The Tudor Health Ecosystem is a fully integrated approach to all aspects of EMR data capture, management, analysis and delivery. Our Ecosystem is designed to overcome the problems associated with legacy EMR data services.
Life science data scientists have many data challenges:
Claims data only work in the USA, Tudor Health EMR data are global.
Legacy EMR data providers have restricted coverage, Tudor Health doesn’t.
Registry data cannot be used for commercial purposes, Tudor Health EMR data can.
In some areas, especially oncology and rare diseases, claims data has been seen to be less value:
Frequently no diagnosis code for rare diseases.
Clustered payment protocols (DRG) make the data harder to analyse.
Timely data delivery is critical and claims approvals processes make the data slow.
First among the major issues with EMR data available for commercial use in the USA market is that the capture rate in many diseases is very low.
For reasons that we shall discuss in detail later in this series, in many rare oncology conditions it is not uncommon for the total number of available patients in an EMR data set to be fewer than 20. This is nowhere near enough data to construct disease models.
In fact, with such small patient numbers the only effective analysis that can be done is an analysis of each individual patient, which is of little value to a commercial team looking to support product-wide decision making.
This situation is also frequently found in non-oncology rare conditions, and the preponderance of rare disease R&D being conducted by the bio-pharma industry would suggest that this results in a significant gap in decision support knowledge and is, therefore, a significant shortcoming of the current sources of EMR data.
Tudor Health has a data capture method that ensures the capture of sufficient patients to permit proper analyses such as line of therapy, patient journey, market shares, brand switching.
Our data capture maintains retrospective data as well as up-to-date data relating to the continuing patient journey – and we capture new patients as well.
Current EMR data are not representative of the entire market for any disease area. Without data representing patients in all treatment locations, including in and out-patient, local and regional treatment centres, primary care as well as specialist offices, it is not possible for EMR data to be representative.
Leaving aside the simple question of how many patients can be found in EMR data, questions of treatment location and geography must also be considered.
The USA has a unique approach to EMR.
Each healthcare provider, from the largest Integrated Network to the smallest solo-practitioner, can choose from a vast array of EMR systems. There is no centrally mandated system for sharing patient data between HCPs.
There are a small number of very large EMR system providers as well as specialist providers. Some of the largest are most widely used in the integrated delivery networks and many are predominantly used in the office environment. Some of these providers make deidentified data from their system available to life science customers, and others don’t (and some make it especially difficult to extract data for analysis).
The one consistent element of the USA market for EMR data is that it is extremely difficult to obtain EMR data that covers multiple treatment locations because the data come from the EMR software vendor and data only come from treatment locations using that software.
Even applications that were previously single-setting (surgical anaesthetics) are now subject to multi-centre issues, to the point where providing data to cover all disease areas now requires a multi-location solution.
What if the data you can obtain comes from community oncology centres but most of your oncology patient population are treated first in academic medical centres? In this situation you cannot monitor what is going on at the most critical step in the treatment journey – the first treatment choice. Also, with data from community oncology centres you aren’t able to stay up to date with the most recent developments, such as the impact of a new competitor.
Tudor Health collect EMR data without being technologically linked to any specific EMR software. Data are obtained from locations using a full range of EMR software, but the Tudor Health data are not affected by this – comprehensive clinical data without going through the EMR software.
Access to all treatment locations is a core benefit of the Tudor Health Ecosystem.
In some oncology indications, and an increasing number of rare diseases, the patient population is so small, and the number of treating locations so few that it becomes essential to ensure that data are collected from specific locations.
EMR data will only do this for those locations using the software of the data vendor and, in most cases, this is not sufficient.
A major issue here is that most of the treating locations involved in these rare diseases are academic medical centres, and most of those are part of larger integrated delivery networks, and most of those use EMR systems that do not make their deidentified data available for commercial use (think EPIC and Cerner). The result is that the clinical data relating to those patients are not available to life science companies.
The Tudor Health methodology makes these data available in two ways:
By collecting EMR data in a way that is not dependent on specific technology, we can collect relevant data from the required HCPs at academic centres.
By engaging with HCPs at specifically named institutions we can ensure that, at the very least, a solid sample of the required treatment centres is included.
Whilst we cannot guarantee the participation of HCPs at specific institutions, we can differentially sample and employ a broad range of professional as well as financial incentives for participation – steps that go beyond a traditional market research model for gaining participation.
In studies completed to date, Tudor Health has achieved a participation rate in excess of 90% from specifically targeted treatment institutions, including large integrated networks, academic medical centres and regional treatment centres.
The USA has a unique market for data at the patient level. Deidentified claims and EMR data are available for most of the US population. Even with the issues and barriers described earlier, patient-level data have become a standard tool for life science commercial departments.
The same cannot be said for other countries.
Considering the major markets for pharmaceutical and biotech treatments, we find several generally consistent characteristics that prevent a market in deidentified PLD:
National health services operate without a claims structure like that of the USA, making these data centrally controlled by the relevant Government ministry, not a commercial company.
National health service control usually means either a single EMR system or a centrally managed mixture of systems operated by companies but reporting to the central authority and with no ability to market PLD.
Centrally available sets of detailed EMR data at the patient level, usually a sample based on population and/or disease, and available for strictly research purposes only.
As a result, patient-level EMR data for commercial use is generally not available outside the USA.
Since the late 1980’s, the life science industry has relied on so-called “chart pull” studies to fill this gap. These studies are particularly powerful when added to strong scientific support, but they still have significant weaknesses compared to what full EMR data would provide.
The Tudor Health ecosystem circumvents these restrictions. By engaging directly with HCPs and not using existing EMR software to collect data, Tudor Health is able to capture detailed longitudinal EMR data in countries outside the USA. Applying the sophisticated engagement methods, targeted sampling and data capture methods, all the other benefits of the Tudor Health approach are also found outside the USA, such as obtaining data from patients treated at target institutions, collecting data relating to patients with highly specific and/or rare conditions in oncology and other areas.