Organization and Performance of US Health Systems.

Importance: Health systems play a central role in the delivery of health care, but relatively little is known about these organizations and their performance. Objective: To (1) identify and describe health systems in the United States; (2) assess differences between physicians and hospitals in and outside of health systems; and (3) compare quality and cost of care delivered by physicians and hospitals in and outside of health systems. Evidence Review: Health systems were defined as groups of commonly owned or managed entities that included at least 1 general acute care hospital, 10 primary care physicians, and 50 total physicians located within a single hospital referral region. They were identified using Centers for Medicare & Medicaid Services administrative data, Internal Revenue Service filings, Medicare and commercial claims, and other data. Health systems were categorized as academic, public, large for-profit, large nonprofit, or other private systems. Quality of preventive care, chronic disease management, patient experience, low-value care, mortality, hospital readmissions, and spending were assessed for Medicare beneficiaries attributed to system and nonsystem physicians. Prices for physician and hospital services and total spending were assessed in 2018 commercial claims data. Outcomes were adjusted for patient characteristics and geographic area. Findings: A total of 580 health systems were identified and varied greatly in size. Systems accounted for 40% of physicians and 84% of general acute care hospital beds and delivered primary care to 41% of traditional Medicare beneficiaries. Academic and large nonprofit systems accounted for a majority of system physicians (80%) and system hospital beds (64%). System hospitals were larger than nonsystem hospitals (67% vs 23% with >100 beds), as were system physician practices (74% vs 12% with >100 physicians). Performance on measures of preventive care, clinical quality, and patient experience was modestly higher for health system physicians and hospitals than for nonsystem physicians and hospitals. Prices paid to health system physicians and hospitals were significantly higher than prices paid to nonsystem physicians and hospitals (12%-26% higher for physician services, 31% for hospital services). Adjusting for practice size attenuated health systems differences on quality measures, but price differences for small and medium practices remained large. Conclusions and Relevance: In 2018, health system physicians and hospitals delivered a large portion of medical services. Performance on clinical quality and patient experience measures was marginally better in systems but spending and prices were substantially higher. This was especially true for small practices. Small quality differentials combined with large price differentials suggests that health systems have not, on average, realized their potential for better care at equal or lower cost.

Uncertainty in lung cancer stage for survival estimation via set-valued classification.

The difficulty in identifying cancer stage in health care claims data has limited oncology quality of care and health outcomes research. We fit prediction algorithms for classifying lung cancer stage into three classes (stages I/II, stage III, and stage IV) using claims data, and then demonstrate a method for incorporating the classification uncertainty in survival estimation. Leveraging set-valued classification and split conformal inference, we show how a fixed algorithm developed in one cohort of data may be deployed in another, while rigorously accounting for uncertainty from the initial classification step. We demonstrate this process using SEER cancer registry data linked with Medicare claims data.

Association of Participation in the Oncology Care Model With Medicare Payments, Utilization, Care Delivery, and Quality Outcomes.

