesses.

COVID-19 is primarily a respiratory illness. Historically, upper and lower respiratory illness has been cared for at home or in the ambulatory primary care setting. It is likely that patients experiencing COVID-19-like symptoms may first contact their primary care provider. The Medical Expenditure Panel Survey (MEPS) is a representative sample of patients from the United States that regularly assesses their use of medical care services. We analyzed 2017 MEPS data to determine the number and proportion of patients who were seen in primary care or family medicine ambulatory settings or hospitalized for upper or lower respiratory illness or pneumonia. In a given year, 19.5 million patients are seen by primary care for an upper respiratory illness, 10.7 million patients for bronchitis, and 9 million for pneumonia. In contrast, 890,000 patients are hospitalized with pneumonia. Given that a primary etiology for respiratory illness in early 2020 was SARS CoV-2 (COVID-19), primary care practices likely were the site of first contact for most patients with COVID-19 illness. Unfortunately, there has been inadequate support for in-person and telehealth visits. Primary care clinicians reported serious shortages of personal protective equipment (PPE) and testing capacity. Inadequate reimbursement for telehealth visits coupled with decreased in-person visits put primary care practices at risk of layoffs and closure. Policies related to primary care payment, federal relief efforts, PPE access, testing and follow-up capacity, and telehealth technical support are essential so primary care can provide first contact and continuity for their patients and communities throughout the COVID-19 pandemic response and recovery.

Capacity of Primary Care to Deliver Telehealth in the United States.

BACKGROUND: Because of the Coronavirus disease 2019 (COVID 19) pandemic, many primary care practices have transitioned to telehealth visits to keep patients at home and decrease the transmission of the disease. Yet, little is known about the nationwide capacity for delivering primary care services via telehealth. METHODS: Using the 2016 National Ambulatory Medical Survey we estimated the number and proportion of reported visits and services that could be provided via telehealth. We also performed cross-tabulations to calculate the number and proportion of physicians providing telephone visits and e-mail/internet encounters. RESULTS: Of the total visits (nearly 400 million) to primary care physicians, 42% were amenable to telehealth and 73% of the total services rendered could be delivered through telehealth modalities. Of the primary care physicians, 44% provided telephone consults and 19% provided e-consults. DISCUSSION: This study underscores how and where primary care services could be delivered. It provides the first estimates of the capacity of primary care to provide telehealth services for COVID-19 related illness, and for several other acute and chronic medical conditions. It also highlights the fact that, as of 2016, most outpatient telehealth visits were done via telephone. CONCLUSIONS: This study provides an estimate of the primary care capacity to deliver telehealth and can guide practices and payers as care delivery models change in a post-COVID 19 environment.

A Cross-Sectional Study of Factors Associated With Pediatric Scope of Care in Family Medicine.

PURPOSE: The objective of this study was to identify demographic and practice characteristics associated with family physicians' provision of care to children including a subgroup analysis of those who see pediatric patients younger or older than 5 years of age. METHODS: This cross-sectional study used data from US family physicians taking the American Board of Family Medicine continuous certification examination registration questionnaire in 2017 and 2018. The outcome of interest was self-reported care of pediatric patients in practice. We performed bivariate and multivariate logistic regression examining the association between various demographic and practice characteristics with the outcome of interest. We performed subgroup analyses for physicians seeing patients under 5 years old and from 5 to 18 years old. RESULTS: Among the 11,674 family physicians included in the final analysis, 9744 (83.8%) saw pediatric patients. Physician- and practice-level factors associated with seeing pediatric patients included rural practice, younger age, non-Hispanic White race/ethnicity, independent practice ownership, nonsolo practice, lower pediatrician density, and higher income geographic area. More family physicians saw 5-to-18-year-olds than < 5-year-olds (83.6% vs 68.2%; P < .001), and the factors associated with pediatric care were similar among these age subgroups. CONCLUSIONS: A majority of continuous certification US family physicians see pediatric patients in practice; however, rates of pediatric care vary widely based on various demographic and practice characteristics. Efforts to maintain a broad scope of practice for US family physicians will require exploration of the underlying mechanisms driving these practice patterns.

Taking a Closer Look at Mental Health Treatment Differences: Effectiveness of Mental Health Treatment by Provider Type in Racial and Ethnic Minorities.

