The Health Sciences Blog covers the latest trends and advances in life sciences and healthcare.

Recent Posts

How FDA’s Guidance on RWE Impacts the Life Sciences and Influences Oracle Health Sciences

If you haven’t yet, please navigate to this link to read this exciting new guidance on Real World Evidence (RWE) released by the FDA in early December 2018!  The guidance document is riveting to read, and it is an exciting element of Oracle Health Sciences’ current strategy to disrupt clinical research with our new Oracle Health Sciences Clinical One platform, Oracle Health Sciences Data Management Workbench (DMW), and Oracle Health Sciences mHealth Cloud Connector Service.  Read on to find out how! First, in the guidance document, the FDA makes clear that coupled with the 21st Century Cures Act, RWE is a critical pillar in the ongoing quest to get the approvals of new drugs and biologics to patients faster and at a lower cost.  This guidance is underscored by FDA Commissioner Scott Gottlieb’s quote, “As the breadth and reliability of RWE increases, so do the opportunities for FDA to make use of this information.” Additional critical points: The FDA’s RWE journey started years ago in the safety arena.  One manifestation that has been very successful is the Sentinel System. This system assesses EHRs and claims-real world data (RWD) sources to generate RWE for drug safety.  The Oracle Empirica Signal Suite is also a byproduct of the FDA’s safety journey. Years ago, the FDA served as an Oracle development partner to create the solution, and the agency continues to use the Empirica Signal Suite in its drug safety program today. Examples of RWE sources include data derived from: EHRs, medical claims and billing, product and disease registries, and patient generated data (including in-home settings) from mobile devices and sensors.  The good news for our customers is that Oracle Helath Sciences knows a lot about RWD sources.  With its healthcare business, Oracle has been successfully integrating EHR data for healthcare provider and government organizations for over 20 years.  Also, because Oracle’s mHealth Connector product supports patient generated data from mobile devices and sensors, electronic clinical outcome assessment (eCOA) and electronic patient reported outcomes (ePRO) data sources can be easily integrated with both Oracle Health Science InForm and Oracle DMW. In addition to using the above-mentioned real world data (RWD) sources to generate RWE, RWD can also be used to improve trial efficiency.  This means that a biopharmaceutical company’s clinical research budget can support exploring more potential targets, not only reducing time to market, but also increasing the number of therapies it can pursue simultaneously.  Examples include:     Generating hypotheses for testing in randomized controlled trials (RCTs) Identifying drug development tools, including biomarker identifiers Assessing trial feasibility by examining the impact of planned inclusion/exclusion criteria in a relevant population Informing prior probability distributions in Bayesian statistical models Identifying patient baseline characteristics for enrichment and/or stratification Assembling geographically distributed research cohorts (particularly for rare diseases and targeted therapeutics) Single-arm trials are a reality!  This development is extremely exciting.  It speaks to manyof the good things they enable, such as: Reducing the number of patients needed for a trial.  Since there isn’t a need to conduct a dual arm trial, a single-arm trial can reduce the number of patients needed by up to 50 percent (50%).  Patient enrollment is the Number One problem with running clinical trials, as only three percent (3%) of patients volunteer to participate.  In addition, the overhead of finding the best sites that can help a sponsor get to the right level of patient enrollment, including getting all the patient data and paperwork required for the regulatory process, adds tremendous time to a successful study startup. (See how our recent goBalto acquisition streamlines this!) Alleviating ethical dilemmas.  For deadly diseases, it’s very difficult, ethically, to have a control group who does not receive the prescribed trial treatment.  In a single-arm trial, the sponsor can mine RWD sources for patients with that disease who have not taken the therapy in past and compare those historic patients to the current studied patient group.​ Speeding the Drug Approval Process. Using RWE, Blincyto, a treatment for acute lymphoblastic leukemia, was approved under the FDA’s accelerated approval process.  The response rate in a clinical trial for Blincyto was compared to historical data from 694 patients. The historical data was extracted from over 2,000 patient records at US and EU clinical study and treatment sites.  Enabling an adaptable learning health system. A dose-response-aspirin-trial compared two common doses (81mg and 325mg) across 20,000 randomized patients who had a history of myocardial infarction (aka heart attack).  This was a long-running trial integrated into regular patient care with minimal inclusion/exclusion criteria and no additional treatment protocol beyond the aspirin regimen.  Primary trial end points such as death, another heart attack, stroke, etc. and secondary endpoints, such as coronary procedures, hospitalization, etc. were captured from electronic health records (EHRs) and claims-data sources.  Arguably, this trial was an example of integrating trials/research directly into patient care processes and/or an instance of the learning health system. Of course, there are many, many more important elements of the FDA’s RWE framework strategy in the document, so please read it! From all of this, it’s easy to see how Oracle Health Sciences’ fundamental Clinical One strategy is really going to advance and re-envision clinical trials. Most critically, the next module for Clinical One is not called EDC (for electronic data capture).   It’s named Data Capture.  Why? Because the game is different now.  Our new data capture capacity isn’t just to get data from one site.  It’s performed to enable data to be captured from all required data sources -- including RWD sources -- in real-time for the trial.  Clinical One allows users to make this fundamental, streamlining, yet disruptive, change to the clinical research process easily. Are you re-thinking your RWE strategy? Take a look at the progress of our Clinical One vision and contact us for a conversation!  

