ACO leakage among gynecologic cancer patients: Incidence, predictors, and impact on annual Medicare expenditure.

OBJECTIVE: To examine patterns of Accountable Care Organizations (ACO) leakage, the receipt of healthcare by ACO-assigned patients from institutions outside assigned ACO network, among patients with gynecologic cancer. ACO leakage was estimated as rates of patients seeking care external to their ACO assignment. Factors associated with ACO leakage were identified and cost differences within the first year of cancer diagnosis described. METHODS: Medicare 5% data (2013-2017) was used to quantify rates of leakage among gynecologic cancer patients with stable ACO assignment. Crude and multivariable adjusted risk ratios of ACO leakage risk factors were estimated using log-binomial regression models. Overall and cancer-specific spending differences by ACO leakage status were compared using Wilcoxon rank-sum test. RESULTS: Overall incidence of ACO leakage was 28.1% with highest leakage for outpatient care and uterine cancer patients. ACO leakage risk was 56% higher among Black relative to White patients, and 77% more for those in higher relative to lowest quintiles of median household income. Leakage decreased by 3% and 8% with each unit increase in ACO size and number of subspecialists, respectively. Healthcare costs were 19.5% higher for leakage patients. CONCLUSIONS: ACO leakage rates among gynecologic cancer patients was overall modest, with some regional and temporal variation, higher leakage for certain subgroups and substantially higher Medicare spending in inpatient and outpatient settings for patients with ACO leakage. These findings identify targets for further investigations and strategies to encourage oncologists to participate in ACOs and prevent increased health care costs associated with use of non-ACO providers.

Neuropsychiatric Comorbidities and Psychotropic Medication Use in Medicare Beneficiaries With Dementia by Sex and Race.

BACKGROUND: Neuropsychiatric symptoms affect the majority of dementia patients. Past studies report high rates of potentially inappropriate prescribing of psychotropic medications in this population. We investigate differences in neuropsychiatric diagnoses and psychotropic medication prescribing in a local US cohort by sex and race. METHODS: We utilize Medicare claims and prescription fill records in a cohort of 100% Medicare North and South Carolina beneficiaries ages 50 and above for the year 2017 with a dementia diagnosis. We identify dementia and quantify diagnosis of anxiety, depression and psychosis using validated coding algorithms. We search Medicare claims for antianxiety, antidepressant and antipsychotic medications to determine prescriptions filled. RESULTS: Anxiety and depression were diagnosed at higher rates in White patients; psychosis at higher rates in Black patients. (P < .001) Females were diagnosed with anxiety, depression and psychosis at higher rates than males (P < .001) and filled more antianxiety and antidepressant medications than males. (P < .001) Black and Other race patients filled more antipsychotic medications for anxiety, depression and psychosis than White patients. (P < .001) Antidepressants were prescribed at higher rates than antianxiety or antipsychotic medications across all patients and diagnoses. Of patients with no neuropsychiatric diagnosis, 11.4% were prescribed an antianxiety medication, 22.8% prescribed an antidepressant and 7.6% prescribed an antipsychotic. CONCLUSIONS: The high fill rate of antianxiety (benzodiazepine) medications in dementia patients, especially females is a concern. Patients are prescribed psychotropic medications at high rates. This practice may represent potentially inappropriate prescribing. Patient/caregiver education with innovative community outreach and care delivery models may help decrease medication use.

Linking Medicare-Medicaid Claims for Patient-Centered Outcomes Research Among Dual-Eligible Beneficiaries.