OBJECTIVES: To estimate racial/ethnic differences in the extent to which mental health treatment is obtained from mental health providers, primary care physicians (PCPs), or both and to examine the effects of provider type on change in mental component scores (MCS) of the SF-12 on different race/ethnic groups. METHODS: Secondary data analysis of 2008 to 2015 Medical Expenditure Panel Survey (MEPS). Non-institutionalized civilian US population, aged 18 to 64 (N = 62 558). Based on counts of all mental health visits in a calendar year, we identified patients who obtained care from PCPs, mental health provider, PCP and mental health providers and other providers and examined changes in MCS by type of care. RESULTS: 9.9% of Non-Hispanic Whites obtained mental health treatment, compared to 5.0% for Hispanics, 5.3% for Blacks and 5.5% for Other Races (P < .001). Non-Hispanic Blacks and non-Hispanic "Other" were more likely than other groups to obtain care from mental health providers only (P = .017). All obtaining care solely from PCP had better mental health (mean (se)) MCS: 43.2(0.28)) than those obtaining care solely from mental health provider (39.8 (0.48)), which in turn was higher than for those obtaining care from both PC and MH providers (38.5 (0.31), (P < .001). CONCLUSION: Even when diagnosed with a mental health disorder, Hispanics and Blacks were less likely to seek mental health treatment than Whites, highlighting the continuing disparity. Future research should focus on understanding how and what aspects of integrated care models and other mental health delivery models that reduce disparities and provide greater accessibility.

Gender Differences in Reported Weekly Work Hours Among Family Physicians.

While women are entering family medicine at higher rates than men, little is known about the present differences in practice patterns between male and female family physicians (FPs). We used 2017 and 2018 American Board of Family Medicine Family Medicine Certification Examination practice demographic questionnaires to assess average weekly total hours and direct patient care hours by age and gender reported by FPs. We found a gender gap between both overall hours worked and direct patient care hours, with female FPs reporting fewer hours across age groups.

Using Machine Learning to Predict Primary Care and Advance Workforce Research.

PURPOSE: To develop and test a machine-learning-based model to predict primary care and other specialties using Medicare claims data. METHODS: We used 2014-2016 prescription and procedure Medicare data to train 3 sets of random forest classifiers (prescription only, procedure only, and combined) to predict specialty. Self-reported specialties were condensed to 27 categories. Physicians were assigned to testing and training cohorts, and random forest models were trained and then applied to 2014-2016 data sets for the testing cohort to generate a series of specialty predictions. Comparing the predicted specialty to self-report, we assessed performance with F1 scores and area under the receiver operating characteristic curve (AUROC) values. RESULTS: A total of 564,986 physicians were included. The combined model had a greater aggregate (macro) F1 score (0.876) than the prescription-only (0.745; P <.01) or procedure-only (0.821; P <.01) model. Mean F1 scores across specialties in the combined model ranged from 0.533 to 0.987. The mean F1 score was 0.920 for primary care. The mean AUROC value for the combined model was 0.992, with values ranging from 0.982 to 0.999. The AUROC value for primary care was 0.982. CONCLUSIONS: This novel approach showed high performance and provides a near real-time assessment of current primary care practice. These findings have important implications for primary care workforce research in the absence of accurate data.

Advancing bibliometric assessment of research productivity: an analysis of US Departments of Family Medicine.

INTRODUCTION Measurement of family medicine research productivity has lacked the replicable methodology needed to document progress. AIM In this study, we compared three methods: (1) faculty-to-publications; (2) publications-to-faculty; and (3) department-reported publications. METHODS In this cross-sectional analysis, publications in peer-reviewed, indexed journals for faculty in 13 US family medicine departments in 2015 were assessed. In the faculty-to-publications method, department websites to identify faculty and Web of Science to identify publications were used. For the publications-to-faculty method, PubMed's author affiliation field were used to identify publications, which were linked to faculty members. In the department-reported method, chairs provided lists of faculty and their publications. For each method, descriptive statistics to compare faculty and publication counts were calculated. RESULTS Overall, 750 faculty members with 1052 unique publications, using all three methods combined as the reference standard, were identified. The department-reported method revealed 878 publications (84%), compared to 616 (59%) for the faculty-to-publications method and 412 (39%) for the publication-to-faculty method. Across all departments, 32% of faculty had any publications, and the mean number of publications per faculty was 1.4 (mean of 4.4 per faculty among those who had published). Assistant Professors, Associate Professors, Professors and Chairs accounted for 92% of all publications. DISCUSSION Online searches capture a fraction of publications, but also capture publications missed through self-report. The ideal methodology includes all three. Tracking publications is important for quantifying the return on our discipline's research investment.