If you haven’t yet, please navigate to this link to read this exciting new guidance on Real World Evidence (RWE) released by the FDA in early December 2018!  The guidance document is riveting to read,...

Health Sciences

Galen to Prix Galien -- from Rome to New York in Two Millennia

Next week, the Galien Foundation hosts the Prix Galien awards at The American Museum of Natural History in Manhattan. The gala event  --  including the ceremony, cocktail party, and dinner  --  recognizes outstanding achievements in pharmaceuticals, biotechnology, and medical technology that improve the human condition. The award is considered the equivalent of the Nobel Prize in life sciences research. Ever wonder how Prix Galien got its name? The award honors Claudius Galenus, an anatomist, physiologist, clinician and researcher. He has been called the father of medical science and modern pharmacology. His work has been considered a reference for over two millenniums. Born in Pergamos in 131 A.D., Galen studied in Smyrna, Corinth and Alexandria, the three centers of medical excellence of the ancient world. According to a legend, Galen dreamt of Aesclapius, the god of medicine in ancient Greek mythology, and this dream inspired the rest of his life.  When he turned 17, Galen worked as a physician at a gladiators’ training center. Marcus Aurelius requested Galen to come to Rome when he was 37. There, he grew in reputation and stature as a healer, teacher, researcher and writer. His ideas on the functioning of the human body were so well received that he became the personal doctor of young Commodus, the Emperor’s heir. During his long, eminent life, Galen completed over 500 works on anatomy, physiology, pathology, medical theory/practice, and forms of therapy. His work formed the basis of Galenism, a medical philosophy that dominated medical thinking until the Renaissance. He also  travelled throughout the world, studying local plants and remedies. He described 473 original drugs and many mineral and plant based substances. He was the first scientist to codify the art of preparing active drugs. His observations, logic, and deduction made him the true successor to Hippocrates and his view that the prime aim of medicine is patient care has formed the very cornerstone of modern pharmacy.  Galen died in 201 AD. Oracle is honored to sponsor an award dedicated to such a remarkable medical pioneer and very proud to celebrate, this year’s Prix Galien nominees for their innovation in science and medicine. Join us in New York on October 26 for the Galien Forum, a day-long event where Nobel laureates and scientists discuss pressing health issues and scientific breakthroughs. And, in the evening, honor the life science innovators of today at Prix Galien.

Next week, the Galien Foundation hosts the Prix Galien awards at The American Museum of Natural History in Manhattan. The gala event  --  including the ceremony, cocktail party, and dinner  -- ...

