Characteristics of Attributes in Patient Preference Studies

Since the theme this month here is patient preference, I figured I would expound on one of the most important features of patient preference (or patient-centered benefit risk) studies – attributes.  The term attribute is defined as the quality or feature regarded as characteristic or inherent part of someone or something.  In this case, we are talking about the attributes of a medical device or procedure or drug.  In other words, when we refer to medical product attributes, we are referring to the characteristics or qualities that are inherent to their design or effectiveness or delivery.  We are defining what makes a medical product unique, so that we can differentiate it from other medical products.

Okay, that’s nifty, and marketing probably likes to display the best attributes of each product in every way they can.  But, how does that help me determine patient preference?  In fact, attributes of medical products are a critical component of preference studies that assess patients’ benefit-risk tolerance, treatment priorities, or most desired outcomes.  The reason? Attributes are the qualities or characteristics on which patients will make their trade-off decisions.  Attributes are not just the sales pitch.  An attribute can be a benefit imparted by the medical product or procedure. But, an attribute can also be a risk associated with the treatment or new drug.  It is the attributes that frame the experiment within which we capture patient preference or risk tolerance.

To illustrate, let’s look at some examples.  In a study published in 2009, authors Reed et al. examined patients’ benefit-risk preferences among those with multiple sclerosis (MS).[1]  The attributes in this experiment included:  the number of relapses in the next 5 years, the time from today until patients’ MS gets worse, the chance of dying from liver failure within 10 years, the chance of dying or severe disability from progressive multifocal leukoencephalopathy, and chance of dying from leukemia in 10 years. In another study published in 2014, Klojgaard et al. examined patient preferences for treatment of back pain.[2]  The attributes in this study included: treatment modality, pain level improvement, problems with activities of daily living, risk of relapse, and time to treatment effect.  In both studies (and many others like them) a selected group of attributes was chosen to define the frame in which patients then make choices between one treatment or another.  The experimental design and number of patients available will determine how many attributes you can have (which we will cover in another blog post another day…)

So, how do you know which attributes to choose?  Determining which attributes to use in your study can be the most difficult and time consuming (and contentious!) part of designing a patient-centered benefit risk study.  Attributes must be clinically meaningful, related to your clinical study or clinical practice outcomes, relevant to regulatory reviewers, and important to patients.  This can be a tall order, and in many cases a sponsor will start out with a laundry list of attributes that must be culled to ensure a powerful experimental design.  Composite attributes can be effective and may be necessary, but they should have a clear clinical relevance.  Attribute descriptions must be clear and understandable to patients.  All attributes should be evaluated by patients in a qualitative setting before being implemented in a quantitative preference study.

Identifying, choosing, and describing attributes for a patient preference study can be challenging.But, there are ways to streamline the process and make it less painful for all involved. Those of us who have done a few of these have learned a few tricks and can make the attribute identification process more efficient and patient-centered.Look for more information about this and other patient-focused policy topics in my blog or in my newsletter, Patient x Design.

[1] Johnson FR et al. (2009).  Multiple sclerosis patients’ benefit-risk preferences:  Serious adverse event risks versus treatment efficacy.  J of Neurology.  256:554-562.        

[2] Klojgaard ME et al. (2014).  Patient preferences for treatment of low back pain—A discrete choice experiment.  Value in Health. 390 – 396.