IMPORTANCE: In 2016, the US Centers for Medicare & Medicaid Services initiated the Oncology Care Model (OCM), an alternative payment model designed to improve the value of care delivered to Medicare beneficiaries with cancer. OBJECTIVE: To assess the association of the OCM with changes in Medicare spending, utilization, quality, and patient experience during the OCM's first 3 years. DESIGN, SETTING, AND PARTICIPANTS: Exploratory difference-in-differences study comparing care during 6-month chemotherapy episodes in OCM participating practices and propensity-matched comparison practices initiated before (January 2014 through June 2015) and after (July 2016 through December 2018) the start of the OCM. Participants included Medicare fee-for-service beneficiaries with cancer treated at these practices through June 2019. EXPOSURES: OCM participation. MAIN OUTCOMES AND MEASURES: Total episode payments (Medicare spending for Parts A, B, and D, not including monthly payments for enhanced oncology services); utilization and payments for hospitalizations, emergency department (ED) visits, office visits, chemotherapy, supportive care, and imaging; quality (chemotherapy-associated hospitalizations and ED visits, timely chemotherapy, end-of-life care, and survival); and patient experiences. RESULTS: Among Medicare fee-for-service beneficiaries with cancer undergoing chemotherapy, 483 319 beneficiaries (mean age, 73.0 [SD, 8.7] years; 60.1% women; 987 332 episodes) were treated at 201 OCM participating practices, and 557 354 beneficiaries (mean age, 72.9 [SD, 9.0] years; 57.4% women; 1 122 597 episodes) were treated at 534 comparison practices. From the baseline period, total episode payments increased from $28 681 for OCM episodes and $28 421 for comparison episodes to $33 211 for OCM episodes and $33 249 for comparison episodes during the intervention period (difference in differences, -$297; 90% CI, -$504 to -$91), less than the mean $704 Monthly Enhanced Oncology Services payments. Relative decreases in total episode payments were primarily for Part B nonchemotherapy drug payments (difference in differences, -$145; 90% CI, -$218 to -$72), especially supportive care drugs (difference in differences, -$150; 90% CI, -$216 to -$84). The OCM was associated with statistically significant relative reductions in total episode payments among higher-risk episodes (difference in differences, -$503; 90% CI, -$802 to -$204) and statistically significant relative increases in total episode payments among lower-risk episodes (difference in differences, $151; 90% CI, $39-$264). The OCM was not significantly associated with differences in hospitalizations, ED visits, or survival. Of 22 measures of utilization, 10 measures of quality, and 7 measures of care experiences, only 5 were significantly different. CONCLUSIONS AND RELEVANCE: In this exploratory analysis, the OCM was significantly associated with modest payment reductions during 6-month episodes for Medicare beneficiaries receiving chemotherapy for cancer in the first 3 years of the OCM that did not offset the monthly payments for enhanced oncology services. There were no statistically significant differences for most utilization, quality, and patient experience outcomes.

Physician and facility drivers of spending variation in locoregional prostate cancer.

BACKGROUND: Prostate cancer is the most common male cancer, with a wide range of treatment options. Payment reform to reduce unnecessary spending variation is an important strategy for reducing waste, but its magnitude and drivers within prostate cancer are unknown. METHODS: In total, 38,971 men aged ≥66 years with localized prostate cancer who were enrolled in Medicare fee-for-service and were included in the Surveillance, Epidemiology, and End Results-Medicare database from 2009 to 2014 were included. Multilevel linear regression with physician and facility random effects was used to examine the contributions of urologists, radiation oncologists, and their affiliated facilities to variation in total patient spending in the year after diagnosis within geographic region. The authors assessed whether spending variation was driven by patient characteristics, disease risk, or treatments. Physicians and facilities were sorted into quintiles of adjusted patient-level spending, and differences between those that were high-spending and low-spending were examined. RESULTS: Substantial variation in spending was driven by physician and facility factors. Differences in cancer treatment modalities drove more variation across physicians than differences in patient and disease characteristics (72% vs 2% for urologists, 20% vs 18% for radiation oncologists). The highest spending physicians spent 46% more than the lowest and had more imaging tests, inpatient care, and radiotherapy spending. There were no differences across spending quintiles in the use of robotic surgery by urologists or the use of brachytherapy by radiation oncologists. CONCLUSIONS: Significant differences were observed for patients with similar demographics and disease characteristics. This variation across both physicians and facilities suggests that efforts to reduce unnecessary spending must address decision making at both levels.

Inference With Difference-in-Differences With a Small Number of Groups: A Review, Simulation Study, and Empirical Application Using SHARE Data.

BACKGROUND: Difference-in-differences (DID) estimation has become increasingly popular as an approach to evaluate the effect of a group-level policy on individual-level outcomes. Several statistical methodologies have been proposed to correct for the within-group correlation of model errors resulting from the clustering of data. Little is known about how well these corrections perform with the often small number of groups observed in health research using longitudinal data. METHODS: First, we review the most commonly used modeling solutions in DID estimation for panel data, including generalized estimating equations (GEE), permutation tests, clustered standard errors (CSE), wild cluster bootstrapping, and aggregation. Second, we compare the empirical coverage rates and power of these methods using a Monte Carlo simulation study in scenarios in which we vary the degree of error correlation, the group size balance, and the proportion of treated groups. Third, we provide an empirical example using the Survey of Health, Ageing, and Retirement in Europe. RESULTS: When the number of groups is small, CSE are systematically biased downwards in scenarios when data are unbalanced or when there is a low proportion of treated groups. This can result in over-rejection of the null even when data are composed of up to 50 groups. Aggregation, permutation tests, bias-adjusted GEE, and wild cluster bootstrap produce coverage rates close to the nominal rate for almost all scenarios, though GEE may suffer from low power. CONCLUSIONS: In DID estimation with a small number of groups, analysis using aggregation, permutation tests, wild cluster bootstrap, or bias-adjusted GEE is recommended.

