What is best when using drugs in chronic disease? A study that uses best-worst scaling to determine patient preferences for the process of drug use in Parkinson Disease.

Managing chronic disease differs from managing acute disease. In acute disease drugs may provide instant relief and even save life, both of which are strong incentives for taking drugs. In contrast, most chronic diseases are slowly progressive and drugs often have subtle effects on the severity of the disease symptoms and/or the progression of disease. The side-effects, complications of long-term use and treatment modality might be reasons for people to discontinue use, because the negative effect of the drugs are more direct and troublesome than the small and sometimes delayed effects of the drug on health. For instance in Parkinson’s Disease, although the drug was proven to be effective, the discontinuation rates of dopamine agonist can be as high as 50%, which indicates that their value to patients is determined by more aspects that effectiveness alone. Long term use of drugs for chronic disease requires, besides maximization of the positive effects, also minimization of the negative aspects of (drug) treatment. In Parkinson’s Disease, newer dopamine agonists are delivered in controlled release formulations or transdermally, thereby limiting the number of daily drug intakes to one. Likewise, MAO-B inhibitor and COMT-inhibitors are used to enhance the half-life of levodopa and dopamine. Continuous infusion of dopaminergic medication is aimed at reducing the late complications of therapy (on-off fluctuations), but the invasive character of continuous infusion is considered a negative aspect of treatment. Determining the relative value of different treatment modalities for PD should take into account the process differences between traditional oral intake of LD, oral long acting drugs, once-daily pills, transdermal formulations, and continuous infusion of LD along with differences in effectiveness, complications and side-effects[1]. The current toolbox used in health technology assessment is limited in its ability to measure the influence of multiple characteristics of drug use on the value (utility) of a drug and its effect on health related quality of life. Often, the cost-effectiveness ratio (CEA) does not discriminate between drugs which differ on the burden of treatment and it is difficult for health authorities and health insurers to determine the relative value of two (or more) therapeutic options for the same disease. A novel multi-criteria methodology for preference elicitation is best worst scaling (BWS). Best-worse scaling (BWS) can be viewed as a specific type of a discrete choice experiment (DCE) that allows for obtaining data to estimate utilities of individual attribute levels or the utility of a profile. In best-worst scaling, health states, health attributes or health profiles consisting of different characteristics of disease are ranked by subsequently selecting the best and the worst out of the available set. In the health profile version of best-worst scaling respondents are presented with a set of hypothetical scenarios which consist of performance levels for each attribute of care. Each scenario offers respondents a combination of attribute levels and asks them to choose which one of the combination of attribute levels is the best (most attractive) and which one of the combination of attribute levels is the worst (least attractive); respondents are essentially asked to choose that pair of attribute levels that maximizes the difference in value between them. Our objective is to 1) determine the validity of BWS to determine stated preferences for multiple aspects of drug use, 2) determine the relationship between the outcomes of BWS and quality of life estimates traditionally used in CEA. The benefits of a method that can estimate the impact of negative aspects of (drug) use and differentiate it from the beneficial effects are multiple. With such a technique, one could (a) estimate the relative influence of multiple characteristics of drug use on the value of a drug, (b) relate these differences to the costs of the drug and estimate the relative cost-effectiveness of reducing the burden of drug use and (c) determine patient preferences for drugs, which could be used in drug development and decision making.