Accuracy incentives and framing effects to minimize the influence of cognitive bias among advanced cancer patients

Many patients with advanced illness have unrealistic survival expectations, largely due to cognitive biases. Studies suggests that when people are motivated to be accurate, they are less prone to succumb to these biases. Using a randomized survey design, we test whether offering advanced cancer patients ( n = 200) incentives to estimate their prognosis improves accuracy. We also test whether presenting treatment benefits in terms of a loss (mortality) rather than a gain (survival) reduces willingness to take up a hypothetical treatment. Results are not consistent with the proposed hypotheses for either accuracy incentives or framing effects.

Exploration Of Theoretical And Application Issues In Using Fully Bayesian Methods For Road Safety Analysis

The Fully Bayesian (FB) approach to road safety analysis has been available for some time, but it is largely unevaluated and untested. This study is trying to bridge the gap by conducting a thorough evaluation of FB method for black spots identification and treatment effect analysis. First, an evaluation is conducted on the univariate FB versus the empirical Bayesian (EB) method for single level severity data through the development of various models, and multivariate FB versus univariate FB for multilevel severity data, as well as the performance of various ranking and evaluation criteria for black spots identification. It is confirmed that the FB method is superior to the EB with respect to key ranking criteria (expected rank, mode rank and median rank of posterior PM, etc.). The multivariate FB method is better than univariate FB for the multilevel severity crashes. Then a teat of the FB before-after method for treatment effect analysis is performed. Two FB testing frameworks were employed. First the univariate before-after fully Bayesian (FB) method was examined using three simulated datasets. Then multivariate Poisson log normal (MVPLN), univariate Poisson log normal (PLN) and PB (Poisson gamma) models were evaluated using two groups of California unsignalized intersections. Hypothetical treatment sites were selected from these datasets such that a significant effect would be estimated by the naive before-after method that does not account for regression to the mean. This study confirmed that FB methods can indeed provide valid results, in that they correctly estimate a treatment effect of zero at these hypothetical treatment sites after accounting for regression to the mean. Finally the EB and the validated FB before after methods were applied to evaluation of two treatments: the conversion of rural intersections from unsignalized to signalized control; and the conversion of road segments from a four-lane to a three-lane cross-section with two-way left turn lanes (also known as road diets). The result indicates that both FB and EB method can provide comparable treatment effect estimates. This would suggest it is still appropriate to conduct treatment effect analysis using the EB method for univariate crash data, but that it is essential in so doing to account for temporal trends in crash frequency.

Exploration Of Theoretical And Application Issues In Using Fully Bayesian Methods For Road Safety Analysis

The Fully Bayesian (FB) approach to road safety analysis has been available for some time, but it is largely unevaluated and untested. This study is trying to bridge the gap by conducting a thorough evaluation of FB method for black spots identification and treatment effect analysis. First, an evaluation is conducted on the univariate FB versus the empirical Bayesian (EB) method for single level severity data through the development of various models, and multivariate FB versus univariate FB for multilevel severity data, as well as the performance of various ranking and evaluation criteria for black spots identification. It is confirmed that the FB method is superior to the EB with respect to key ranking criteria (expected rank, mode rank and median rank of posterior PM, etc.). The multivariate FB method is better than univariate FB for the multilevel severity crashes. Then a teat of the FB before-after method for treatment effect analysis is performed. Two FB testing frameworks were employed. First the univariate before-after fully Bayesian (FB) method was examined using three simulated datasets. Then multivariate Poisson log normal (MVPLN), univariate Poisson log normal (PLN) and PB (Poisson gamma) models were evaluated using two groups of California unsignalized intersections. Hypothetical treatment sites were selected from these datasets such that a significant effect would be estimated by the naive before-after method that does not account for regression to the mean. This study confirmed that FB methods can indeed provide valid results, in that they correctly estimate a treatment effect of zero at these hypothetical treatment sites after accounting for regression to the mean. Finally the EB and the validated FB before after methods were applied to evaluation of two treatments: the conversion of rural intersections from unsignalized to signalized control; and the conversion of road segments from a four-lane to a three-lane cross-section with two-way left turn lanes (also known as road diets). The result indicates that both FB and EB method can provide comparable treatment effect estimates. This would suggest it is still appropriate to conduct treatment effect analysis using the EB method for univariate crash data, but that it is essential in so doing to account for temporal trends in crash frequency.

