Comparison of real-world clinical and economic outcomes in patients receiving oral anticoagulants: A retrospective claims analysis.

BACKGROUND: Direct oral anticoagulants (DOACs) have become widely used for the prevention of stroke in nonvalvular atrial fibrillation (AF) and for the treatment of venous thromboembolism (VTE). Warfarin, the standard of care prior to DOACs, requires monitoring and dose adjustment to ensure patients remain appropriately anticoagulated. DOACs do not require monitoring but are significantly more expensive. We sought to examine real-world effectiveness and costs of DOACs and warfarin in patients with AF and VTE. OBJECTIVE: To examine clinical and economic outcomes. The clinical objectives were to determine the bleeding and thrombotic event rates associated with DOACs vs warfarin. The economic objectives were to determine the cost associated with these events, as well as the all-cause medical and pharmacy costs associated with DOACs vs warfarin. METHODS: This analysis was an observational, propensity-matched comparison of retrospective medical and pharmacy claims data for members enrolled in an integrated health plan between October 1, 2015, and September 30, 2020. Members who were older than 18 years of age with at least 1 30-day supply of warfarin or a DOAC filled within 30 days of a new diagnosis of VTE or nonvalvular AF were eligible for the analysis. Cox hazard ratios were used to compare differences in clinical outcomes, where paired t-tests were used to evaluate economic outcomes. RESULTS: After matching, there were 893 patients in each group. Among matched members, warfarin was associated with increased risk of nonmajor bleeds relative to apixaban (hazard ratio [HR] = 1.526; P = 0.0048) and increased risk of pulmonary embolism relative to both DOACs (apixaban: HR = 1.941 [P = 0.0328]; rivaroxaban: HR = 1.833 [P = 0.0489]). No statistically significant difference was observed in hospitalizations or in length of stay between warfarin and either DOAC. The difference-in-difference (DID) in total costs of care per member per month for apixaban and rivaroxaban relative to warfarin were $801.64 (P = 0.0178) and $534.23 (P = 0.0998) more, respectively. DID in VTE-related cost for apixaban was $177.09 less, relative to warfarin (P = 0.0098). DID in all-cause pharmacy costs for apixaban and rivaroxaban relative to warfarin were $342.47 (P < 0.0001) and $386.42 (P < 0.001) more, respectively. CONCLUSIONS: Warfarin use was associated with a significant decrease in total cost of care despite a significant increase in VTE-related costs vs apixaban. Warfarin was also associated with a significant increase in other nonmajor bleeds relative to apixaban, as well as a significant increase in pulmonary embolism relative to both DOACs. Warfarin was associated with a significant reduction in all-cause pharmacy cost compared with either DOAC. DISCLOSURES: The authors of this study have nothing to disclose.

A pilot study of oral treprostinil pharmacogenomics and treatment persistence in patients with pulmonary arterial hypertension.

BACKGROUND AND AIMS: Treprostinil is a prostacyclin analog used to treat pulmonary arterial hypertension. Dosing is empiric and based on tolerability. Adverse effects are common and can affect treatment persistence. Pharmacogenomic variants that may affect treprostinil metabolism and transport have not been well-characterized. We aimed to investigate the pharmacogenomic sources of variability in treatment persistence and dosing. METHODS: Patients were prospectively recruited from an IRB approved biobank registry at a single pulmonary hypertension center. A cohort of patients who received oral treprostinil were screened for participation. Pharmacogenomic analysis was for variants in CYP2C8, CYP2C9, and ABCC4. A retrospective review was conducted for demographics, clinical status, dosing, and response. Fisher's exact test was used for categorical data and Kruskal-Wallis test or Wilcoxon rank sum were used for continuous data. RESULTS: A total of 15 patients received oral treprostinil and were consented. Their median age was 53 years, 73% were female, and 93% were White. The median total daily dose was 22.5 mg (13.5, 41) at last clinical observation. 40% of patients discontinued treatment with a majority due to adverse effects. Approximately 27% of patients had a loss-of-function variant in CYP2C8 (*1/*3 or *1/*4), whereas 47% of patients had a loss-of-function variant in CYP2C9 (*1/*2, *1/*3, or *2/*2). Minor allele frequencies for ABCC4 (rs1751034 and rs3742106) were 0.17 and 0.43, respectively. Survival analysis showed that increased CYP2C9 activity score was associated with decreased risk for treatment discontinuation [hazard ratio (HR): 0.13; 95% confidence interval (CI): 0.02, 0.91; p = 0.04]. Genetic variants were not significantly associated with dosing. CONCLUSION: Genetic variants responsible for the metabolism and transport of oral treprostinil were common. Increased CYP2C9 activity score was associated with decreased risk for treatment discontinuation. However, dosing was not associated with genetic variants in metabolizing enzymes for treprostinil. Our findings suggest significant variability in treatment persistence to oral treprostinil, with pharmacogenomics being a potentially important contributor.The reviews of this paper are available via the supplemental material section.

Influence of Viscosity and Consumer Use on Accuracy of Oral Medication Dosing Devices.

Background: In response to numerous reports of overdoses with over-the-counter (OTC) liquid medications, the Food and Drug Administration has recommended that all OTC liquid drug products contain a measuring device but provided no recommendation on the type of device to be included. Objective: To evaluate the accuracy of liquid medication dosing devices (cup, dropper, syringe) in dispensing medications of varying viscosity in the laboratory and clinical settings. Methods: This experimental study evaluated dosing device accuracy. A pharmaceutics laboratory was used to evaluate accuracy under ideal conditions and subjects ≥18 years of age were recruited from community pharmacies to evaluate accuracy when used by consumers. Results: In the laboratory setting, the syringe was the most accurate for the more viscous formulations (cherry and grape suspension; 1% error, 1.2% error, respectively), and the dropper was the most accurate for the less viscous formulation (solution; 0.8% error). A volunteer sample of 320 participants was enrolled from the clinical setting. In the clinical setting, the syringe was most accurate, followed by the cup and then the dropper for all formulations (mean error, 2%, 14%, 33%, respectively). The cup was the most likely to overdose (mean, 5.7 mL), while the dropper was most likely to underdose (mean, 3.3 mL). Conclusions: The results of this study suggest that medication viscosity, consumer use, and dosing device contribute to dosing accuracy. The syringe appears to be the most accurate dosing device, accounting for differences in medication viscosity and the impact of consumer use.