Enumeration System Research Articles

Led By Vascular Surgery, Vascular Interventions Are Increasingly Performed By Women.

Historically, the medical profession is a male dominated field. While the number of women entering surgical specialties is rising, this increase is not proportionate to the composition of medical school graduates, which are now 50% female. This study aimed to investigate the specialty and gender of practitioners performing common vascular procedures. Medical claims data was obtained from the Center for Medicare and Medicaid Services (CMS). Claims were linked to provider characteristics in the National Plan and Provider Enumeration System (NPPES) using national physician identifier (NPI). The study included final billing records from inpatient claims 2017-2021. Procedures of interest were identified by the primary International Classification of Disease 10 Procedure Coding System (ICD-10 PCS) codes. Provider taxonomy, gender, tenure, and region were derived from National Plan and Provider Enumeration System using their NPI. Among all specialties, board certified vascular surgeons (VS) performed the highest percentage of open infrarenal aortic interventions (68.6%), endovascular aorta repairs (EVAR, 65.2%), carotid endarterectomies (CEA, 59.1%), endovascular carotid interventions (32.4%), open arteriovenous fistulas (AVF, 60.3%), endovascular AVF (59.3%), open infrainguinal interventions (71.4%), and endovascular infrainguinal interventions (45.7%). For the two categories where VS performed less than half of the procedures, the next most common specialties were neurology/neurosurgery (17.9%) and interventional cardiology (10.8%) for endovascular carotid interventions and interventional cardiology (11.1%) and interventional radiology (10.3%) for endovascular infrainguinal interventions. Over the five-year period, the percentage of procedures performed by vascular surgeons increased for all categories except endovascular AVF. Analyzing by gender, the majority of all procedures were performed by male physicians, which ranged from 88.3% (endovascular AVF) to 94.7% (endovascular carotid intervention). Over the five-year interval, however, all specialties had a significant increase in the percentage of vascular procedures performed by females (P<0.05) except thoracic/cardiac surgery. Among all specialties, vascular surgery had the greatest absolute increase in female performed interventions (P<0.05). While the majority of vascular interventions are still performed by male practitioners, the proportion of procedures conducted by females is increasing. Among all specialties, vascular surgery has both the highest proportion and greatest absolute increase of vascular procedures performed by women. Future recruitment efforts should focus on continuing to reduce this disparity.

Evaluating access to psychosocial services for the medicaid-insured children in Georgia

BackgroundEvaluating access to psychosocial services can inform policy decision-making on ways to address shortages in the availability of mental health (MH)-specialized providers. The objective of the study was to assess how the mental health (MH)-specialized workforce met the demand for psychosocial services of Medicaid-insured children in Georgia, with direct relevance in establishing quantitative network adequacy.MethodsWe used the 2018 Medicaid (TAF) claims data, the 2018 National Plan and Provider Enumeration System database, and the 2019 Georgia school-based program data to estimate community-level demand and practice-level supply of psychosocial services. We evaluated the availability of services using mathematical models. The outcome measures were met demand and travel distance. We explored the impact of increasing in-home care delivery, expanding Medicaid participation, or increasing caseload for the Medicaid-participating providers on improving met demand for psychosocial services.ResultsOur findings showed that 34% of the demand from Medicaid-insured children in Georgia remained unmet, and 25% of the Georgia census tracts (rural 79%; urban 16%) had < 50% service coverage. The travel distance for in-clinic services was 3.84 miles on average. Increasing provider Medicaid caseload or expanding Medicaid participation, resulting in a 5–40% supply increase, would reduce unmet demand to 7% and decrease the number of unserved and underserved census tracts to 3% and 2% respectively. Meeting 75% of the demand required a 15% increase in the supply.ConclusionsThe main source of network inadequacy was the scarcity of MH providers available to Medicaid-insured children in Georgia, coming from both the limited caseload of existing MH providers and low Medicaid participation, rather than travel constraints. Increasing provider caseload and expanding Medicaid participation were found to reduce unmet demand. Interventions increasing caseloads were the most effective intervention since existing Medicaid-participating providers already had sufficient network coverage geographically.

