Working from Claims Data - Text Mining Techniques for Healthcare Provider Quality Determination - page 341

Chapter 9
Working from Claims Data
IntroductIon
Claims data are more difficult to work with to extract the necessary information about patient condi-
tions in relationship to costs. There can be multiple claims for the same patient episode from different
sources. A physician visit after an inpatient claim can be followed up for the inpatient stay rather than to
consider the inpatient stay as the start of a new patient episode or a new patient problem. Therefore, in
addition to analyzing patient conditions as represented by ICD9 codes, we must also attempt to define
an episode and to distinguish between new problems and follow up of old problems.
The second major difference in using claims data as opposed to the data we have been using up to this
point is that there exists a one-to-many relationship in the data that must be taken care of. For example,
one inpatient stay may have several follow up visits, and one or two of those visits may be classified
as outpatient stays while others are classified as physician visits. We need to find a way to convert the
one-to-many relationship to a one-to-one so that we can investigate outcomes. Another difference is
that not all medical services are necessarily allowable, so that the medical record may be incomplete.
For example, until recently, Medicare disallowed almost all medications. Therefore, such information