Publications

BACKGROUND & AIMS: Ustekinumab is a monoclonal antibody against interleukin 12 and interleukin 23 that has been approved by the Food and Drug Administration for treatment of Crohn's disease (CD). We sought to identify the ideal position for ustekinumab in treatment algorithms for CD.

METHODS: We constructed a Markov model to identify an optimal treatment sequence for CD that included ustekinumab for 1 year or more. The base case was a 35-year old male with moderate to severe CD who had not previously received biologic or immunomodulator therapy. The standard of care treatment algorithm was defined as initial therapy with infliximab and azathioprine, followed by adalimumab and azathioprine, vedolizumab, and lastly surgical resection. The model assessed positions for ustekinumab before standard of care, ustekinumab after infliximab and azathioprine but before the remaining treatments, after infliximab, azathioprine, and adalimumab but before vedolizumab and surgery, or after the other biologics but before surgery. We derived transition probabilities and quality adjusted life years (QALYs) from relevant trials, observational studies, and time trade-off analyses. Primary analyses consisted of first order Monte Carlo simulation of 100 trials of cohorts of 100,000 individuals.

RESULTS: Ustekinumab as first-line therapy yielded the greatest QALYs (incremental effectiveness, 0.016-0.020 QALYs), resulting in 10% more patients in remission or response, and 2% fewer surgeries at 1 year, compared with other algorithms. The model was not sensitive to 25% variation in transition probabilities.

CONCLUSIONS: In a simulation based on a 35-year old male patient with moderate to severe CD, we found that ustekinumab as the first-line biologic therapy yields greater QALYs at the end of 1 year than compared with use later in the CD treatment algorithm.

BACKGROUND: Medicare Part B pharmaceutical spending has increased rapidly, more than doubling in 2006-2017. Yet, it is unclear whether this increase was driven by increased utilization or increased cost per claim. OBJECTIVE: To evaluate the relative impact of changes in drug utilization and cost per claim on changes in Medicare Part B pharmaceutical spending in 2008-2016 overall, by drug type (specialty and nonspecialty) and therapeutic category. METHODS: In this retrospective descriptive study, we extracted all claims in 2008-2016 for separately payable Part B drugs from a 5% random sample of Medicare beneficiaries. Our study included 3 outcomes calculated annually for all included drugs: (1) spending, defined as the sum of total payments; (2) utilization, defined as total number of claims; and (3) cost per claim, defined as spending divided by the number of claims. Estimates of spending and utilization were expressed per beneficiary-year. Spending and cost per claim were adjusted for inflation. For each outcome, we calculated relative changes in 2008-2016. We repeated analyses stratifying by drug type (specialty and nonspecialty) and therapeutic class. RESULTS: Pharmaceutical spending in Medicare Part B increased by 34% from 2008-2016, driven by a 53% increase in the cost per claim. Utilization decreased by 12%. Spending on specialty drugs increased by 56%, driven by a 48% increase in the cost per claim and a 6% utilization increase. Spending on nonspecialty drugs decreased by 32% driven by an 18% reduction in the cost per claim and a 17% reduction in utilization. Spending on ophthalmic preparations increased by 281%, driven by a 238% utilization increase and a 13% increase in the cost per claim. Spending on antiarthritic and immunologic agents increased by 159%, driven by a 117% increase in the cost per claim and a 19% utilization increase. CONCLUSIONS: Medicare Part B pharmaceutical spending grew in recent years, despite decreased utilization, driven by an overall increase in the cost per claim. This was a product of rising drug prices and increased utilization of more expensive specialty drugs. These findings support the development of policies that aim to spur competition and control price growth of provider-administered drugs. DISCLOSURES: The authors acknowledge funding from the Myers Family Foundation. Hernandez was funded by the National Heart, Lung and Blood Institute (grant number K01HL142847). Shrank is an employee of Humana. Good is an employee of the UPMC Health Plan Insurance Services Division. There are no other potential conflicts of interest to disclose.

