Since the start of the pandemic, the health care and life sciences sectors have relied more heavily on real-world data (RWD) and real-world evidence (RWE) as they work to develop new drugs and devices and survey the world outside of clinical trials. Hospitals and health systems are key to these efforts.
The FDA defines real-world data as data relating to a patient’s health status and/or the delivery of health care routinely collected from a variety of sources, such as electronic health records, claims and billing activities, and product and disease registries. Real world evidence is the clinical evidence regarding the use and potential benefits — or risks — of a medical product derived from analysis of real world data.
The U.S. Food and Drug Administration has used real world data from Israel to make decisions on Covid-19 boosters based on evidence of fading vaccine immunity over time. Real world data and real world evidence are also currently being used to examine the effectiveness of treatments for conditions ranging from sickle cell disease to atopic dermatitis, metastatic renal cell carcinoma, and many others.
As President Biden pushes for advancing biomedical and health research through the creation of the Advanced Research Projects Agency for Health (ARPA-H), the demand for real world data will continue to grow rapidly. While much potential exists in all forms of data for research, the foundation of real world data are the clinical data collected by hospital and health systems during patient interactions. As such, there needs to be a focus on how clinical datasets are produced and how they are used, as well as how data are sourced for research.
Several barriers prevent the wider adoption of clinical real-world data for research, including by biopharma, clinical research organizations, and other entities. First, the ability of hospitals and health systems to meet the heightened demand for RWD is uncertain as they are already spread thin with the ongoing health care worker shortages and depleted resources following the pandemic. Second, the proliferation of real world data companies has created an opaque chain of data transactions leading to datasets with credibility and quality questions because the provenance of these data has been lost.
Hospitals and health systems are the most consistent and reliable sources of real world data, yet are often not involved in how data generated at their facilities may be routed for other purposes by third parties, losing sight of the data management practices that may be occurring as data leaves the walls of the hospitals. To support the continued widespread use of real world data in clinical decision making, the health care and life sciences industry should focus on establishing scalable, consistent models for generating high-fidelity clinical data from hospitals and health systems in an auditable fashion. The gap between hospitals and health systems and biopharma companies must also be bridged to better facilitate sharing RWD and improve patient outcomes, which starts by hospitals taking an active role in how to manage and distribute the data they are producing.
Real world data start at hospitals and health systems
Hospitals and health systems are significant contributors to the real world data industry. Without the generation of clinical data, it’s impossible to understand a patient’s clinical condition and their medical journey.
Hospitals employ two important sources of real world data — the electronic medical records they use to manage patients’ care and various clinical registries — special databases that store detailed information about people with a specific disease or condition — to which they submit data. The two are connected, as data from the electronic medical record is often reviewed in producing registry submissions. For example, cancer registries collect real world data on clinical care provided to cancer patients and their outcomes in order to help evaluate the efficacy of treatment modalities, calculate statistics on cancer incidences, and conduct research.
The FDA has recently acknowledged that clinical registries have the potential to support medical product development and offer an advantage over other data sources because of the insights that can be gathered on defined patient populations with information about disease history, medical care, and complications. But for electronic health record data and, in turn, registry data to be useful for real world data applications, they must be of the highest quality, which is determined by hospitals’ data management processes.
Hospitals and health systems need support to generate more RWD
Collecting and interrogating clinical data is a time-intensive task that requires the experience of trained clinical experts, such as nurses or medical assistants, to identify trends and other valuable information within the data that will be useful for tracking outcomes. This is because the most important information lies within unstructured data, such as imaging studies or doctors’ notes, which must be clinically interpreted and structured to be leveraged most efficiently.
While the structured data in electronic medical records present basic health information, only the complete record, including notes and other unstructured information, contain sufficient detail on disease history, outcomes, and other nuanced details about a patient’s journey. Electronic medical records were designed for billing, not for research purposes, which is why the data they collect and store must be curated and enriched with unstructured information coming from the various documents driving a patient’s journey to be useful as a real world data source for a vast majority of diseases and conditions. This requires full medical chart abstractions by skilled clinical experts who can interpret all the data collected throughout a patient’s interaction with the health system. This method ensures that consistency and formal processes are followed during data entry. Too often, research datasets at hospitals are built in an underfunded fashion, with medical fellows and other patient-facing staff reviewing patient records in their spare time, leading to difficulties in establishing consistent processes to develop high-quality data, and in turn, high-quality research.
