Applications of Digital Health to Clinical Research

Health Update

The clinical research landscape is changing rapidly. The future will be characterized by changes in trial and study design, continued shifts in the funding environment, and new types of data and data collection methods, among other transformational developments. The chart below captures what we view as the likely characteristics of clinical research in the future across six key areas—trial and study design; data acquisition and analysis; strategic alliances; the funding environment; recruitment, engagement and follow-up; and organization and culture.

Domain Characteristic   Domain Characteristic
Trial and Study Design
 
(See note under chart for definitions of trial designs.)
  • Virtual trials
  • Adaptive trials
  • Pragmatic trials
  • N-of-1 trials
  • Clinical trials “on the fly”
  • Registry-based trials
  • Community-engaged research
Funding Environment
  • Growth in private foundation funding
  • Continued shift to industry funding and industry partnerships
Data Acquisition and Analysis
  • Passive data acquisition in electronic medical records (EMR) for pragmatic trials
  • Patient-generated health data and patient-reported outcomes from consumer digital health products
  • Widespread deep phenotyping methods
  • Genomic data acquisition and integration with clinical/phenotypic sources
  • Continuous tracking of longitudinal outcomes
  • Seamless data aggregation and normalization across multiple sites
Recruitment, Engagement and Follow-Up
  • Platforms to engage and empower participants as partners (e.g., creating communities, not cohorts)
  • Seamless recruitment (e.g., universal consent)
  • Modern patient engagement methods (e.g., mobile, interactive, real-time) and measurement of patient “study” satisfaction
  • Regular and meaningful follow-up and results dissemination
Strategic Alliances
  • With industry, technology companies and health systems
  • Across the continuum from engagement, recruitment, data collection, study implementation and dissemination
  • With local patient-advocacy organizations
Organization and Culture
  • Promotion of clinical research “hygiene”—reporting all clinical trial results, promoting data transparency
  • Development, promotion and incentives to support a culture of patient empowerment, engagement and data sharing
 

Note: Virtual clinical trials do not require travel to a clinical research facility or doctor’s office but allow patients to use mobile devices or wearable sensors to link to the study and transmit information. An adaptive clinical trial design allows modifications to the trial and/or its statistical procedures after the trial has been initiated based on observed outcomes without compromising the trial’s integrity or validity. Pragmatic clinical trials focus on the correlation between treatment and outcomes in a real-world health system practice. N-of-1 or single-subject clinical trials consider a single patient as the sole participant being observed in a study investigating the efficacy or side effect profile of different interventions. Registry-based randomized clinical trials include a randomization module in a large and inclusive clinical registry.    

Application of Digital Health in Clinical Research

Clinical research (and clinical trials in particular) are lengthy, expensive and laden with inefficiencies in multiple areas, from participant enrollment to confirming eligibility to data collection and medication adherence to outcomes reporting and regulatory approval.

While traditional technology and data resources, such as electronic medical records, clinical research management systems and clinicaltrials.gov, are necessary tools and enablers, there are many new health tech companies that are seeking to significantly enhance how clinical research is conducted. Their vision is to make it easier to match patients to studies, improve enrollment processes, increase adherence, reduce dropout rates and improve data collection. We have highlighted some of the companies working in each of the domains below:

  • Improving Clinical Study Design: Fifty percent of rejected drugs fail, because they don’t promptly follow Food and Drug Administration (FDA) feedback, adhere to FDA timelines or file forms on time. Companies such as Trials.ai, Protocols.io, ProofPilot and Medaptive Health use artificial intelligence (AI) to optimize research study design and protocol adherence, as well as open source platforms to share study protocols and engage participants.
  • Enhancing Patient Matching: A Tufts study found that two-thirds of clinical trial sites don’t meet enrollment requirements for individual trials. Companies like Deep 6 AI use AI tools to mine structured and unstructured clinical data to match participants to clinical trial criteria. With companies like PatientWing, researchers can build online recruitment platforms that are easy for participants to use (mobile-friendly, SEO-optimized, embedded forms).
  • Improving the Enrollment Process: Determining whether patients meet inclusion or exclusion criteria can be a long, expensive and labor-intensive process. Deep 6 AI, PatientWing and SubjectWell help with enrollment by determining if participants meet inclusion criteria or making it easier for patients to go through the enrollment process. Verified Clinical Trials has a database that prevents patients from enrolling in multiple clinical trials and determines if a participant could have protocol violations. While it doesn’t help with the inclusion and exclusion criteria matching problem, DocuSign makes it easier to sign documents online.
  • Increasing Patient Adherence: Forty percent of patients become nonadherent to investigational medical products after 150 days in a clinical trial. Companies like Towerview Health, Wellth, Pillsy, MedMinder, AdhereTech and Medisafe are focused on improving medication adherence with smart pillboxes or pill bottles, virtual pillboxes, or incentives based on behavioral economics. Solutions like emocha Mobile Health and AiCure use digital forms of directly observed therapy (DOT), which involves a person or an AI application watching patients take their medication. This approach results in 86-90% of participants completing their treatment on time versus 61% for self-administered therapies.
  • Reducing Patient Dropout: The average dropout rate across all clinical trials is 30%. Brite Health analyzes structured and unstructured patient data and then sends personalized messages and notifications to encourage people to continue participating in the trial. It also predicts the patients who are likely to drop out and notifies staff members, so they can intervene. In addition, companies such as ProofPilot, PatientWing and Medaptive Health have participant engagement tools. However, this area is still ripe for innovation.
  • Improving Data Collection: A tool like Trials.ai provides immediate feedback to clinical research organizers, so they can correct data collection obstacles at specific trial sites. Companies such as Medidata1 offer cloud storage, management software and data analytics services for clinical trials. Patient-generated outcomes are starting to become more commonly utilized with platforms such as Google’s Project Baseline and smartphone/wearable-based platforms. These are the building blocks for fully remote virtual trials. Estimates suggest that there is an opportunity to shift 25% of trials to virtual.2

1https://www.mobihealthnews.com/content/depth-rise-digital-clinical-trial

2https://medcitynews.com/2018/02/whats-future-clinical-trial-design-medidata-sees-blend-virtual-person-components/?rf=1