Last November, Tom Krohn announced that we would begin collaborating with Novartis and Pfizer to build a consistent framework for clinical trial eligibility criteria by developing target profiles and implementing them in our API. Our shared mission was to make it easier for people to find clinical trials that are right for them or their loved ones.
Not long after that announcement, I was thrilled to be given the task of helping to lead such an important project. And, right now, I’m excited to share with you that some great strides have been made. In the last six months we’ve consulted with our partners from Pfizer and Novartis, along with other outside experts, to determine the best possibilities for presenting clinical trial eligibility criteria in a way that’s clear and consistent. From these discussions, we were able to form our first target profiles, which were made available to the public in late July through a major update to the LCOI-API. And, we’ve only just begun…
So, What is a Target Profile?
A target profile, in the most basic sense, is a consistent, streamlined, and simplified representation of a clinical trial’s eligibility criteria which are human-readable and developer-friendly through our LCOI-API. In his November blog post, Tom explained how the lack of consistent language for eligibility criteria can make it difficult for patients who are hoping to join a clinical to find the information they need:
“Imagine you’re a patient who has just received a diagnosis: Cancer, or Alzheimer’s, or a debilitating autoimmune disease. You set out to find information about related clinical trials. You find ClinicalTrials.gov, a service of the National Institutes of Health. It contains a wealth of information, but it overwhelms you. You can’t sort out exactly which trials you’re eligible for, or how to compare them. You fear you’ll never find an appropriate study.
What is it that makes the information on the site seem so confusing? Part of the problem is a lack of consistency, especially in representing eligibility.”
So, how does the eligibility criteria on the site end up being so inconsistent? ClinicalTrials.gov provides a free-form text field for entering eligibility, which allows study sponsors to represent their eligibility criteria in a lot of different ways. The biggest challenge in developing target profiles was in creating a consistent language to use across the three different sponsoring organizations when each one tends to represent the same basic criteria in their own unique ways, multiple times across various trials.
To address this problem, representatives from Lilly, Pfizer, and Novartis, along with some outside experts, began working together to create a new and uniform way of representing basic eligibility criteria. The groups have worked very closely together to decipher the different ways eligibility is represented on ClinicalTrials.gov and to make the kinds of choices they believe will help make information from ClinicalTrials.gov more usable, more understandable, more patient-friendly, and more developer-friendly.
We started by taking a closer look at the information ClinicalTrials.gov lists as core eligibility criteria and creating what we call a “simplified subset of core eligibility criteria” from it. The types of data we chose to include in this subset and in the target profiles are those that are not screening visit related. In other words, information included in a target profile can either be easily matched to data in an electronic medical record, like Blue Button Plus, or is based on a question that a patient could answer him or herself.
Since target profiles are designed to help patients pre-screen themselves, information that can only be obtained through a screening appointment with the investigators is not included. The goal is to help get the right patients to the right trials faster. By helping them to access information that can help them understand immediately whether the trial is a good match for them, we believe that we can save them the time and inconvenience of setting up unnecessary appointments, while at the same time, helping them access to a wider range of research options.
How Do Electronic Medical Records Come Into Play?
Target profiles are set up to support matching against Blue Button Plus (HL7 CCDA) format. Blue Button is an initiative to allow patients to access much of their own health information. It’s intended to give individuals ownership of their health records so that they can regularly check them for accuracy and/or share with other healthcare providers electronically if they so choose. Since the target profiles in our API support matching with Blue Button, developers can build apps that allow patients to seamlessly match their electronic health information to open clinical trials.
Though our initial focus has been on making target profiles Blue Button compatible, there are ways a savvy developer could make the information we’re offering compatible with a variety of other online medical records systems. The information in our API is presented in a JSON format, which can be downloaded and converted into a different format and mapped with different systems. As time goes on, we will continue to look at ways to make the data even more flexible.
What Does All of This Mean for Patients?
Many companies and patient advocacy groups have already developed very effective and unique platforms to help people connect to clinical trials. But, they struggle with inconsistencies in the data they are provided through public sources like ClinicalTrials.gov. Our goal is not to reinvent the wheel by building a new clinical trial matching app. We don’t want to move patients away from communities that they know and trust. Instead, we want to work directly with those communities to help them enhance their already existing platforms, and to build new ones if needed, by enabling them with access to better data and technology.
So, if patients are using a platform that is powered by the LCOI-API and target profiles, they maybe begin to notice that their search results include more robust eligibility criteria and that they are able to get a clearer understanding of what each trial requires. They may also be given more opportunities to sign up on a platform where they can create a personal profile and allow the software to continually and automatically search for trials on their behalf.
I’m a Developer. How Do I Get Access to Target Profiles?
The target profiles that have been developed at this point are currently available through Version 2 of our API. To learn how to connect to our API and get access to the latest target profiles go to developer.lillycoi.com. The site is free and open to the public. Once you’re there you’ll be able to register and obtain a key to access the API.
One thing to note: Target profiles are still actively being developed and evolving. So, you can expect the information in the API to be added and updated on a regular and fairly frequent basis.
We couldn’t be happier with the success we’ve had so far. Each company and expert has come to the table with the same goal in mind: Making the public information available on ClinicalTrials.gov more accessible, useable, and patient-friendly. And, each member has contributed many valuable ideas and insights in the true spirit of collaboration and open innovation.
However, there’s still plenty of room to grow. The majority of our target profiles currently feature open Phase II and Phase III trials with at least one site located in the U.S. We are hoping to expand to include a wider range of trials and therapeutic areas.
We also know that even once we’ve managed to convert every Lilly, Pfizer, and Novartis study into target profiles, that still leaves out a lot of other viable options for patients. To solve this, we’re looking into ways to expand the scope of target profiles to include even more valuable information for patients and caregivers.
We are looking forward to continuing the conversations we’ve been having with our colleagues, stakeholders, and patient advocates to bring even more value and utility to clinical trial information. As always, we’d love to hear some of your thoughts on how we can effectively help patients find the right trials faster. Please leave a comment below or send us a tweet.