Understanding Treatment Strategies and Preferences in Nonmetastatic Castration-Resistant Prostate Cancer From the Japanese Physician Perspective

PURPOSE Sixteen percent (16%) of patients with castration-resistant prostate cancer (CRPC) show no bone metastasis at diagnosis. However, 33% will become metastatic within 2 years. The goal of treatment in patients with nonmetastatic CRPC (nmCRPC), therefore, is to delay symptomatic metastases without undue toxicity. With novel antiandrogen treatments of different strengths and limitations available, physician preferences for nmCRPC treatment in Japan should be understood. METHODS A discrete choice experiment was conducted. Physicians chose between two hypothetical treatments in nmCRPC defined by six attributes: risk of fatigue, falls or fracture, cognitive impairment, hypertension, rashes as side effects of treatment, and extension of time until cancer-related pain occurs. Relative preference weights and relative importance were estimated by hierarchical Bayesian logistic regression. Physicians were also asked to make treatment decisions based on four hypothetical patient profiles to understand the most important factors driving decision making. RESULTS A total of 151 physicians completed the survey. Extension of time until cancer-related pain occurs was the most important attribute (relative importance, 32.3%; CI, 31.3% to 33.3%). Based on summed preference weights across all attributes, preferences for hypothetical treatment profiles I, II, and III were compared. A hypothetical treatment profile with better safety though shorter extension time was preferred (I: mean [standard deviation] = 1.7 [1.6 to 2.1]) over treatment profiles with lower safety but longer extension time (II: −2.7 [−2.8 to −2.6] and III: −0.2 [−0.3 to −0.1]). Treatment characteristics were more important factors for physicians' decision making than patient characteristics in prescribing treatment. CONCLUSION Physicians preferred a treatment with better safety profile, and treatment characteristics were the most important factors for decision making. This might have implications in physicians' decision making for nmCRPC treatment in the future in Japan.

Instrument-Assisted Soft Tissue Mobilization Forces Applied by Trained Clinicians During a Simulated Treatment

Context: Instrument-assisted Soft Tissue Mobilization (IASTM) is a therapeutic intervention used by clinicians to identify and treat myofascial dysfunction or pathology. However, little is known about the amount of force used by clinicians during an IASTM treatment and how it compares to reports of force in the current literature. Objective: To quantify the range of force applied by trained clinicians during a simulated IASTM treatment scenario. Design: Experimental. Setting: University research laboratory. Participants: Eleven licensed clinicians (physical therapist = 2, chiropractor = 2, and athletic trainer = 7) with professional IASTM training participated in the study. The participants reported a range of credentialed experience from 1 to 15 years (mean = 7 [4.7] y; median = 6 y). Intervention: Participants performed 15 one-handed unidirectional sweeping strokes with each of the 5 instruments for a total of 75 data points each. Force data were collected from a force plate with an attached skin simulant during a hypothetical treatment scenario. Main Outcome Measures: Peak force and average forces for individual strokes across all instruments were identified. Averages for these forces were calculated for all participants combined, as well as for individual participants. Results: The average of peak forces produced by our sample of trained clinicians was 6.7 N and the average mean forces was 4.5 N. Across individual clinicians, average peak forces ranged from 2.6 to 14.0 N, and average mean forces ranged from 1.6 to 10.0 N. Conclusions: The clinicians in our study produced a broad range of IASTM forces. The observed forces in our study were similar to those reported in prior research examining an IASTM treatment to the gastrocnemius of healthy individuals and greater than what has been reported as effective in treating delayed onset muscle soreness. Our data can be used by researchers examining clinically relevant IASTM treatment force on patient outcomes.

How Do Progression-Free Survival and Toxicity Risks Influence Oncologist and Patient Preferences for Novel Agents in Chronic Lymphocytic Leukemia?