BACKGROUND: Evaluation of Medicare-Medicaid integration models' effects on patient-centered outcomes and costs requires multiple data sources and validated processes for linkage and reconciliation. OBJECTIVE: To describe the opportunities and limitations of linking state-specific Medicaid and Centers for Medicare & Medicaid Services administrative claims data to measure patient-centered outcomes for North Carolina dual-eligible beneficiaries. RESEARCH DESIGN: We developed systematic processes to (1) validate the beneficiary ID linkage using sex and date of birth in a beneficiary ID crosswalk, (2) verify dates of dual enrollment, and (3) reconcile Medicare-Medicaid claims data to support the development and use of patient-centered outcomes in linked data. PARTICIPANTS: North Carolina Medicaid beneficiaries with full Medicaid benefits and concurrent Medicare enrollment (FBDE) between 2014 and 2017. MEASURES: We identified need-based subgroups based on service use and eligibility program requirements. We calculated utilization and costs for Medicaid and Medicare, matched Medicaid claims to Medicare service categories where possible, and reported outcomes by the payer. Some services were covered only by Medicaid or Medicare, including Medicaid-only covered home and community-based services (HCBS). RESULTS: Of 498,030 potential dual enrollees, we verified the linkage and FBDE eligibility of 425,664 (85.5%) beneficiaries, including 281,174 adults enrolled in Medicaid and Medicare fee-for-service. The most common need-based subgroups were intensive behavioral health service users (26.2%) and HCBS users (10.8%) for adults under age 65, and HCBS users (20.6%) and nursing home residents (12.4%) for adults age 65 and over. Medicaid funded 42% and 49% of spending for adults under 65 and adults 65 and older, respectively. Adults under 65 had greater behavioral health service utilization but less skilled nursing facility, HCBS, and home health utilization compared with adults 65 and older. CONCLUSIONS: Linkage of Medicare-Medicaid data improves understanding of patient-centered outcomes among FBDE by combining Medicare-funded acute and ambulatory services with Medicaid-funded HCBS. Using linked Medicare-Medicaid data illustrates the diverse patient experience within FBDE beneficiaries, which is key to informing patient-centered outcomes, developing and evaluating integrated Medicare and Medicaid programs, and promoting health equity.

Dementia and Parkinson's disease diagnoses in electronic health records vs. Medicare claims data: a study of 101,980 linked patients.

BACKGROUND: Medicare claims and electronic health record data are both commonly used for research and clinical practice improvement; however, it is not known how concordant diagnoses of neurodegenerative diseases (NDD, comprising dementia and Parkinson's disease) are in these data types. Therefore, our objective was to determine the sensitivity and specificity of neurodegenerative disease (NDD) diagnoses contained in structured electronic health record (EHR) data compared to Medicare claims data. METHODS: This was a retrospective cohort study of 101,980 unique patients seen at a large North Carolina health system between 2013-2017, which were linked to 100% North and South Carolina Medicare claims data, to evaluate the accuracy of diagnoses of neurodegenerative diseases in EHRs compared to Medicare claims data. Patients age > 50 who were enrolled in fee-for-service Medicare were included in the study. Patients were classified as having or not having NDD based on the presence of validated ICD-CM-9 or ICD-CM-10 codes associated with NDD or claims for prescription drugs used to treat NDD. EHR diagnoses were compared to Medicare claims diagnoses. RESULTS: The specificity of any EHR diagnosis of NDD was 99.0%; sensitivity was 61.3%. Positive predictive value and negative predictive value were 90.8% and 94.1% respectively. Specificity of an EHR diagnosis of dementia was 99.0%, and sensitivity was 56.1%. Specificity of an EHR diagnosis of PD was 99.7%, while sensitivity was 76.1%. CONCLUSIONS: More research is needed to investigate under-documentation of NDD in electronic health records relative to Medicare claims data, which has major implications for clinical practice (particularly patient safety) and research using real-world data.

Racial/ethnic disparities in dementia incidence, outcomes, and health-care utilization.

INTRODUCTION: Racial/ethnic disparities exist in many aspects of health care, but data on racial/ethnic disparities for neurodegenerative diseases (NDDs), such as dementia and Parkinson's disease (PD), are limited. METHODS: We used North and South Carolina Medicare claims from 2013 to 2017 to evaluate disparities in incidence of NDDs and in health-care utilization and outcomes for patients with NDDs. RESULTS: Disparities in incidence of NDD between Black and White beneficiaries narrowed by 0.37 per 100 person-years from 2014 to 2017. After thorough covariate adjustment, Black beneficiaries had a 4% higher risk of all-cause hospitalization, spent 8% more days in skilled nursing facilities and 14% fewer days in hospice facilities, were 38% less likely to receive physical/occupational therapy services, were 8% less likely to receive dementia medications, and were 19% less likely to receive PD medications than White beneficiaries. DISCUSSION: Effective system-level approaches to promote health equity in NDD diagnosis, treatment, and outcomes are clearly needed. HIGHLIGHTS: Racial disparities in neurodegenerative disease incidence narrowed between 2014 and 2017. Black patients were less likely than White patients to receive hospice services. Black patients were less likely than White patients to receive physical therapy. Black patients were less likely than White patients to receive Alzheimer's disease or Parkinson's disease medications. There is a shortage of neurologists in counties with high dementia incidence.