Health Sciences

Data Models & Their Importance for Clinical Data Management

A data model is quite simply a 'description of the structure' of stored data.  For example, a clinical trial in Oracle Health Sciences InForm is described by an InForm Clinical Trial Data Model.  Data models are created by applications to store data and accessed by end users to manipulate data.   Data models are important because they help end users to access data that is easily understandable, meaningful, and useable.  For example, an Oracle Health Sciences InForm Clinical Trial Data Model is easily understandable to a study manager.  But, it is not immediately useable for submission to the FDA because it is not in a submission ready structure.  The FDA requires clinical data to be provided as a "Study Data Tabulation Model” (SDTM) dataset. So the business need is for end users to easily create data models that can be used: To capture data (ELECTRONIC DATA CAPTURE data models) To review data (REVIEW data models) To submit data to regulatory authorities (SDTMs) To visualize data (ANALYTICS data models) For any other purpose Rather like the alchemists of old who were trying to convert lead into gold, Clinical R&D is currently challenged with slow, complex, manual processes that must transform raw clinical data sets into high value clinical data models and datasets. What This Means for Sponsors and CROs: Sponsors and CROs need to take raw clinical data and provide it to their internal stakeholders in real time to: Fix data collection and quality issues Identify clinical safety issues Monitor trial progress Support interim analysis However, many data managers work with multiple systems, with many CROs, and many internal clinical groups.  Often they use processes that are disconnected, fragmented, expensive, and complicated to manage. How Oracle Health Sciences Data Management Workbench (DMW) Can Help Oracle DMW is the only industry solution that provides a unified platform in which data models can easily be created, stored, re-used, and transformed from one model to another. The value in this is that end users can easily connect to a data model to access and use clinical data in real time. This month at the Society of Clinical Data Management Annual Conference (SCDM), sponsors and CROs will be looking for ways to enable the transparency and analytic capabilities of the data within the clinical R&D process.  They can meet Oracle there (Booth #310) to discuss: The ease of creating clinical data models  The ease of transforming clinical data models from one to another The ease of connecting to a data model to visualize data The ease of automating clinical data flow to provide end users access to more accurate, higher quality data faster, and at lower cost Oracle Health Sciences breaks down barriers and opens new pathways to unify people and processes to help bring new drugs to market faster. Join us at SCDM, Booth #310, for a demonstration of Oracle Health Sciences Data Management Warehouse to add value to your patient-centric trials. ### Srinivas Karri is Director of Life Sciences Product Strategy for Oracle Health Sciences.

A data model is quite simply a 'description of the structure' of stored data.  For example, a clinical trial in Oracle Health Sciences InForm is described by an InForm Clinical Trial Data Model.  Data...

