Pharmaceutical companies have long been using real-world evidence (RWE) to inform their decision making, respond to requests from external stakeholders, improve their therapies and, more recently in some jurisdictions, to gain regulatory approval for new treatments. The use of advanced analytics to draw insights from RWE takes the potential benefits to a whole new level. It is estimated that the world’s top 20 pharma companies could each gain, on average, more than $300 million a year over the next three to five years by using advanced RWE analytics across the value chain. This could help pharma companies to not only save costs but to also understand how patient characteristics and behaviors affect health outcomes, ensuring that the right patient receives the right treatment at the right time.
The extent to which companies are exploring advanced RWE analytics differs by country and is dependent on a wide range of factors, including capabilities, the availability of data, regulations, and a supporting ecosystem. As such, advanced RWE analytics have been deployed more widely in Europe and the United States but are now gaining momentum in Japan.
Some companies in Japan have already seen results. Few, however, are deploying advanced RWE analytics at scale, and some are yet to begin. Wherever they fall on the deployment spectrum, there are valuable lessons to be learned from the experience of those operating in more mature markets.
Why now?
Around the world, advances in analytical techniques such as predictive models, machine learning, probabilistic causal models, and unsupervised algorithms are helping to make RWE an increasingly powerful resource for pharma companies. Applied to RWE, such techniques can help provide a better understanding of patient outcomes—for example, which patient subsegments respond best to a particular therapy. Such insights can inform decision making across the value chain.
Other developments, however, are more specific to the Japanese market, namely wider acceptance by regulators of studies other than randomized controlled trials, government initiatives to develop and allow aggregated data, an emerging ecosystem, and more offerings from data providers.
Increased regulatory acceptance. Japan’s MID-NET database, which became fully operational in April 2018, was developed by the Ministry of Health, Labour and Welfare (MHLW) and the Pharmaceuticals and Medical Devices Agency (PMDA), which integrates electronic medical records (EMR), medical claims, and diagnosis procedure combination (DPC) data from ten sentinel sites at 23 hospitals. Pharmaceutical companies can use the MID-NET database to assess drug safety and to perform research for public benefit.1 Japan has also revised the ministerial ordinance regarding the postmarketing study of drugs, and in April 2018 authorized the use of databases such as MID-NET for postmarketing surveillance.2 Moreover, last year, the MHLW published basic policy outlining examples of how registry data might be used for regulatory filings—as an external control arm in cases where it proved hard to conduct randomized controlled trials, for example. Data from medical records could potentially be used too, but to a lesser extent, according to the basic policy.3
Government initiatives for the development and use of aggregated data. Various government initiatives are under way to develop the aggregated data needed for advanced RWE analysis and to permit its usage, including the following:
- Consolidation of healthcare data. Japan’s National Database of Health Insurance Claims and Specific Health Checkups (NDB) holds the electronic medical claims data of the entire population. As of October 2020, revised legislation allows for the interlinkage of NDB with the country’s long-term-care insurance database, creating one of the world’s largest population-wide, longitudinal data sets, and further allows the data to be used by private companies for public benefit.4 In addition, the government is developing a personal health record (PHR) system using “My Number” ID cards.5 Although primarily intended as a health record for individuals and their healthcare providers, private PHR service providers can access the data with patient consent. The use of this database could potentially be expanded for secondary use by pharma companies and academia.6
- Accreditation to accelerate data aggregation and information exchange across healthcare institutions. Provided patient opt in is obtained, or on an opt-out basis when conditions are met on notifying the patient beforehand, the Next-Generation Medical Infrastructure Act allows certified organizations to collect personal medical information from healthcare providers, aggregate a patient’s data from multiple data sources, then anonymize it and provide access to researchers and private companies.7 As of May 2022, three data providers have been certified to produce and provide such data.8
A growing ecosystem. Ecosystems that include data providers, healthcare providers, pharmaceutical companies, analytics services providers, software companies, and government bodies will be key to obtaining the full potential of advanced RWE analytics. In Japan, some pharma companies have entered partnerships to help establish such ecosystems. Pfizer Japan, Life Data Initiative, and NTT Data are conducting joint research on using RWE to assess cancer outcomes, for example.9 But pharma companies in Japan, unlike those in the United States, have not yet made big acquisitions of data providers or formed large-scale joint ventures with them. Companies that make bolder moves to build differentiated capabilities could create distinctive advantages.
