RxDataScience Patient Journey App

Kx Use Case: The RxDataScience Patient Journey Application

23 Oct 2017 | , , , , ,
Share on:


RxDataScience is a data science software solutions company focused on the pharmaceutical and healthcare industries.  They offer a complete suite of products across the pharmaceutical value chain from drug discovery research to drug development to commercialization to manufacturing. A partner company to Kx, RxDataScience builds their products on the same core technology that Kx uses. This article is about its Patient Treatment Journey application.

The challenge

Longitudinal patient data is a key tool pharmaceutical companies use for drug research and health monitoring, as well as for deriving business insights. These studies can often have extremely large datasets, which pose analytical challenges for researchers.  Consider that there are approximately 280 million patients in the US representing 80% of the total population.  A single longitudinal study capturing daily health events, like doctor visits, lab tests, prescriptions, blood draws and hospital visits, will quickly generate billions of records.

Drawing on patient histories further adds to the size of the data, especially when a population has been tracked over a significant period of time. Patterns in disease progression related to environmental factors, for example, may only begin to emerge when the minimum sample size is three or four million patients and the time frame is over five years. The amount of data analyzed in a single study is often staggering.

An additional challenge for researchers is the complexity and unpredictable nature of the data itself. Because each major provider of healthcare data has their own table structure and classification system for the elements in their datasets, the process of conforming data from diverse sources into a single format can be extremely time consuming. Deep expertise is needed to understand the nuanced differences from source to source, and to know how to most effectively prepare and load the data before analysts can begin to draw meaningful insights from it.

The RxDataScience Solution

Pharmaceutical data scientists who work with longitudinal patient data typically use statistical analysis tools like SAS, R or MatLab which operate on very large datasets stored in industry-standard relational databases or distributed computing environments.

RxDataScience applications replace those backend databases with a time-series in-memory data management solution powered by Kx technology that is particularly suited for the chronological data used in longitudinal studies. Compared to traditional solutions, this data management technology has been found to be 100 times faster at a quarter of the cost. Because of its efficient design, delivery and installation takes half the time of standard solutions, and is often operational within 30 days on premises or in the cloud. The time-series technology allows new ways to combine data in multiple tables that are not supported in standard SQL.

The speed of the technology means that pharma data scientists can easily create tables and quickly switch parameters in the Patient Treatment Journey application. Features include: (i) dynamic visualization that reflects sub-population cohort selection; (ii) the ability to isolate and analyze specific starting or ending drug regimens; (iii) sophisticated business rules for defining new patient starts, switches, concomitant drug regimens and discontinuation; (iv) an insights module that identifies product share differences for key patient segments across sales regions; (v) filter and pivot functions on multiple variables to explore differences in treatment, and (vi) data extracts to facilitate additional charting and reporting options.

With the Patient Journey application, pharma data scientists have a front-end dashboard that lets them, for example, analyze tens of billions of rows of records in an insurance claims dataset broken down by patient to see how a disease progresses in different sub-groups. The underlying Kx technology enables easy, sub-notational analytics in large datasets. Flexible widgets allow them to drag and drop parameters where they need them on their dashboard.


Leading pharmaceutical companies invest tens of millions of dollars in data and technology every year. Much of that software is industry standard, mandated officially, or unofficially, by the FDA and other regulators. RxDataScience provides these companies with business solutions that complement their existing software stack.

With the Patient Journey application they are able to easily identify and quantify market opportunities where drug utilization may improve patient outcomes. It has also proven to be an invaluable tool for strategic planning, tactical promotion deployment and new product positioning.


Kdb+ Transitive Comparisons

6 Jun 2018 | , ,

By Hugh Hyndman, Director, Industrial IoT Solutions. A direct comparison of the performance of kdb+ against InfluxData and, by transitivity, against Cassandra, ElasticSearch, MongoDB, and OpenTSDB

Kx provides rapid access to unstructured data

Insights into kdb+ 3.6

5 Jun 2018 | , ,

At Kx25, the international kdb+ user group conference held on May 18th, we released kdb+ version 3.6. With this release, we are opening up new possibilities for kdb+ application developers.
It provides a lot of new capabilities and streamlines previously complex processes so that programmers will be able to take a simpler approach to writing systems compared to what they did in the past.