Leveraging Data to Improve Patient Recruitment Pain Points

Clinical Researcher—August 2021 (Volume 35, Issue 6)

RECRUITMENT & RETENTION

Jason Bhan, MD

 

At every level, data silos remain commonplace throughout healthcare.{1} These silos represent unconnected repositories of segmented information that not only complicate communications across organizations and health facilities, but can also slow or prevent the successful recruitment process for crucial clinical trials. Identifying the gaps and bringing these data together are critical to overcoming isolated silos and improving clinical trial outcomes.

Clinical research is often an expensive and time-consuming undertaking. When a company develops a new drug, the average cost is estimated at $2.6 billion.{2} Even at that price, only a small fraction of drugs under development eventually gain regulatory approval. A significant portion of the company’s tab comes from conducting clinical trials, especially if a contract research organization (CRO) involved in the process is forced to restart recruitment efforts due to an initial lack of patient retention.

Further, the phenomenon of participants dropping out of clinical trials isn’t rare; typically, 30% of patients withdraw from a study before its completion.{3} It typically costs organizations $6,533 to recruit one patient to a clinical study, and recruiting a new patient if one withdraws costs $19,533.{4} Harnessing the power of real-world patient data can help sponsors and CROs minimize delays, increase participant retention, rein in costs, and increase operational efficiencies.

The recruitment phase of a clinical trial is probably the most important differentiator among CROs, which are continuously seeking more sophisticated and accurate methods to identify patients. Leveraging data and analysis technologies provides that insight. Technology analytics can help process and survey data to make it easier to recruit and retain the right patients for the right trial, ensuring lower costs and faster results.

The Importance and Challenge of Data Access

In addition to the costly process, the average clinical trial generates up to 3 million datapoints{5} that typically remain siloed and inaccessible. Clinical trials aren’t alone in this; there are more than 6,000 hospitals in the U.S.{6} and each implements its own, specific data record management system. Within these databases, fields are labeled differently and extracting the data isn’t easy. This is true of any electronic health record (EHR), whether in a hospital system, laboratory, radiology department, or medical office.

A survey conducted by the U.S. Office of the National Coordinator for Health Information Technology{7} found that nearly 70% of hospitals reported integrating patient data into their EHRs, but only one in 10 facilities used exclusively electronic methods to send or receive secure electronic health messages to and from outside organizations. Enabling global data flows and greater interoperability would liberate information across healthcare systems. The right data analysis technology enables CROs to leverage real-world patient data, which translates to improved quality performance, advanced care for patients, and more accurate participant recruitment for clinical trials.

The COVID-19 pandemic has demonstrated at scale the inadequacies within EHR systems in the U.S.{8} The inharmonious electronic records left Americans vulnerable as teams worked to target outbreaks, chronicle patient recoveries, and identify which treatments were widely successful. Pulling COVID-19 patient data from EHRs in hospitals across the U.S. proved a technical nightmare,{9} largely due to the various EHR software being unable to retrieve and integrate information with competing software designs.

Realizing the pandemic called for a more centralized data system, the National Institutes of Health created the National COVID Cohort Collaborative, which extracted siloed data to answer key questions and track successful treatments. It houses 6.3 million de-identified patient records from 56 institutions and counting. Experts believe it is one of the most promising tools for studying the disease now and in the future.{10}

To make clinical research and recruitment as effective as possible and develop the most widely successful treatments, CROs need to abandon the inflexible, non-interoperable, siloed data approach of the past and gain consistent access to real-world data.

Breaking Down the Silos Ensures Greater Success

Powerful analytics tools can mine millions of EHRs, claims data, and other resources to home in on specific patient groups faster than other methods. These Health Insurance Portability and Accountability Act–compliant solutions allow CROs to break down the data silos and easily integrate with systems to help ease clinical trial burdens. Clinical investigators can then find patients that meet their trial criteria more easily, which speeds up the timeline and increases the odds of recruitment success.

Once the organization resolves the challenge surrounding data access, it can begin to review patient data while patient recruitment is organized into two key parts: 1) the criteria for the target patient; 2) locations where key patients can be recruited.