Data providers expanding offerings. Organizations from healthcare and beyond are investing to expand their data offerings. For example, Medical Data Vison, a healthcare data provider that initially provided anonymized patient data from acute-care hospitals, now also provides medical claims data from employer-based payers.10 The company also announced a collaboration with DeNA to expand to payer data including self-employed in National Health Insurance and elderly in Senior Elderly Insurance schemes.11 NTT established PRiME-R as a joint venture with Kyoto University, aiming to collect EMR and other data from 100 hospitals, initially focusing on oncology.12
The opportunity
Against this backdrop, many pharma companies in Japan are already starting to obtain value from advanced RWE analytics.
The Japanese affiliate of one multinational pharma company used RWE data to throw light on the treatment journey of over 50,000 patients and revealed previously unrecognized issues such as a low treatment rates, late initiation, and the omission of large patient segments that the company could address through its engagement with healthcare providers. Another global company that wanted to understand the patient journey in Japan for one of its main therapeutic areas used more than ten machine learning models to analyze payer- and provider-based RWE data covering DPC and medical claims, each addressing specific business questions regarding the market position and usage of its products. As a result, the company was able to identify dozens of key patient characteristics and opportunities to improve patient outcomes and reflected these findings in its business strategy.
Third-party databases cover a wide range of data but have limitations
Many third-party databases, built from data sourced from healthcare providers, payers, and pharmacies, are available. The data can include electronic medical records (EMR) and diagnosis procedure combination (DPC) data, inpatient and outpatient care in hospitals and clinics covered by public health insurance, pharmacy claims, health checkup data—a blood-sugar-level test, for example—and drug sales data.
However, all databases have potential limitations, including inability to track patients when they switch providers in provider-based databases, limited coverage of the elderly population outside of the working age group in data sets from employer-based payers, and a lack of detailed information such as diagnosis of patients in pharmacy databases (exhibit).
Databases will therefore have to be assessed carefully to understand the degree to which they can support a use case.
For the time being, the availability of data might dictate which use cases are viable. Though there are currently many databases covering major areas in the healthcare ecosystem, the source and scope of the data differ, and each may have certain limitations. Databases will therefore need to be carefully assessed to understand the degree to which they can support a use case (see sidebar, “Third-party databases cover a wide range of data but have limitations”). Companies should also consider that Japan’s drug pricing, reimbursement, and access systems and the regulations governing them will limit the number of viable use cases compared with the United States. But there are many opportunities for R&D, commercial, and medical, including the identification of unmet need, the improved targeting of underdiagnosed patients, and improved pharmacovigilance.
Key areas of focus
Whether a pharma company intends to start deploying advanced RWE analytics, is already deploying the techniques in some parts of the organization, or is widely adopting them, certain actions in key areas will help it progress (exhibit).
Intention to start
Companies intending to start deploying advanced RWE analytics will likely understand its potential but might struggle to know where to begin given existing organizational capabilities. To get going, they will need to be clear about their strategy and vision on RWE as well as the value they are targeting, then move quickly to build momentum.
Strategy, vision, and value. Companies can easily find themselves mired in continuous debate as to where value might lie and what are the best use cases. Everyone has different views and priorities, making alignment hard. To help overcome this, a team that includes commercial, medical, and R&D personnel should work together to align on the overarching strategy and vision for RWE, to identify areas likely to provide the most benefit across the value chain, and to specify what value the company wants to create, how, and by when. When choosing and prioritizing use cases, the team should be guided by both business priorities and impact. It’s important for everyone to understand what advanced RWE analytics can achieve, so early lighthouse use cases can be both feasible and move the value needle. Successful completion could then raise support for more projects.
Capabilities. Companies new to RWE analytics may already have data scientists but often lack local data engineers who, in our experience, are some of the most important contributors to a successful program. Most databases are written in Japanese and require an in-depth understanding of the data structure as well as Japanese healthcare systems. It is also important to have business translators who can convert business requirements into executable directives for the technical team and understand the business implications of analytical insights. To what degree these capabilities are developed in-house, rather than outsourced, is an organizational decision, but they should not be outsourced entirely. All companies will need some of their own experts—data scientists, data engineers, and translators. Decisions as to which future capabilities will be required and an action plan for building or acquiring them don’t need to be made immediately, as it can be hard to envisage what might be needed at the outset. Instead, companies should regard implementation of the initial use case as a means of reaching a clearer view and accept that some outsourcing might be required for a while. Building and/or acquiring the full team will take time.