Bimodal distribution of atrial fibrillation burden in 3 distinct cohorts: What is 'paroxysmal' atrial fibrillation?

BACKGROUND: Burden of atrial fibrillation (AF), as a continuous measure, is an emerging alternative classification often assumed to increase linearly with progression of disease. Yet there are no descriptions of AF burden distributions across populations. METHODS: We examined patterns of AF burden (% time in AF) across 3 different cohorts: outpatients with AF undergoing Holter monitoring in a national registry (ORBIT-AF II), routine outpatients undergoing Holter monitoring in a tertiary healthcare system (UHealth), and patients >= 65 years with cardiac implantable electronic devices (Merlin.netTM linked to Medicare). RESULTS: We included 2,058 ORBIT-AF II patients, 4,537 UHealth patients, and 39,710 from Merlin.net. Mean age ranged from 56 to 77 years, sex ranged from 40% to 61% male, and mean CHA2DS2-VASc scores ranged from 2.2 to 4.9. Across all cohorts, AF burden demonstrated skewed frequency towards the extremes, with the vast majority of patients having either very low or very high AF burden. This bimodal distribution was consistent across cohorts, across clinically-documented AF types (paroxysmal v persistent), patients with or without a known AF diagnosis, and among patients with different types of cardiac implantable electronic devices. CONCLUSIONS: Across 3 broad, diverse cohorts with continuous monitoring, distribution of AF burden was consistently skewed towards the extremes without an even, linear distribution or progression. As AF burden is increasingly recognized as a descriptor and potential risk-stratifier, these findings have important implications for future research and patient care.

Association between economic and arrhythmic burden of paroxysmal atrial fibrillation in patients with cardiac implanted electronic devices.

BACKGROUND: As the prevalence of atrial fibrillation (AF) increases, a greater understanding of the costs associated with AF care is required. While individuals with greater arrhythmic burden may interact with the health system more frequently, the relationship between AF burden and costs remains undefined. METHODS: In a longitudinal patient cohort with paroxysmal AF and newly implanted cardiovascular implantable electronic devices (CIED) (2010-2016), we linked CIED remote-monitoring data with Medicare claims to assess the association between the 30-day averaged device-detected daily percentage of time in AF ("AF burden") and healthcare costs. RESULTS: Among 39,345 patients, the mean age was 77.1 ± 8.7 years, 60.7% were male, and the mean CHA2DS2-VASc score was 4.9 ± 1.3. The mean total 1-year costs were $18,668 ± 29,173, driven by hospitalization costs where two-thirds of admissions were due to heart failure. Increasing AF burden was associated with increasing costs: $17,860 ± 28,525 for 0% daily AF burden; $18,840 ± 29,104 for 0-5% daily AF burden; and $20,968 ± 31,228 for 5-98% daily AF burden. There was a linear relationship between AF burden expressed as a continuous variable and 1-year costs (adjusted cost rate ratio 1.031 per 10% daily duration in AF, 95% confidence interval 1.023-1.038; P < .001). CONCLUSIONS: Among older patients with paroxysmal AF and CIEDs, increasing AF burden is associated with higher healthcare costs. As the observational study design is unable to determine causal relationships, prospective study is required to explore the intriguing hypothesis that targeted AF strategies, including comorbidity management, that reduce AF burden may also reduce the high annual costs associated with AF.

Impact of risk factor control on peripheral artery disease outcomes and health disparities.