Determining inclusion and exclusion criteria for clinical participants is a key practice when designing high-quality research trial protocols.{11} To determine the inclusion criteria for the trial, investigators will narrow down the key features of the target population needed to conduct the trial.

Typical inclusion criteria include diagnosis, demographic, clinical, and geographic characteristics. In contrast, exclusion criteria are the characteristics of potential participants that could interfere with the success of the study. The most common exclusion criteria that disqualify otherwise eligible patients include traits that could make them highly likely to be lost to follow-up, miss scheduled trial appointments, or have comorbidities that could increase their risk for adverse effects. During this process, it’s crucial to make sure the inclusion and exclusion criteria give the team a more accurate funnel from the start.

Inclusion and exclusion criteria impact the team’s recruitment funnel and it’s important to have an understanding of them up front. Many times, teams are collecting these data blind and cannot incorporate insight from real-world patient information. Without being able to broadly sort through population data, research teams end up making the funnel too narrow from the start, thereby limiting recruitment options. Gaining access to laboratory or healthcare data that are more readily accessible can be a game changer and open doors to larger recruitment opportunities.

Once the necessary criteria have been selected, teams must begin considering how they’ll find the patients. This may present as a highly complicated problem, especially without access to broadscale, real-world data. However, once CROs have access to de-identified patient information, the recruitment team can begin locating areas with qualifying patients. Investigators may not be able to identify patients down to the individual level, but can identify doctors who have high concentrations of the appropriate patients.

To gain the most impact from the accessible data, investigators can connect with the identified doctors to enroll their facility as a clinical trial site or patient recruiting center. This approach allows doctors to directly connect with their own patients and recommend the trial to those who are the best fit.

Impact of Real-World Data on Clinical Trials

The use and application of real-world data and advanced analytics is increasingly important to inform clinical trial work. Enhanced analytical models inform faster decisions around trial execution, which saves time and money. By breaking down the siloed real-world data using powerful analytics tools, clinical trial coordinators can gain access to the crucial data necessary to develop the strongest roster of clinical participants.

A clinical trial’s recruitment phase plays the largest role in determining if the study will succeed. CROs continue to seek sophisticated methods for leveraging data. Analysis technology and real-world patient data provides clinical trial investigators with more transparent access to the team’s recruitment funnel. They still may not be able to review data down to the individual level, but the teams can highlight geographical areas and healthcare facilities with qualifying patients.

By gaining this insight, CRO teams can establish connections with the doctors, hospitals, and patients who most need access to the trials. In this way, clinical trial coordinators can better ensure the retention of their patient participants and develop better results for the healthcare community and world.

References

  1. https://www.healthitoutcomes.com/doc/how-to-bridge-the-gap-between-healthcare-data-silos-0001
  2. https://cen.acs.org/articles/92/web/2014/11/Tufts-Study-Finds-Big-Rise.html
  3. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3684189/
  4. https://pubmed.ncbi.nlm.nih.gov/30264133/
  5. https://www.fiercebiotech.com/sponsored/how-to-make-most-your-clinical-trial-data-all-it
  6. https://www.aha.org/statistics/fast-facts-us-hospitals
  7. https://www.healthit.gov/sites/default/files/page/2021-03/Hospital%20Use%20of%20Certified%20HIT_Interop%20v10_1.pdf
  8. https://hbr.org/2020/06/its-time-for-a-new-kind-of-electronic-health-record
  9. https://www.statnews.com/2020/03/12/covid-19-huge-stress-test-electronic-health-record-systems/
  10. https://www.beckershospitalreview.com/data-analytics/how-the-pandemic-forced-the-us-to-centralize-its-healthcare-data.html
  11. https://assessment-module.yale.edu/human-subjects-protection/protocol-design-inclusion-and-exclusion-criteria


Jason Bhan, MD,
is a family physician and serves as the Chief Medical Officer at Prognos with a focus on the applications of technology to healthcare and medicine. He previously worked with Clinovations and managed several projects for large hospital systems involving EHR implementations, outcomes measurements, and data analyses.