Innovations in Patient Matching

Two years ago, Applied Clinical Trials looked at the technologies intending to close the ever-elusive patient recruitment gap. In that time, other innovative approaches have emerged, and while there are no hard and fast numbers that any have in fact sped up the recruitment process, many have re-imagined electonic health records (EHRs), and offer insight into the future of personalized medicine, as well as the physician’s role in clinical trials.

At HIMSS16 in March, Microsoft announced the winners of its Health Innovation Award. Among them was West Ridge Obstetrics & Gynecology and Patient identification Platform, or Patient iP, which won in the category of Building the Intelligent Cloud. The Platform securely de-identifies and aggregates EHR data so that clinical trial protocols can be automatically processed to more quickly identify where and how many patients match the inclusion/exclusion criteria requirements.

Michael J. Margiotta, CEO, said the inspiration for Patient iP came from his long history in healthcare IT, close familiarity with the strengths and limitations of EHRs/EMRs, and observing that the clinical trial recruitment system was broken because of manual processes.

Margiotta says, “EMRs are just a repository of patient data. Those systems don’t capture data in a way that can be aggregated or analyzed and perform data mining on the patient populations.” According to Margiotta, even the top three EHR vendors-Epic, Cerner and Allscripts-still face significant challenges in the markets they serve. “65% of all larger institutions, your typical study sites, can’t get data usage rates or analytics, even in a de-identified manner, to do anything with the information they have in their EHRs.”  

As the healthcare industry moves into Meaningful Use 3, significant changes around the use of EHR and the disconnect between its use and promise are expected to be addressed. Most of that is out of scope to this article, suffice to say that EHR interoperability and data sharing potential are considered a limitation. And that is where Margiotta stepped in-to provide a platform that would be able to leverage EMR data in a way the software currently can’t. In 2014, he launched his company to be able to match patients to specific criteria based on aggregated information including genetic markers, blood values, medications, and more to find those exact patients very quickly. Think of it as an EMR booster.

For CROs and sponsors, they can use Patient iP for protocol modeling-making sure patients actually exist for the protocol they have designed; as well as site feasibility-are there patients in their sites that are applicable to the study.

For sites, it works two ways. If the site is already conducting clinical trials, they can quickly know how many patients in their networks are potential participants through the EHR. And, as in the case with the West Ridge Obstetrics & Gynecology, find out how many patients in their practice were eligible for a current protocol and then decide if they wanted to join the world of clinical research. They could pursue a study to offer cutting edge therapies to their patients, expand their revenue potential with clinical studies and/or ensure that current patients weren’t referred out of their care into a clinical trial.

The application for Patient iP spans clinical trials. For example, Margiotta says, an accountable care organization could use the platform to identify an at-risk population with the potential of going to the next stage of disease, or a healthy population at-risk to a condition, and intervene by applying patient matching and inclusion/exclusion criteria. Others that receive grants, would be able to show actual successes from the aggregated data in the population. And also the ability to share clinical outcomes and aggregate successful clinical care options within a network.

But for now, Margiotta says they will prove their platform in the clinical trials world. “It seems the pre-process of the clinical trial is the toughest part and the steps leading to success are manual,” explained Margiotta. “We have bridged the gaps between pharma, CRO and sites and automated it for success.”

Another solution that incorporates EHRs in the mix is ePatientfinder. ePatientfinder crossed our radar later in 2014 and Tom Dorsett, President and CEO, explained the history of the company that he founded in 2010. Since that interview, the company has experienced growth among both its clients in the life sciences world and its network of EHR partners, which provides the access to physicans and patients. Dorsett believes that though many solutions for patient recruitment in clinical trials have emerged, there exists a lack of actionable models for getting those patients into clinical trials. And here is where his solution comes in-to provide what Dorsett calls “the last mile solution”-the puzzle piece missing between pharma, CRO, medical device sponsor, the EHR, the physician, and the patient.

That last mile, as explained in the previous article, is ePatientfinder’s three-tier funnel or level of screening to find the highest quality of referrals. The funnel includes ePatientfinder sending potential trials with patients to a physician through the EHR. If the physician opts in, ePatientFinder reaches out to patients initially to see if they are interested, then provides an IVR pre-screen survey to uncover any subjective issues that may not be in an EHR. Those patients are then referred to the opted-in physician for a consultation to see if they can go into the clinical trial.

According to Dorsett, the platform builds on the trust inherently found between patient and doctor, and is a process that keeps the physician in the drivers’ seat, which Dorsett says they appreciate. In addition, the company has been achieving the best quality referrals to sites, and has feedback from the sites themselves that the three-tier screening provides very high conversion rates.

Recently, Tufts CSDD released results of a survey it conducted on the attitudes and practices of physicians and nurses and clinical trials. The top three reasons these healthcare providers gave for not referring patients into trials were “lack access to the information,” “unsure of where to refer” and “not enough time to learn.”

Dorsett explained that their solution hits each of these points, and more, through its business model. “We are an information delivery platform, we educate physicians on trials, and then they decide whether they want to refer their patients or not.” Dorsett’s team curates information, including geographical site locations, so the information given to physicians on the clinical trials is concise, specific, available to their patients, and is tailored to that physician’s proclivities. “We do all the heavy lifting on their behalf.” Which also hits on the remaining four reasons why HCPs don’t refer patients into trials: “trials are not ‘appropriate,’” “lack of time to discuss,” “proximity to research center” and “fear of losing a patient.”

