Data Integration: Changing the Pharma and Healthcare Landscape
Data Integration: Changing the Pharma and Healthcare Landscape
Complete the form below and we will email you a PDF version of "Data Integration: Changing the Pharma and Healthcare Landscape"
The integration paradigm fueling data insightsDespite the data explosion in the past 10-15 years, this has largely been “data unrealized” – its value has not been exploited. Subsequent investments in data integration strategies, technology, and analytics have transformed a medley of free-floating data points into an integrated, coherent message. Pharma has leveraged data integration across the value chain, from discovery through development to commercialization. Healthcare is also leveraging data integration strategies to drive value-based healthcare models.
The integration of this data has created the opportunity to delve deep into the real world and generate meaningful insights, optimize patient recruitment, develop a better understanding of various therapeutics, and directly improve patient outcomes.
Data integration is triggering significant change in the healthcare industry
Evolving business models in healthcare
Healthcare is evolving from a fee-for service (FFS) model to a “patient-centric”, data-driven ‘value-based’ healthcare model and this transition will only be possible through the effective integration of multiple data sources to assess patient outcomes, measure performance improvements, and correlate cost control measures (Horner B et al, 2019).
Interoperability would significantly ease the data integration process. Fast Healthcare Interoperability Resources (FHIR) is an open healthcare data exchange and information modeling standard created by Health Level Seven, an international health-care standards organization. It provides data formats and elements and an application programming interface (API) to drive interoperability within our highly fragmented healthcare ecosystem.
Not only has interoperability been limited owing to the disparate systems that exist, but also because some players have built barriers to data sharing. To surmount this, the Centers for Medicare & Medicaid Services (CMS) and the Office of the National Coordinator of Health IT proposed a rule in 2019 to prevent information blocking to enable patients to access their electronic health information (Terry, 2019). FHIR’s internet-based approach, called Substitutable Medical Applications and Reusable Technologies (SMART)on FHIR, utilizes web standards that allow developers to adopt a standard set of app models that would work across the different health IT systems.
Innovative projects lead the way
Data integration has been driven by various FHIR initiatives involving key stakeholders, such as the payer-focusedDa Vinci Project, and the provider-centered Argonaut initiative. Technology initiatives include Certified EHR Technology (CEHRT) and SMART APP have been launched. CEHRT, which became effective in 2019, provides an assurance that an EHR system offers the necessary technological capability, functionality, and security to help them meet the meaningful use criteria. To avoid a downward payment adjustment, health care providers are required to use the 2015 Edition CEHRT (Bresnick, 2018, CMS.gov). The SMART APP Launch provides a framework that connects third-party applications to EHRs. However, despite these initiatives, industry challenges still do exist in terms of integration with legacy approaches and shifting from transaction-oriented standards to FHIR-based interaction-oriented standards (Heath, 2016).
Medicare’s Blue Button 2.0 leverages FHIR-based APIs which allow subjects to download their personal health data directly from CMS and share it with providers or retain it for their own records. End-users such as developers, providers, and researchers can also mine that data and generate actionable insights from the same at no cost and the sandbox has been created in such a way that the data cannot be traced back to patients (Bresnick, 2018).
Data integration is triggering significant change in the pharma industry
Integration can power precision medicine and transform clinical trials
Getting the right medicine to the right person at the right time and place requires the ability to pool multiple data sources and derive meaningful insights. The importance of leveraging social media to drive patient recruitment, especially for subjects with rare diseases (Capone, 2017), and the ability to tap genomic data to identify biomarkers for oncology studies (Davi et al, 2018, Cavlan, 2018) are possible only as a result of the effective integration and analysis of the data. The increasing acceptance of using real world data from outside the restrictive inclusion/exclusion criteria of controlled clinical trials to generate crucial intelligence on the actual safety and the efficacy of investigational drugs in the real-world environment is also only feasible with smart integration practice.
Project Data Sphere, an independent initiative of theCEO Roundtable on Cancer Inc.’s Life Sciences Consortium, is an open-access, data sharing research platform launched in April 2014 that integrates patient-level data from over a hundred thousand patients involved in completed cancer clinical trials and is being used to accelerate the development of critical new therapies for cancer patients (Cary, 2019).
The NIH’sPrecision Medicine Initiative’s (PMI) All of Us Research Program is integrating patient-reported health data and EHR data of a cohort of 1 million volunteers in a centralized national database to support precision medicine research. It has launched a Data Browser tool that provides access to a data dashboard that will allow researchers to access approved health data from consenting participants (Bresnick, 2019).
