ABSTRACT
It is nearly impossible for these population-level questions to be answered by a single data-collection study. Such a study would need to provide evidence on the populations of interest (there could be many) under the relevant conditions (which may be highly local and variable) by comparing against appropriate alternative interventions (also subject to local variations in practice) to obtain unbiased information on all health and economic aspects of interest for a long enough period to reasonably cover the relevant outcomes, and do all of these in a timely way to inform the actual decisions. Because of the limitations just listed and others, rather than rely on a single study, health care decisions must be informed by data obtained from a variety of sources. Some of these data sources may be more relevant or valid than others, and combining them in a single analysis forces the decision maker to make various assumptions regarding how the data fit together when estimating the possible impacts of each course of action the policy makers may choose.