Health Sciences

Realizing Clinical Trial Value and Business Efficiency in the Cloud

One of the most interesting aspects of organizations moving to the cloud has been increased scrutiny around Value.   Increasingly, many of our customer interactions feature one or more conversations around business value and operational efficiencies.   Value is important because it allows organizations to focus on what is important with a new IT project. It helps manage and maintain scope, cost and risk.   Value  can be quantified  at any stage using business objectives during and post implementation to ensure continued, focused delivery through organizational alignment. This article discusses value in relation to multiproduct platform solutions in the Oracle cloud.   Value Assuming that a customer is considering deploying one or more Oracle products and associated services, it becomes necessary to quantify the opportunity cost for the change and balance the potential derived ROI benefits. The following describes the value associated with Oracle Health Sciences Data Management Workbench (DMW). Additionally, there are significant operational benefits associated with moving to the cloud that also should be considered and quantified. Figure 1.  One definition of value can be represented as the freeing up of cash flow as a result of decreasing direct and indirect costs, increasing revenue, or decreasing investment in fixed assets.   By increasing incremental free cash flow, value is generated within the enterprise. Business value is at the heart of any business case and is, in essence, the key driver for any new implementation.  It is evident that the business case at the very least requires three components: Understanding what product capabilities are provided by the solution and the business benefits:  For example, Oracle DMW helps organizations automatically aggregate, clean and transform data, which they may currently perform manually.  The business benefit is automation and making data more available to a larger number of downstream consumers.  Understanding how the business benefits translate to operational improvements, which are characterized by operational metrics.   For the example, using Oracle DMW, an organization can expect a decrease in the operational costs associated with study setup, data cleaning, and data transformation. Examples of operational metrics that capture these improvements are: Cost per change order, $ Cost per data load per study, $ Cycle time from raw to review model, days Number of man hours spent on validation and testing for new study build, man hours For example, a study build in Oracle DMW can be performed in less than an hour. Some organizations take over eight weeks to perform the same! Understanding how operational metrics translate into financial metrics.  Operational metrics, such as the cost of operations, translate directly into financial metrics, such as direct and indirect operations costs.   However, the link between operational metrics and financial metrics is not always clear and may even overlap between different operational metrics.  Usually the activity to translate operational metrics into financial metrics is performed by the customer using proprietary data.      Figure 2.  Operational metrics are intimately tied to financial metrics that appear on financial statements. In addition to these elements, the business case should also describe how the benefits would be realized through the implementation and subsequent go-live timeline.  Value may continue to be derived well after go live and may even increase through greater organizational adoption of the solution. Measuring Value by Implementing Oracle DMW So, given this approach to quantifying value, how can this be applied to implementing DMW?     Figure 3.  Value categories impacted by implementing Oracle DMW. To help our customers navigate the value realization roadmap, Oracle Health Sciences (OHS) has developed a comprehensive tool that describes product features, business benefits, operational metrics, and  mapping to quantifiable value categories.  With this tool, OHS customers can create a pragmatic, focused approach to a successful DMW implementation that can form the foundation of their clinical data collection and management strategy. If you would like to know more about this approach and would like to use the tool with your customers, please don’t hesitate to get in touch. If you have any questions or suggestions based on what you’ve just read or what you would like to read about, I’d love to hear from you. Contact me at: srinivas.karri@oracle.com

One of the most interesting aspects of organizations moving to the cloud has been increased scrutiny around Value.   Increasingly, many of our customer interactions feature one or more conversations...

