In recent years, both regulators and pharmaceutical companies have become increasingly willing to incorporate real-world evidence (RWE) across all phases of drug development. The process of developing and deploying a new medication has long been complex, challenging and expensive. Using RWE alongside clinical trial data has the potential to streamline time-to-insight while generating a fuller and more patient-centric view of a given therapy’s efficacy.
With electronic health record (EHR) systems and the growing adoption of wearables, biosensors and remote patient monitoring (RPM) devices, the healthcare sector’s ability to capture, store and analyze health-related data has increased dramatically.
Because today’s healthcare providers, payers and clinical researchers have access to large volumes of data, the desire to extract value from it is only natural. After all, such data has the potential to allow researchers to not only better design and conduct clinical trials but also answer valuable research questions that were previously infeasible.
The 21st Century Cures Act cites electronic health record (EHR) and insurance claims data to support regulatory decision-making. As such, there is growing interest in leveraging previously unused reserves of both structured and unstructured healthcare data. The potential for scale is enormous as billions of health records exist, covering vast numbers of patient encounters across the globe. The evidence that these records contain may be more relevant and representative than what’s produced by many clinical trials.
However, extracting usable RWE from many EHR systems poses both technical and practical challenges. There are a wide array of EHR solutions on the market today (over 500 different products are available in the U.S. alone), and not all are compatible or interoperable with other platforms. This means that there will be inherent variability in how health records are organized, and which fields are included—variability that can make it challenging to conduct reliable and consistent analysis.
Further, healthcare providers may record information differently. They may use inconsistent coding so that what one physician describes as a “heart attack,” another may call a “myocardial infarction.” Finally, gold-standard measures such as clinician-reported outcomes may be absent from both EHR and claims data.
Interestingly, specialty-specific EHRs represent one possible solution to address some of the aforementioned challenges, especially when applied on a smaller scale. These specialty specific EHRs can be further augmented with prospective data collection, including direct-to-patient measures such as patient-reported outcomes (PROs).
Finally, claims data can provide yet another dimension of information and is widely used in health economics and outcomes research to better understand aspects of medication adherence and healthcare resource utilization. While this type of data may be limited to providing demographic, procedural or pharmacy script data, linking or combining data sets offers a strategic benefit of expanded patient insights.