BACKGROUND: Peripheral artery disease (PAD) is associated with modifiable atherosclerotic risk factors like hypertension, diabetes, hyperlipidemia, and smoking. However, the effect of risk factor control on outcomes and disparities in achieving control is less well understood. METHODS: All patients in an integrated, regional health system with PAD-related encounters, fee-for-service Medicare, and clinical risk factor control data were identified. Component risk factors were dichotomized into controlled and uncontrolled categories (control defined as low-density lipoprotein < 100 mg/dL, hemoglobin A1c < 7.0%, SBP < 140 mmHg, and current nonsmoker) and composite categories (none, 1, ⩾ 2 uncontrolled RFs) created. The primary outcome was major adverse vascular events (MAVE, a composite of all-cause mortality, myocardial infarction, stroke, and lower-extremity revascularization and amputation). RESULTS: The cohort included 781 patients with PAD, average age 72.5 ± 9.8 years, of whom 30.1% were Black, and 19.1% were Medicaid dual-enrolled. In this cohort, 260 (33.3%) had no uncontrolled risk factors and 200 (25.6%) had two or more uncontrolled risk factors. Patients with the poorest risk factor control were more likely to be Black (p < 0.001), Medicaid dual-enrolled (p < 0.001), and have chronic limb-threatening ischemia (p = 0.009). Significant differences in MAVE by degree of risk factor control were observed at 30 days (none uncontrolled: 5.8%, 1 uncontrolled: 11.5%, ⩾ 2 uncontrolled: 13.6%; p = 0.01) but not at 1 year (p = 0.08). risk factor control was not associated with outcomes at 1 year after adjustment for patient and PAD-specific characteristics. CONCLUSIONS: risk factor control is poor among patients with PAD. Significant disparities in achieving optimal risk factor control represent a potential target for reducing inequities in outcomes.

Underutilization of Guideline-based Abdominal Aortic Aneurysm Screening in an Academic Health System.

OBJECTIVES: The US Preventive Services Task Force (USPSTF) recommends a 1 time screening for AAA with ultrasonography in men aged 65-75 who have ever smoked. Our objectives were to identify the AAA screening rates in a large academic health system and assess factors associated with receipt of screening. METHODS: Data were extracted from electronic health records from the Duke University Health System and the US Census Bureau. Index screening eligibility date was defined as the 65th birthdate for male patients with a history of smoking. Patients with an index screening eligibility date between January 1, 2016 and December 31, 2018 were included in the study population and followed through December 31, 2019. Screened patients were identified by procedure codes for ultrasonography, CT or MRI. RESULTS: Among 6,682 eligible patients who turned 65 years old between January 1, 2016 and December 31, 2018 with at least 1 year of follow-up, only 463 (6.9%) received AAA screening during the study period. The odds of receiving AAA screening within 1 year of index eligibility were 27% lower for Black patients compared to whites [OR = 0.73, 95% CI (0.58,0.93)]. Patients who visited a PCP or were diagnosed with hypertension had 75% and 41% greater odds of receiving screening, respectively [OR 1.75, 95% CI (1.36,2.25)] and [OR 1.41 95% CI (1.11,1.80)] compared with patients who did not. Among 4,580 men with 2 years of follow-up, AAA screening rate increased to 13.0%. Patients who visited a PCP had 64% greater odds of receiving screening within 2 years of index eligibility compared to those who did not [OR = 1.64, 95% CI (1.30,2.06)]. CONCLUSIONS: Screening for AAA per USPSTF guidelines is underutilized with evidence of a racial disparity. Although PCP visit is the most consistent predictor of screening, provider screening rates are low.

eConsults' Impact on Care Access and Wait Times in Rheumatology.

BACKGROUND/OBJECTIVE: A growing number of health systems have implemented eConsults to improve access to specialty advice, but few studies have described their use in rheumatology or impact on visit wait times. We evaluated the uptake of an eConsult program and its impact on wait times for in-person rheumatology visits. METHODS: In this quality improvement project, we analyzed electronic health record data from 4 intervention clinics and 4 comparison clinics, 12 months before and after implementation of an eConsult program. We compared median wait time for rheumatology appointments using a pre-post difference-in-differences analysis and quantile regression, adjusting for patient age, race, sex, clinic pair, and primary insurance payer. We also interviewed 11 primary care providers from the intervention clinics and conducted a rheumatology provider focus group (n = 4) to elucidate experiences with the program. RESULTS: Rheumatologists recommended management in primary care or referral to another specialty for 41% of eConsults, reducing initial demand for in-person visits. The median wait times dropped in the intervention and the comparison clinics (42 and 25 days, respectively). Intervention clinic median wait time dropped 17 days more than comparison clinics, and this was nonstatistically significant (p = 0.089). eConsults fit provider care tasks best for triage or initial workup for diagnosis, and less well when tests required interpretation, or when back and forth communication was needed to manage the patient's condition. CONCLUSIONS: Implementation of eConsults for rheumatology was associated with reduced wait times for rheumatology appointments and supported primary care providers in the triage and workup for a substantial portion of patients.

The association of healthcare disparities and patient-specific factors on clinical outcomes in peripheral artery disease.