CLINICIAN TYPE AND SETTING OF CARE DELIVERED TO MEDICARE BENEFICIARIES WITH DEMENTIA IN THE U.S

Abstract Improving care coordination for persons living with dementia (PLWD) requires understanding of the types of clinicians delivering care and the settings in which they practice. We identified all beneficiaries with dementia in traditional Medicare in 2019. We used the Medicare Carrier file—i.e., where bills for care provided by individual clinicians are recorded—to identify clinicians providing care to PLWD, which we then linked to the National Plan &amp; Provider Enumeration System (NPPES). Using the Carrier and NPPES files, we characterized the workforce providing care to PLWD, including specialty and licensure (e.g., primary care physician, neurologist, nurse practitioner), setting of care delivery, and whether the clinician included a dementia diagnosis in the billing information. Overall, 1,934,318 PLWD received care from 783,225 unique clinicians. The most common settings of care were office (74.8% of PLWD), emergency room (63.9%), inpatient hospital (52.1%), and skilled nursing facility (37.1%). 87.0% of PLWD received care from a primary care physician, 62.9% from a nurse practitioner, and 33.1% from a physician assistant. Among all the clinicians that care for PLWD, 20.3% are primary care physicians, 15.3% nurse practitioners, and 9.0% physician assistants. 2.4% are psychiatrists, 1.7% neurologist, and just 0.5% are geriatric subspecialist physicians. For clinicians providing care to PLWD, the majority provide care in the outpatient setting, followed by hospital-based care. Programs to improve care coordination for PLWD, such as GUIDE Model programs, must account for delivery across multiple health care settings, recognizing that geriatric subspecialists comprise an extremely small proportion of the PLWD-serving workforce.

Enhancing representativeness of patient-reported outcomes in routine radiation oncology care: a quality improvement protocol to address non-response

IntroductionNon-response significantly undermines the representativeness of patient-reported outcome (PRO) data, thereby compromising its utility for facilitating high-value, equitable, patient-centred care in cancer clinics. Quality improvement studies are needed to assess...

Impact of Physician Specialty on Treatment Costs of Invasive Melanoma.

Melanoma is a deadly type of skin cancer that develops from melanocytes and can manifest on the skin or other regions of the body. Its incidence is increasing rapidly, with approximately 100,000 diagnoses and 7000 deaths per year in the US alone. We conducted a cross-sectional study with the aim of determining an association between the cost of care for invasive melanoma and the specialty involved in the treatment to adequately guide future treatment. We analyzed data from 3817 patients (2013-2018) using the Florida inpatient/outpatient dataset, CMS cost reports, and the National Plan and Provider Enumeration System. Covariates included age, sex, race/ethnicity, insurance type, region, county rurality, the number of procedures, the comorbidity index, obesity, metastatic cancer presence, hospital size, and physician volume. Multivariable mixed linear regression was used to analyze the data, and the cost was adjusted to the 2019 USD. Dermatology had the largest decrease in the overall and outpatient costs compared to general surgery, followed by plastic surgery. The inpatient costs for dermatology and plastic surgery were lower than those for general surgery, but not significantly so. The costs associated with surgical procedures may vary depending on the specialty of the physician treating the patient. Dermatology was associated with lower treatment costs for invasive melanoma compared to other specialties, indicating that physician specialty influences the cost of care.

Industry-Sponsored Research Funding to Urologists in the United States Between 2014 and 2022.

