Premium
Economic Evaluation of Antiepileptic Drug Therapy: A Methodologic Review
Author(s) -
Levy Pierre
Publication year - 2002
Publication title -
epilepsia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.687
H-Index - 191
eISSN - 1528-1167
pISSN - 0013-9580
DOI - 10.1046/j.1528-1157.2002.46401.x
Subject(s) - antiepileptic drug , medicine , vigabatrin , epilepsy , clinical neurology , drug , pharmacotherapy , anticonvulsant , psychiatry , psychology , neuroscience
Summary: Purpose: The increasing number of antiepileptic drugs (AEDs) fostered the development of economic studies in epilepsy. We reviewed this literature to identify and discuss methodologic issues. Methods: We included all studies devoted to cost‐based evaluation in epilepy, published in English from 1989 to 2001, and identified via a Medline search. Results: We identified a series of methodologic problems. First, we reconsidered heterogeneity of concepts and estimating methods, often cited as the most critical problem, as they do not necessarily result from a failure to apply standard methods. One must distinguish “natural” sources of heterogeneity arising from the many unconstrained choices left open in the implementation of economic evaluation on the one hand, and imperfect information and observation‐based sources of heterogeneity leading to constrained choices on the other hand. By their very nature, cost‐of‐illness studies are subject to this variety of choices and were used to illustrate our purpose. Second, cost‐minimization studies were reviewed, as they raise additional problems related to study design and choice of an outcome measure. Finally, deficiencies were also identified in cost‐effectiveness and cost–utility studies concerning attempts to incorporate patient's point of view in outcome measurement. Conclusions: We agreed with previous reviews on the difficulty of compare results from economic studies in epilepsy due to heterogeneity in methods and concepts used. This is partly due to imperfect information and limits in observation as sources for data collection, as well as to unavailability of refined outcome measures. Therefore, improvements are possible in this field.