BACKGROUND: PAD increases the risk of cardiovascular mortality and limb loss, and disparities in treatment and outcomes have been described. However, the association of patient-specific characteristics with variation in outcomes is less well known. METHODS: Patients with PAD from Duke University Health System (DUHS) between January 1, 2015 and March 31, 2016 were identified. PAD status was confirmed through ground truth adjudication and predictive modeling using diagnosis codes, procedure codes, and other administrative data. Symptom severity, lower extremity imaging, and ankle-brachial index (ABI) were manually abstracted from the electronic health record (EHR). Data was linked to Centers for Medicare and Medicaid Services data to provide longitudinal follow up. Primary outcome was major adverse vascular events (MAVE), a composite of all-cause mortality, myocardial infarction (MI), stroke, lower extremity revascularization and amputation. RESULTS: Of 1,768 patients with PAD, 31.6% were asymptomatic, 41.2% had intermittent claudication (IC), and 27.3% had chronic limb-threatening ischemia (CLTI). At 1 year, patients with CLTI had higher rates of MAVE compared with asymptomatic or IC patients. CLTI and Medicaid dual eligibility were independent predictors of mortality. CLTI and Black race were associated with amputation. CONCLUSIONS: Rates of MAVE were highest in patients with CLTI, but patients with IC or asymptomatic disease also had high rates of adverse events. Black and Medicaid dual-eligible patients were disproportionately present in the CLTI subgroup and were at higher risk of amputation and mortality, respectively. Future studies must focus on early identification of high-risk patient groups to improve outcomes in patients with PAD.

Sacubitril/Valsartan Initiation and Postdischarge Adherence Among Patients Hospitalized for Heart Failure.

BACKGROUND: We investigated associations between timing of sacubitril/valsartan initiation and postdischarge adherence among patients hospitalized for heart failure with reduced ejection fraction (HFrEF). Clinical trials support initiation of sacubitril/valsartan among patients hospitalized with HFrEF. The association between timing of initiation and postdischarge adherence is unknown. METHODS AND RESULTS: We analyzed patients hospitalized for HFrEF (EF of ≤40%) within the Get With The Guidelines Heart Failure registry linked with Medicare claims between October 2015 and September 2017 who were eligible for sacubitril/valsartan. Follow-up was through December 2018. Patients were grouped by timing of sacubitril/valsartan initiation. Sacubitril/valsartan adherence at 90 and 365 days after discharge was assessed by calculating proportion of days covered (PDC) using medication fills. Among 4666 patients, 108 (2.3%) were continued on sacubitril/valsartan (on sacubitril/valsartan at admission and discharge), 191 (4.1%) were initiated as inpatients, 130 (2.8%) were initiated at discharge, and 4237 (90.1%) were discharged without sacubitril/valsartan. Median (25th, 75th) proportion of days covered through 90 days among those continued, initiated as inpatients, and initiated at discharge was 0.9 (0.6-0.1), 0.3 (0.0-0.7), and 0.0 (0.0-0.7), respectively (P < .001). Patients discharged without sacubitril/valsartan had very low rates of any sacubitril/valsartan fills within 90 and 365 days of discharge (2.1% and 7.7% of surviving patients, respectively). CONCLUSIONS: In 2015-2017 US clinical practice, more than 90% of eligible patients hospitalized for HFrEF were discharged without sacubitril/valsartan. Patients initiated as inpatients had a higher postdischarge proportion of days covered than patients initiated at discharge. Patients discharged without sacubitril/valsartan were unlikely to receive it during follow-up. These findings highlight the importance of initiating sacubitril/valsartan during hospitalization to improve the quality of care.

Shared-patient physician networks and their impact on the uptake of genomic testing in breast cancer.

