Warsaw, Poland – Convincing decision makers to reimburse a health technology usually requires economic rationale proving that the cost needed to achieve an additional health effect is relatively low and the technology provides good value for money. Health effects can be measured either in life-years gained (through cost-effectiveness analysis (CEA)) or quality-adjusted life-years (through cost-utility analysis (CUA)). Researchers from
HealthQuest spółka z ograniczoną odpowiedzialnością Sp. K and the
Warsaw School of Economics wondered if drug companies select the type of analysis that offers more favourable results when submitting economic evaluations.
In the study, “
Cost-effectiveness versus Cost-Utility Analyses: What Are the Motives Behind Using Each and How Do Their Results Differ?—A Polish Example,” the researchers analyzed economic evaluations submitted to AHTAPol from 2007-2011 to examine what factors lead to choosing CEA or CUA and how the precision (measured as range of values in sensitivity analyses) of cost-effectiveness or cost-utility measures differed.
No clear signs of bias seemed to be present in past submissions to the
Polish Agency for Health Technology Assessment (AHTAPol), and so the strong preference of CUA imposed in 2012 seems not to have been necessary. Results from submissions to AHTAPol, however, demonstrate that CUAs usually lead to greater precision of estimates (even though they require additional parameters for health-related quality of life weighting).
CEAs and CUAs disagree more often than previously shown in the literature (in about 13% cases), but there are still no clear signs of bias. (For example: in submissions for drugs used in oncology, CUA is used less frequently and is also less favourable (not showing drug’s good value for money). On the other hand, CEA is less favourable for drugs in submissions that use long-time horizon modelling and is still used more often. Although performing CUA requires additional parameters, the aggregate outcome (incremental cost-utility ratio) is estimated more precisely than its counterpart, CEA.
Michał Jakubczyk, PhD, Assistant Professor at the Warsaw School of Economics, and a co-author on the study says: “Reassuringly, introducing additional parameters into modelling [utilities] does not seem to worsen the precision of the estimates. Luckily then, it seems that we may not have to choose between being ‘roughly right’ or ‘precisely wrong’, but may be a little bit more precisely right [in health technology assessment] with CUA.”