The role of families in decisions regarding cancer treatments.

BACKGROUND: Shared decision-making is an important component of patient-centered care and is associated with improved outcomes. To the authors' knowledge, little is known concerning the extent and predictors of the involvement of a patient's family in decisions regarding cancer treatments. METHODS: The Cancer Care Outcomes Research and Surveillance (CanCORS) Consortium is a large, multiregional, prospective cohort study of the cancer care and outcomes of patients with lung and colorectal cancer. Participants reported the roles of their families in decision-making regarding treatment. Multinomial logistic regression was used to assess patient factors associated with family roles in decisions. RESULTS: Among 5284 patients, 80 (1.5%) reported family-controlled decisions, with the highest adjusted rates (12.8%) noted among non-English-speaking Asians. Among the 5204 remaining patients, 49.4% reported equally sharing decisions with family, 22.1% reported some family input, and 28.5% reported little family input. In adjusted analyses, patients who were married, female, older, and insured more often reported equally shared decisions with family (all P <.001). Adjusted family involvement varied by race/ethnicity and language, with Chinese-speaking Asian (59.8%) and Spanish-speaking Hispanic (54.8%) patients equally sharing decisions with family more often than white individuals (47.6%). Veterans Affairs patients were least likely to report sharing decisions with family, even after adjustment for marital status and social support (P <.001). CONCLUSIONS: The majority of patients with newly diagnosed lung or colorectal cancer involve family members in treatment decisions. Non-English-speaking Asians and Hispanics rely significantly on family. Further studies are needed to determine the impact of family involvement in treatment decisions on outcomes; until then, physicians should consider eliciting patients' preferences for family involvement.

Developing and Evaluating Composite Measures of Cancer Care Quality.

BACKGROUND: Composite measures are useful for distilling quality data into summary scores; yet, there has been limited use of composite measures for cancer care. OBJECTIVE: Compare multiple approaches for generating cancer care composite measures and evaluate how well composite measures summarize dimensions of cancer care and predict survival. STUDY DESIGN: We computed hospital-level rates for 13 colorectal, lung, and prostate cancer process measures in 59 Veterans Affairs hospitals. We computed 4 empirical-factor (based on an exploratory factor analysis), 3 cancer-specific (colorectal, lung, prostate care), and 3 care modality-specific (diagnosis/evaluation, surgical, nonsurgical treatments) composite measures. We assessed correlations among all composite measures and estimated all-cause survival for colon, rectal, non-small cell lung, and small cell lung cancers as a function of composite scores, adjusting for patient characteristics. RESULTS: Four factors emerged from the factor analysis: nonsurgical treatment, surgical treatment, colorectal early diagnosis, and prostate treatment. We observed strong correlations (r) among composite measures comprised of similar process measures (r=0.58-1.00, P<0.0001), but not among composite measures reflecting different care dimensions. Composite measures were rarely associated with survival. CONCLUSIONS: The empirical-factor domains grouped measures variously by cancer type and care modality. The evidence did not support any single approach for generating cancer care composite measures. Weak associations across different care domains suggest that low-quality and high-quality cancer care delivery may coexist within Veterans Affairs hospitals.

Geographic variation in cancer-related imaging: Veterans Affairs health care system versus Medicare.