BACKGROUND: Factors influencing the adoption of genomic testing are poorly understood, which may lead to inequitable and suboptimal treatment in cancer patients. Oncotype DX (ODX) is one of the first and most widely used genomic assays to stratify risk in women with early-stage breast cancer (BC). Physician networks have emerged as a significant and modifiable driver of emerging medical technology adoption. OBJECTIVE: To investigate the association between physician network connections and the use of ODX testing. METHODS: A retrospective study of women diagnosed with BC using SEER-Medicare from 2008 to 2012 was used. Medical oncologists were "connected" if they shared two or more patients during the early-adoption period (2008-2009). Parallel physician- and patient-level analyses employed logistic mixed models to determine the impact of being "connected" to an early-adopting oncologist on ODX use in 2011-2012. RESULTS: 24,463 women met study criteria; 12,874 were diagnosed with BC in the early-adoption time period. 2129 medical oncologists treated these patients from 2008 to 2009. Medical oncologists had a median number of peer connections of 4 (IQR: 2-7). Peer connection to an early-adopting provider in 2008-2009 was associated with a 1.7-fold increase in providers' adopting ODX (95% CI: 1.1-2.6) and a 1.5-fold increase in their patients receiving ODX (95% CI: 1.1-2.0) in 2010-2012. CONCLUSIONS: Peer connectedness to an early-adopting physician predicts ODX adoption in both physician-level and patient-level analyses. Provider networks may provide a potent and modifiable means to modulate the diffusion of emerging medical technologies. Efforts to increase testing, where appropriate, may benefit from peer-to-peer-based connection strategies.

Use and 1-year outcomes with conventional and drug-coated balloon angioplasty in patients with lower extremity peripheral artery disease.

BACKGROUND: With the growing use of drug-coated balloons for the treatment of peripheral artery disease, information regarding the safety and effectiveness of drug-coated balloons in current practice is needed. We examined patient, physician, and procedural characteristics as well as cardiovascular and limb events in patients who underwent peripheral vascular intervention with drug-coated balloons. METHODS: This is a retrospective cohort analysis utilizing Medicare data for 100% of fee-for-service beneficiaries from 2015 to 2016 who had a claim for femoropopliteal intervention. The use of drug-coated balloons was identified via specific transitional pass-through codes. All-cause mortality, all-cause hospitalization, repeat femoropopliteal intervention, and major lower extremity amputation at 1 year were the clinical outcomes of interest. RESULTS: In total, 83,225 patients underwent femoropopliteal intervention, and drug-coated balloons were utilized in 29% of all procedures. Patients treated with drug-coated balloons had a lower cumulative incidence of all-cause hospitalization, all-cause mortality, and major lower extremity amputation, but were more likely to undergo repeat femoropopliteal intervention when compared with patients treated with conventional balloon angioplasty. After adjustment for measured confounders, patients treated with drug-coated balloons had lower rates of hospitalization (HR 0.91 (0.88, 0.93), P < .001), all-cause mortality (HR 0.89 [0.84, 0.94], P < .001), and major amputation (HR 0.93 [0.88, 0.99], P = .017). CONCLUSIONS: Patients who underwent femoropopliteal intervention with drug-coated balloons had lower observed rates of all-cause mortality, all-cause hospitalization, and major amputation at 1 year. Interestingly, there was not a reduction in rates of repeat revascularization, and further work is required to understand this finding. Nevertheless, the use of drug-coated balloons appears to be safe in this large study of contemporary patients in the United States.

Neck circumference and cardiovascular outcomes: Insights from the Jackson Heart Study.

BACKGROUND: Emerging data suggest that neck circumference (NC) is associated with cardiometabolic risk factors. Limited research is available regarding the association between NC and cardiovascular outcomes in African Americans. METHODS: Using data from the Jackson Heart Study, we included participants with recorded NC measurements at baseline (2000-2004). Baseline characteristics for the included population were summarized by tertiles of NC. We then calculated age- and sex-adjusted cumulative incidence of clinical cardiovascular outcomes and performed Cox proportional-hazards with stepwise models. RESULTS: Overall, 5,290 participants were categorized into tertiles of baseline NC defined as ≤37 cm (n = 2179), 38-40 cm (n = 1552), and >40 cm (n = 1559). After adjusting for age and sex, increasing NC was associated with increased risk of heart failure (HF) hospitalization (cumulative incidence = 13.4% [99% CI, 10.7-16.7] in the largest NC tertile vs 6.5% [99% CI, 4.7-8.8] in the smallest NC tertile), but not mortality, stroke, myocardial infarction, or coronary heart disease (all P ≥ .1). Following full risk adjustment, there was a nominal increase in the risk of HF hospitalization with increasing NC, but this was not statistically significant (hazard ratio per 1-cm increase, 1.04 [99% CI, 0.99-1.10], P = .06). CONCLUSIONS: In this large cohort of African American individuals, a larger NC was associated with increased risk for HF hospitalization following adjustment for age and sex, but this risk was not statistically significant after adjusting for other clinical variables. Although NC is not independently associated with increased risk for cardiovascular events, it may offer prognostic information particularly related to HF hospitalization.