Limited adoption
Companies at this stage will have begun to adopt advanced RWE analytics in specific use cases, will likely have executives pushing its adoption, and will have started to build internal analytical capabilities. But a common challenge is keeping the momentum going and disseminating the learnings from initial use cases across the organization. Creating the right organizational structure, operating model, and culture will help, as will a data strategy that helps deliver impact at scale.
Organization, operating model, and culture. Companies can drive the wider adoption of advanced RWE analytics by creating a new, dedicated role to drive adoption across the organization. It will entail liaising across departments to identify opportunities, shaping a portfolio of work, and challenging brands and functions to adopt innovative approaches. Also, companies typically need to set up a central interdisciplinary team to orchestrate cross-departmental dialogs and establish internal processes regarding all RWE-related initiatives—for example, how use cases are sequenced, funds allocated, or personnel assigned. The interdisciplinary team with business, analytics, domain knowledge and/or medical expertise can also help resolve challenges that might arise in the event of process and/or cultural changes.
These interdisciplinary functions could sit within a country, global, or regional organization, but it is essential that they have decision-making authority in Japan, including for funding, as there will be aspects of the Japanese market that those less close to it might not appreciate.
It is also worth noting that even with the best head of RWE and an interdisciplinary team, a change management program might be needed to bring the organization on board and to adopt advanced RWE analytics more widely. Executives need to see RWE as a key element in strategic discussions and development programs, be willing to adopt novel approaches, and accept a degree of risk when taking actions based on new insights. A change management program—one that perhaps includes mandatory advanced-analytics learning modules—can help drive cultural change across the company.
Data. Given the current limitations of available databases, pharma companies will need to give careful thought to their data strategy if they intend to use advanced RWE analytics across the company. Data coverage is less comprehensive in Japan than in the United States, for example, and there is limited information on healthcare institutions or geography where patients are treated. Companies should therefore consider and develop a data strategy for each asset and therapeutic area. Some disease areas, such as rare diseases, may require innovative data partnerships, for example, as commercially available databases may not provide sufficient breadth or depth. Pharma companies must be able to identify which database to use and which use case to implement. They should also continuously scan the fast-evolving data landscape and invest in new data sources that will help deliver on an advanced RWE analytics strategy.
Wide adoption
More advanced companies are able to systematically drive the design and execution of advanced RWE analytics. Senior employees are accountable for its success, and a designated team ensures insights are operationalized. But these companies also work to identify new opportunities and overcome barriers to further expansion, seeking to establish a lead in advanced RWE analytics. Here, investments in long-term partnerships as well as tools and the technological environment are key.
Partnerships. More advanced companies may find that the limited availability of longitudinal data is slowing further use of advanced RWE analytics, especially if they wish to expand beyond common disease areas, evidencing the need to build relationships with data providers to secure privileged access and develop proprietary data sets. Such assets will likely become the source of significant value in Japan. Commercial data providers are not the only potential partners, however. Local governments, start-ups, academic consortia, analytics companies, and healthcare providers are among those who can help pharmaceutical companies advance their capabilities.
Tools and environment. Companies aspiring to scale up their RWE analytics will, at a minimum, need a “sandbox” for conducting basic experiments with use cases and delivery models. More advanced companies should, however, consider investing in their own state-of-the-art platform and/or consolidate existing platforms to support automated pipelines, repositories of analytical assets, and visualizations. By continuously converting real-world data input to in-depth insights on the platform, companies can make informed decisions based on up-to-date RWE. Finally, companies should establish governance principles to ensure rapid adoption of new data and collaboration with external parties.
By committing to impact and scale, pharmaceutical companies in Japan have the potential to become key players in the field of advanced RWE analytics. Ultimately, as Japan is already the most advanced market for advanced RWE analytics in Asia, they might use the expertise they build to spread its adoption across the region. It will likely take several years of effort to obtain the discipline’s full value given the new capabilities that need to be built and the new ways of working, but there are early signs that a few companies are exceling in this area. Companies that have been slower to adopt advanced RWE analytics should therefore consider accelerating their efforts—or be left behind.