Urologists face challenges in obtaining public research funding, leading to increasing reliance on the industry for research support. This study aimed to examine the extent and trends in industry-sponsored research payments to urologists from 2014 to 2022 in the United States. We identified all US urologists using the Centers for Medicare and Medicaid Services National Plan and Provider Enumeration System database and extracted their industry-sponsored research payments data from the Centers for Medicare and Medicaid Services Open Payments Database. We performed descriptive analyses of the payments data. Among 13,902 US urologists, 1330 (9.6%) received at least one industry-sponsored research payment. Urologists received $605.1 million between 2014 and 2022. Of all research payments, 98.7% ($597.4 million) were associated research payments for research where urologists served as principal investigators. The top 10% of urologists in research payments received 79.3% ($480.0 million) of total research payments. Only 0.4% ($2.3 million) of research payments were for preclinical research, whereas registered clinical trials totaled $159.0 million (26.3% of all research payments). The annual value of research payments increased from $35.2 million in 2014 to $101.7 million in 2022. The average percentage change in total payments showed a significant annual increase of 13.9% (95% CI, 11.6%-16.3%; P < .001) in value. There was no significant trend in the number of urologists receiving research payments. Industry-sponsored research payments to urologists are substantial and have increased in both payment amount and number over time.

Fast enumeration of all cost-bounded solutions for combinatorial problems using ZDDs

We propose a fast method for exactly enumerating a large number of all cost-bounded solutions for combinatorial problems using Zero-suppressed Binary Decision Diagrams (ZDDs). Enumerating cost-bounded solutions is an important task to know the exact ranking of any given solution, and it enables statistical p-value evaluation and quality-controlled random sampling. Our method is based on a backtracking search for a given ZDD representing the set of all feasible solutions. The key idea is to integrate “BDD operation cache” technique and “interval-memoizing” technique with a depth-first manner of backtracking procedure. The computation time of the algorithm is empirically bounded by the input and output sizes of ZDDs. It can be significantly faster than existing enumeration systems, especially when the input and output ZDDs are well-compressed and the range of the cost constraint is large. As application examples, we applied our method to a set of difficult instances of the Hamiltonian cycle/path problems. Although it is NP-hard to find any one feasible solution without cost constraints, we tried more difficult problems to enumerate all cost-bounded solutions, and succeeded in implicitly enumerating all cost-bounded solutions for many nontrivial instances which contain billions of solutions. As cost constraints often appear in many real-life situations, our method will be useful in various fields, including data mining, statistical testing, hardware/software design, and many other operations research problems.

Clinician Specialty and HIV PrEP Prescription Reversals and Abandonments

Clinicians are a key component of preexposure prophylaxis (PrEP) care. Yet, no prior studies have quantitatively investigated how PrEP adherence differs by clinician specialty. To understand the association between prescribing clinician specialty and patients not picking up (reversal/abandonment) their initial PrEP prescription. This cross-sectional study of patients who were 18 years or older used pharmacy claims data from 2015 to 2019 on new insurer-approved PrEP prescriptions that were matched with clinician data from the US National Plan and Provider Enumeration System. Data were analyzed from January to May 2022. Clinician specialties included primary care practitioners (PCPs), infectious disease (ID), or other specialties. Reversal was defined as a patient not picking up their insurer-approved initial PrEP prescription. Abandonment was defined as a patient who reversed and still did not pick their prescription within 365 days. Of the 37 003 patients, 4439 (12%) were female and 32 564 (88%) were male, and 77% were aged 25 to 54 years. A total of 24 604 (67%) received prescriptions from PCPs, 3571 (10%) from ID specialists, and 8828 (24%) from other specialty clinicians. The prevalence of reversals for patients of PCPs, ID specialists, and other specialty clinicians was 18%, 18%, and 25%, respectively, and for abandonments was 12%, 12%, and 20%, respectively. After adjusting for confounding, logistic regression models showed that, compared with patients who were prescribed PrEP by a PCP, patients prescribed PrEP by ID specialists had 10% lower odds of reversals (odds ratio [OR], 0.90; 95% CI, 0.81-0.99) and 12% lower odds of abandonment (OR, 0.88; 95% CI, 0.78-0.98), while patients prescribed by other clinicians had 33% higher odds of reversals (OR, 1.33; 95% CI, 1.25-1.41) and 54% higher odds of abandonment (OR, 1.54; 95% CI, 1.44-1.65). The results of this cross-sectional study suggest that PCPs do most of the new PrEP prescribing and are a critical entry point for patients. PrEP adherence differs by clinician specialties, likely due to the populations served by them. Future studies to test interventions that provide adherence support and education are needed.

