COVID-19 RESPONSE  Learn More

2015: The State of Big Data for Patient Recruitment, Part 1

by Dan McDonald

Focus on: Study Feasibility, Site Identification, and Patient Recruitment
A 3-Part Series

It is well known that successful patient recruitment is part art and part science. Though, as we are seeing today in many other aspects of health care, the science is evolving at a much faster pace than the art.

Healthcare companies, insurance companies and employers are leveraging the expertise of behavioral economists, and information contained in claims databases, electronic medical records and a variety of other data sources to understand more about patients. This knowledge base is helping us understand things like:

  • where patients present for care
  • how they learn about treatment options
  • how they are referred to a healthcare provider
  • why they make decisions for or against pursuit of treatment options
  • and much more

In addition, there is a vast and growing spectrum of internet-based pathways that can be utilized in a push-and-pull manner to drive traffic to study sites. In aggregate, there is a mountain of interesting ways that companies are leveraging science to improve the efficiency of clinical trials.

A solution that has existed for more than a decade, but has improved dramatically in terms of richness, affordability and adoption, is the use of health insurance claims data. Specifically, in the process of using such data in the planning and execution of pivotal-stage clinical trials. The use of this data is helping to address several long-standing deficiencies in that process. Here we address some of the most common deficiencies and uses:

– Protocol Feasibility: The scientific and medical leaders on the protocol development and planning teams of biopharmaceutical companies are using claims data to determine the impact of adding certain inclusion and exclusion criteria on patient availability. Not only are companies able to look at diagnosis coding (ICD-9 coding), but they are also looking at procedural coding/claims, prescription coding/claims and diagnostic coding/claims. For example, if patients cannot have been on a certain medication in the previous twelve months, how does that impact the availability of the target patients? The data can help answer those and similar questions so that sponsors understand whether or not their protocol is realistic.

– Geo-Targeting: Through examining where claims are being filed geographically and which physicians and hospitals are filing those claims, you essentially understand where patients are located who have certain diseases and conditions. If you continue to connect the dots, you also can understand pockets of disease prevalence from a geographic perspective. This is invaluable from a patient recruitment standpoint, especially as it relates to the use of broad-based marketing channels such as radio, television, print and also internet-based advertising. Using this data helps to minimize the risk of wasted advertising dollars, historically a big driver of cost on certain recruitment programs.

big data for clinical trials

 

In a couple weeks, I’ll look at several other areas where the use of insurance claims data (and increasingly, electronic health records/electronic medical records) is being used to improve the effectiveness of patient recruitment initiatives. [View Part 2 Here]

 

I recently hosted a podcast on the evolving perspectives and uses of big data. I interviewed Bill Gwinn, VP of Clinical Infomatic Solutions at OptumInsight. If you’re interested, I encourage you to give it a listen!

You may also like

Leave a Comment