How Real-world Evidence is Changing Evidence-based Medicine
According to Dan Riskin, MD, CEO, and Founder of Verantos, in a Forbes article, “The system of evidence-based medicine, representing the foundation of modern healthcare, is crumbling. He points to traditional methodologies’ unsustainability and failure rates to evidence-based medicine. He continues, “Fortunately, a new pathway is available in the modern world of digitized health data, powerful artificial intelligence technology, and scalable compute power.” Real-world evidence is key to a more robust, less costly, and more inclusive approach to better healthcare through research and a sustainable life sciences market. In this insight, we will cover the recent drivers of RWE adoption, provide a concrete definition, and share five trends impacting real-world evidence use.
Why is Real-world Data Needed for Evidence-based Medicine?
Within the pharmaceutical sector of life sciences, creating new and safe medicines is fraught with challenges. The time it takes to develop new therapies takes an average of 17 years, costing $2.5 billion. After all that investment and effort, only 50 percent of patients adhere to their prescription in the real world. Plus, their failure rate is incredibly high, with under 10 percent of medicines making it into Phase I trials. And only ~14 percent receive FDA approval. Couple that with 80 percent of clinical trials that derail for failing to meet enrollment timelines and requirements.
Traditionally, regulatory approvals of new drugs are primarily based on randomized clinical trials (RCT). RCT studies mandate subject inclusion and exclusion eligibility criteria to ensure homogeneity and representativeness within the study findings. This can exclude patients with multiple comorbidities because inclusion can compromise data validity. While this provides rigor to the clinical evidence, it does limit insights into how treatment could affect population subsets.
Real-world Evidence Benefits
Unlike most industries, pharma companies are struggling to design trials that are scalable and repeatable. Most industries use large datasets and apply artificial intelligence to optimize business operations. However, with drug trials, the data is rarely touched once the single trial is completed and submitted.
Real-world evidence (RWE) studies are growing in adoption because of the added efficiency and accuracy of using real-world data. These studies demonstrate value in planning, recruitment, trial design, and post-marketing surveillance for usage and safety data.
What is Real-world Evidence Created from Real-world Data?
- Real-world Data (RWD) relate to patient health status and the delivery of healthcare that is routinely collected from a variety of sources.
- Real-world Evidence (RWE) is clinical evidence about the usage and potential benefits or risks of a medical product derived from analysis of RWD.
RWD comes from a number of sources, including EHRs, claims and billing information, product and disease registries, patient-generated data, and data gathered from other sources that provide health status information. Some refer to RWD as administrative byproducts of care delivery that can be repurposed for research. RWE may be generated from various study designs or analyses, including randomized trials, large simple trials, pragmatic trials, and observational studies.
Recent Drivers for RWE & RWD
Numerous factors have accelerated the adoption of RWE, including an industry-wide shift from volume- to value-based payment models, personalized medicine, and the need to adapt clinical trials during the pandemic. These factors have influenced regulatory bodies globally and fuel interest in using RWE to “understand and demonstrate the value of pharmaceutical and medical device innovations.”
The 21st Century Act of 2016
The bipartisan 21st Century Cures Act, signed by President Obama in late 2016, charged the Food and Drug Administration (FDA) with evaluating the expanded use of RWE, “including its potential to support the approval of new indications for previously approved drugs.”
The act established an RWE program within the FDA to support several objectives:
- Generating hypotheses for testing in randomized controlled trials
- Identifying drug development tools, including biomarker identification
- Assessing trial feasibility by examining the impact of planned inclusion/exclusion criteria in the relevant population, both within a geographical area or at a particular trial site
- Informing prior probability distributions in Bayesian statistical models
- Identifying prognostic indicators or patient baseline characteristics for enrichment or stratification
- Assembling geographically distributed research cohorts as in drug development for rare diseases or targeted therapeutics
RWE and the FDA
According to the FDA, RWE trials are crucial to understanding newly-approved products in real-world settings, providing clinical evidence on long-term safety and drug efficacy. In September 2021, the FDA published draft guidance on the RWE to support regulatory decision-making. This was an essential step toward a nationwide plan to achieve the “accuracy, completeness, provenance, and traceability” needed to generate quality evidence from RWD.
RWE trials in 2021 reached the highest number to date. The Global Data Clinical Trials database listed 194 RWE trials and 160 trials in 2019. The FDA also published an analysis of 90 examples using RWE to support regulatory decision-making in MedTech.
What’s the Potential for RWE?
According to a 2021 RWD panel, “The holy grail is not just to extend medication indications and labels, but to also get an approval for new medications using RWE sources, as it has the potential to be cheaper and to better account for real-world practice than conducting an RCT (randomized controlled trial).” Investigators would like to have a feedback loop where patient results and usage could be helpful for clinical decision support.
