Three absolute truths exist about clinical research trials: 

  • If we don’t have the right data, we will not find the right patients.
  • When we don’t find patients, we miss recruitment goals, and trials can fail.
  • When trials fail, we significantly waste costs and resources and, more importantly, we deny patients much-needed treatment options.

Many oncology studies with complex criteria continue to fall short of enrollment goals. One promising solution is using Electronic Medical Record (EMR) tools for recruitment. Interrogating the EMR can be an effective and efficient method for clinical trial recruitment, but many challenges exist. Over the next few weeks, we will post a three-part series examining:

  1. The Challenges of EMR Interrogation.
  2. How and When to Interrogate EMRs to Find Patients.
  3. Tools Available for EMR Interrogation and How Best to Evaluate Them.

EMR Interrogation for Patient Recruitment: Much Potential & Many Challenges

In a 2019 review of the impact of EMR recruitment methods:

  • One study showed enrollment jump to 4.86% (176 of 3621) of those screened as compared with 2.38%.
  • Another study showed the number of patients enrolled per month doubled after the implementation of automatic EMR screening tools with automated alerts.
  • A third one showed that the recruitment yield increased from 6% to 86% when alerts were generated from EMRs.

Although there has been success, it is not a complete solution when you consider three main challenges.

1. Data Gaps

Estimates suggest that only 60% of the data needed to match patients to oncology clinical trials exists within EMRs. Many EMRs do not contain longitudinal patient data, which is important to understand eligibility criteria. With the complex oncology eligibility criteria data not readily available, research coordinators must be creative to optimize the efficiency of patient screening. It becomes more involved when you consider how increasingly complex inclusion/exclusion criteria are becoming. Add to that complexity the rapid pace of change taking place in oncology.

For instance, enrollment may require a patient to be treatment-naïve, at a specific disease stage, or have a combination of alterations in order to qualify. While many such data points are considered “unstructured data” found in physician notes or pathology reports, EMRs are designed around structured data — claims and codes. What’s missing is the actual, free-form text used by the physicians to understand an entire patient. Estimates suggest up to 80% of all EMR data includes such unstructured data, which is difficult to easily search in an automated manner.

Currently, clinical research staff must manually search patient charts for this information written in such notes, as well as manually review pathology reports to find the information they need — a daunting, labor-intensive, and inefficient process.

2. Lack of Interoperability / Conformity

There is limited interoperability among EMR systems to begin with, and some healthcare systems have multiple EMR systems, which can make it exceedingly difficult, if not impossible, to search across them. Weaving in the lack of conformity between clinical trial eligibility criteria and EMR data structures creates an environment where automated searching becomes nearly impossible, leaving clinical research staffs to, once again, revert to inefficient manual chart reviews.

3. Lack of Specific Oncology Data

Consider that NCCN recently broadened guidelines for genetic testing. Oncology continuously discovers new remedies, such as targeted therapeutics and immune checkpoint inhibitors. When evaluating patients, trial inclusion may be contingent upon on the presence or absence of a single biomarker, molecular alteration, or other nuanced genomic information. Many EMR systems lack the ability to collect such oncology-specific information, let alone be modified and reconfigured to keep up with new developments. The result is a dearth of oncology-eligibility screening data.

Challenges exist, but so do methods to more effectively interrogate an EMR. Our next post “How and When to Interrogate EMRs to Find Patients” will discuss how to fine tune and tailor processes to meet the workflow requirements and find the best timing for EMR interrogation.

© Copyright 2020 Inteliquet, Inc. All rights reserved.