Biomedical Engineering Reference
In-Depth Information
In this example, objective treatment outcomes, i.e., the occurrence of upper
gastrointestinal events and ischemic cerebrovascular events, have been collected to
measure the relative effectiveness and safety of the two treatments. However, how
participants evaluate and make trade-off between effectiveness and side effects have
not been studied. Realizing this limitation, researchers have recently included addi-
tional measures that refl ect participants' preferences, along with the traditional
treatment outcome data, in the clinical trials analysis.
Bordeleau et al. (
2010
) studied a multicenter, randomized, crossover clinical trial
to compare venlafaxine, a SNRI (selective noradrenergic reuptake inhibitor) antide-
pressant, with gabapentin, another existing drug, in treating hot fl ashes for women
with breast cancer. Though there has been evidence that both are effective and well
tolerated, the question of which drug is preferred by the patients remains unan-
swered. To answer this question, participant preferences were measured along with
the self-reported incidence of hot fl ashes and the occurrences of various side effects.
Sixty-six participants were randomly assigned to receive either venlafaxine or gaba-
pentin, for the fi rst 4-week period; after a washout period they would be crossed
over to the other treatment for a second 4-week period. At the end of the study,
participants were asked to complete a questionnaire regarding their preferred treat-
ment. The crossover design and after-study questionnaire bring researchers with
additional information that is unavailable in traditional clinical trials. Analysis
results showed that both drugs reduced hot fl ash scores to a similar extent. However,
side effects profi les were different between these two treatments: while venlafaxine
was associated with increased nausea, appetite loss, constipation, and reduced nega-
tive mood change, gabapentin was associated with increased dizziness and appetite.
Out of the 58 patients who provided preferred treatment data, a majority of 38 chose
venlafaxine, 18 chose gabapentin, and 2 had no preference for either drug.
Researchers also found that participants' preferences are correlated with the treat-
ment effectiveness, since their preferred treatment is usually associated with a larger
reduction in hot fl ash.
King et al. (
2007
) used a discrete choice experiment embedded in clinical trial to
study patient preferences across various treatment attributes, including symptom
control, its effect on daily activities, side effects, and convenience and cost of taking
medication. The purpose of their multicenter, crossover, randomized controlled trial
was to compare the effi cacy and safety of three preventive asthma medications,
including formoterol, a long-acting beta agonist (LABA), montelukast, a leukotri-
ene antagonist (LTRA), and fl uticasone, an inhaled corticosteroid (ICS). The study
contained three treatment phases, each with a 6-week duration followed by a 2-week
washout period. The discrete choice experiment was conducted after participants
received either formoterol or montelukast in the fi rst treatment phase. Participants
were then randomly assigned to read alternative profi les of a hypothetical drug,
which varied in treatment attributes, before they made the choice whether to con-
tinue with the current drug, to switch to the hypothetical medication, or to take no
medication. Patients also reported experienced symptoms, side effects, and their
daily activities while they were on one of the medications. Researchers estimated
a random-coeffi cient multinomial logit model using the reported choice data.
Search WWH ::
Custom Search