Publications

INTRODUCTION: The COVID-19 pandemic has had a significant impact on healthcare delivery. Although others have documented the impact on new cancer diagnoses, trends in new starts for oncology drugs are less clear. We examined changes in new users of oral oncology medications in the US following COVID-19 stay-at-home orders in 2020 compared to prior years.

METHODS: We examined prescription data for members enrolled with a national pharmacy benefits manager in the US from January 1-October 31 of 2018, 2019, and/or 2020. This is a retrospective, observational study comparing new users per 100,000 members per month for all oral oncology drugs, and separately for breast, lung, and prostate cancer, leukemia, and melanoma oral drugs. We performed a difference-in-differences analysis for change in new users from pre-period (prior to pandemic-induced disruption, January-March), to post-period (following pandemic-induced disruption, April-October), between 2020 and 2019, and 2020 and 2018.

RESULTS: New oral oncology drug users per 100,000 members per month declined by an additional 11.3% in the 2020 post-period compared to 2019 (p = 0.048). New oral breast cancer drug starts declined by an additional 14.0% in the 2020 post-period compared to 2019 (p = 0.040). Similar but non-significant trends were found between 2020 and 2018. No significant differences were found between post-period monthly new starts of leukemia, melanoma, lung or prostate cancer disease-specific oral medications.

CONCLUSIONS: Long-term implications of delays in cancer treatment initiation are unclear, although there is concern that patient outcomes may be negatively impacted.

OBJECTIVE: The aim of the study is to assess the impact of ≥15% body mass index (BMI) reduction on employees' health expenditures.

METHODS: We retrospectively analyzed health risk assessment surveys combined with insurance claims from January 2014 to December 2019. We compared costs of employees with baseline BMI > 30 who reported ≥15% BMI reduction in subsequent health risk assessment reports with employees who lost ≤5% BMI within the same period, matching the two cohorts on demographics and costs.

RESULTS: The study cohort of 197 lost an average of 23% of their BMI from baseline. The average age was 44 years with majority females (approximately 80%). Group health insurance payments were similar at baseline; at year 1, the study cohort had a 33% payment reduction compared with 10% reduction in the control group.

CONCLUSIONS: A ≥15% BMI reduction was associated with a substantial medical cost savings.

Background PCSK9is (proprotein convertase subtilisin/kexin type 9 inhibitors) are well tolerated, potently lower cholesterol, and decrease cardiovascular events when added to statins. However, statin adherence may decrease after PCSK9i initiation and alter clinical outcomes. We evaluate the association of PCSK9i initiation on statin discontinuation and adherence. Methods and Results In this retrospective pre-post difference-in-difference analysis, new PCSK9i claims were propensity matched with statin-alone users (April 2017-September 2019). The primary outcomes were statin adherence (proportion of days covered) and statin discontinuation (absence of statin coverage for at least 60 days) 12 months following PCSK9i initiation. Secondary outcomes included low-density lipoprotein cholesterol levels after 1 year. A total of 220 538 statin users and 700 PCSK9i users were identified, from which 178 on PCSK9i were included and matched to 712 on statins alone. At 12 months, mean statin proportion of days covered decreased from 67% to 48% in the PCSK9i group but increased from 68% to 86% in the statin-alone groups (P<0.0001). Statin discontinuation rates increased from 11% to 39% in the PCSK9i group and from 7% to 9% in the statin-alone group (P=0.0041). Patients with low-density lipoprotein cholesterol <70 mg/dL increased from 5% to 68% with PCSK9i but increased from 16% to 24% with statins alone (P<0.0001). Changes in hospitalization rates were similar between both groups during the follow-up period. Conclusions PCSK9i initiation was associated with decreased low-density lipoprotein cholesterol, higher statin discontinuation, and reduced statin adherence.

PURPOSE: Costs of hospitalization due to severe adverse drug reactions (ADRs) were previously estimated within the Veterans Health Administration (VHA), but additional analyses are needed to infer potential interventions to mitigate these negative outcomes. The objective of this study was to compare specific adverse reaction-related hospitalization costs between medications with similar indications.

METHODS: Mean hospitalization costs associated with the same ADR symptom were compared for different drugs with similar indications using adjusted generalized linear models with a Bonferroni correction for multiple comparisons as well as a gamma distribution.

