Lowering the Stakes of a Real World Data Purchase
- March 08, 2021
- Posted in Marketplace
Written by Melissa Leonhauser, VP, Product Strategy – Life Sciences
A real world data purchase is rife with inefficiencies and challenges. We regularly hear from pharmaceutical customers about the frustration of managing multiple data vendors to acquire broad and diverse healthcare data to drive commercial, RWE, outcomes, and clinically-focused analytics. The standard way of acquiring this data is a time consuming and costly aspect of doing business. We knew there had to be a better way.
Prognos Marketplace, part of the new Prognos Factor® platform, is an ecosystem with hundreds of sources and more than 45 billion healthcare records on 325 million de-identified patients. Marketplace is a new way of acquiring data that addresses many of the most common pain points in the healthcare industry related to purchasing, storing, integrating, and analyzing de-identified patient data.
Pain Point: a real world data purchase is expensive, time-consuming, inflexible, and opaque
The current industry standard for purchasing real-world data is a lengthy and expensive process. For larger pharma companies the process of securing data typically involves a drawn out RFP process every few years. During the process, purchasers aim to deduce exactly what each vendor can provide in a data set, but are typically left to trust what the vendor is telling them about the data with no option to preview the data in advance. This is especially problematic because the company is typically then locked in to a multi-year contract where any additional requests for data or alterations incur upcharges. To try to preempt these charges, many larger companies buy extra data upfront to avoid paying more for it later. In addition, they may also buy additional data sets from other sources with largely redundant content in order to gain incremental value. Both scenarios result in a need to manage, store and analyze more data than is actually needed in the company’s data lakes.
For smaller pharmaceutical manufacturers the process may look different, but the results are similar – an inefficient, time consuming process. And, for these companies the stakes are even higher. With tighter budgets the purchase has to be correct the first time since they will use the data repeatedly to make the most of their investment.
To address these challenges, Prognos Marketplace is built to be transparent, flexible, and fast, as well as to provide a high level of specificity in the data selection process. With de-identified patient data including the largest collection of harmonized lab sources integrated with multiple large sources of open prescription and medical claims, one of the largest sources of closed-system medical and prescription claims – and more being added regularly – the pool of data to select from is robust.
When evaluating the data for purchase, Marketplace provides transparency and specificity showing significant demographic detail and information such as top therapies, tests, comorbidities, and more for the selected patient cohort. This allows companies to iterate and further refine their selections before purchasing. The result is that companies get only the data they need and all the data they need.
For example, instead of buying all available data for patients with diabetes, Marketplace allows for additional levels of detail including tests and test results, treatments, and more, to further refine a search. Plus, when a patient fits within the selected cohort criteria, all records for that patient are provided, not just the ones deemed as relevant for the identified issue (i.e. diabetes). This patient-centric approach provides a greater level of detail and a better understanding of the whole patient, including co-morbidities and other related issues.
All of this is possible at speeds previously unheard of. Cohorts can be pulled for evaluation in less than a minute, and from query to pricing to purchase to delivery, the entire process can take less than one week.
Pain Point: Managing multiple data vendors is challenging and time consuming
Another common challenge we hear from the industry is that traditional methods of managing multiple vendors to acquire the necessary data lead to wasted time and resources. In many cases, the analysts – the experts tasked with leveraging the data – spend their valuable time managing vendors, unharmonized data sources, and contracts instead.
By bringing together a robust set of harmonized sources within Marketplace, we created a single point of contact and single contracting process to streamline a real world data purchase process.
Pain Point: Data purchased from multiple sources is difficult to link
With data coming from multiple sources or vendors, challenges frequently arise related to linking of patient data. Inconsistent data tags, different use rights, privacy concerns, and more, may mean that companies are unable to, or choose not to, link decentralized data purchases. For those who do, the process of formatting data from different sources to make them easy to use and store is a challenge. Yet, having non-integratable assets frequently leads to ineffective use of the data.
Within the Marketplace, Prognos does the hard work so the data delivers optimized value. All data is harmonized to address any concerns about inconsistencies between sources and to ensure it is ready to analyze. If a company is buying record-level data they receive harmonized data lakes ready to transfer and analyze. And for those purchasing use of the data within Prognos applications, they receive ready-to-use data with all the speed and capabilities of Prognos Factor.
Pain Point: Maintaining patient privacy requires added investment
Patient privacy is always a top priority, but it also presents added challenges. When acquiring data from multiple sources it may not be possible to link patient data across datasets if different methods are used to de-identify patients. In addition, often companies will avoid linking patient data across sources out of concern for re-identification risk.
Prognos Marketplace uses Datavant as its de-identification technology so that every source within Marketplace is de-identified in the same manner, ensuring patients can be linked across datasets. For customers with their own data – previously acquired or data they created – they can link it to Marketplace data if they also use Datavant for encryption.
Prognos’ data privacy experts ensure that any data acquired and integrated through Marketplace is de-identified and that risk of re-identification is appropriately low. This reduces the need for pharma companies to invest in privacy experts to conduct this work on acquired data, saving time and money.
Pain Point: Source-driven data often falls short
With historical methods of acquiring data, the information received is driven by what data is available from a given source. This puts the emphasis on what is available, instead of what patient data is most needed.
With Marketplace, the data is patient-driven, not source-driven. Once the criteria is set and patient cohorts are defined, Prognos identifies which sources have records for those patients and prices based on necessary patient data. With this patient-centric approach, the buyer receives records and a well-rounded view of patient events from multiple sources.
A new industry standard for a real world data purchase is born
By overhauling the process to address these challenges, a new industry standard for evaluating and purchasing patient data has been born. To learn more or request a demo, contact firstname.lastname@example.org.