Patient forecasting needs innovation since traditional methods, such as using past experience and estimates from investigators, consistently fail sponsors and CROs. In fact, data show that:

  • Nearly 80% of clinical trials don’t meet enrollment timelines.*
  • Up to half of all research sites only enroll one or zero patients.**
  • Half of all sites underperform by not delivering the number of patients they expect.*

These challenges occur in part because sponsors and CROs lack access to the kind of information that can better inform which patients can be viable candidates for complex oncology clinical trials.

For example, there is information useful that can be found in “non-discrete” EHR data. This is usually found in handwritten notes on paper (filed in a standalone chart and scanned into an EHR), voice recognition software text reports (often standalone text in an “open notes field”), and static documents scanned into an electronic filing system (i.e., interfaced records, patient’s history) within an EHR. This non-discrete data can include histology and patient performance, or molecular and pathology testing and cancer staging for solid tumors, as well as co-morbidity data – all important data points that improve patient forecasting accuracy.

Nearly 80% of clinical trials don’t meet enrollment timelines. Such challenges occur in part because sponsors and CROs lack access to the kind of information that can better inform which patients may be viable candidates for complex oncology clinical trials.

Filling Information Gaps and Improving Trial Design from the Outset
Natural Language Processing (NLP) technologies can find clinical data points that can bring a new level of insight from data oncology patients across leading academic and community institutions. These discrete and non-discrete sources include notes and qualitative data found in EMRs, biobanks, cancer registries and laboratory information and pathology management systems. Many non-discrete data points can improve digital screening and patient matching.

In addition, such real world evidence (RWE) can improve clinical trial retrospective and prospective insights to better informs future trial design, which can:

  • Improve clinical trial design from the outset. For example, knowing the age of diagnosis is a large factor that can help sponsors and CROs target a specific section of a patient population (e.g., patients under 50) , which can save time and resources.
  • Set better expectations about enrollment and potential timelines before a study opens, which can benefit sponsors and CROs up front.
  • Remove guesswork from site predictions, benefiting CROs and clinical sites.
  • Identify clinical sites that may not be appropriate for a particular clinical trial. This benefits sites and CROs by removing burdens from sites to produce patients they don’t have, saving thousands of dollars in costs associated with opening sites that will not enroll patients.

Technology is One Part of the Full Inteliquet Story
But technology, such as NLP and machine learning, is just one part of Inteliquet and our approach to improve the way clinical research is conducted. Along with innovative technology, we provide expertise across the entire clinical trial process, including clinical engagement specialists who are dedicated to training, supporting and working side by side with providers to ensure optimal success with the Inteliquet Software Platform. We also employ enhanced data and insights and expertise to help sponsors and CROs improve clinical trial feasibility, increase patient accrual rates and elevate visibility of trials to clinicians.

At ASCO 2019, we made a presentation about how a combination of cancer diagnosis with just one laboratory test found patients who could be considered for trials even though there were deemed ineligible due to a narrowly missed requirement. By using RWE at the time of trial design, we found clinical trials can be optimized to improve patient recruitment and expand availability to patients while ensuring clinical suitability.

For more information about Inteliquet and its complete role in improving clinical trial feasibility as well as initial clinical trial design, click here for a webinar on machine learning and its benefits in patient forecasting.

—–

*Clinical trial delays: America’s recruitment dilemma, Drug Development Technology. www.drugdevelopment-technology.com/features/featureclinical-trial-patient-recruitment/

**Tufts Center for the Study of Drug Development: Impact Report January/February 2013. “89% of Trials Meet Enrollment, But Timelines Slip, Half of Sites Under-enroll.”

© Copyright 2020 Inteliquet, Inc. All rights reserved.