BACKGROUND: Because of improved clinical outcomes, recent American Diabetes Association guidelines recommend the use of newer antidiabetic agents-glucagon-like peptide-1 receptor agonists (GLP-1RA) and sodium-glucose cotransporter 2 inhibitors (SGLT2i)-by those with cardiovascular disease. It is unclear, however, how switching to these newer agents affects health care utilization and costs. OBJECTIVE: To compare health care utilization and costs between users of dipeptidyl peptidase-4 inhibitors (DPP-4i) who switch to GLP-1RA or SGLT2i and nonswitchers. METHODS: We used claims data from a large pharmacy benefit manager. Patients included were commercially insured adults with type 2 diabetes and a prescription claim for DPP-4i in 2016 or 2017. Using propensity score methods, we matched patients who switched to SGLT2i or GLP-1RA with those who remained on DPP-4i. Among matched samples, we conducted multivariable negative binomial regression to examine differences in the incidence of inpatient and emergency room (ER) visits and generalized linear regression to examine differences in health care costs. RESULTS: Among 47,953 patients who used DPP-4i in 2016 and 2017, 507 switched to SGLT2i and 808 switched to GLP-1RA. Propensity score matching of 1:6 resulted in 3,042 nonswitchers/507 switchers for the SGLT2i cohort and 4,848 nonswitchers/808 switchers for the GLP-1RA cohort. Switchers to SGLT2i experienced a 39% reduction (incidence rate ratio [IRR] = 0.61, 95% CI = 0.38-0.96), and GLP-1RA switchers experienced a 29% reduction (IRR = 0.71, 95% CI = 0.52-0.97) in inpatient hospitalizations. ER visit rates did not differ significantly between switchers and nonswitchers. Switchers to SGLT2i did not have statistically significant differences in medical or pharmacy costs compared with DPP-4i users, while switchers to GLP-1RA had significantly higher total pharmacy costs (adjusted difference of $2,453.10, 95% CI = $1,837.20-$3,069.00). CONCLUSIONS: Switching from DPP-4i to GLP-1RA or SGLT2i was associated with fewer hospitalizations; however, higher pharmacy costs may outweigh savings from reduced hospitalizations, especially for GLP-1RAs. As newer diabetes guidelines steer specific populations to these drug classes, it is important to optimize drug pricing to realize their true value. DISCLOSURES: No outside funding supported this study. Neilson, Good, Swart, and Huang are employees of UPMC Center for Value-Based Pharmacy Initiatives and High-Value Care. Parekh reports employment at UPMC until July 2019. Munshi and Henderson are employed by Express Scripts. Newman has no disclosures to report.

BACKGROUND: Although medication therapy management (MTM) has specific eligibility criteria and is mandated for specific Medicare Part D enrollees, some health plans have expanded MTM eligibility beyond the minimum criteria to include other Medicare Part D enrollees, Medicaid, and commercial health plan patients. Differences exist in the mode of delivery, location of services, type of personnel involved in managing the service, and the subsequent outcomes. The type and intensity of MTM services delivered have evolved with time to more streamlined and robust interventions, necessitating ongoing evaluation of the effect on clinical and economic outcomes. OBJECTIVE: To assess the effect of changes to an existing MTM program on cost of care, utilization, and medication adherence. METHODS: UPMC Health Plan made changes to an existing MTM program by expanding eligibility (customized by the type of health plan), intervention types, pharmacist involvement, and patient followup contacts. After matching our intervention cohort (identified January 2017-June 2018) with the pre-2016 MTM historical controls (patients identified January 2014-June 2015 who would have been eligible if we used the intervention cohort eligibility criteria), we estimated that the effect of the program changes with a difference-in-difference model (preintervention [2014-2016] and postintervention [2017-2019]). Outcomes of interest included cost (total cost of care including medical, pharmacy, and unplanned care [i.e., unscheduled health care use such as emergency department visits] in 2017 U.S. dollars); utilization; medication adherence (proportion of days covered); and return on investment (ROI). Target population included continuously enrolled patients aged ≥ 21 years in the commercial, Medicare, and Medicaid health plans. RESULTS: Total propensity score-matched members was 10,747, 55% of which were in the historic control group. The average (SD) ages after matching the groups were similar (historical control group: 57.08 years [14.23], intervention group: 56.79 years [14.21]) and the majority was female (57%). Comorbidities identified most for patients included hypertension (77%), dyslipidemia (70%), and diabetes (52%). Forty-one percent were in the commercial, 37% in the Medicaid, and 23% in the Medicare health plans. Proportion of care activities undertaken in the intervention period compared with the control period were significantly different: "sent letter to physician" (67% vs. 87%), "sent letter to member" (15% vs. 0%), "pharmacist phone call to physician" (15% vs. 0.1%), and "pharmacist phone call to member" (13% vs. 7%). There were statistically significant reductions in unplanned care across all health plans especially in the Medicare population, in total cost of care, and increases in medication adherence in 4 therapeutic classes: anticoagulants (OR = 1.25, P = 0.005), cardiac medications (OR = 1.20, P < 0.001), statins (OR = 1.21, P < 0.001), and antidepressants (OR = 1.15, P < 0.001). There was a positive ROI of $18.50 per dollar spent, which equated to a cumulative net savings of $11 million over 24 months. CONCLUSIONS: In a large health plan, expanding MTM eligibility, intensifying patient follow-up contact and pharmacist involvement, and improving provider awareness had favorable clinical and economic benefits. DISCLOSURES: There was no funding for this project except employees' time. All authors are employees of UPMC and have no conflicts of interest to report.

