Comparison of Oral Anticoagulants for Stroke Prevention in Nonvalvular Atrial Fibrillation: A Multicriteria Decision Analysis.

BACKGROUND: Decision on the most appropriate oral anticoagulation therapy for stroke prevention in patients with nonvalvular atrial fibrillation is difficult because multiple treatment options are available, and these vary in their clinical effects and relevant nonclinical characteristics. OBJECTIVES: To use a multicriteria decision analysis (MCDA) to compare the oral anticoagulants apixaban, dabigatran, edoxaban, rivaroxaban, and vitamin K antagonist (VKAs; specifically warfarin) in patients with nonvalvular atrial fibrillation. METHODS: We identified the evaluation criteria through a targeted literature review and clinical judgment. The final evaluation model included nine clinical events and four other criteria. We ranked possibly fatal clinical event criteria on the basis of the differences in risks of fatal events and the corresponding window of therapeutic opportunity, as observed in clinical trials. Clinical judgment was used to rank other criteria. Full criteria ranking was used to calculate centroid weights, which were combined with individual treatment performances to estimate the overall value score for each treatment. RESULTS: Using such an MCDA, dabigatran yielded the highest overall value, approximately 6% higher than that of the second-best treatment, apixaban. Dabigatran also had the highest first-rank probability (0.72) in the probabilistic sensitivity analysis. Rivaroxaban performed worse than the other non-VKA oral anticoagulants, but better than VKAs (with both having 0.00 first-rank probability). The results were insensitive to changes in model structure. CONCLUSIONS: When all key oral anticoagulant value criteria and their relative importance are investigated in an MCDA, dabigatran appears to rank the highest and warfarin the lowest.

MCDA swing weighting and discrete choice experiments for elicitation of patient benefit-risk preferences: a critical assessment.

PURPOSE: Multiple criteria decision analysis swing weighting (SW) and discrete choice experiments (DCE) are appropriate methods for capturing patient preferences on treatment benefit-risk trade-offs. This paper presents a qualitative comparison of the 2 methods. METHODS: We review and critically assess similarities and differences of SW and DCE based on 6 aspects: comprehension by study participants, cognitive biases, sample representativeness, ability to capture heterogeneity in preferences, reliability and validity, and robustness of the results. RESULTS: The SW choice task can be more difficult, but the workshop context in which SW is conducted may provide more support to patients who are unfamiliar with the end points being evaluated or who have cognitive impairments. Both methods are similarly prone to a number of biases associated with preference elicitation, and DCE is prone to simplifying heuristics, which limits its application with large number of attributes. The low cost per patient of the DCE means that it can be better at achieving a representative sample, though SW does not require such large sample sizes due to exact nature of the collected preference data. This also means that internal validity is automatically enforced with SW, while the internal validity of DCE results needs to be assessed manually. CONCLUSIONS: Choice between the 2 methods depends on characteristics of the benefit-risk assessment, especially on how difficult the trade-offs are for the patients to make and how many patients are available. Although there exist some empirical studies on many of the evaluation aspects, critical evidence gaps remain.

In the Opponent's Shoes: Increasing the Behavioral Validity of Attackers' Judgments in Counterterrorism Models.

A key objective for policymakers and analysts dealing with terrorist threats is trying to predict the actions that malicious agents may take. A recent trend in counterterrorism risk analysis is to model the terrorists' judgments, as these will guide their choices of such actions. The standard assumptions in most of these models are that terrorists are fully rational, following all the normative desiderata required for rational choices, such as having a set of constant and ordered preferences, being able to perform a cost-benefit analysis of their alternatives, among many others. However, are such assumptions reasonable from a behavioral perspective? In this article, we analyze the types of assumptions made across various counterterrorism analytical models that represent malicious agents' judgments and discuss their suitability from a descriptive point of view. We then suggest how some of these assumptions could be modified to describe terrorists' preferences more accurately, by drawing knowledge from the fields of behavioral decision research, politics, philosophy of choice, public choice, and conflict management in terrorism. Such insight, we hope, might help make the assumptions of these models more behaviorally valid for counterterrorism risk analysis.