BACKGROUND: Geographic variations in use of medical services have been interpreted as indirect evidence of wasteful care. Less overuse of services, however, may not be reliably associated with less geographic variation. OBJECTIVE: To compare average use and geographic variation in use of cancer-related imaging between fee-for-service Medicare and the Department of Veterans Affairs (VA) health care system. DESIGN: Observational analysis of cancer-related imaging from 2003 to 2005 using Medicare and VA utilization data linked to cancer registry data. Multilevel models, adjusted for sociodemographic and tumor characteristics, were used to estimate mean differences in annual imaging use between cohorts of Medicare and VA patients within geographic areas and variation in use across areas for each cohort. SETTING: 40 hospital referral regions. PATIENTS: Older men with lung, colorectal, or prostate cancer, including 34,475 traditional Medicare beneficiaries (Medicare cohort) and 6835 VA patients (VA cohort). MEASUREMENTS: Per-patient count of imaging studies for which lung, colorectal, or prostate cancer was the primary diagnosis (each study weighted by a standardized price), and a direct measure of overuse-advanced imaging for prostate cancer at low risk for metastasis. RESULTS: Adjusted annual use of cancer-related imaging was lower in the VA cohort than in the Medicare cohort (price-weighted count, $197 vs. $379 per patient; P < 0.001), as was annual use of advanced imaging for prostate cancer at low risk for metastasis ($41 vs. $117 per patient; P < 0.001). Geographic variation in cancer-related imaging use was similar in magnitude in the VA and Medicare cohorts. LIMITATION: Observational study design. CONCLUSION: Use of cancer-related imaging was lower in the VA health care system than in fee-for-service Medicare, but lower use was not associated with less geographic variation. Geographic variation in service use may not be a reliable indicator of the extent of overuse. PRIMARY FUNDING SOURCE: Doris Duke Charitable Foundation and Department of Veterans Affairs Office of Policy and Planning.

Racial disparities in cancer care in the Veterans Affairs health care system and the role of site of care.

OBJECTIVES: We assessed cancer care disparities within the Veterans Affairs (VA) health care system and whether between-hospital differences explained disparities. METHODS: We linked VA cancer registry data with VA and Medicare administrative data and examined 20 cancer-related quality measures among Black and White veterans diagnosed with colorectal (n = 12,897), lung (n = 25,608), or prostate (n = 38,202) cancer from 2001 to 2004. We used logistic regression to assess racial disparities for each measure and hospital fixed-effects models to determine whether disparities were attributable to between- or within-hospital differences. RESULTS: Compared with Whites, Blacks had lower rates of early-stage colon cancer diagnosis (adjusted odds ratio [AOR] = 0.80; 95% confidence interval [CI] = 0.72, 0.90), curative surgery for stage I, II, or III rectal cancer (AOR = 0.57; 95% CI = 0.41, 0.78), 3-year survival for colon cancer (AOR = 0.75; 95% CI = 0.62, 0.89) and rectal cancer (AOR = 0.61; 95% CI = 0.42, 0.87), curative surgery for early-stage lung cancer (AOR = 0.50; 95% CI = 0.41, 0.60), 3-dimensional conformal or intensity-modulated radiation (3-D CRT/IMRT; AOR = 0.53; 95% CI = 0.47, 0.59), and potent antiemetics for highly emetogenic chemotherapy (AOR = 0.87; 95% CI = 0.78, 0.98). Adjustment for hospital fixed-effects minimally influenced racial gaps except for 3-D CRT/IMRT (AOR = 0.75; 95% CI = 0.65, 0.87) and potent antiemetics (AOR = 0.95; 95% CI = 0.82, 1.10). CONCLUSIONS: Disparities in VA cancer care were observed for 7 of 20 measures and were primarily attributable to within-hospital differences.

Combining information from two data sources with misreporting and incompleteness to assess hospice-use among cancer patients: a multiple imputation approach.

Combining information from multiple data sources can enhance estimates of health-related measures by using one source to supply information that is lacking in another, assuming the former has accurate and complete data. However, there is little research conducted on combining methods when each source might be imperfect, for example, subject to measurement errors and/or missing data. In a multisite study of hospice-use by late-stage cancer patients, this variable was available from patients' abstracted medical records, which may be considerably underreported because of incomplete acquisition of these records. Therefore, data for Medicare-eligible patients were supplemented with their Medicare claims that contained information on hospice-use, which may also be subject to underreporting yet to a lesser degree. In addition, both sources suffered from missing data because of unit nonresponse from medical record abstraction and sample undercoverage for Medicare claims. We treat the true hospice-use status from these patients as a latent variable and propose to multiply impute it using information from both data sources, borrowing the strength from each. We characterize the complete-data model as a product of an 'outcome' model for the probability of hospice-use and a 'reporting' model for the probability of underreporting from both sources, adjusting for other covariates. Assuming the reports of hospice-use from both sources are missing at random and the underreporting are conditionally independent, we develop a Bayesian multiple imputation algorithm and conduct multiple imputation analyses of patient hospice-use in demographic and clinical subgroups. The proposed approach yields more sensible results than alternative methods in our example. Our model is also related to dual system estimation in population censuses and dual exposure assessment in epidemiology.