The company recently raised $8.2 million in Series B funding for its future growth plans, increasing the company’s total funding to $11 million. Dorsett said that it’s taken three years to build the partnerships and technology together to make it work for them and their clients. Future plans include other data services such as site selection, and protocol design and refinement, an expansion into Europe in early 2017, and an extension of therapeutic areas.

Dorsett says the platform, which focused on chronic diseases initially, has perfected its processes enough so the technology can now address oncology. He explained, “Oncology is nuanced. You have to get the patient when they present with the disease, when they’ve been diagnosed. So it wasn’t a large pivot for us, but we did need to make a few technological adjustments.” Many of ePatientfinder’s clients, as well as numerous cancer centers, have been asking for the service, and Dorsett confirms they very excited to be undertaking oncology in Q4 2016.

While getting patients matched and into all clinical trials is a challenge, it’s been shown that recruitment for oncology studies is very low. In March 2016, MolecularMatch, a cloud-based, clinical informatics company that works with labs, hospitals, genomic cores and physicians to connect cancer patients to treatment options, launched its MM LAB software  This software allows pathology labs and others to match patients’ test results to personalized cancer treatments, including clinical trials and experimental drugs.

MolecularMatch offers a public-facing web site for people looking for oncology treatments, searchable by diagnosis, specific gene mutation, comorbidities and more. The data behind the search is culled from web-based information sources including clinicaltrials.gov, registries, institutions, PubMed abstracts, COSMIC and more. It is fully automated to create structured data from unstructured sources.

According to Xuan Shirley Li, PhD, Chief Scientific Officer of Molecular Match, the MM LAB software was a natural next step for the company’s offerings. MM LAB generates a customized report based on the specific markets that come from tumor testing. “The variant data is quite important for all patients and that comes from diagnostic labs. Labs using our software can offer physicians and patients a tailored report with more information on specific trials and treatments that is culled from our data.” She continued, “We can align what labs generate to what physicians can offer their patients.” Basically, for labs, the software can be used to generate a value-add service for those physicians or health networks.

Li told Applied Clinical Trials, “Diagnostics traditionally found only what you were looking for. But now with Next-Generation Sequencing (NGS), it also comes with discovery.” She explained that the need for information management in the lab is a direct result of NGS testing. “For patients that relapse or have an aggressive cancer, they are looking at NGS testing and larger work-ups on the lab side. Targeted treatments for gene mutations are mostly in the pipeline, which is where we see clinical trials as the best option. Physicians need to consider clinical trials in complex cancer cases or patients looking for new options if the Standard of Care pathways are not compatible with their life or lifestyle.”

Also in the realm of oncology is Quintiles’ precision enrollment model, which is comprised of a network of more than 80 U.S.-based oncology centers, designed to speed up recruitment using pre-identified patients based on study and biomarker criteria, across broad geographic areas, using electronic health records (EHRs) and other data sources. In this newly-launched model, patients upon entering the network have their tumors tested. The genomic analysis and alterations of these tumors are reported back to the patient and site and can be matched to protocols using the genomic alteration criteria for the protocol. It isn’t until a patient is identified that the site is activated. In this article, Jeff Ventimiglia, Director, Site & Patient Networks, Quintiles, explains that study start-up is reduced because the site previously joins the Quintiles network and fills out all the documentation and service agreements and joins the Quintiles Infosario Site Gateway. A site is activated once the patient is identified and the remaining start-up activities take 21 days.

Ventimiglia shared in the article the results of a small-scale pilot study targeting 50 metastatic colorectal cancer (mCRC) patients, sponsored by Quintiles, suggests that genomic profiling may increase clinical trial participation among cancer patients from the current level of 3-5% to as much as 35%. This was due to treating physicians recommending a clinical trial in 35% of cases that reported actionable mutations. This pilot suggests that there is potential to increase screening rates and shorten timelines for clinical trials by providing a broad genomic panel rather than using a single biomarker.

WIRB-Copernicus ReferralPlus. In April 2014, ePharmaSolutions announced positive results of studies using its patient matching and triaging solution, ReferralPlus, its tool to match patients who disqualify for one study with other studies they might qualify for using a geo-therapeutic matching algorithm listed on its technology. The pilot data came from three pharmaceutical companies who listed their studies on the CenterWatch web listing service and used the ReferralPlus creening and matching solution.

• The positive results included 33% of the total patients who screened for one study were eligible for and referred to a study site

• 18% of patients who disqualified from the first study, pre-qualified and were referred to another study

• 50% of patients who disqualified for any of the studies listed, registered to be contacted about future studies

In September 2014, ePharmaSolutions was acquired by WIRB-Copernicus. As of this writing, the company does not have an update on the pilot as it was on tap for focus in 2017.

ICON/IBM Watson. In September 2015, ICON announced that it would be using IBM’s Watson Clinical Trial Matching in its breast, lung, colon and rectal cancer trials. In the initial six-month pilot, ICON was using the matching in 25 studies in those four oncologic areas. The ICON sites would have access to Watson and Watson has access to the inclusion/exclusion criteria for those studies and access to the records at the sites. Watson will use the records and inclusion/exclusion criteria to match the up with the EMRs in its database. The pilot is concentrated in the Midwest, primarily because of Watson’s access the Mayo Clinic and Cleveland Clinic’s records. Initial feedback on the pilot was expected by end of first-quarter 2016, however, at the writing of this article, the company reported that the pilot program was still underway, and then ICON and IBM will review the results and discuss next steps.