Managing risks to drive recruitment
Conventional double-blind, controlled trials pose multiple challenges. Apart from the cost factor associated with multi-arm studies and the challenges associated with recruiting subjects in the case of rare diseases, there remains the ethical issue of placing subjects with serious conditions in a controlled, standard-of-care group. This has resulted in the development of synthetic control arms which pull raw data from past studies to create unbiased control arms, carefully matched at the subject level to a trial’s inclusion/exclusion criteria (Capone, 2017). The Blue Button 2.0 sandbox is another excellent data source that is being screened to generate synthetic control arms (Bresnick J, 2018).
Risk-based monitoring is an alternative to the conventional on-site monitoring approach. It centralizes a significant portion of the on-site monitoring activity and leverages analytics to span across integrated data points. It provides a helicopter view of the progression of the site’s performance and the subject’s health parameters across the course of the study. It identifies outliers and uses risk triggers to drive on-site visits on an as-needed basis. This may not only reduce study costs by 15 – 20%, but also help proactively manage study risk, add quality, and enhance data integrity (Limaye N and Jaguste V, 2016; ACRO).
During safety signal detection, safety reports from diverse sources and products are integrated through the FDA’s Adverse Reporting System (FAERS). This includes adverse event reports, medication error reports and product quality complaints resulting in adverse events. Reports come in from manufacturers, healthcare professionals and consumers. FAERS supports the FDA's post-marketing safety surveillance program. The FDA’s Sentinel Initiative is a national electronic system and the largest multisite distributed database in the world dedicated to medical product safety. It integrates healthcare data from 18 partner institutions and the meta-analysis of this pooled data allows the FDA to evaluate safety risk signals from this real world evidence (FDA, 2008, Lyapustina S, 2019).
Integration across the whole supply chain
Another critical area is the real-time monitoring of the entire supply chain. Data captured from IoT-like devices is being integrated to ensure that temperatures are controlled across the entire supply chain. Radio-frequency identification (RFID) tags with electronic product code (EPC) technology are used for the electronic tracking, tracing, and authentication of pharmaceuticals to ensure the integrity of the supply chain and detect counterfeit drugs (Syrma Technology, 2019). In fact, big data and predictive analyticshave even been utilized to analyze vast amounts of weather data to predict seasonal variations and associated conditions such as hay fever, triggering a ramp up in manufacturing and inventory control (Hughes, 2018).
Interoperability, technology and compliance – seeking a "single source of truth"These improved data integration practices result in data being pulled from multiple sources, integrated in a data lake (a raw data dump) or a data repository (an organized structured data repository where data with a pre-defined end use has been stored) and analyzed to generate new insights. The FAIR (Findable, Accessible, Interoperable, Re-usable) data principles established in 2016 help outline best practices for storing data. They focus upon the importance of metadata to allow for the traceability of data and of “interoperability” to allow for a “single source of truth” (Fox B, 2019).
While there is a lot of data to the order of exabytes being gathered by pharma today, there is a need to transition from big data to deep data. Deep data is a method not only of processing data in large amounts, but of putting raw data into context and drawing further comparisons and deriving deep insights. Integrating data from devices, wearables, EDC, etc., complemented by therapeutic area expertise, analytics and technology has enabled the industry to run truly ‘patient-centric’ virtual clinical trials (Limaye R et al, 2018).
Along with integrating data from multiple sources comes the requirement to ensure the highest standards of data governance, traceability, data quality, data privacy and data security (Walker N, 2018), while ensuring compliance with requirements such as Electronic Records and Electronic Signatures (ERES) / 21CFR part 11, Health Insurance Portability and Accountability Act (HIPAA) and General Data Privacy Regulation (GDPR) (Hughes, 2018). Machine learning, artificial intelligence and natural language processing (Limaye N and Limaye R, 2018, Walker N, 2018) are being used extensively to derive meaningful insights from large volumes of integrated data. Mixed reality, virtual reality and augmented reality are being used to create interactive data visualizations of integrated genomic, transcriptomic, metabolomic, and proteomic data to drive drug discovery in biopharma (Limaye N, 2019).
The importance of being able to derive intelligence not only from structured data, but from unstructured data such as clinical notes as well, cannot be discounted and FHIR 4.0.1 offers the capability to integrate this data (HL7 FHIR® US Core Implementation Guide CI Build).