Diabetes-Focused Health Care Utilization Among Adolescents and Young Adults With Type 1 Diabetes.

OBJECTIVE: To describe the current rates of health services use with various types of providers among adolescents and young adults (AYA) with type 1 diabetes (T1D) and evaluate which patient factors are associated with rates of service use from different provider types. METHODS: Using 2012-16 claims data from a national commercial insurer, we identified 18,927 person-years of AYA with T1D aged 13 to 26 years and evaluated the frequency at which: 1) AYA skipped diabetes care for a year despite being insured; 2) received care from pediatric or non-pediatric generalists or endocrinologists if care was sought; and 3) received annual hemoglobin A1c (HbA1c) testing as recommended for AYA. We used descriptive statistics and multivariable regression to examine patient, insurance, and physician characteristics associated with utilization and quality outcomes. RESULTS: Between ages 13 and 26, the percentage of AYA with: any diabetes-focused visits declined from 95.3% to 90.3%; the mean annual number of diabetes-focused visits, if any, decreased from 3.5 to 3.0; receipt of ≥2 HbA1c tests annually decreased from 82.3% to 60.6%. Endocrinologists were the majority providers of diabetes care across ages, yet the relative proportion of AYA whose diabetes care was endocrinologist-dominated decreased from 67.3% to 52.7% while diabetes care dominated by primary care providers increased from 19.9% to 38.2%. The strongest predictors of diabetes care utilization were younger age and use of diabetes technology (pumps and continuous glucose monitors). CONCLUSIONS: Several provider types are involved in the care of AYA with T1D, though predominate provider type and care quality changes substantially across age in a commercially-insured population.

Representativeness of participants in the cancer care outcomes research and surveillance consortium relative to the surveillance, epidemiology, and end results program.

BACKGROUND: The research goals of the Cancer Care Outcomes Research and Surveillance (CanCORS) Consortium are to determine how characteristics and beliefs of patients, providers, and health care organizations influence the treatments and outcomes of individuals with newly diagnosed lung and colorectal cancers. As CanCORS results will inform national policy, it is important to know how they generalize to the United States population with these cancers. RESEARCH DESIGN: This study assessed the representativeness of the CanCORS cohort of 10,547 patients with lung cancer (LC) or colorectal cancer (CRC) enrolled between 2003 and 2005. We compared characteristics (sex, race, age, and disease stage) with the Surveillance, Epidemiology, and End Results (SEER) population of 234,464 patients with new onset of these cancers during the CanCORS recruitment period. RESULTS: The CanCORS sample is well matched to the SEER Program for both cancers. In CanCORS, 41% LC/47% CRC were female versus 47% LC/49% CRC in SEER. African American, Hispanic, and Asian cases differed by no more than 5 percentage points between CanCORS and SEER. The SEER population is slightly older, with the percentage of patients older than 75 years 33.1% LC/37.3% CRC in SEER versus 26.9% LC/29.4% in CanCORS, and also has a slightly higher proportion of early stage patients. We also found that the CanCORS cohort was representative within specific SEER regions that map closely to CanCORS sites. CONCLUSIONS: This study demonstrates that the CanCORS Consortium was successful in enrolling a demographically representative sample within the CanCORS regions.

Adverse clinical events among medicare beneficiaries using antipsychotic drugs: linking health insurance benefits and clinical needs.

