OR WAIT 15 SECS
Study examines the growing integration of real-world data and evidence and the remaining roadblocks to adoption.
This past year has seen fervent attention on the use of real-world data (RWD) and real-world evidence (RWE) to support drug development, patient safety, and commercialization activity. The number of internal discussions, conferences, and publications focusing on RWD/RWE have increased dramatically, largely in response to the 21st Century Cures Act of 2016 requiring the FDA to implement a framework for RWE’s
role in drug development within two years.
The Tufts Center for the Study of Drug Development (Tufts CSDD) conducted a recent study to gather baseline data on current and planned uses of RWD and RWE, operational approaches that support the use of this data, and return-on-investment metrics.
Thirteen pharmaceutical, biotechnology, and contract research organizations (CROs) participated in a working group study to identify key areas of inquiry and to develop an in-depth survey instrument. And 30 distinct companies provided complete responses to the online survey.
Among the key insights from this study:
Underlying experience and operating models
Based on sponsor and CRO reports, RWD and RWE are used widely to support R&D, health economics, and outcomes research. The majority of responding companies have functions that have worked with real-world data and evidence, on average, for seven years. Nearly two-thirds (63%) of organizations report that their primary centers managing RWE reside within commercial functions, including epidemiology health economics and outcomes research and medical affairs. Slightly less than four-out-of-10 companies report that their primary RWE hub operates within the R&D division.
Most companies report that RWD and RWE use is managed centrally, but variation in operating approaches is observed between large versus small and mid-sized companies. Nearly 70% of large companies (annual sales greater than $11 billion and R&D spend greater than $2 billion) and 58% of small and mid-sized companies report that their RWE function is centralized and globally supporting other functional areas.
Across all companies, average fixed headcount dedicated to RWE departments are nearly double the average variable contract services headcount (mean of 19.1 fixed fulltime equivalents vs. 10.7 contract FTEs). Large companies have an average of 88.8 FTEs compared with 12.6 FTEs at small and mid-sized companies.
Responding companies expect a 25% increase, on average, in fixed and contract FTEs by 2020. Large companies project a 34.5% increase in staff by 2020. Small and mid-sized companies project a 16.8% increase in total FTEs dedicated to managing RWD and RWE use in that time frame.
Diverse data types
Pharmaceutical and biotechnology companies and CROs responding to the survey report using a variety of data elements to support a new drug application (NDA), including: claims data (used by 95% of survey respondents), electronic health record clinical data (71%), prescription data (67%), patient-reported outcomes data (48%), and demographic data (48%). Approximately four-out-of-10 (38%) and three-out-of-10 (29%) companies report using biomarker and genomics data and protocol feasibility data, respectively.
RWD and RWE are becoming essential to evaluating clinical and financial value. Respondent companies indicate that the foremost use of RWE is to evaluate and improve the economic value of their drug products (75%) and to strengthen product positioning in the marketplace (75%).
Companies also report using RWD and RWE to support critical decision-making associated with product effectiveness (63%), patient recruitment (50%), and improved completion of post-marketing requirements (46%). Less than one-third of companies report using RWD and RWE at this time to support portfolio decisions (34%), investigative site identification (29%), signal detection for risk management (29%), reduction in overall R&D development resource and financial investment (25%), and to capture product quality measures (21%).
Companies project the highest relative growth in the use of social media data (up 42%) by 2020, as more reliable and efficient means of gathering these data grow. Use of data from wearable and mobile devices is also expected to see higher relative levels of growth in usage by 2020, as more of these data sources are validated. Use of claims and prescription data-two relatively established and mature data types-is expected to decline during the next three years.
Six-out-of-10 companies report that the availability of RWD and RWE data poses the greatest challenge at this time. Lack of external stakeholder trust in RWE (35%) and the cost of acquiring data (25%) were the second- and third-most cited challenges. Other challenges identified include determining causation (20%), and quality and reliability of claims and electronic health record data (15%).
One-in-five organizations (20%) cite the cost and effort of data integration. This finding echoes that observed in another recent Tufts CSDD study looking at the high volume and diversity of clinical research data now being handled by the data management function. In this study, of nearly 257 unique sponsors and CROs, the majority reported that the primary electronic data capture (EDC) system is managing electronic case report form data and lab data. But all other data types, including biomarker, electronic clinical outcomes data, patient-reported outcomes data, pharmacokinetic and pharmacodynamic data, and mobile health data represent a very small proportion of the total data captured in the primary EDC. This study also found that the cycle time from last patient last visit to data lock (now averaging 36.3 days) is longer than it was 10 years ago in part due to integration and data loading challenges.
A high proportion of companies cite the lack of trust among regulatory agencies, health authorities, and payers as a major challenge to adoption. Moving forward, steps to improve receptivity and acceptance among these stakeholders will go far in helping to realize the tremendous potential of RWD and RWE.
Also, growing demand for RWD and RWE will not be fully realized given the challenges associated with the high cost and effort required at this time to collect, integrate, and use this data. Most organizations concede that they lack the in-house expertise to manage the volume and diversity of data that organizations are eyeing to support robust inferential and predictive analyses. These conditions favor the emergence of new technologies that integrate disparate data sources such as HL7’s FHIR® and the semantic web. They also favor consultants and contract service providers well positioned to assist sponsors and CROs in achieving the dynamic level of data integration required.
Ken Getz, MBA, is the Director of Sponsored Research at the Tufts CSDD and Chairman of CISCRP, both based in Boston, MA. email: firstname.lastname@example.org