Five good reasons to use MarketScan Treatment Pathways

Visualize a treatment pathway

With IBM® MarketScan® Treatment Pathways, you can build a sequence of events with intuitively drawn treatment flow diagrams. You can also create nodes and arcs to determine the events leading up to a diagnosis, time to treatment, switching pattern and outcome events. The visual interface is designed to help you quickly cut the data by diagnosis, procedures, treatments, events and treatment locations. You can move forward or backward in time, or toggle between tables and graphs.

Get insights quickly

By visualizing robust patient-level data such as those found in the MarketScan Research databases, you can efficiently see different patient pathways. Standard reports include demographics, annualized costs, comorbidities, procedures, top drugs and proximity maps for pre- and post-event comparisons. You can build study groups from nodes to perform more sophisticated analyses built into the tool.

Pursue opportunities with greater flexibility

MarketScan Treatment Pathways helps you to explore your options using an intuitive interface overlaying real-world patient-level data. You can explore a hypothesis, test an analytic strategy for a full-blown study, calculate market size or even determine patient segments.

Gain access to powerful MarketScan data

Virtually any healthcare dataset can be analyzed with MarketScan Treatment Pathways. Among those datasets are the MarketScan Research Databases. These patient-level claims databases have been used since 1992 for healthcare research and have an excellent publication track record in peer-reviewed journals. They are the longest-running and largest proprietary US claims databases available for healthcare research, with data from more than 185 million unique patients.

Take advantage of boosted propensity scoring

Boosted propensity scoring allows you to better control confounding and bias using advanced boosting technology compared to traditional logistic regression. Treatment Pathways Boosted Propensity Score is a machine learning capability that can handle correlated explanatory variables and non-linear relationships to estimate propensity score weights. This is accomplished by using a modern method with three layers of protection to control confounding.

MarketScan Treatment Pathways: Databases and tools

Read the brochure