BACKGROUND: List prices of tumor necrosis factor (TNF) inhibitors drastically increased during the last decade, but previous research has shown that half of these increases were offset by rising manufacturer discounts. It remains unclear to what extent manufacturers' discounts have offset increases in list prices of each self-administered injectable TNF inhibitor. Evaluating trends in net prices and discounts at the product level will be paramount in understanding the role of competition in the biologic market. OBJECTIVES: To (a) describe product-level changes in net prices of each self-administered injectable TNF inhibitor available in 2007-2019 and (b) quantify to what extent manufacturer discounts have offset increases in list prices. METHODS: We obtained 2007-2019 pricing data for etanercept, adalimumab, certolizumab, and golimumab from the investment firm SSR Health, which uses company-reported sales to estimate net prices and discounts for brand products manufactured by publicly traded companies. For each drug and year, we calculated annual costs of treatment for patients with rheumatoid arthritis based on list and net prices and discounts in Medicaid and other payers. RESULTS: From 2007-2019, list prices of etanercept and adalimumab increased by 293% and 295%, respectively; however, discounts offset 47% and 45% of these increases, leading to net price increases of 171% and 203%. List prices of golimumab and certolizumab increased by 183% and 182%, respectively, but with discounts offsetting 58% and 59% of these increases, net prices increased by 103% and 109%. Net prices of golimumab started to decrease after 2016, while net prices of adalimumab and certolizumab experienced their first drop in 2019. Across the study period, discounts in Medicaid and in other payers increased, respectively, from 21% to 85% and 6% to 32% for etanercept; from 26% to 88% and 19% to 35% for adalimumab; from 28% to 63% and 22% to 46% for golimumab; and from 29% to 83% and 27% to 47% for certolizumab. CONCLUSIONS: Despite growing manufacturer discounts, net prices of self-administered injectable TNF inhibitors still increased at a mean annual rate of 9.6% in 2007-2019. This led to net prices tripling for adalimumab and more than doubling for etanercept, golimumab, and certolizumab. DISCLOSURES: This study was funded by the Myers Family Foundation. Hernandez is funded by the National Heart, Lung and Blood Institute (grant number K01HL142847). Funding sources had no role in design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication. Hernandez has served on Pfizer's scientific advisory board. The other authors have nothing to disclose.

IMPORTANCE: Atrial fibrillation is a common cardiac rhythm disturbance causing substantial morbidity and mortality that disproportionately affects racial/ethnic minority groups. Anticoagulation reduces stroke risk in atrial fibrillation, yet studies show it is underprescribed in racial/ethnic minority patients.

OBJECTIVE: To compare initiation of anticoagulant therapy by race/ethnicity for patients in the Veterans Health Administration (VA) system with atrial fibrillation.

DESIGN, SETTING, AND PARTICIPANTS: This retrospective cohort study included 111 666 patients within the VA system with incident atrial fibrillation between January 1, 2014, and December 31, 2018. Data were analyzed between December 1, 2019, and March 31, 2020.

EXPOSURES: Any anticoagulation was defined as receipt of warfarin or direct-acting oral anticoagulants, apixaban, dabigatran, edoxaban, or rivaroxaban.

