Comparative Effectiveness and Economic Evaluations of Anticoagulation Strategies in Patients with Left Ventricular Assist Devices

Abstract

Left ventricular assist devices (LVADs) are surgically implanted devices that support the circulation system in patients with advanced heart failure. Patients with LVADs often require bridging anticoagulation in the chronic therapy phase following hospital discharge to prevent life-threatening thromboembolic complications. Given a lack of clinical trial data and clinical management guidelines, bridging practice varies widely between different LVAD centers. This dissertation evaluated the comparative effectiveness and cost-effectiveness of outpatient management with low-molecular-weight heparin (LMWH) vs hospitalization with unfractionated heparin (UFH) for LVAD patients. We also quantified the potential value of a new trial assessing the relative safety and efficacy of these alternative bridging anticoagulation strategies and determined the optimal design of such a trial.

To assess the comparative effectiveness of LMWH vs UFH bridging, we conducted a retrospective cohort study of adults with LVAD implantation between January 2014 and December 2018 from two academic medical centers. Data were collected from 269 patients and 1364 bridging episodes where either UFH or LMWH was administered. Records were reviewed for 30 days after bridging UFH or LMWH was discontinued, assessing for bleeding and/or thromboembolic events. Multivariable logistic regression analysis adjusted for site- and patient- level clustering along with LVAD type and the HAS-BLED score for bleed risk. We found that the rate of major bleeding or thromboembolism was non-statistically significantly lower for patients receiving LMWH as compared to UFH.

We then projected health and economic outcomes for LMWH vs UFH bridging strategies using a decision analytic model parameterized using data on rates of bleeding and thrombotic events and deaths from our retrospective cohort study of LVAD patients and the published literature. The primary outcome was the incremental cost-effectiveness ratio in 2021 US dollars per quality-adjusted life year (QALY) gained. Using a healthcare sector perspective, the base-case cost-effectiveness analysis showed that outpatient management with LMWH was cost-saving. In probabilistic sensitivity analyses, LMWH remained cost-saving in 98.0% of iterations at a willingness to pay of $100,000/QALY.

Finally, expected value of perfect, partial perfect and sample information (EVPI, EVPPI and EVSI) analyses were conducted using a probabilistic model and the probabilities of bleeding and stroke associated with LMWH and UFH bridging for low INR from a retrospective cohort study examining the 3-month follow-up period after LVAD implantation. EVSI was quantified with net monetary benefit (assuming willingness to pay for health as $100,000/QALY). We calculated discounted population-level EVSI by multiplying per-episode EVSI by the annual number of bridging procedures in the United States and assuming a 10-year time frame over which improved anticoagulation would be used. Study costs were based on administrative costs and LMWH/UFH costs. The discounted population-level EVPI and EVPPI were $5.6 million and $2.8 million, respectively. The VOI from future trials collecting data on adverse event rates was outweighed by the costs of these trials across study sample sizes.

In conclusion, using data on adverse event rates associated with LMWH vs UFH bridging from our retrospective cohort study, we found that there appears to be little uncertainty that outpatient management with LMWH is cost-saving for LVAD patients, as compared to inpatient UFH bridging. Despite the statistical uncertainty in the adverse events associated with the bridging strategies, a trial in which more information on the relative safety and efficacy of LMWH vs UFH bridging anticoagulation is collected would not represent good value for information.