Data cannot rest in silos. It needs to be integrated. Integration needs to be supported not only by analysis, but by interpretation as well, to derive real-time meaningful insights. As data is exploding in this day and age, cognitive computing has served as a game changer, allowing one to rapidly sift through the data and detect the needle in the haystack.
Cavlan O, Chilukuri S and Evers M. Real-world evidence: From activity to impact in healthcare decision making. (2018) McKinsey & Company https://www.mckinsey.com/industries/pharmaceuticals-and-medical-products/our-insights/real-world-evidence-from-activity-to-impact-in-healthcare-decision-making#
2. Limaye N and Limaye R. Why Pharma Is Rethinking Its Data Integration Strategy. (2018) Technology Networks https://www.technologynetworks.com/informatics/articles/why-pharma-is-rethinking-its-data-integration-strategy-the-criticality-of-data-integration-307918
Horner B, WV Leeuven, M Larkin, Baker, J and S Larsson. Paying for Value in Health Care. (2019) BCG https://www.bcg.com/en-us/publications/2019/paying-value-health-care.aspx
Bresnick J. FHIR is Blazing a Path to Patient-Centered, Data-Driven Healthcare. (2018) Health IT Analytics https://healthitanalytics.com/features/fhir-is-blazing-a-path-to-patient-centered-data-driven-healthcare
Campbell M. The All of Us Research Project: A Historic Effort to Sequence One Million Genomes. (2019) Technology Networks https://www.technologynetworks.com/genomics/articles/the-all-of-us-research-project-a-historic-effort-to-sequence-one-million-genomes-323460
6. Certified EHR Technology. https://www.cms.gov/Regulations-and-Guidance/Legislation/EHRIncentivePrograms/Certification
7. How SMART on FHIR Grew Vendor Support for Interoperable HIT Appshttps://ehrintelligence.com/news/how-smart-on-fhir-grew-vendor-support-for-interoperable-hit-apps
Walker N, The Evolution of Pharma Depends on Data Integration and Application. (2018) Pharma’s Almanac https://www.pharmasalmanac.com/articles/the-evolution-of-pharma-depends-on-data-integration-and-application
Bresnick J. “All of Us” Offers Online Tool for Participant Health Data Analytics. (2019) Health IT Analytics https://healthitanalytics.com/news/all-of-us-offers-online-tool-for-participant-health-data-analytics
14. Limaye N and Jaguste V. Risk-Based Monitoring (RBM) Implementation: Challenges and Potential Solutions. (2018) Therapeutic Innovation and Regulatory Sciences, 53.2
15. Establishing risk-based monitoring within a quality-based system as “best practice” for clinical studies. https://www.acrohealth.org/rbqm-report/
17. Lyapustina S. FDA Expands Big Data Use with Sentinel System. (2019) The National Law Review https://www.natlawreview.com/article/fda-expands-big-data-use-sentinel-system
18. RFID Prevents Counterfeit Drug Distribution. Syrma Technology (2019) Syrma https://www.syrmatech.com/drugs/
19. Hughes, A. Top 5 Trends Affecting the Pharma/Life Sciences Supply Chain. (2018) Cleo https://www.cleo.com/blog/life-sciences-supply-chain-trends
20. HL7 FHIR® US Core Implementation Guide CI Build. https://build.fhir.org/ig/HL7/US-Core-R4/clinical-notes-guidance.html
21. FAIR Principles https://www.go-fair.org/fair-principles/
Fox B. Leveraging the FAIR principles of data in pharma. (2019). Pharmaphorum https://pharmaphorum.com/views-analysis-digital/leveraging-the-fair-principles-of-data-in-pharma/
Limaye R, Kumar L and Limaye N. Fourth generation technologies in pharmaceuticals-Revolutionizing healthcare. Journal of Systems Biology & Proteome Research (Special Issue -2018). https://www.alliedacademies.org/articles/fourth-generation-technologies-in-pharmaceuticalsrevolutionizing-healthcare-9615.html
24. Limaye N. Data Visualization in Biopharma: Leveraging AI, VR and MR to Support Drug Discovery. Technology Networks. https://www.technologynetworks.com/biopharma/articles/data-visualization-in-biopharma-leveraging-ai-vr-and-mr-to-support-drug-discovery-320108