Geographic variability of Medicaid acceptance among allergists in the US.

To determine the geographic variability of Medicaid acceptance among allergists in the US. Geospatial analysis predicted Medicaid acceptance across space, and a multivariable regression identified area-level population demographic variables associated with acceptance. We used the National Plan & Provider Enumeration System database to identify allergists. Medicaid acceptance was determined from lists or search engines from state Medicaid offices and calls to provider offices. Spatial analysis was performed using the empirical Bayesian kriging tool. Multivariate logistic regression was used to identify county-level characteristics associated with provider Medicaid acceptance. Of 5694 allergists, 55.5% accepted Medicaid. Acceptance in each state ranged from 13% to 90%. Washington, Arizona, and the Northeast had lowest predicted proportion of both Medicaid acceptance and Medicaid acceptance per 10,000 enrollees. Overall, county-level characteristics were not associated with the likelihood of accepting Medicaid in multivariate analyses. Only the percentage of individuals living in poverty was associated with a higher likelihood of providers accepting Medicaid (OR, 1.245; 95% CI, 1.156-1.340; P < .001). A barrier to accessing allergy-related health care is finding a provider who accepts a patient's insurance, which is largely variable by state. Lack of access to allergy care likely affects health outcomes for children with prevalent atopic conditions such as food allergy.

Behavioral Health Workforce Distribution in Socially Disadvantaged Communities.

This study sought to understand the geographic distribution of three behavioral health clinician (BHC) types in disadvantaged communities in the U.S. across a standardized index of area disadvantage. CMS National Plan and Provider Enumeration System's data were used to identify BHC practice addresses. Addresses were geocoded and mapped to census block groups across Area Disadvantage Index (ADI) scores. Differences in the proportion of BHCs per 100k people in a block group by ADI, clinician type, and rurality were compared. Zero-inflated negative binomial models assessed associations between ADI score with any amount, and expected count, of BHC type in a block group. The sample included 836,780 BHCs (51.5% counselors, 34.5% social workers, 14.0% psychologists). Results indicated there were fewer BHCs in areas of high disadvantage with 351 BHCs in the lowest need versus 267 BHCs in highest need areas, per 100k people. BHCtype was differently associated with the rate of clinicians per 100k by ADI and block groups that were both rural and high ADI had the least BHCs located. Findings suggest the maldistribution of BHCs by ADI underscores how some BHCs may be better positioned to meet the needs of vulnerable communities. Increasing access to behavioral health care requires a workforce equitably positioned in high-need areas. Reforms to payment and practice regulations may support BHCs to deliver services in socially disadvantaged neighborhoods.

DuMato: An efficient warp-centric subgraph enumeration system for GPU

Subgraph enumeration is a heavy-computing procedure that lies at the core of Graph Pattern Mining (GPM) algorithms, whose goal is to extract subgraphs from larger graphs according to a given property. Scaling GPM algorithms for GPUs is challenging due to irregularity, high memory demand, and non-trivial choice of enumeration paradigms. In this work we propose a depth-first-search subgraph exploration strategy (DFS-wide) to improve the memory locality and access patterns across different enumeration paradigms. We design a warp-centric workflow to the problem that reduces divergences and ensures that accesses to graph data are coalesced. A weight-based dynamic workload redistribution is also proposed to mitigate load imbalance. We put together these strategies in a system called DuMato, allowing efficient implementations of several GPM algorithms via a common set of GPU primitives. Our experiments show that DuMato's optimizations are effective and that it enables exploring larger subgraphs when compared to state-of-the-art systems.

Trends in medical device company payments in neurosurgery: a nationwide, multidatabase, geospatial analysis.

