Biomedical Engineering Reference
In-Depth Information
Challenges of Evaluation in
Biomedical Informatics
This chapter develops, in a general and informal way, the many topics that
are explored in more detail in later chapters of this topic. It gives a first def-
inition of evaluation, describes why evaluation is needed, and notes some
of the problems of evaluation in biomedical informatics that distinguish it
from evaluation in other areas. In addition, it lists some of the many types
of information systems and resources, the questions that can be asked about
them, and the various perspectives of those concerned.
First Definitions
Most people understand the term “evaluation” to mean measuring or
describing something, usually to answer questions or help make decisions.
Whether we are choosing a holiday destination or a web browser, we eval-
uate the options and how well they fit key objectives or personal prefer-
ences. The form of the evaluation differs widely according to what is being
evaluated and how important the decision is. Therefore, in the case of
holiday destinations, we may ask our friend which Hawaiian island she
prefers, and why, before searching for more specific information about
hotels. For a web browser, we may focus on more technical details, such as
the time to open a web site or its compatibility with our disability. Thus, the
term “evaluation” describes a range of data-collection activities designed
to answer questions ranging from the casual “What does my friend think
of Maui?” to the more focused “Is web browser
A
quicker than web browser
B
on my computer?”
In biomedical informatics, we study the collection, processing, and com-
munication of information related to health care, research, and education.
We build “information resources,” usually consisting of computer hardware
and software, to facilitate these activities. For healthcare applications, such
information resources include systems to collect, store, and communicate
data about specific patients (e.g., clinical workstations and electronic patient
records) and systems to assemble, store, and reason using medical knowl-
1