Introduction: Treatment for chronic lymphocytic leukemia (CLL) has changed with the approval of novel agents, which have different toxicity profiles. The aim of the current analysis was to determine how incremental changes in efficacy and toxicity profile impact treatment selection among patients with CLL and oncologists. Methods: In this US-based study, oncologists and CLL patients completed an online survey that included a discrete choice experiment (DCE) to quantify preferences for first line (1L) treatment with novel agents. Self-reported data on oncologist and patient characteristics were also collected. In the DCE, respondents selected between hypothetical treatment profiles consisting of 8 attributes with varying levels. The attributes were selected based on the findings of qualitative interviews assessing treatment priorities among oncologists and patients. The attribute levels were abstracted from clinical trials and published literature (Table 1). Hierarchical Bayesian regression models were used to estimate preference weights for each attribute level. The preference weights were used to generate a base case hypothetical treatment profile, against which other hypothetical profiles varying in adverse event (AE) risk and 2-year progression-free survival (PFS) were evaluated to understand which attributes and levels drive treatment selection. The overall mean summed preference weight for the collective set of alternative profiles was then compared with the base case, with higher positive values indicating the more preferred profile. Results: Oncologists (N=151) reported a mean of 16.3±7.0 years in practice (Table 2). Most practiced in a community setting (72%) and in a major metropolitan/urban area (64%). Among patients (N=220), median age was 56.0 years, with a mean disease duration of 2.0±3.1 years at time of study (Table 2). Most patients were in or had completed at least 1L therapy (68%). Figure 1 illustrates the impact of changing various attribute levels on the overall preference of an alternative profile, relative to a base case profile. Decreasing 2-year PFS from 95% to 75% (Profile A) and reducing the risks of atrial fibrillation (A-fib) from 20% to 5% (Profile B) and infection from 30% to 7% (Profile C) had the greatest influence on treatment preferences for oncologists when holding all other attribute levels constant. Oncologists preferred the profile with reduced 2-year PFS and reduced risk of AEs (Profile G) to the base case profile, corresponding to a 45% difference in preference share. Similarly, for patients, reducing 2-year PFS from 95% to 75% (Profile A) and reducing the risks of infection from 30% to 7% (Profile C) and A-fib from 20% to 5% (Profile B) were most influential in treatment choice when all other attribute levels were held constant (Figure 1). However, patients preferred the base case profile with higher PFS and higher risk of AEs to Profile G, corresponding to a 38% difference in preference share. Conclusions: For oncologists and patients, decreasing 2-year PFS and reducing the risks of either A-fib or infection had the most influence on 1L treatment preference, relative to the base case in which 2-year PFS was the highest and the AE risks were set to the worst levels reported in literature for novel therapies. The pattern and direction of treatment preferences were generally consistent among oncologists and patients, relative to the base case, across all alternative profiles when individually examined. However, for patients, the positive impact of reducing the risk of AEs was outweighed by the negative impact of reducing PFS. Consequently, all other things being equal, patients preferred a 1L profile with higher PFS even if it came with higher AE risks. In contrast, oncologists were willing to accept a treatment with less efficacy in exchange for the most favorable safety profile. This difference may be due to a patient's lack of understanding of the potential severity of AEs or a physician's lack of awareness of the patient's treatment priorities. This is an area for future research and highlights the importance of oncologists explicitly communicating the known efficacy benefits and AE risks associated with each treatment option because these factors may influence patients' treatment choice in a way that differs from their own. Disclosures Ryan: AstraZeneca: Current Employment, Current equity holder in private company. Le:AstraZeneca: Current Employment, Current equity holder in private company. Wahlstrom:AstraZeneca: Current Employment, Current equity holder in private company. Beusterien:Kantar: Other: Employee of Kantar, which received funding from AstraZeneca to conduct this study. Will:Kantar: Other: Employee of Kantar, which received funding from AstraZeneca to conduct this study. Maculaitis:Kantar: Other: Employee of Kantar, which received funding from AstraZeneca to conduct this study. Leblanc:UpToDate: Patents & Royalties: Royalties; Agios, AbbVie, and Bristol Myers Squibb/Celgene: Speakers Bureau; AstraZeneca: Research Funding; AbbVie, Agios, Amgen, AstraZeneca, CareVive, BMS/Celgene, Daiichi-Sankyo, Flatiron, Helsinn, Heron, Otsuka, Medtronic, Pfizer, Seattle Genetics, Welvie: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees; American Cancer Society, BMS, Duke University, NINR/NIH, Jazz Pharmaceuticals, Seattle Genetics: Research Funding.