Industry partnerships help advance the field of neurosurgery. Given the nature of the field and its close relationship with innovation, neurosurgeons frequently partner with the medical device industry to advance technology and improve outcomes. However, this can create important ethical concerns for patients. In this paper, the authors sought to comprehensively study how physician payments from medical device companies have changed and what geographic parameters influence the trends observed over the years. The authors queried and merged several large databases, including Medicare and Medicaid provider usage data and databases from the Open Payments Program, National Plan and Provider Enumeration System, and US Census Bureau. Geospatial analysis was performed using Moran's I and II clustering. Univariate and multivariable analyses were performed using the Mann-Whitney U-test and geospatially weighted multivariable regression for hot spot and cold spot membership. Data for 952 counties across the continental United States were analyzed. Ninety-seven counties constituted geographic hot spots. These hot spots were primarily concentrated in Florida, the New York-Pennsylvania region, central Colorado, and southwestern United States. Independent predictors of hot spot membership included greater unemployment rates, the percentage of White patients, the presence of mobile homes, and the percentage of county Hispanic and Black populations. Company-based differences were examined. The vast majority of Medtronic's payments were in the form of royalties and licensing (86.6%). Royalties and licensing accounted for the majority of payments for DePuy (69.4%), Globus Medical (62%), and NuVasive (77.1%). In contrast, other companies, such as Boston Scientific, opted to pay physicians in the form of ownership and investment interests (42.1%). The impact of the COVID-19 pandemic was also assessed. During the onset of the pandemic in 2020, physician payments fell or remained the same across all regions with the exception of the South Atlantic region. However, it was observed that nearly all regions rebounded, with stark elevations in physician payments immediately in 2021. This analysis demonstrates that there are national hot spots and cold spots of physician payments, and offers some social, economic, and company-dependent predictors that may influence the magnitude of payments. Further analysis is needed to better understand this clinical-commercial partnership in healthcare, specifically within neurosurgical practice.

Perspectives of Oncologists on the Ethical Implications of Using Artificial Intelligence for Cancer Care

Artificial intelligence (AI) tools are rapidly integrating into cancer care. Understanding stakeholder views on ethical issues associated with the implementation of AI in oncology is critical to optimal deployment. To evaluate oncologists' views on the ethical domains of the use of AI in clinical care, including familiarity, predictions, explainability (the ability to explain how a result was determined), bias, deference, and responsibilities. This cross-sectional, population-based survey study was conducted from November 15, 2022, to July 31, 2023, among 204 US-based oncologists identified using the National Plan & Provider Enumeration System. The primary outcome was response to a question asking whether participants agreed or disagreed that patients need to provide informed consent for AI model use during cancer treatment decisions. Of 387 surveys, 204 were completed (response rate, 52.7%). Participants represented 37 states, 120 (63.7%) identified as male, 128 (62.7%) as non-Hispanic White, and 60 (29.4%) were from academic practices; 95 (46.6%) had received some education on AI use in health care, and 45.3% (92 of 203) reported familiarity with clinical decision models. Most participants (84.8% [173 of 204]) reported that AI-based clinical decision models needed to be explainable by oncologists to be used in the clinic; 23.0% (47 of 204) stated they also needed to be explainable by patients. Patient consent for AI model use during treatment decisions was supported by 81.4% of participants (166 of 204). When presented with a scenario in which an AI decision model selected a different treatment regimen than the oncologist planned to recommend, the most common response was to present both options and let the patient decide (36.8% [75 of 204]); respondents from academic settings were more likely than those from other settings to let the patient decide (OR, 2.56; 95% CI, 1.19-5.51). Most respondents (90.7% [185 of 204]) reported that AI developers were responsible for the medico-legal problems associated with AI use. Some agreed that this responsibility was shared by physicians (47.1% [96 of 204]) or hospitals (43.1% [88 of 204]). Finally, most respondents (76.5% [156 of 204]) agreed that oncologists should protect patients from biased AI tools, but only 27.9% (57 of 204) were confident in their ability to identify poorly representative AI models. In this cross-sectional survey study, few oncologists reported that patients needed to understand AI models, but most agreed that patients should consent to their use, and many tasked patients with choosing between physician- and AI-recommended treatment regimens. These findings suggest that the implementation of AI in oncology must include rigorous assessments of its effect on care decisions as well as decisional responsibility when problems related to AI use arise.