Ultimately, RWD could revolutionize every clinical research stage, from trial design to outcomes measurement. High-quality information could be used to determine which treatments should be approved, reimbursed, and prescribed. RWE could also drive new standards for new sources of high-validity and quality evidence to be adopted by regulators, payers, and providers.
RWE Market Size
The global market for real-world evidence solutions was valued at $37.2B in 2020 and is expected to grow at 7.6 percent CAGR between 2021-2028. Regulatory bodies’ support for RWE solutions and an increase in R&D spending are anticipated to boost market growth.
RWE Critical to Success of Future Randomized Clinical Trials
Compared with traditional RCTs, RWE is playing a greater, complementary role, providing needed information throughout the product lifecycle. International guidelines state that “RWE generated by pragmatic trial that embeds randomization within EHR and claims data could provide insight into post-market safety issues, inform clinical care practices and avoid adverse events.”
As RWE is used increasingly to inform regulatory decisions, it becomes clearer that RWD is necessary to understand what happens when therapies are used with patients in everyday life, indicating effectiveness in a broader range of patient cohorts than is achievable with RCTs. RWE can complement internal validity efforts, and using both can bridge the “clinical-trial-to-real-world divide.”
Top Trends in RWE & RWD
A 2020 industry survey report highlighted key trends related to RWD and RWE:
- Partnerships: More than 80 percent of companies indicated they are entering into strategic partnerships to access new sources of RWD.
- Investment areas: Almost all companies expected to increase investments in talent, technology, and external partnerships to strengthen RWE capabilities.
- Maturing capabilities: 90 percent reported they have already established or are currently investing in building RWE capabilities for use across the entire product lifecycle.
- However, only 45 percent felt that their capabilities are mature enough to do so.
Here are six key trends we see emerging or gaining traction now:
Trend #1: Expanding Where and When RWE is Used
Regulatory approvals: Like other healthcare areas, the pandemic made gathering clinical evidence through RCTs virtually impossible. RWE played a larger role in regulatory approvals for new therapies and is expected to continue. Seventy-five percent of new drug applications and biologic license applications in 2020 included evidence from RWD.
Pharma is tapping RWE in new and expanded ways. While some use it for clinical trial design and recruitment, others use RWE for financial analysis related to payer reimbursement or to monitor post-market safety and adverse events. Some are creating continuous learning approaches within their product development lifecycle.
A recent RWE trends webinar highlighted what the top pharma organizations are prioritizing. During the webinar, attendees participated in several Zoom polls, supporting this trend. When asked, “For which of these commercial and medical activities do you rely on RWE the most today?” here were the results.
Participants were then asked, “What types of SDoH data do you use today?”
In addition to patient recruiting and other uses mentioned above, life sciences organizations use RWE to inform patient and physician engagement, expand clinical trials through control arms and streamline data aggregation and the continuum of care. Using RWD is promising to improve early discovery decisions, enhance clinical trial design and execution, as well as reduce the resources required for post-market surveillance and studies.
Trend #2: RWD Expanding Sources & Quality
The expanding sources and quality of RWD used for RWE are expanding and improving as more life science organizations, and the FDA realize increasing value.
Digital health and wearables: In March 2021, the FDA published an analysis of 90 examples of RWE used to support decisions, and, of those 90 examples, three involved digital health devices and two featured patient-generated data, while most of the other examples featured registries or medical records. Computers, mobile devices, biosensors, and wearables are increasingly used to gather and store large amounts of health-related data.
“Digital health Innovation has significantly enhanced the capabilities of wearable devices, which has resulted in the increased use of the technology in clinical trials.” The ClinicalTrials.gov database was cited with more than 1,000 clinical studies completed or ongoing that include wearables, particularly across therapeutic areas like oncology, neuroscience, as well as respiratory and metabolic disorders.
Decentralized or hybrid trials: For 2022, it was estimated that decentralized clinical trials would increase by 28 percent over last year, and three-fourths of clinical trials will use a hybrid or decentralized approach. And in phase two and three trials, there is a 69 percent increase in the number of procedures, as well as three times more data points being collected. Protocol complexity has also increased significantly over the past two years.
Population health management: “RWE is being widely used to gain an understanding of patient populations and subpopulations, as well as the patient journey and when and how treatments are used and any resulting gaps in care.” PHM data is being leveraged across the lifecycle of drug development from planning and early development to include business and commercial activities like market analysis, health technology assessments, and contracting.
