Insights from Triangulation of Two Purchase Choice Elicitation Methods to Predict Social Decision Making in Healthcare

Article excerpt

Key points for decision makers

* Predicting the purchase probability of healthcare products and services and the factors that might affect social decision making is of interest to both suppliers and end users of healthcare products

* This study suggests inconsistency in the capacity of different methods to predict social decision making for pharmaceutical funding, particularly in the mid-range of purchase probability where decisions may be challenging

* It is important to use multiple methods to obtain a more complete picture of purchase probability in social decision-making contexts

Introduction

For many products, market success not only relates to end users of the product, it also relies on key decision makers. For example, consumable goods require buying managers and/or committees to decide to stock a product in their stores. In healthcare, it is common for decision makers to make purchase recommendations as agents for healthcare consumers. Prominent cases include recommendations related to the public subsidy of health technologies,[1] or choice of provider for healthcare services made by a healthcare purchaser, such as Primary Care Trusts in the UK or Health Maintenance Organizations in the US.[2] Predicting the purchase probability of healthcare products and services and the factors that might affect social decision making is of interest both to suppliers who wish to gain initial product listings or increase market share for their product or service, and to the end users for whom affordable access often depends on the social purchase decisions that are made.

The aim of this paper is to gain an overview of social decisions for a decision-making population through data triangulation using two methods for predicting purchase probability in social decision making: the discrete choice experiment (DCE) and the Juster purchase probability scale. We focus on insights related to (i) the consistency of the methods in predicting purchase choice, and (ii) the ability of the methods to capture heterogeneity in the decision choice. To illustrate our discussion, we present data from a social decision-making study of pharmaceutical subsidy in Australia. This paper starts by reviewing the literature on methods that can be used to elicit social decision-maker preferences or purchase probability. We then provide some background information on the social decision-making context of pharmaceutical subsidy in Australia. The methods for the exploratory study are outlined, and the resulting purchase probabilities are compared and contrasted both at the aggregate and individual decision-maker level using both methods for the same hypothetical pharmaceutical funding decisions. Finally, we discuss the preliminary findings of our study and their implications for the applicability of the two methods to the social decision-making context, making some recommendations for future research avenues.

Methods for Predicting Purchase Probability and their Application in Health Care

Predicting Purchase Probability Indirectly from Social Decision-Maker Preferences

The 'gold standard' of observing actual purchase decisions to explain purchase behaviour and demand that might be used in other fields of economics and marketing is difficult to apply in healthcare, due to market failures and a lack of revealed preference data.[3,4] Instead, the predominant approach has been to indirectly estimate preferences for hypothetical scenarios, and use them to predict the key factors that might impact demand and the likelihood of future choice or purchase decisions.

A number of methodological techniques, varying in complexity, have been used to elicit social preferences (that is, preferences for the availability of products/services for the treatment of others as opposed to oneself)[5] in healthcare. These include qualitative methods such as focus groups,[6] semi-structured interviews[7-9] and citizens' juries;[10] and quantitative methods, such as the person trade-off technique[11-14] and DCEs. …

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.