Comparison of industry payments to psychiatrists and psychiatric advanced practice clinicians in the USA, 2021: a cross-sectional study

ObjectivesTo compare industry payment patterns among US psychiatrists and psychiatric advanced practice clinicians (APCs) and determine how scope of practice laws has influenced these patterns.DesignCross-sectional study.SettingThis study used the publicly...

The Association of Oncology Provider Density With Black-White Disparities in Cancer Mortality in US Counties.

Black-White racial disparities in cancer mortality are well-documented in the US. Given the estimated shortage of oncologists over the next decade, understanding how access to oncology care might influence cancer disparities is of considerable importance. We aim to examine the association between oncology provider density in a county and Black-White cancer mortality disparities. An ecological study of 1048 US counties was performed. Oncology provider density was estimated using the 2013 National Plan and Provider Enumeration System data. Black:White cancer mortality ratio was calculated using 2014-2018 age-standardized cancer mortality rates from State Cancer Profiles. Linear regression with covariate adjustment was constructed to assess the association of provider density with (1) Black:White cancer mortality ratio, and (2) cancer mortality rates overall, and separately among Black and White persons. The mean Black:White cancer mortality ratio was 1.12, indicating that cancer mortality rate among Black persons was on average 12% higher than that among White persons. Oncology provider density was significantly associated with greater cancer mortality disparities: every 5 additional oncology providers per 100000 in a county was associated with a .02 increase in the Black:White cancer mortality ratio (95% CI: .007 to .03); however, the unexpected finding may be explained by further analysis showing that the relationship between oncology provider density and cancer mortality was different by race group. Every 5 additional oncologists per 100000 was associated with a 1.6 decrease per 100000 in cancer mortality rates among White persons (95% CI: -3.0 to -.2), whereas oncology provider density was not associated with cancer mortality among Black persons. Greater oncology provider density was associated with significantly lower cancer mortality among White persons, but not among Black persons. Higher oncology provider density alone may not resolve cancer mortality disparities, thus attention to ensuring equitable care is critical.

Factors influencing geographic gender disparity in neurosurgery: a nationwide geospatial clustering analysis.

Women neurosurgeons (WNs) continue to remain a minority in the specialty despite significant initiatives to increase their representation. One domain less explored is the regional distribution of WNs, facilitated by the hiring practices of neurosurgical departments across the US. In this analysis, the authors coupled the stated practice location of WNs with regional geospatial data to identify hot spots and cold spots of prevalence and examined regional predictors of increases and decreases in WNs over time. The authors examined the National Provider Identifier (NPI) numbers of all neurosurgeons obtained via the National Plan and Provider Enumeration System (NPPES), identifying the percentage of WNs in each county for which data were appended with data from the US Census Bureau. Change in WN rates was identified by calculating a regression slope for all years included (2015-2022). Hot spots and cold spots of WNs were identified through Moran's clustering analysis. Population and surgeon features were compared for hot spots and cold spots. WNs constituted 10.73% of all currently active neurosurgical NPIs, which has increased from 2015 (8.81%). Three hot spots were found-including the Middle Atlantic and Pacific divisions-that contrasted with scattered cold spots throughout the East Central regions that included Memphis as a major city. Although relatively rapidly growing, hot spots had significant gender inequality, with a median WN percentage of 11.38% and a median of 0.61 WNs added to each respective county per year. The authors analyzed the prevalence of WNs by using aggregated data from the NPPES and US Census Bureau. The authors also show regional hot spots of WNs and that the establishment of WNs in a region is a predictor of additional WNs entering the region. These data suggest that female neurosurgical mentorship and representation may be a major driver of acceptance and further gender diversity in a given region.

