Seda Erdem

Accounting for attribute-level non-attendance in a health choice experiment: does it matter?

An extensive literature has established that it is common for respondents to ignore attributes of the alternatives within choice experiments. In most of the studies on attribute non-attendance, it is assumed that respondents consciously (or unconsciously) ignore one or more attributes of the alternatives, regardless of their levels. In this paper, we present a new line of enquiry and approach for modelling non-attendance in the context of investigating preferences for health service innovations. This approach recognises that non-attendance may not just be associated with attributes but may also apply to the attributes levels. Our results show that respondents process each level of an attribute differently: while attending to the attribute, they ignore a subset of the attributes levels. In such cases, the usual approach of assuming that respondents either attend to the attribute or not, irrespective of its levels, is erroneous and could lead to misguided policy recommendations. Our results indicate that allowing for attribute-level non-attendance leads to substantial improvements in the model fit and has an impact on estimated marginal willingness to pay and choice predictions.

Prioritising health service innovation investments using public preferences: a discrete choice experiment

BackgroundPrioritising scarce resources for investment in innovation by publically funded health systems is unavoidable. Many healthcare systems wish to foster transparency and accountability in the decisions they make by incorporating the public in decision-making processes. This paper presents a unique conceptual approach exploring the public’s preferences for health service innovations by viewing healthcare innovations as ‘bundles’ of characteristics. This decompositional approach allows policy-makers to compare numerous competing health service innovations without repeatedly administering surveys for specific innovation choices.MethodsA Discrete Choice Experiment (DCE) was used to elicit preferences. Individuals chose from presented innovation options that they believe the UK National Health Service (NHS) should invest the most in. Innovations differed according to: (i) target population; (ii) target age; (iii) implementation time; (iv) uncertainty associated with their likely effects; (v) potential health benefits; and, (vi) cost to a taxpayer. This approach fosters multidimensional decision-making, rather than imposing a single decision criterion (e.g., cost, target age) in prioritisation. Choice data was then analysed using scale-adjusted Latent Class models to investigate variability in preferences and scale and valuations amongst respondents.ResultsThree latent classes with considerable heterogeneity in the preferences were present. Each latent class is composed of two consumer subgroups varying in the level of certainty in their choices. All groups preferred scientifically proven innovations, those with potential health benefits that cost less. There were, however, some important differences in their preferences for innovation investment choices: Class-1 (54%) prefers innovations benefitting adults and young people and does not prefer innovations targeting people with ‘drug addiction’ and ‘obesity’. Class- 2 (34%) prefers innovations targeting ‘cancer’ patients only and has negative preferences for innovations targeting elderly, and Class-3 (12%) prefers spending on elderly and cancer patients the most.ConclusionsDCE can help policy-makers incorporate public preferences for health service innovation investment choices into decision making. The findings provide useful information on the public’s valuation and acceptability of potential health service innovations. Such information can be used to guide innovation prioritisation decisions by comparing competing innovation options. The approach in this paper makes, these often implicit and opaque decisions, more transparent and explicit.

Investigating Heterogeneity in the Characterization of Risks Using Best Worst Scaling

This research proposes and implements a new approach to the elicitation and analysis of perceptions of risk. We use best worst scaling (BWS) to elicit the levels of control respondents believe they have over risks and the level of concern those risks prompt. The approach seeks perceptions of control and concern over a large risk set and the elicitation method is structured so as to reduce the cognitive burden typically associated with ranking over large sets. The BWS approach is designed to yield strong discrimination over items. Further, the approach permits derivation of individual-level values, in this case of perceptions of control and worry, and analysis of how these vary over observable characteristics, through estimation of random parameter logit models. The approach is implemented for a set of 20 food and nonfood risks. The results show considerable heterogeneity in perceptions of control and worry, that the degree of heterogeneity varies across the risks, and that women systematically consider themselves to have less control over the risks than men.

Elimination and selection by aspects in health choice experiments: Prioritising health service innovations

Priorities for public health innovations are typically not considered equally by all members of the public. When faced with a choice between various innovation options, it is, therefore, possible that some respondents eliminate and/or select innovations based on certain characteristics. This paper proposes a flexible method for exploring and accommodating situations where respondents exhibit such behaviours, whilst addressing preference heterogeneity. We present an empirical case study on the publics preferences for health service innovations. We show that allowing for elimination-by-aspects and/or selection-by-aspects behavioural rules leads to substantial improvements in model fit and, importantly, has implications for willingness to pay estimates and scenario analysis.

Including Opt-Out Options in Discrete Choice Experiments: Issues to Consider

Providing an opt-out alternative in discrete choice experiments can often be considered to be important for presenting real-life choice situations in different contexts, including health. However, insufficient attention has been given to how best to address choice behaviours relating to this opt-out alternative when modelling discrete choice experiments, particularly in health studies. The objective of this paper is to demonstrate how to account for different opt-out effects in choice models. We aim to contribute to a better understanding of how to model opt-out choices and show the consequences of addressing the effects in an incorrect fashion. We present our code written in the R statistical language so that others can explore these issues in their own data. In this practical guideline, we generate synthetic data on medication choice and use Monte Carlo simulation. We consider three different definitions for the opt-out alternative and four candidate models for each definition. We apply a frequentist-based multimodel inference approach and use performance indicators to assess the relative suitability of each candidate model in a range of settings. We show that misspecifying the opt-out effect has repercussions for marginal willingness to pay estimation and the forecasting of market shares. Our findings also suggest a number of key recommendations for DCE practitioners interested in exploring these issues. There is no unique best way to analyse data collected from discrete choice experiments. Researchers should consider several models so that the relative support for different hypotheses of opt-out effects can be explored.