MAIN OUTCOMES AND MEASURES: Initiation of any anticoagulation (or direct-acting oral anticoagulant therapy in those who initiated any anticoagulation) was examined within 90 days of an index atrial fibrillation diagnosis.

RESULTS: Our final cohort comprised 111 666 patients (109 386 men [98.0%] and 95 493 White patients [85.5%]; mean [SD] age, 72.9 [10.4] years). A total of 69 590 patients (62.3%) initiated any anticoagulant therapy, varying 10.5 percentage points by race/ethnicity (P < .001); initiation was lowest in Asian (52.2% [n = 676]) and Black (60.3% [n = 6177]) patients and highest in White patients (62.7% [n = 59 881]). Among anticoagulant initiators, 45 381 (65.2%) used direct-acting oral anticoagulants, varying 7.2 percentage points by race/ethnicity (P < .001); initiation was lowest in Hispanic (58.3% [n = 1470]), American Indian/Alaska Native (59.8% [n = 201]), and Black (60.9% [n = 3763]) patients and highest in White patients (66.0% [n = 39 502). Compared with White patients, the odds of initiating any anticoagulant therapy were significantly lower for Asian (adjusted odds ratio [aOR], 0.82; 95% CI, 0.72-0.94) and Black (aOR, 0.90; 95% CI 0.85-0.95) patients. Among initiators, the adjusted odds of direct-acting oral anticoagulant initiation were significantly lower for Hispanic (aOR, 0.79; 95% CI, 0.70-0.89), American Indian/Alaska Native (aOR, 0.75; 95% CI, 0.57-0.99), and Black (aOR, 0.74; 95% CI 0.69-0.80) patients.

CONCLUSIONS AND RELEVANCE: This cohort study found that in patients with incident atrial fibrillation managed in the VA system, race/ethnicity was independently associated with initiating any anticoagulant therapy and direct-acting oral anticoagulant use among anticoagulant initiators. Understanding the reasons for these treatment disparities is essential to improving equitable atrial fibrillation management and outcomes among racial/ethnic minority patients treated in the VA system.

INTRODUCTION: Although opioid use disorder (OUD) is common in patients with cirrhosis, it is unclear how medication treatment for OUD (MOUD) is used in this population. We aimed to assess the factors associated with MOUD and mortality in a cohort of Veterans with cirrhosis and OUD.

METHODS: Within the Veterans Health Administration Corporate Data Warehouse, we developed a cohort of Veterans with cirrhosis and active OUD, using 2 outpatient or 1 inpatient International Classification of Diseases, ninth revision codes from 2011 to 2015 to define each condition. We assessed MOUD initiation with methadone or buprenorphine over the 180 days following the first OUD International Classification of Diseases, ninth revision code in the study period. We fit multivariable regression models to assess the association of sociodemographic and clinical factors with receiving MOUD and the associations between MOUD and subsequent clinical outcomes, including new hepatic decompensation and mortality.

RESULTS: Among 5,600 Veterans meeting criteria for active OUD and cirrhosis, 722 (13%) were prescribed MOUD over 180 days of follow-up. In multivariable modeling, MOUD was significantly, positively associated with age (adjusted odds ratio [AOR] per year: 1.04, 95% confidence interval (CI): 1.01-1.07), hepatitis C virus (AOR = 2.15, 95% CI = 1.37-3.35), and other substance use disorders (AOR = 1.47, 95% CI = 1.05-2.04) negatively associated with alcohol use disorder (AOR = 0.70, 95% CI = 0.52-0.95), opioid prescription (AOR = 0.51, 95% CI = 0.38-0.70), and schizophrenia (AOR = 0.59, 95% CI = 0.37-0.95). MOUD was not significantly associated with mortality (adjusted hazards ratio = 1.20, 95% CI = 0.95-1.52) or new hepatic decompensation (OR = 0.57, CI = 0.30-1.09).

DISCUSSION: Few Veterans with active OUD and cirrhosis received MOUD, and those with alcohol use disorder, schizophrenia, and previous prescriptions for opioids were least likely to receive these effective therapies.

This cohort study used Medicaid data to examine trends in the use of and spending for hemophilia pharmaceuticals from 2005 to 2020.