Life Sciences

The FHIR RESTful Services Standard

Real world data generates real world evidence.  Many use cases drive the biopharma organization’s needs to do this.  These use cases can aid many areas within the biopharma organization including: the commercial, health economics and outcomes research (HE&OR), early development, translational research, corporate strategy, safety, and clinical R&D groups.  Therefore, real world evidence substantially enhances the effectiveness of the overall biopharma organization.  This post discusses how a biopharma organization’s clinical R&D team can populate case report forms (CRFs) in an electronic data collection (EDC) application from electronic health record (EHR) systems using in Oracle Health Sciences InForm and the emerging HL7 standard, known as Fast Healthcare Interoperability Resources (FHIR). The FHIR standard is a specification for implementation of RESTful Services* (a technology approach to building APIs). It enables access to patient data in EHR systems in support of system to system communication (e.g. interoperability).  The FHIR standard is actively under development by the HL7 standards organization and is maturing rapidly from investment over the past 5+ years. Some of its specifications are implemented in several EHR vendors’ systems.  So, code can be written once to the interface standard; and data can be accessed from supporting electronic health record (EHR) systems.  While it is not a fully mature standard, yet, it certainly has substantial momentum in the patient care technology arena.  Over the last couple of years, there has been a focus on using FHIR RESTful Services to integrate patient care data into the clinical research process.  A key use case has been populating electronic data collection (EDC) system case report forms (CRFs) from EHR systems.  This use case can help the biopharma company and its sites participating in a clinical trial to: 1.       Reduce overall data entry volume for each clinical trial 2.       Improve quality of entered data for each clinical trial 3.       Reduce clinical trial costs as data populating the CRF from the EHR system does not have to be    source data verified.  Source data verification is typically done at each site participating in the clinical trial by biopharma-retained employees known as contract research associates (CRAs). Exploring FHIR right now, there is clear evidence that biopharma companies are experimenting with EHR to EDC integration.  Recent discussions with several top biopharma organizations indicate interest as companies look to exploit the benefits of this approach.  HL7 Connectathon Connectathons support quick and easy experimental approaches to developing the various standards the HL7 organization produces.  HL7 continues to use this technique, now over 20 years old, as it develops the FHIR RESTful Service Standard.  A connectathon is a weekend “hacking” session. It   includes members of healthcare organizations who attend and work to “prove” the standard.  They hack together “quick and dirty” working code (a FHIR RESTful Services working prototype) demonstrating that the standard actually will work in the real world. These events occur two to three times per year as the standard is developed.  Oracle Health Sciences (OHS) is participating in the 2017 FHIR Hackathon this autumn.  Currently OHS is building an integration using the HL7 FHIR RESTful Services with InForm. The integration enables mapping patient data from an EHR system into the appropriate trial visit CRF schedule. It can include information such as demography, vital signs, and more.  The integration initiative will help OHS participate in the upcoming HL7 FHIR Connectathon this fall.    The InForm Portal is a key component enabling a user to configure basic mapping metadata required for the integration of: 1.      Which fields in each CRF form map to which fields in the source EHR system 2.      Code list mapping for CRF filed referenced in #1, above 3.       Which patient care visit data ranges (aka Encounter) map to which visits in the visit schedule 4.       The initiation of the data movement from the source EHR to the targeted InForm system This is all accomplished by using FHIR RESTful Services in the standard as they interact with InForm’s Clinical APIs. Customer Focus This summer, OHS is running the first meeting of a newly formed patient EHR data driven Clinical R&D Customer Focus Group and plans to showcase the FHIR RESTful Services to InForm integration.   We’d like to gather customer feedback for this as a very powerful, high value-add, real world data/ evidence use case.  In addition, we will share these advances with our customers as we continue to help them understand what is possible with the HL7 FHIR Standard and InForm’s robust clinical API capability.  OHS is at the beginning of its FHIR RESTful Services journey with biopharma customers.  The early steps include working with customer partners running initial pilots with participating clinical trial sites.  In 2016, OHS worked with a large healthcare organization and a big pharma company on developing pilots in this area. Additionally, two other large biopharmas are considering running FHIR RESTful Services pilots on Phase 1 oncology sites and a  COPD observational study.  Stay tuned for progress updates in this exciting, new innovation area. Greg Jones is responsible for Enterprise Architecture Strategy for Oracle's Health Sciences business.    *Representational state transfer (REST) or RESTful Web services are one way of providing interoperability between computer systems on the Internet.

Real world data generates real world evidence.  Many use cases drive the biopharma organization’s needs to do this.  These use cases can aid many areas within the biopharma organization including: the...

Life Sciences

Oh No! Johnny's Getting Hypertension Again

Imagine if we could remotely connect to subjects in clinical trials and measure their vitals. Imagine if we could take those measurements, in real time, and deliver them to an investigator’s Oracle Health Sciences InForm desktop solution for his/her review. Imagine if a subject felt less burdened by staying home and having his six weekly Investigator Meetings from the comfort of his home, instead of making a two hour commute to the clinic.  Imagine when the subject arrives at the clinic, he can’t find a car parking space. So by the time he reaches the clinic, he is really frustrated and fuming mad that he has to be on this “damn clinical trial in the first place.”         “Sit down Johnny.  Let me take your blood pressure. It seems a little high.  I think you have hypertension…”           “NO! I DO NOT!!! I tell you what I do have.  It’s A PARKING TICKET!  I am quitting this clinical trial right now. GOODBYE!” Imagine keeping Johnny as a subject on the trial because he used a remote wearable sensor device. Imagine his data flowing seamlessly into Oracle InForm for site review, and simultaneously, into Oracle Health Sciences Data Management Workbench (DMW). Imagine the data instantly transformed into SDTM data formats and made immediately available to data managers for downstream, actionable, medical reviews and clinical monitors. Well, the good news is, you don’t need to imagine these things for much longer. The Oracle Health Sciences (OHS) team has just released its first, initial component for its mHealth stack. This library allows app developers to create  IOS/Android mHealth mobile apps that connect to the Oracle Cloud. The OHS team welcomes the opportunity to leverage this exciting new offering to enable connected scenarios.                   For more info please watch this video. Jonathan Palmer is Oracle Health Sciences Senior Director, Product Strategy.