RWD sources: In another poll during the previous-mentioned webinar on pharma trends, the audience was asked, “what types of real-world data are most important to demonstrate value to payers and providers in addition to clinical trial data?” (They could select multiple choices).
These top three or four data sources support this trend, helping to address specific research questions and enabling greater richness, indicating future expansion into genomics, consumer data, and even social media.
Trend #3: RWE + RCT = Better End-to-End Clinical Evidence
Designing repeatable, scalable clinical trials is becoming a major burden to pharma. While most industries use large datasets combined with AI and machine learning to assess and optimize processes and operations, data from clinical trials is rarely accessed once completed and submitted.
RWE studies are seeing adoption growth because of the efficiency and accuracy gleaned from RWD. RWD gathered from virtual trials expands synthetic control arms for rare therapy areas and indications. Through patient personalization, early diagnosis, and remote patient monitoring within trials, costs can be optimized.
RWD holds promise to optimize results because it allows more homogeneous and targeted patient populations, clustering of cohorts with similar profiles, and highlighting those who’ve “responded in the same way to a particular drug or have biologic signals pointing to a similar likelihood of disease progression.” When. Combined with AI, researchers can more deeply utilize clinical datasets that may be siloed with critical information like biomarkers, lab results, and images. More precise study designs yield more robust treatment vetting.
Trend #4: Expanding Equity & Diversity Representation with RWE
Only three percent of patients and providers participate in clinical trials. This requires researchers to work with smaller datasets, exacerbating an existing lack of diversity in clinical trials, leading to therapeutics that may not be as effective for all populations. That’s where RWD and RWE offer an opportunity to produce racially- and ethnically-representative datasets that can expose inherent biases in research.
Additionally, many people don’t have ready access to clinical trials or social determinants barriers that keep patients from realizing treatment success. RWE offers the chance to expand the diversity of patient cohorts while focusing on decreasing the risk of bias.
Trend #5: RWD Data Sharing and RWE Interoperability Partnerships
RWE is dependent on RWD access and linkage, which requires interoperability. Because no single healthcare system or organization has all the data and federated analysis is frequently used due to privacy concerns, interoperability or a shared system based on trust is crucial. In Massachusetts, they passed a law called Chapter 55 to interconnect multiple databases and inform state policy to help manage opioid overdose-related programs.
New partnerships in data sharing, sparked by the pandemic, created greater cross-sector collaboration as demonstrated in the creation of the COVID-19 Evidence Accelerator, which is an initiative launched in June 2020 by the Reagan-Udall Foundation for the FDA, “creating a forum for data holders and researchers to pool data and share insights to speed understanding of COVID-19 treatment and response.”
Trend #6: AI for RWE Analysis
RWD requires the most work in preparing the data and deriving meaningful variables. Employing AI to enhance data anomaly detection, standardization, and quality checking at the pre-processing stage. While data linkage and interoperability are required to provide real-time access and feedback necessary to harness the potential of AI for clinical decision-making, AI and machine learning will continue to have a growing value in creating RWE.
Throughout different stages of drug development — from early discovery, clinical trials, and post-marketing — more applications are accessing EHRs to inform drug discovery. One author reviewed 20 years of research and “found that a lot of EHR data has been applied in the drug discovery stage for identifying biomarkers—not so much about novel targets but merely about the biomarkers.” Trial recruitment seems to be where AI is used frequently with RWD.
In post-marketing research, adverse events reporting is an area for AI use, creating greater automation and efficiency in a historically manual-heavy area. Techniques like natural language processing (NLP)n enable AI to scan tens of thousands of records and quickly find adverse events details.
Even the speed with which COVID-19 vaccines were developed and deployed was mainly due to new drug development platforms that tap into AI/ML for modeling protein and cellular interactions. Traditional laboratory testing and retesting can be supplemented with AI to enable simulations on computers rather than testing in live conditions, processing thousands or millions of simulations to identify high-potential “candidates for treatment consideration and subsequent analysis.” These AI-enabled discovery models could eliminate months and years from the overall research timeline, accelerating time-to-treatment.
AI-enabled analytics and automation provide access to insights from historical clinical trial data and RWE, expanding end-to-end clinical trial capabilities:
- Data ingestion – historical, publicly available, RWD
- Text extraction – NLP used to extract key entities from clinical trial documents
- Data transformation & standardization – pre-built models for data standardization
- AI model deployment – predict trial design impacts on costs, cycle times, feasibility, and quality risk
Harness the Value of RWD and RWE through Technology
In a competitive life sciences industry, creating an evidence ecosystem that follows RCT and RWE across the product lifecycle is the goal. These trends demonstrate the growing sophistication of RWD and the value of RWE, as well as the technologies that enable them.