OBJECTIVE: Medicare Part D provides formulary protections for antipsychotics but does not exempt these drugs from cost-sharing. We investigated the impact of Part D coverage on antipsychotic drug spending, adherence, and clinical outcomes among beneficiaries with varying indications for use. METHODS: We conducted a historical cohort study of Medicare Advantage beneficiaries who received antipsychotic drugs, with diagnoses of schizophrenia or bipolar disorder or with no mental health diagnoses (N=10,190). Half had a coverage gap; half had no gap because of low-income subsidies. Using fixed effects regression models, we examined changes in spending and adherence as beneficiaries experienced cost-sharing increases after reaching the gap. We examined changes in hospitalizations and emergency department visits using proportional hazard models. RESULTS: Across all diagnostic groups, total monthly expenditure on antipsychotic drugs decreased with cost-sharing increases in the gap compared with those with no gap (eg, schizophrenia: -$123 95% confidence interval [-$138, -$108]), and out-of-pocket spending increased (eg, schizophrenia: $104 [$98, $110]). Adherence similarly decreased, with the largest declines among those with schizophrenia (-20.6 percentage points [-22.3, -18.9] in proportion of days covered). Among beneficiaries with schizophrenia and bipolar disorder, hospitalizations and emergency department visit rates increased with cost-sharing increases (eg, schizophrenia: hazard ratio=1.32 [1.06, 1.65] for all hospitalizations), but did not among subjects without mental health diagnoses. Clinical event rates did not change among beneficiaries with low-income subsidies without gaps. CONCLUSIONS: There is evidence of interruptions in antipsychotic use attributable to Part D cost-sharing. Adverse events increased among beneficiaries with approved indications for use, but not among beneficiaries without such indications.

Propensity score weighting with multilevel data.

Propensity score methods are being increasingly used as a less parametric alternative to traditional regression to balance observed differences across groups in both descriptive and causal comparisons. Data collected in many disciplines often have analytically relevant multilevel or clustered structure. The propensity score, however, was developed and has been used primarily with unstructured data. We present and compare several propensity-score-weighted estimators for clustered data, including marginal, cluster-weighted, and doubly robust estimators. Using both analytical derivations and Monte Carlo simulations, we illustrate bias arising when the usual assumptions of propensity score analysis do not hold for multilevel data. We show that exploiting the multilevel structure, either parametrically or nonparametrically, in at least one stage of the propensity score analysis can greatly reduce these biases. We applied these methods to a study of racial disparities in breast cancer screening among beneficiaries of Medicare health plans.

Adoption of Innovative Therapies Across Oncology Practices-Evidence From Immunotherapy.

Importance: Immunotherapies reflect an important breakthrough in cancer treatment, substantially improving outcomes for patients with a variety of cancer types, yet little is known about which practices have adopted this novel therapy or the pace of adoption. Objective: To assess adoption of immunotherapies across US oncology practices and examine variation in adoption by practice type. Design, Setting, and Participants: This cohort study used data from Medicare fee-for-service beneficiaries undergoing 6-month chemotherapy episodes between 2010 and 2017. Data were analyzed January 19, 2021, to September 28, 2022, for patients with cancer types for which immunotherapy was approved by the US Food and Drug Administration (FDA) during the study period: melanoma, kidney cancer, lung cancer, and head and neck cancer. Exposures: Oncology practice location (rural vs urban), affiliation type (academic system, nonacademic system, independent), and size (1 to 5 physicians vs 6 or more physicians). Main Outcomes and Measures: The primary outcome was whether a practice adopted immunotherapy. Adoption rates for each practice type were estimated using multivariate linear models that adjusted for patient characteristics (age, sex, race and ethnicity, cancer type, Charlson Comorbidity Index, and median household income). Results: Data included 71 659 episodes at 1732 oncology practices. Of these, 264 practices (15%) were rural, 900 (52%) were independent, and 492 (28%) had 1 to 5 physicians. Most practices adopted immunotherapy within 2 years of FDA approval, but there was substantial variation in adoption rates across practice types. After FDA approval, adoption of immunotherapy was 11 (95% CI, -16 to -6) percentage points lower at rural practices than urban practices and 27 (95% CI, -32 to -22) percentage points lower at practices with 1 to 5 physicians than practices with 6 or more physicians. Adoption rates were similar at independent practices and nonacademic systems; however, both practice types had lower adoption than academic systems (independent practice difference, -6 [95% CI, -9 to -3] percentage points; nonacademic systems difference, -9 [95% CI, -11 to -6] percentage points). Conclusions and Relevance: In this cohort study of Medicare claims, practice characteristics, especially practice size and rural location, were associated with adoption of immunotherapy. These findings suggest that there may be geographic disparities in access to important innovations for treating patients with cancer.

Specialist use among privately insured children with disabilities.

