Industries

Health & Life Sciences

Wolters Kluwer Health says,

"We needed to break the cycle of more data, more requirements, more money and the answer was Netezza."

Electronic Health Record Analytics & Secondary Use

EHR Secondary Use is a compelling investment benefiting pharmaceutical companies, as well as payers, corporate insurers, regulators, and even patients. Secondary use can generate insights toward better drugs, safety and outcomes leveraged toward patients by third parties as well as providers. Value propositions in secondary use include:

• Population studies on outcomes using longitudinal analysis and hazard models
• Participation in the US FDA Sentinel program for pharmacovigilance
• Population segment analytics for more targeted drug development
• Enables joint pharmaceutical and payer partnerships
• Clinical outcomes for pharmacoeconomics and pharmacoepidemiology
• Group risk factors and predictive interventions
• Identification of clinical trial prospect groups enabling faster drug approvals and rejections

For providers with successful electronic health record solutions, there is an increasing trend to have a single data repository outside the EHR, a—'data warehouse'—for analyzing electronic health record data. If the EHR provider offers an analytic 'module' accompanying the EHR solution, why pull the information into a second system? Three reasons:

• First, most providers have more than one EHR. EHR solutions are often not sliced evenly between, for example, ambulatory and inpatient—instead they cross facilities, payers, disease vectors, IT support organizations, physician groups or departments. As a result, to get a true cross-network view of patients, care, and outcomes, only a single target database accepting and joining data from key EHR systems will suffice.
• Second, while analytics based exclusively on data inside a single EHR is valuable, the value increases when that data is merged with data only found outside the EHR. A traditional EHR often does not have the ability or need to leverage many types of external data, let alone adequately manage that information. A centralized data warehouse is well suited to merging data from multiple EHR with other data like: health network key metadata, classifications, and descriptive attribute data around its own physicians, treatment venues, and more. Further, 3rd party data purchased for gathering insights on prescriptions, populations, events and outcomes have even more value when merged with EHR data.
• Third, the analytics offered by the EHR provider may be limited. Some advanced end users may want to tap directly into the valuable EHR data via simple dashboard reporting one day; ad-hoc multi-pass SQL queries the next; and complex data mining algorithms after that. Only by taking data out of the EHR(s) and allowing end-users to use best-in-breed BI and analytic tools against a high-performance centralized data warehouse can, IT support all their needs.

Why Netezza?

Analyzing EHR data, without even considering the added complexity of merging it with valuable outside data, has three challenges. Why Netezza?

• First, the data sizes are vast. If one million lives requires up to a terabyte of data, then the quantity of lives for secondary use is going to mean data sizes that make most data warehouse environments untenable, unsupportable, and unresponsive.
• Speed of insight must be nearly instant. As data is pouring in daily, rules-based risk and safety models are ready to trigger for many drugs and hundreds or thousands of patients. With the lives of patients at risk, there is no greater cost to delaying answers to questions.
• The science can be complex. With the opportunity to do longitudinal outcomes; the need for propensity of risk scores, hazard and other multi-state regression models; and use in the typically complex and sometimes statistically-mandated world of regulation and academia, the math in secondary EHR to drive predictive analytics is not trivial.