Information Technology Reference
In-Depth Information
may not contain this code at all. Each claim will have a service date. We start by creating a clustering
for each patient based upon date and DRG codes. Not every patient claim will be clustered successfully.
From there, predictive modeling will be used for the unclustered values to predict membership into each
cluster.(Putten, Kok, & Gupta, 2002; Xue, 2006)
To examine the sequence of episodes, we can define a time series with multiple time endpoints. The
initial time point will define the initial treatment and beginning of a chronic problem. The additional time
points will be defined as either the end of the episode, or a change in condition, where the chronic illness
gets better, or worse. We can use both fixed and dynamic regressors to investigate the patient outcomes.
These regressors can represent a different medication, or a decision to perform surgery, or a change from
outpatient to inpatient status. They can also represent a new, ongoing treatment. The fixed regressors
will represent patient demographic information, and the initial severity of the patient's condition. The
time series will be transactional in nature as the changes in treatment will not necessarily occur at fixed
intervals. We can start by defining a time series for each patient, and then consolidating them into a series
of forecasted outcomes. Once we have the likelihood of various outcomes defined by the time series, we
can create a decision tree to look at the probability of each outcome given treatment choices.
In addition, it will be important to detect outliers either as they occur, or before they occur in terms
of both cost and outcomes. Therefore, the claims data can be considered streaming data, with changes in
treatments indicative of future outcomes that can be costly either to payer or patient. Once the model is
developed, it can be scored and used in future data. In that way, we can continue to examine treatment
pathways for optimal choices even as treatments and protocols change with the development of new
drugs, devices, and procedures.
What we will focus on in the chapter is the necessary SAS preprocessing code that is needed to define
episodes in the claims data so that the data can be investigated as discussed in chapters 2-8. In particular,
we will discuss ways to convert the one-to-many to a one-to-one relationship for examination.
mePs data
We will first examine the MEPS dataset. It has one major difference compared to more general claims
data in that it has translated all patient encounters into ICD9 codes. Therefore, we do not have to worry
about conversion of codes, nor do we need to consider multiple coding systems. However, we do have
to consider the fact that different types of patient utilization of the healthcare system is contained within
different datasets. It has an additional advantage in that there is extensive information on patients, in-
cluding employment, income, and geographic region.
We will start with a problem first discussed in Chapter 2. We investigated patient compliance with
medication prescriptions for osteoporosis. We did not examine which of the medications prescribed
optimize outcomes because the medications file contains no outcome measures. However, we can con-
sider the inpatient, outpatient, and physician visits files to determine which of the patients suffered from
orthopedic-related injuries, specifically fractures. Then we can see if there is a relationship between
compliance, patient severity, medications, and adverse events to see if one of the medications reduces
the occurrence of these events. We use the measure defined in Chapter 2 relating patient compliance
to the number of prescriptions. Since we discussed the inpatient and outpatient episodes, and showed
that patients with higher levels of medication compliance in fact had more episodes of treatment, we
examine the relationship of medications to physician visits here. In this example, we investigate the
Search WWH ::
Custom Search