OBJECTIVE: To investigate primary care practice ownership and specialist-use patterns for commercially insured children with disabilities. DATA SOURCES AND STUDY SETTING: A national commercial claims database and the Health Systems and Provider Database from 2012 to 2016 are the data sources for this study. STUDY DESIGN: This cross-sectional, descriptive study examines: (1) the most visited type of pediatric primary care physician and practice (independent or system-owned); (2) pediatric and non-pediatric specialist-use patterns; and (3) how practice ownership relates to specialist-use patterns. DATA COLLECTION/EXTRACTION METHODS: This study identifies 133,749 person-years of commercially insured children with disabilities aged 0-18 years with at least 24 months of continuous insurance coverage by linking a national commercial claims data set with the Health Systems and Provider Database and applying the validated Children with Disabilities Algorithm. PRINCIPAL FINDINGS: Three-quarters (75.9%) of children with disabilities received their pediatric primary care in independent practices. Nearly two thirds (59.6%) used at least one specialist with 45.1% using nonpediatric specialists, 28.8% using pediatric ones, and 17.0% using both. Specialist-use patterns varied by both child age and specialist type. Children with disabilities in independent practices were as likely to see a specialist as those in system-owned ones: 57.1% (95% confidence interval [95% CI] 56.7%-57.4%) versus 57.3% (95% CI 56.6%-58.0%), respectively (p = 0.635). The percent using two or more types of specialists was 46.1% (95% CI 45.4%-46.7%) in independent practices, comparable to that in systems 47.1% (95% CI 46.2%-48.0%) (p = 0.054). However, the mean number of specialist visits was significantly lower in independent practices than in systems-4.0 (95% CI 3.9%-4.0%) versus 4.4 (95% CI 4.3%-4.6%) respectively-reaching statistical significance with p < 0.0001. CONCLUSIONS: Recognizing how privately insured children with disabilities use pediatric primary care from pediatric and nonpediatric primary care specialists through both independent and system-owned practices is important for improving care quality and value.

Organization of Cancer Specialists in US Physician Practices and Health Systems.

PURPOSE: To describe the supply of cancer specialists, the organization of cancer care within versus outside of health systems, and the distance to multispecialty cancer centers. METHODS: Using the 2018 Health Systems and Provider Database from the National Bureau of Economic Research and 2018 Medicare data, we identified 46,341 unique physicians providing cancer care. We stratified physicians by discipline (adult/pediatric medical oncologists, radiation oncologists, surgical/gynecologic oncologists, other surgeons performing cancer surgeries, or palliative care physicians), system type (National Cancer Institute [NCI] Cancer Center system, non-NCI academic system, nonacademic system, or nonsystem/independent practice), practice size, and composition (single disciplinary oncology, multidisciplinary oncology, or multispecialty). We computed the density of cancer specialists by county and calculated distances to the nearest NCI Cancer Center. RESULTS: More than half of all cancer specialists (57.8%) practiced in health systems, but 55.0% of cancer-related visits occurred in independent practices. Most system-based physicians were in large practices with more than 100 physicians, while those in independent practices were in smaller practices. Practices in NCI Cancer Center systems (95.2%), non-NCI academic systems (95.0%), and nonacademic systems (94.3%) were primarily multispecialty, while fewer independent practices (44.8%) were. Cancer specialist density was sparse in many rural areas, where the median travel distance to an NCI Cancer Center was 98.7 miles. Distances to NCI Cancer Centers were shorter for individuals living in high-income areas than in low-income areas, even for individuals in suburban and urban areas. CONCLUSION: Although many cancer specialists practiced in multispecialty health systems, many also worked in smaller-sized independent practices where most patients were treated. Access to cancer specialists and cancer centers was limited in many areas, particularly in rural and low-income areas.

Racial and Ethnic Disparities in Opioid Access and Urine Drug Screening Among Older Patients With Poor-Prognosis Cancer Near the End of Life.