An inversion statistic on the generalized symmetric groups

In this paper, we construct a mixed-base number system over the generalized symmetric group G(m,1,n), which is a complex reflection group with a root system of type Bn(m). We also establish one-to-one correspondence between all positive integers in the set {1,⋯,mnn!} and the elements of G(m,1,n) by constructing the subexceedant function in relation to this group. In addition, we provide a new enumeration system for G(m,1,n) by defining the inversion statistic on G(m,1,n). Finally, we prove that the flag-major index is equi-distributed with this inversion statistic on G(m,1,n). Therefore, the flag-major index is a Mahonian statistic on G(m,1,n) with respect to the length function L.

CottonSense: A high-throughput field phenotyping system for cotton fruit segmentation and enumeration on edge devices

High-throughput phenotyping (HTP) has become a powerful tool for gaining insights into the genetic and environmental factors that affect cotton (Gossypium spp.) growth and yield. With the recent advances in the field of computer vision, namely the integration of deep learning algorithms, the accuracy and efficiency of HTP systems have improved dramatically, enabling them to automatically quantify such fundamental phenotypic traits as fruit identification and enumeration. However, there is currently no HTP system available for counting all the reproductive phases of cotton crop that can be deployed in agronomic field conditions throughout the growing season. This study presents CottonSense, an advanced HTP system that overcomes the challenges of deployment across multiple growth periods by effectively segmenting and enumerating cotton fruits at four stages of growth, including square, flower, closed boll, and open boll. Consequently, CottonSense enhances agronomic management through increased opportunities for data collection and analysis. Using RGB-D cameras, it captures and processes both two and three-dimensional data, facilitating a wider range of phenotypic trait extractions such as crop biomass and plant architecture. To segment the cotton fruits, a Mask-RCNN model is trained and optimized for faster inference using TensorRT. The model yields an average AP score of 79% in segmentation across the four fruit categories. Moreover, the model's accuracy in estimating total fruit count per image is validated by a strong agreement with the counts given by ten domain experts, as reflected by an R2 value of 0.94. Furthermore, to accurately count the segmented fruits over large populations of plants, an enumeration algorithm based on a tracking strategy is developed that achieves an R2 value of 0.93 when compared to hand-counted fruits in the field. The proposed HTP system, which is implemented entirely on an edge computing device, is cost-effective and power-efficient, making it an effective tool for high-yield cotton breeding and crop improvement. The code for CottonSense is publicly available at https://github.com/FeriBolour/CottonSense.

Pharmacists Colocated With Primary Care Physicians: Understanding Delivery of Interprofessional Primary Care.

While evidence supports interprofessional primary care models that include pharmacists, the extent to which pharmacists are working in primary care and the factors associated with colocation is unknown. This study aimed to analyze the physical colocation of pharmacists with primary care providers (PCPs) and examine predictors associated with colocation. This is a retrospective cross-sectional study of pharmacists and PCPs with individual National Provider Identifiers in the National Plan and Provider Enumeration System's database. Pharmacist and PCP practice addresses of the health care professionals were geocoded, and distances less than 0.1 miles were considered physically colocated. In all, 502,373 physicians and 221,534 pharmacists were included. When excluding hospital-based pharmacists, 1 in 10 (11%) pharmacists were colocated with a PCP. Pharmacists in urban settings were more likely to be colocated than those in rural areas (OR=1.32, CI: 1.26-1.38). Counties with the highest proportion of licensed pharmacists per 100,000 people in the county had higher colocation (OR=1.38, CI: 1.32-1.45). Colocation was significantly higher in states with an expanded scope of practice (OR 1.37, CI: 1.32-1.42) and those that have expanded Medicaid (OR 1.07, CI: 1.03-1.11). Colocated pharmacists more commonly worked in larger physician practices. Although including pharmacists on primary care teams improves clinical outcomes, reduces health care costs, and enhances patient and provider experience, colocation appears to be unevenly dispersed across the United States, with lower rates in rural areas. As the integration of pharmacists in primary care continues to expand, knowing the prevalence and facilitators of growth will be helpful to policymakers, researchers, and clinical administrators.