RESULTS: Overall, hospitalization costs between medications with similar indications were not significantly different for specific adverse reactions. However, gastrointestinal hemorrhage-associated costs were higher for warfarin versus nonsteroidal anti-inflammatory drugs (model estimate of mean cost, $18,114 [range of lower and upper model estimates, $12,522-$26,202] vs $14,255 [estimate range, $9,710-$20,929]). Similarly, the estimated mean hospitalization cost associated with angioedema was higher for losartan versus lisinopril or lisinopril/hydrochlorothiazide: $14,591 (range, $9467-$22,488) versus $8,935 (range, $6,301-$12,669) and $8,022 (range, $5,424-$11,865), respectively.

CONCLUSION: Although we found few differences in the cost of hospitalization when comparing drugs with similar indications and the same adverse reaction, there were specific drug-ADR pairs that merit attention and consideration of interventions to improve safe and appropriate medication use. Evaluation of the effect of those interventions on the incidence of ADRs is an area for future study.

BACKGROUND: Outcomes-based agreements (OBA) are performance-based risk-sharing agreements between manufacturers and payers which provide the opportunity for collection and evaluation of real-world outcomes to supplement clinical trials.

OBJECTIVES: To describe an OBA comparing ticagrelor to clopidogrel in patients admitted with acute coronary syndrome (ACS) and proportion of recurrent myocardial infarction (MI) in a real-world setting.

METHODS: Commercial (CM) and Medicare (MC) insurance patients of a large regional health plan, who presented with ACS and were prescribed either ticagrelor or clopidogrel were prospectively analyzed. The cohort consisted of adults (18-85 years) discharged between January 1, 2019, and December 31, 2020, who were adherent to the study medications, within the confines of the OBA. The primary outcome of interest was the proportion of recurrent MI hospitalizations within one year of discharge.

RESULTS: There were 500 patients who met inclusion criteria in the ticagrelor cohort and 648 in the clopidogrel cohort. The mean age of patients in the ticagrelor cohort was 61.5 ± 10.5 years old and 66.5 ± 10.2 years in the clopidogrel cohort. The proportion of patients with type 2 diabetes, hypertension, or a history of congestive heart failure at baseline in the ticagrelor cohort was 31%, 85%, 14% respectively, and 43%, 90%, and 32% respectively in the clopidogrel cohort. The overall proportion of hospitalization for recurrent MI was 1.00% in the ticagrelor and 3.13% in the clopidogrel cohorts. In the follow-up propensity-matched analysis, although recurrent MI hospitalization was higher in the clopidogrel cohort (1.69% vs 1.21%) it was not statistically significant (p-value 0.5242).

CONCLUSION: Patients presenting with ACS and treated with ticagrelor had a lower rate of hospitalization for recurrent MI compared to patients treated with clopidogrel cohort within the confines of an OBA in a real-world setting.

As genomic medicine becomes increasingly complex, pharmacists need to work collaboratively with other healthcare professionals to provide genomics-based care. The core pharmacist competencies in genomics were recently updated and mapped to the entrustable professional activities (EPAs). The new competency that is mapped to the "Interprofessional Team Member" EPA domain emphasizes the role of pharmacists as the pharmacogenomics experts in an interprofessional healthcare team. Interprofessional education (IPE) activities involving student pharmacists and students from other healthcare disciplines are crucial to prepare student pharmacists for a team-based approach to patient-centered care. This commentary discusses the pharmacogenomics-focused IPE activities implemented by 3 programs, the challenges faced, and the lessons learned. It also discusses strategies to develop pharmacogenomics-focused IPE activities based on existing resources. Developing pharmacogenomics-focused IPE activities will help prepare pharmacy graduates with the knowledge, skills, and attitudes to lead collaborative, interprofessional teams in the provision of pharmacogenomics-based care, consistent with the standards described in the genomics competencies for pharmacists.

Infrastructure system in the U.S. have been shown to be linked to social and health inequities. We calculated driving distance to the closest health care facility for a representative sample of the U.S. population using ArcGIS Network Analyst and a national transportation dataset, and identified areas where Black residents have a longer driving distance to the closest facility than White residents. Our data demonstrated that racial disparities in access to health care facilities presented large geographic variation. Counties with significant racial disparities were concentrated in the Southeast and did not correspond to counties with a greater proportion of the overall population >5 miles to the closest facility, which were concentrated in the Midwest. This geographic variation demonstrates the need to adopt a spatially explicit data driven approach in the design of equitable health care facility establishment that address the specific limitations of the local infrastructure.