PURPOSE: To characterize racial and ethnic disparities and trends in opioid access and urine drug screening (UDS) among patients dying of cancer, and to explore potential mechanisms. METHODS: Among 318,549 non-Hispanic White (White), Black, and Hispanic Medicare decedents older than 65 years with poor-prognosis cancers, we examined 2007-2019 trends in opioid prescription fills and potency (morphine milligram equivalents [MMEs] per day [MMEDs]) near the end of life (EOL), defined as 30 days before death or hospice enrollment. We estimated the effects of race and ethnicity on opioid access, controlling for demographic and clinical factors. Models were further adjusted for socioeconomic factors including dual-eligibility status, community-level deprivation, and rurality. We similarly explored disparities in UDS. RESULTS: Between 2007 and 2019, White, Black, and Hispanic decedents experienced steady declines in EOL opioid access and rapid expansion of UDS. Compared with White patients, Black and Hispanic patients were less likely to receive any opioid (Black, -4.3 percentage points, 95% CI, -4.8 to -3.6; Hispanic, -3.6 percentage points, 95% CI, -4.4 to -2.9) and long-acting opioids (Black, -3.1 percentage points, 95% CI, -3.6 to -2.8; Hispanic, -2.2 percentage points, 95% CI, -2.7 to -1.7). They also received lower daily doses (Black, -10.5 MMED, 95% CI, -12.8 to -8.2; Hispanic, -9.1 MMED, 95% CI, -12.1 to -6.1) and lower total doses (Black, -210 MMEs, 95% CI, -293 to -207; Hispanic, -179 MMEs, 95% CI, -217 to -142); Black patients were also more likely to undergo UDS (0.5 percentage points; 95% CI, 0.3 to 0.8). Disparities in EOL opioid access and UDS disproportionately affected Black men. Adjustment for socioeconomic factors did not attenuate the EOL opioid access disparities. CONCLUSION: There are substantial and persistent racial and ethnic inequities in opioid access among older patients dying of cancer, which are not mediated by socioeconomic variables.

Practice-Level Variation in Molecular Testing and Use of Targeted Therapy for Patients With Non-Small Cell Lung Cancer and Colorectal Cancer.

Importance: All patients with newly diagnosed non-small cell lung cancer (NSCLC) and colorectal cancer (CRC) should receive molecular testing to identify those who can benefit from targeted therapies. However, many patients do not receive recommended testing and targeted therapies. Objective: To compare rates of molecular testing and targeted therapy use by practice type and across practices. Design, Setting, and Participants: This cross-sectional study used 100% Medicare fee-for-service data from 2015 through 2019 to identify beneficiaries with new metastatic NSCLC or CRC diagnoses receiving systemic therapy and to assign patients to oncology practices. Hierarchical linear models were used to characterize variation by practice type and across practices. Data analysis was conducted from June 2019 to October 2022. Exposures: Oncology practice providing care. Outcomes: Primary outcomes were rates of molecular testing and targeted therapy use for patients with NSCLC and CRC. Secondary outcomes were rates of multigene testing for NSCLC and CRC. Results: There were 106 228 Medicare beneficiaries with incident NSCLC (31 521 [29.7%] aged 65-69 years; 50 348 [47.4%] female patients; 2269 [2.1%] Asian, 8282 [7.8%] Black, and 91 215 [85.9%] White patients) and 39 512 beneficiaries with incident CRC (14 045 [35.5%] aged 65-69 years; 17 518 [44.3%] female patients; 896 [2.3%] Asian, 3521 [8.9%] Black, and 32 753 [82.9%] White patients) between 2015 and 2019. Among these beneficiaries, 18 435 (12.9%) were treated at National Cancer Institute (NCI)-designated centers, 8187 (5.6%) were treated at other academic centers, and 94 329 (64.7%) were treated at independent oncology practices. Molecular testing rates increased from 74% to 85% for NSCLC and 45% to 65% for CRC. First-line targeted therapy use decreased from 12% to 8% among patients with NSCLC and was constant at 5% for patients with CRC. For NSCLC, molecular testing rates were similar across practice types while rates of multigene panel use (13.2%) and targeted therapy use (16.6%) were highest at NCI-designated cancer centers. For CRC, molecular testing rates were 3.8 (95% CI: 1.2-6.5), 3.3 (95% CI, 0.4-6.1), and 12.2 (95% CI, 9.1-15.3) percentage points lower at hospital-owned practices, large independent practices, and small independent practices, respectively. Rates of targeted therapy use for CRC were similar across practice types. After adjusting for patient characteristics, there was moderate variation in molecular testing and targeted therapy use across oncology practices. Conclusions and Relevance: In this cross-sectional study of Medicare beneficiaries, molecular testing rates for NSCLC and CRC increased in recent years but remained lower than recommended levels. Rates of targeted therapy use decreased for NSCLC and remained stable for CRC. Variation across practices suggests that where a patient was treated may have affected access to recommended testing and efficacious treatments.