Select Page

Tigermed Insight

The Potential of Real-World Data.


January, 2021

In the past decade, the use of real-world evidence has expanded from its use for post-market surveillance of medical products to use throughout the drug development life cycle, including the development and market access phases. Various players in healthcare and drug development are actively using or encouraging the use of real-world data, including the FDA and the European Commission.

In this blog post we will first discuss what real-world data and evidence is, what kind of sources are used for it, give examples of purposes it can be used for, and outline what needs to be taken into account when planning to collect and use Real-Word Data.

What are Real-World Data & Real-World Evidence?

Let’s start by asking ourselves what we consider Real-world data (RWD) to be. It is an umbrella term for data used in healthcare and clinical settings coming from a variety of sources other than randomized clinical trials. Real-world evidence (RWE) in turn is the clinical evidence derived from analysis of RWD relating to the usage, potential benefits and risks of medical products.

What are the common sources of Real-World Data? Usually, the following are used to collect the data:

  • Observational studies, including case reports, case-control studies and cohort studies.
  • Pragmatic trials assessing the effectiveness of medical products in real-life settings.
  • Electronic Medical Records (EMR) containing e.g., diagnoses, test results, lab values and demographic data.
  • Administrative databases, most commonly of health insurance claims and bills. This usually contains data on performed tests and treatments and the associated costs.
  • Patient and disease registries with data on e.g., patient sample sizes and treatment protocols.
  • Patient-generated data, e.g., health surveys and data from wearable devices.
  • Other sources, such as social media, online patient organization forums etc.


What can Real-World Data be used for?

Traditionally, RWD has been used by pharmaceutical companies and regulators for pharmacovigilance studies to confirm the efficacy and safety of a drug after market introduction. Nowadays, other players such as Healthcare Providers (HCP), payers, and Health Technology Assessment (HTA) and regulatory bodies also make use of RWD for various purposes. We will describe some valuable and interesting examples of RWE use below.

The different sources of RWD, and particularly EMR databases, provide valuable data on clinical outcomes, such as overall survival, progression-free survival and quality of life. One example where data like this is used by pharmaceutical companies and regulatory bodies is in the accelerated regulatory approval of medical products, for instance in the case of some of the current SARS-CoV-2 vaccines, which allow a substantial part of clinical evidence to be provided by RWD after-market authorization. Facilitated regulatory approval is also common for rare diseases and in the fields of oncology and paediatry, where it would be very difficult or unethical to perform a randomized clinical trial. In these cases, efficacy and safety can better be confirmed after treatment has started, and EMR data from previous patients can be used as historical controls.

By combining EMR databases and patient registries from several HCPs, as well as claims databases from payers, large datasets can be created which can enable identification and more accurate estimation of unmet medical needs, patient population sizes, disease burdens and cost effectiveness of treatment options.


Patient Selection

RWD from these sources also allows for more specific patient selection. This can be done on a study level, where a subset of patients with particular characteristics is selected. On a treatment level, for instance data on biomarkers can be used as an indication for whether a particular treatment will be effective for an individual patient, thereby enabling a more personalized-medicine approach. Payers can also use this to implement outcome-based reimbursement schemes, where the cost of treatment is only paid for if the clinical outcome measure has indeed improved.


Pharmacoeconomic Analysis

Further pharmacoeconomic analyses can be performed after market access, when payers and HTA bodies use RWD to compare the efficacy, safety and cost-efficiency of a medical product with alternative and competitor products. Similar analyses can be done comparing the clinical outcomes of a new drug versus the standard treatment options.

Furthermore, RWD on the use of a drug for off-label medical indications and populations might provide valuable data on whether an application for expanding the approved indications and patient target group would be feasible. Moreover, when a medicine has been approved in one country, and RWE supports the efficacy and safety claims there, approval can be obtained more easily in a different country with a similar patient population and similar treatment standards.


How to successfully implement Real-World Data use?

To effectively use RWE for any purpose, one needs to be aware of the potential pitfalls and limitations of RWD, to carefully plan ahead and to allocate resources early on. We will discuss some of the principles of RWD good practice here.


Strategic planning of Real-World Data generation

Pharmaceutical companies, RWD service providers and other players should not sit around and wait until a question arises, to then look for data sources to answer that question. Instead, making the generation of RWD an integral part of business and of the life cycles of medical products is essential. This requires regular market analyses to anticipate future needs, and the early allocation of manpower and financial resources to the appropriate departments. 


Central RWD team

One should not work from the perspective of individual studies and projects, but rather a central RWD team, for instance as part of the Medical Affairs department, should oversee the presence of capabilities in study design, data collection, data management, statistics, maintaining quality standards and collaboration with partners.  Some of these capabilities can also be established through specialist partners.


Investment in database management and data-analytics

Investment in hardware and skilled workforce is important for data management and processing, which can be done in standardized ways, thereby improving cost efficiency.

Since common variables between different data sources can be rather limited, skilled statisticians that can perform complex analyses can partly offer solutions. More importantly, it is required to carefully plan ahead and focus on increasing the number of common variables and of clinical populations between data sources that have already been and that will be collected.


Appropriate data source selection

Feasibility analyses

It is important to perform feasibility analyses on each potential data source to determine whether and which one(s) can provide the desired data for specific purposes or decision-making processes. For example, to assess the cost-efficacy for treatment of a particular disorder, solely using healthcare insurers’ claims databases is probably not an option, since information on diagnosis is often missing.

Availability and access to data

Patient and disease registries and claims databases differ widely between countries, not only in content and structure of the databases, but also in terms of who can access them. In a large portion of countries, databases can only be accessed by academics, so involvement of an independent academic researcher is required. Moreover, access to particular databases might be highly regulated and time consuming. It should therefore be considered whether it is worthwhile to include specific data sources, as well as specific countries where regulation might be a huge burden. 

On a national, international and regulatory level, this also means that more guidelines should be provided for the standardization of data sources, and for the regulations for accessing these sources.


Data processing

Data quality and integrity

Virtually all data sources have erroneous entries and missing or incomplete data. For this reason, data cleansing and data transformation/imputation is essential, where errors and missing entries are either fixed or removed. It is important to document these procedures to maintain transparency. Data should also be stored in a format that is easily accessible and transformable across different platforms.

Linking different data sources

Most likely, several databases with different structures and content have to be combined to provide a sufficient data volume. Therefore, processing of data on an individual database level should be avoided whenever possible, as procedural differences can lead to inaccurate data statements. Furthermore, it decreases the transparency and traceability of the data by making it harder to see where the original data came from.

Privacy and data security

De-identification of EMRs and other identifiable data is necessary to adhere to privacy laws. This can require advice from privacy experts, as it entails more than simply removing names, but also other factors that could potentially match, also between databases where names have not been removed.


Establishing analysis standards

There are currently no quality guidelines for processing and analyzing RWD. This is a striking difference with the Good Clinical Practice standards for randomized trials. Transparency about analysis methods is therefore crucial. On an international level, most logically at least within the EU, the establishment of standardized analysis methods for RWD is desirable.


Creating a partner ecosystem

Pharmaceutical companies should be in close contact with HCPs, payers, HTA and regulatory bodies to ensure that interests and opinions are aligned. For instance, if an HTA body expects certain data to be able to assess the efficacy of a medical product, it is crucial to know this at an early stage to ensure this data is appropriately collected.  Furthermore, a pharmaceutical company should have a network of HCPs and other data providers where data standards have been assessed, are standardized and quality controlled.