Outcomes Associated With Extraction Versus Capping and Abandoning Pacing and Defibrillator Leads.

BACKGROUND: Lead management is an increasingly important aspect of care in patients with cardiac implantable electronic devices; however, relatively little is known about long-term outcomes after capping and abandoning leads. METHODS: Using the 5% Medicare sample, we identified patients with de novo cardiac implantable electronic device implantations between January 1, 2000, and December 31, 2013, and with a subsequent lead addition or extraction ≥12 months after the de novo implantation. Patients who underwent extraction for infection were excluded. Using multivariable Cox proportional hazards models, we compared cumulative incidence of all-cause mortality, device-related infection, device revision, and lead extraction at 1 and 5 years for the extraction versus the cap and abandon group. RESULTS: Among 6859 patients, 1113 (16.2%) underwent extraction, whereas 5746 (83.8%) underwent capping and abandonment. Extraction patients tended to be younger (median, 78 versus 79 years; P<0.0001), were less likely to be male (65% versus 68%; P=0.05), and had shorter lead dwell time (median, 3.0 versus 4.0 years; P<0.0001) and fewer comorbidities. Over a median follow-up of 2.4 years (25th, 75th percentiles, 1.0, 4.3 years), the overall 1-year and 5-year cumulative incidence of mortality was 13.5% (95% confidence interval [CI], 12.7-14.4) and 54.3% (95% CI, 52.8-55.8), respectively. Extraction was associated with a lower risk of device infection at 5 years relative to capping (adjusted hazard ratio, 0.78; 95% CI, 0.62-0.97; P=0.027). There was no association between extraction and mortality, lead revision, or lead extraction at 5 years. CONCLUSIONS: Elective lead extraction for noninfectious indications had similar long-term survival to that for capping and abandoning leads in a Medicare population. However, extraction was associated with lower risk of device infections at 5 years.

Association between physician characteristics and the use of 21-gene recurrence score genomic testing among Medicare beneficiaries with early-stage breast cancer, 2008-2011.

PURPOSE: We sought to determine whether physician-level characteristics were associated with 21-gene recurrence score (RS) genomic testing to evaluate recurrence risk and benefit of adjuvant chemotherapy in patients with estrogen receptor-positive, node-negative breast cancer. METHODS: Retrospective cohort study of a nationally representative sample of Medicare beneficiaries using Surveillance, Epidemiology, and End Results program-Medicare data linked with the American Medical Association physician master file. The main outcome was receipt of genomic testing within 1 year of diagnosis as a function of physician-level factors. RESULTS: A total of 24,463 patients met the study criteria; they received care from 3172 surgeons and 2475 medical oncologists. Of 4124 tests ordered, 70% were ordered by a medical oncologist and 16% by a surgeon. In multivariable regression models, multiple variables were associated with receipt of testing, including having a medical oncologist (odds ratio [OR] 2.77; 95% CI 2.00-3.82), a surgeon specializing in surgical oncology (OR 1.20; 95% CI 1.09-1.31), and a female medical oncologist (OR 1.10; 95% CI 1.02-1.20). Having a medical oncologist with 5 or more years in practice was associated with lower odds of testing (OR 0.83; 95% CI 0.76-0.92). Surgical procedures performed at academic centers were associated with higher odds of testing (OR 1.11; 95% CI 1.02-1.20). CONCLUSIONS: Although most RS testing was ordered by medical oncologists, physicians in other specialties ordered roughly one-third of the tests. Physician characteristics, including gender and time in practice, were associated with receiving testing, creating opportunities for targeting interventions to help patients receive optimal care.

Early impact of guideline publication on angiotensin-receptor neprilysin inhibitor use among patients hospitalized for heart failure.