A National Survey of Oncologists' Perspectives on the Ethical Implications of Artificial Intelligence in Cancer Care

Introduction Artificial Intelligence (AI) is an emerging technology that promises advancements across blood cancer discovery and care delivery. There are already FDA-approved uses of AI in hematology and oncology but few data available about its use or utility. Concerns have been raised over AI's potential biases, “explainability,” and who is responsible for its use. Oncologists' familiarity with AI and perspectives on these issues are not well known and are critical to understanding the ethical deployment of AI in blood cancer care. Methods From January to July 2023, we performed a national cross-sectional survey of practicing oncologists. A draft survey instrument was developed by a team with content and methodological expertise. Iterative refinement occurred through cognitive debriefing with 5 practicing hematologist/oncologists. The final instrument included 24 questions covering domains of AI familiarity, predictions, explainability, and responsibilities. A random sample of US oncologists was identified using the National Plan &amp; Provider Enumeration System. Paper surveys were mailed with $25 incentivizing gift cards. Reminder letters and phone calls were used for non-responders with an electronic survey option provided. Responses were analyzed descriptively. X2 testing assessed bivariate associations between responses and participant practice setting, with odds ratios (OR) reported. Results Of 399 mailed surveys, 12 were undeliverable and 201 were completed for a response rate of 51.9%. Question missingness was &amp;lt;1%. Participants were 63.2% White, 16.9% Indian Asian, 10.0% Eastern Asian or Pacific Islander, and 4.5% Black; 6.0% were Hispanic. They represented 37 states. The majority were male (63.7%), 40-59 years old (54.2%), and had no specific training in AI (52.2%); 68.2% had at least 11 years in practice and 29.9% practiced at a primary academic hospital. Questions assessing overall familiarity with and predictions related to AI are shown in Table 1. Those at primary academic hospitals were more likely to be familiar with radiology models (OR 2.82, p&amp;lt;0.001) but not pathology, prognostic, or clinical decision models (all p&amp;gt;0.05). They were more likely to predict that AI would improve side effect management (OR 2.01, p=0.03) and end-of-life decision making (OR 2.07, p=0.02). High proportions of respondents thought that AI prognostic (67.1%) and clinical decision models (62.1%) needed to be explainable by clinicians and researchers to be used in clinic. When posed with a scenario where an FDA-approved AI model selects a different regimen than the oncologist planned to propose (Figure 1), the largest proportion reported they would present both options and let the patient decide (37.3%). Of those recommending a regimen, respondents from primary academic hospitals more frequently recommend the AI-generated regimen (56.1%), while respondents from elsewhere recommended their original regimen (69.0%; OR 2.85, p=0.007). Acceptability of some direct-to-patient AI model applications is shown in Table 1. A majority of respondents (90.5%) thought that AI developers should be responsible for medico-legal problems caused by AI use; 47.8% thought that physicians using the tool should be responsible. Though most (76.1%) agreed that it was oncologists' responsibility to protect patients from biased AI tools, few (27.3%) were confident in their ability to identify how representative the data used in an AI model was relative to the population with that cancer (including 66.7% of those who agreed it was oncologists' responsibility). More respondents thought that patients should need to consent to the use of AI tools in cancer treatment decisions (81.1%) than cancer diagnosis (55.7%); those from primary academic hospitals were less likely to respond that patients should consent for treatment or diagnosis (OR 0.43, p=0.02 and OR 0.44, p=0.009). Conclusions In this national survey of oncologists' views on AI, there was optimism about its impact on care with notable caveats related to end-of-life care, patient trust, and care disparities. Despite many approving use of AI models too complex for patients to explain, a substantial proportion deferred to patient choice when AI and physician recommendations differed. There was also dissonance between accepting responsibility to protect patients from bias and respondents' self-perceived ability to do so.