Assessing the Potential Impact of a Long-Acting Skin Disinfectant in the Prevention of MRSA

Background: Healthcare-associated transmission of methicillin-resistant Staphylococcus aureus (MRSA) remains a persistent problem despite advances in prevention. The use of chlorhexidine gluconate (CHG) as a means of decolonizing patients, either through targeted decolonization or daily bathing, is a frequently used measure to supplement other MRSA reduction interventions. However, there is room for new and innovative decolonizing agents. We explored the potential utility of a long-acting CHG-like disinfectant with a persistent protective effect as well as an immediate decolonizing action in the prevention of MRSA acquisition as well as the subsequent development of clinical illness and MRSA-related mortality. Methods: We modeled MRSA transmission throughout an 18-bed intensive care unit, based on previously published models. A baseline model with no daily decolonizing protocol was used as a baseline and was compared to a scenario assuming that patients were bathed with CHG, which decolonizes them but provides no ongoing protection, as well as a scenario involving a hypothetical treatment that both decolonizes and provides ongoing protection from subsequent colonization. We varied the duration and efficacy of this protection to fully explore the potential utility of such a treatment. Results: The results of the simulations are shown in Fig. 1, where duration and efficacy of protection varied. The number of MRSA acquisitions from each combination is depicted as a single point, with blue points indicating correspondingly fewer MRSA acquisitions. Overall, improved efficacy of the hypothetical disinfectant resulted in immediate improvements in MRSA acquisition rates when compared to the baseline. To see major improvements in the MRSA acquisition rate due to the duration of infection, that duration must be well above 10 hours in many scenarios. There is also little evidence of synergy between the two. Conclusions: Based on recent results suggesting CHG has a relatively modest per-use efficacy (<.20), there is room for improvement in the formulation and administration of decolonizing agents. Although there has been considerable excitement about the possibility of long-acting agents that not only decolonize but provide long-acting protection against colonization, these results suggest that such protection would only result in markedly decreased acquisition rates only if that duration of protection was extremely long, or if the agent itself was also considerably more efficacious than CHG. These results may be used to help consider the necessary study size for clinical studies of these agents in the future, or to set research priorities and properly calibrate expectations.Funding: NoneDisclosures: None

Mechanisms of a sustained anti-inflammatory drug response in alveolar macrophages unraveled with mathematical modeling

Both initiation and suppression of inflammation are hallmarks of the immune response. If not balanced, the inflammation may cause extensive tissue damage, which is associated with common diseases, e.g. asthma and atherosclerosis. Anti-inflammatory drugs often come with severe side effects driven by high and fluctuating drug concentrations. To remedy this, drugs with sustained anti-inflammatory responses are desirable. However, we still lack a quantitative mechanistic understanding of such sustained effects. Here, we study the anti-inflammatory response to a common glucocorticoid drug, Dexamethasone. We find a sustained response 22 hours after drug removal. With hypothesis testing using mathematical modeling, we unravel the underlying mechanism – a slow release of Dexamethasone from the receptor-drug complex. The developed model is in agreement with time-resolved training and testing data, and is used to simulate hypothetical treatment schemes. This work opens up for a more knowledge-driven drug development, to find sustained anti-inflammatory responses and fewer side effects.

Assessing Pharmacists’ Preferences towards Efficacy Attributes of Disease-Modifying Therapies in Relapsing-Remitting Multiple Sclerosis

Introduction: Hospital pharmacists are increasingly playing a critical role in the care of patients with multiple sclerosis (MS). However, little is known about their preferences and perspectives towards different attributes of disease-modifying therapies (DMTs). The objective of this research was to assess pharmacists´ preferences for DMT efficacy attributes. Methods: A multicenter, non-interventional, cross-sectional, web-based study was conducted. Preventing relapses, delaying disease progression, controlling radiological activity, and preserving health-related quality of life (HRQoL) and cognition were the attributes selected based on a literature review and a focus group with six hospital pharmacists. Conjoint analysis was used to determine preferences in eight hypothetical treatment scenarios, combining different levels of each attribute and ranking them from most to least preferred. Results: Sixty-five hospital pharmacists completed the study (mean age: 43.5 ± 7.8 years, 63.1% female, mean years of professional experience: 16.1 ± 7.4 years). Participants placed the greatest preference on delaying disease progression (35.7%) and preserving HRQoL (21.6%) and cognition (21.6%). Importance was consistent in all groups of pharmacists stratified according to demographic characteristics, experience, research background, and volume of patients seen per year. Conclusions: Understanding which treatment characteristics are meaningful to hospital pharmacists may help to enhance their synergistic role in the multidisciplinary management of patients with MS.