BACKGROUND: On May 20, 2016, US professional organizations in cardiology published joint treatment guidelines recommending the use of angiotensin-receptor neprilysin inhibitor (ARNI) for eligible patients with heart failure with reduced ejection fraction (HFrEF). Using data from the Get With The Guidelines-Heart Failure registry, we evaluated the early impact of this update on temporal trends in ARNI prescription. METHODS: We analyzed patients with HFrEF who were eligible for ARNI prescription (EF ≤40%, no contraindications) and hospitalized from February 20, 2016, through August 19, 2016-allowing for 13weeks before and after guideline publication. We quantified trends in ARNI use associated with guidelines publication with an interrupted time-series design using logistic regression and accounting for correlations within hospitals using general estimating equation methods. RESULTS: Of 7,200 eligible patient hospitalizations, 51.9% were discharged in the period directly preceding publication of the guidelines, and 48.1% were discharged after. Odds ratios of ARNI prescription at discharge were significantly higher in the postguideline period compared with the preguideline period in adjusted models (adjusted odds ratio 1.29, 95% CI 1.06-1.57, P=.01). However, there was no significant interaction between observed and expected ARNI use after guideline publication (Pinteraction=.14). Results were consistent using a 6-month before and after time frame. CONCLUSIONS: The model suggested a small increase in ARNI use in HF patients being discharged from the hospital immediately after guideline release. However, the publication of national guidelines recommending ARNI use seemed to have little influence on the adoption of this evidence-based medication in the first 3 to 6months.

Intentional and unintentional medication non-adherence in African Americans: Insights from the Jackson Heart Study.

BACKGROUND: Non-adherence to medications is common and leads to suboptimal outcomes. Non-adherence can be intentional (e.g., deciding to skip dosages) or unintentional (e.g., forgetting), yet few studies have distinguished these reasons. An improved understanding of the reasons for non-adherence could inform the development of effective interventions. METHODS AND RESULTS: We analyzed data from African Americans in the Jackson Heart Study who were prescribed medications for one or more chronic conditions. Participants were grouped by patient-reported adherence with non-adherence categorized as being intentional, unintentional or both. We used modified Poisson regression models to examine the factors associated with types of non-adherence. Of 2933 participants taking medication, 2138 (72.9%) reported non-adherence with 754 (35.3%) reporting only unintentional non-adherence, 263 (12.3%) only intentional non-adherence, and 1121 (52.4%) both. Factors independently associated with intentional non-adherence included female sex and depressive symptoms while factors associated with unintentional non-adherence included younger age and separated relationship status. Unintentional and intentional non-adherence was more common among participants taking anti-arrhythmic and anti-asthmatic medications, respectively. Higher levels of global perceived stress was associated with both types of non-adherence. The adjusted models for intentional and unintentional non-adherence had c-statistics of 0.65 and 0.66, respectively, indicating modest discrimination. CONCLUSION: Specific patient factors and individual medication classes were associated with distinct patterns of intentional and unintentional non-adherence, yet the overall modest discrimination of the models suggests contributions from other unmeasured factors. These findings provide a construct for understanding reasons for non-adherence and provide rationale to assess whether personalized interventions can improve adherence.

Overtreatment and Deintensification of Diabetic Therapy among Medicare Beneficiaries.

BACKGROUND: Deintensification of diabetic therapy is often clinically appropriate for older adults, because the benefit of aggressive diabetes treatment declines with age, while the risks increase. OBJECTIVE: We examined rates of overtreatment and deintensification of therapy for older adults with diabetes, and whether these rates differed by medical, demographic, and socioeconomic characteristics. DESIGN, SUBJECTS, AND MAIN MEASURES: We analyzed Medicare claims data from 10 states, linked to outpatient laboratory values to identify patients potentially overtreated for diabetes (HbA1c < 6.5% with fills for any diabetes medications beyond metformin, 1/1/2011-6/30/2011). We examined characteristics associated with deintensification for potentially overtreated diabetic patients. We used multinomial logistic regression to examine whether patient characteristics associated with overtreatment of diabetes differed from those associated with undertreatment (i.e. HbA1c > 9.0%). KEY RESULTS: Of 78,792 Medicare recipients with diabetes, 8560 (10.9%) were potentially overtreated. Overtreatment of diabetes was more common among those who were over 75 years of age and enrolled in Medicaid (p < 0.001), and was less common among Hispanics (p = 0.009). Therapy was deintensified for 14% of overtreated diabetics. Appropriate deintensification of diabetic therapy was more common for patients with six or more chronic conditions, more outpatient visits, or living in urban areas; deintensification was less common for those over age 75. Only 6.9% of Medicare recipients with diabetes were potentially undertreated. Variables associated with overtreatment of diabetes differed from those associated with undertreatment. CONCLUSIONS: Medicare recipients are more frequently overtreated than undertreated for diabetes. Medicare recipients who are overtreated for diabetes rarely have their regimens deintensified.