The volume of clinical data has been growing exponentially due to broad uptake of electronic health records (EHRs), innovation from consumer technology companies, and quality reporting requirements from payers, to name just a few of the industry-transforming factors that are driving the increased availability of clinical data for diverse uses. As in many industries, clinical healthcare data is highly valuable and can offer novel insights on the practice of medicine to interested stakeholders throughout the healthcare ecosystem. As the volume and availability of clinical data have increased, so has commercial interest in clinical data. In particular, pharmaceutical and life sciences companies are increasingly using real-world data from clinical settings to inform their efforts throughout the drug development pipeline, including identifying unmet patient needs, engaging with clinicians, supporting clinical trial recruitment, market access and pricing strategy, and drug safety monitoring, just to name a few use cases.
Several high-profile and large-scale data commercialization initiatives have been announced by large technology companies in recent months, ranging from Oracle, which recently acquired EHR company Cerner and plans to create a Healthcare Data Repository, to Walgreens, which has recently announced a move into clinical trials, leveraging its wealth of pharmacy and community data, with many other smaller companies and startups working in the space as well.
The appeal of clinical data commercialization is clear: leverage real-world clinical data to support the use cases named above with data and insights that are nearer to real time than can be accessed through historical billing data, for example. We have seen significant willingness to pay from pharmaceutical and life sciences companies that have a meaningful need that is difficult to meet with clinical trial findings or billing data alone. With this interest from the market, a large number of data commercialization vendors have emerged, each claiming to have a superior product based on volume of data, source of data, diversity and representativeness of data, applicability to a specific or niche use case, and more.
The value propositions in clinical and other real-world data for both data sellers and buyers are clear, but the wide array of data-sharing initiatives makes it difficult for pharmaceutical companies to identify and select the right source or sources of data to deliver on their needs. In our experience, there are at least three key factors pharmaceutical and life sciences companies should consider as they select and engage with clinical data providers.
While there are several interoperability standards like HL7 FHIR in use in the clinical data space, interoperability of clinical data remains a major challenge when thinking of data commercialization use cases. Especially applicable to large-scale data sharing efforts, such as the above-mentioned Oracle example, interoperability – ensuring that data in different formats and from different sources can “talk to” each other and facilitate analyses – is a major challenge. Broadening access to clinical data to the wider healthcare ecosystem is not the core challenge of interoperability, but interoperability has a large impact on the potential success of data sharing initiatives. A 2022 publication in the Journal of the American Medical Association studied 3,928 hospitals in the United States and their progress between 2014 and 2018 toward data interoperability. The study found that less than half of hospitals in the study had made significant progress toward interoperability, as assessed by four interoperability domains: electronically finding (querying), sending, receiving, and integrating information into EHRs.
This slow progress toward interoperability means that large-scale data-sharing efforts that are attractive to buyers based on the purported size of their databases are likely to struggle to deliver meaningful insights from across clinical settings. It remains unknown and unproven whether and how data from different sources and in different formats can operate together to deliver any insight to the pharmaceutical or life sciences community. Though consistently applied interoperability standards have been under development for several years, none has yet emerged as the “winner” in a crowded category. We urge potential buyers or licensors of clinical data to clarify the sources and standards of the data included in databases to ensure that they may benefit from the databases’ large scale.
Clinical data registries have an advantage over large-scale data-sharing initiatives in that they deliver data that has already been rendered interoperable. Often managed by specialty healthcare associations as a value-driver for members, clinical data registries offer a voluntary data-sharing platform for clinicians to contribute their data in support of research, quality improvement, and other clinical applications. Clinical data registries accept data from diverse sources and clinical settings and map it to a central and consistent data schema, shortcutting the problem of interoperability and offering healthcare industry stakeholders a more accessible data asset that suffers less from the challenges of interoperability standards, or lack thereof.
2. Data Security
A survey of pharmaceutical industry professionals published in 2021 found that 35% of respondents indicated that digital privacy and data security were their “central concern” related to digital transformation. Interestingly, that figure represented a decrease from 2019, perhaps indicating more openness to data sharing caused by pervasive data sharing efforts throughout the COVID-19 pandemic.
Organizations that make data available for commercial use must ensure that data is being gathered, held, and transferred in accordance with all applicable regulations that prioritize patient and clinician privacy. Regulations like HIPAA and GDPR will remain central qualifications to meet, as well as those that are more niche, like ISO27001, which provides technical specifications for information security management systems. Companies that are looking to buy or license real-world clinical data should also prioritize collaborations with organizations that demonstrate a strong sense of ethical use of the data.
3. Use Case Specificity
The huge volumes of clinical data that are accessible to pharmaceutical and life sciences organizations make it tempting for potential buyers to turn to the data sellers and ask, “what can it do?,” seeking insight in a vacuum. Different sources of clinical data have different strengths, depending on their source, the use case, and the context in which the data was collected or in which it will be applied.
As pharmaceutical and life sciences leaders seek real-world or clinical data to augment their knowledge and insight, we strongly recommend that they have or seek out a clear set of hypotheses that drive the analytics and inquiry required to draw meaningful conclusions from the data. This type of specificity can be achieved through strong understanding of the patient or clinician journey or a clear set of requirements that the data user has when selecting a data source. Large-scale clinical databases may not have the level of specificity and granularity that an organization requires to meet their needs. Alternatively, smaller-scale, more niche databases may be too detailed to answer broader questions about health behaviors and outcomes. Furthermore, if an organization is seeking insight on health outcomes or exposures within a specific clinical domain or medical specialty, long-term billing data from a pooled database will not deliver the type or quality of information that’s required.
To identify the optimal data sets, data leaders should have a clear set of objectives and use cases for which they are seeking insight. For example, it is more fruitful to seek validation of a specific hypothesis like “adults who report consuming more than 2,500 milligrams of sodium per day are more likely to develop hypertension than adults who report consuming 2,000 milligrams or fewer of sodium per day” than to blindly search a database for patients with hypertension and discover their dietary patterns. Not all data sources will be able to answer all questions about real-world patient behavior or deliver the same clinical insight, so organizations that seek to use clinical data should take care to seek out the sources that can deliver the insights that are most crucial to their work.
In our work with the American College of Emergency Physicians to broaden data access and deliver insight to a larger group of healthcare system stakeholders, we start with a set of hypotheses – we call them use cases – that guide our inquiry. We start from macro topics, like drug safety monitoring, and work with our clients and collaborators to refine the question, resulting in a testable hypothesis that we can bring to our data asset, the Emergency Medicine Data Institute Registry, the largest clinical data registry in the United States. We create specific hypotheses like “patients with an active prescription of medication X are less likely to visit the emergency department than similar patients who were not taking medication X.” Using custom selections of the data, we can identify trends, note anomalies or differences in the data, and deliver meaningful data insights that quickly add value.
As clinical data becomes more widely available for commercial analytics and insight, there are massive opportunities for pharmaceutical companies, health insurers, and other healthcare industry stakeholders to make use of the data to add value, from product development to go-to-market strategies to post-market surveillance and beyond. Potential data users should be aware that the industry is evolving rapidly and select a data provider with credible credentials to deliver the most value and insight.
Jenna Phillips is an innovation expert at PA Consulting