Life Sciences (Pharmaceutical, Biotechnology and Medical Device) and Healthcare (Providers and Payors) organizations need to be able to handle and analyze many diverse sets of longitudinal health data in order to manage long-term and often complex patient health journeys and optimize outcomes. Complex and mission-critical data models for demographics, clinical and medical records, diagnoses, therapeutic treatments and medications, laboratory results, institutional resources, and more are needed to handle these data. Numerous, industry-specific disparate systems have been developed over decades to manage these data, leading to significant challenges in inter-industry and extra-industry interoperability and data sharing. Even with the development of industry-focused data standards, such as Clinical Data Interchange Standards Consortium (CDISC) for Life Sciences and HL7's Fast Healthcare Interoperability Resources (FHIR) for Healthcare, data sharing challenges remain in legacy systems, databases and standards. Clearly, new methods to manage and share these data silos are needed.
Cambridge Semantics' Anzo platform bridges these data challenges with the implementation of enterprise-scale Health Sciences data fabrics. A Health Sciences data fabric integrates and unifies complex data structures and formats from Life Sciences Research and Development (R&D) systems, healthcare provider Electronic Medical / Health Records (EMRs/EHRs), healthcare payer Claims and Prescription data, and various industry data standards. These data fabrics enable powerful exploratory analytic and support many data-on-demand applications, including Artificial Intelligence / Machine Learning (AI/ML) techniques.
This blog post is one of a 3-part series. Please visit the other posts to get a complete picture of how two industries, relying on similar data, manage their data and have implemented ways to share it between industries.
Part 1 – What health data and why share it?
Part 2 – Health sciences data landscape – now and future
Part 3 – Addressing the remaining data challenges
Part 1 - What health data and why share it?
At the core of both industries is the patient, their health concerns, and possible treatments leading to better outcomes. Improving patient health outcomes requires lots of data. The Life Sciences and Healthcare industries collect, manage, and analyze vast amounts of patient data in different ways. In this blog post, we review the types of data each industry focuses on and summarizes the benefits of sharing data between both industries.
What health data?
There is a wide variety of data focused on the typical patient journey – from the initial complaint to an evidential health outcome. Industry-specific uses of these data help improve individual patient outcomes and lead to more effective standards of care for patient populations.
Starting with a patient complaint, care providers and health insurers (payors in the US) record the episode and begin a cascade of data capture activities in several systems such as Electronic Medical Records (EMRs) and Insurance Benefits and Claims Management. Patient demographic data are verified along with an accumulation of diagnostic data including medical history, family history, social determinants of health (lifestyle), lab results and genomics leading to a treatment plan. Upon treatment, additional data are collected based on Standard of Care (SoC), recommended interventions for therapies and medications, additional lab results and expert opinions and notes, and the monitoring of adverse reactions, if any. The patient's overall health and wellbeing outcome is monitored for Quality of Life (QoL) and ongoing adverse reactions. And finally, cost of care data accumulates throughout the entire patient journey, is aggregated, and analyzed by both provider and payor and used to refine treatment and insurance plans.
Healthcare industry data
During the patient journey, data are collected, aggregated, and analyzed by health insurers (payors), healthcare providers and laboratories using many different systems. At the center of these data collection systems is one or more Electronic Medical Record / Electronic Health Record (EMR / EHR) solutions. Individual patient records are entered, reviewed, and assigned industry standard codes, such as ICD10 procedure codes, to help unify data and improve data quality. Provider facility systems are used to monitor the use and consumption of facility resources and record patient costs for submission of claims to a payor. Data in these systems are often duplicated and in different formats – truly siloed data.
Life sciences industry data
Although the Life Science industries' primary goal is to improve patient outcomes, the process of interacting with patients is more proactive and does not start with a patient-initiated complaint. Instead, patients are recruited to participate in clinical trials for specific indications within a therapeutic area based on their demographic, medical history, lifestyle, and genomic profile. The clinical trial protocol defines the number of patients, their overall profile based on inclusion and exclusion criteria, intervention regime according to their assigned study arm, and overall trial endpoint goals. Through the clinical trial process, Life Science companies expect to prove to regulatory agencies that their new or improved intervention is safe, is efficacious, and establishes a better Standard of Care. Clinical trial unique data are specifically defined for diagnostic and treatment activities are collected in Electronic Data Capture (EDC) systems for every patient visit and reviewed for further analysis. Data aggregation and analysis are completed on the captured data to determine if protocol endpoints have been attained and culminate in either a successful trial (endpoints are attained) or an unsuccessful trial. Even if unsuccessful, useful scientific and medical information is obtained and may be useful for subsequent trials or new interventions.
The time and cost of setting up, executing, analyzing, and reporting on clinical trials is substantial. Numerous advanced data management techniques have been applied to improve clinical trial efficiency including the use of longitudinal healthcare as Real-World Data (RWD).
Why share it?
Although healthcare providers and payors largely have their own self-contained systems and data, there are several reasons to consider utilizing Life Science clinical data. For the healthcare industry, the benefits are:
Better Patient outcomes. New and improved interventions are constantly being developed, leading to better / faster Standard of Care (SoC), increased drug efficacy and reduced Adverse Reactions.
Lower costs. Improved interventions can often provide substantial savings for patients in terms of reduced frequency or dosage and fewer adverse reactions requiring additional treatments.
Translational Medicine. Greater exposure to ‘bedside' outcomes from leading practitioners involved in clinical trials and Key Opinion Leaders (KOLs).
Better Payor acceptance. Improved interventions leading to lower costs or improved outcomes often become preferred treatments by healthcare insurers.
Off-label Therapies. Although not usually recommended, off-label interventions can provide novel treatments when usual Standards of Care are not effective for an individual patient.
For years, Life Science companies have conducted standalone clinical trials to carefully control study conditions and collect trial data. However, pervasive adoption of EMR/EHR systems and the time and expense for these trials have led to increasing use of longitudinal healthcare data with some of the following clear benefits:
Real-World Data (RWD) / Real-World Evidence (RWE). Longitudinal healthcare Real-World Data (RWD) for drug interventions typically underperform clinical trials with overstated efficacy and understated adverse reactions. RWD shows patient populations are sicker, with more comorbidities, and have more complex medical histories, with more concomitant medications than carefully sampled clinical trial cohorts. Life Science companies are researching RWD to understand the drivers in these data disparities and are using insights gained to develop future clinical trials.
Translational Medicine. The focus on ‘bedside' health outcomes in translational medicine is driven by RWD and collaboration with front-line healthcare providers.
Observational / Retrospective studies. Observational and retrospective clinical studies driven by RWD are providing excellent patient cohorts analyses and enable early screening of potentially promising new drug interventions.
Support Health Economic & Outcomes Research (HEOR). Longitudinal health data are enabling research into population, economic, and outcomes for long-term treatment of chronic conditions and quality-of-life assessments.
Outcomes-based Reimbursement. RWD offers the opportunity to develop results-oriented drug reimbursement options based on health outcomes and an estimate of lifetime cure pricing versus the price of ongoing treatment of symptoms.
The need to share data across the healthcare and Life Science industries is clear and compelling. Collaboration will directly benefit patients with improved treatment options at better prices. However, data interoperability within and across industries remains challenging and will be investigated further in the next blog posts in this series.