This is the stage many hospitals struggle with today. They don’t have the resources to train and hire clinical experts whose sole focus is data abstraction. Instead, they rely on clinicians who need to prioritize direct patient care, while also needing to spend time helping to curate datasets that then need to be submitted to registry bodies, payers, and the government. This was never sustainable, and the well-documented health care worker shortage and pandemic fatigue is decreasing hospitals’ ability to produce high-quality clinical data. Hospitals are increasingly unable to keep up with their clinical data management or produce the high-quality data necessary for registries and beyond.
When hospitals and health systems have the resources necessary to prioritize their clinical data, it leads to improved outcomes not only within their own facilities, but worldwide. Since clinical data are the basis of RWD, more resources dedicated to the generation of high-fidelity clinical data will only advance the health care industry through the development of more effective treatments, progression of precision medicine and improved health equity.
The increasing importance of clinical data
The rise of RWD is attributed to the weaknesses inherent within clinical trials. While they remain the gold standard to test drug efficacy, through a precision medicine approach, drugs are ultimately designed to target very specific cohorts of patients that increasingly represent smaller portions of the patient population, thus making it harder to enroll adequately sized clinical trials.
It’s extremely difficult for biopharma and hospitals to find patients that match these descriptions, and because of the barriers associated with participating in clinical trials (proximity to academic medical centers, ability to take time off work, requirements to participate, and the like), many clinical trials are approved using fewer patients, most of which are white and healthier than the rest of the population. Oftentimes, the sickest patients or those with existing medical conditions are excluded from trials to produce a more homogenous trial population. Many trials rule out participants who have an autoimmune condition, which often is not reflective of the patient population.
Diversifying participation in clinical trials is critical to improving patient outcomes and achieving health equity. The National Academies of Sciences, Engineering and Medicine released a report in May citing the urgent need for diversity in clinical trials, with an economic analysis commissioned as part of the report finding billions could be saved by reducing the harms caused by diabetes and heart disease if just 1% of health disparities were reduced though more diverse clinical trials. Hospitals and sponsors can work together to leverage the real-world data generated with each patient encounter to work toward the goal of building trial populations that are reflective of the true patient population and closing these gaps in care.
With improved data management practices, hospitals and health systems may be able to better support patients at the decision points that matter along their journeys, matching their data to clinical trials or the newest treatment guidelines. By using clinical data, the medical community is able to understand the efficacy of drugs and treatments for people of all backgrounds and medical histories. This approach allows the more accurate capture of trends within patient populations and a better understanding of which treatment modality may be most suitable.
Clinical data: the future of medicine
There’s broad consensus that a myriad of potential use cases exist for real world data, and clinical data in particular, but there’s less agreement on the models and means to produce these types of datasets. Hospitals and health systems should lean into their role as the definitive source of high-quality clinical data and leverage their data to help unlock this potential across a range of use cases, from hospital-driven research to biopharma studies and clinical trial needs.
The more available research-grade clinical data become, the easier it will be for the medical and scientific communities to capture trends more accurately within patient populations of all backgrounds and medical histories, develop more effective treatments and better understand which drugs may be most suitable for a specific patient.
Hospitals and health systems are key in the evolution of medicine, yet receive little to no attention and investment to produce higher quality data for their own organizations and the biopharmaceutical companies that rely on quality data. The health care industry needs to prioritize overcoming the barriers currently associated with the collection and interpretation of clinical data. Doing so will help unlock the critical insights within the data and streamline the use of RWD throughout the industry, which will ultimately improve the health of patients everywhere.
Victor Wang is the senior vice president of data and research at Q-Centrix.
Published in STAT First Opinion. Read the article here.