Imagine if we could remotely connect to subjects in clinical trials and measure their vitals. Imagine if we could take those measurements, in real time, and deliver them to an investigator’s Oracle...

Life Sciences

EMR to EDC for RWE

A new real world data/real world evidence (RWD/RWE) industry trend is emerging.  That is, electronic medical records (EMRs) to electronic data collection (EDC) integration.   Here, instead of trial research staffers entering patient history data information, the sponsor implements the “wiring” to take the data directly from the site’s EMR system. Several members of the Oracle HSGBU team have been working on this with our biopharma and our healthcare provider customers including myself, Paul Boyd, Kim Rejndrup, and Chris Huang.  Here’re a few of the things we’ve learned: -In the summer of 2016 the FDA published guidance around considerations for biopharma organizations as they consider moving their EMR to EDC integration. -Every EMR system out there will have only a subset of the information necessary for a particular clinical trial and that subset will vary by clinical trial and therapeutic area.  It’s too early to know precisely what percentage of coverage will be available.  So right now, the expectancy is 40-70 percent (40-70%) on average, per trial.  Again, this is an evolving learning experience in relation to our work and our partnerships with our customers. -Here are some powerful numbers Paul Boyd put together. 1) There are 745,000 data points in a production trial. 2) Approximately 300,000 data entry strokes could be saved, if 40 percent (40%) of the data from EMR could be mapped.  There can be substantial savings of time and energy with EMR to EDC integration!  These numbers also produce savings on the monitoring side of the equation.  Data fields in the InForm customer report forms (CRFs) sourced from the EMR system don’t have to be source data verified (SDV’d) by sponsor employees monitoring the clinical trial at each site (usually known as contract research associates).  Therefore, there will be substantial savings as a major benefit, as well. -Academic Medical Centers (AMCs) that participate in clinical trials as sponsors, have their research staff perform the dual data entry as explained above.  In recent discussions with AMCs, they’ve also indicated that sometimes they’ll set up a “shadow” EDC system for what they own and run for a particular trial, and input the data, yet again (in triplicate!).  They want to capture the data they are providing to the sponsor in their own, internal, research database to advance their various research programs. -Our customers are interested in the APIs and various technology interfaces Oracle Health Sciences supports on the InForm platform.  These interfaces are key to allowing them to bring data from sites’ EMR systems into their InForm clinical trial instances to capture the above stated benefits. -Our customers are interested in Oracle Healthcare Foundation (OHF).   Imagine this scenario. A large biopharmaceutical organization is running 100 plus clinical trials.  It has piloted EMR to EDC integration in several trials over a couple of years.  Now the organization is ready to ramp up this capability as a standard element for many of the clinical trials in its portfolio.  Each trial on average has about 65 sites (that number comes from our experience with production trials run in our Health Sciences Cloud over the years).  Plus a number of sites the biopharma deals with will likely participate in multiple trials with it – not just the one trial.  Some percentages of the sites in the trial (an educated guess is 50-80 percent (50-80%) of them) have the ability to provide EMR data to the biopharma’s EDC system for that clinical trial.           Here’s the challenge. Does the biopharma “wire” each individual InForm instance for each trial to every site?  That’s definitely doable from a brute force perspective. But probably not idea, as it will be very expensive.  Or does the org use a technology like OHF as a platform to integrate receiving data from sites and doing all the wiring and plumbing to those sites with the platform.  Then as new trials start up, they “wire” InForm to the OHF platform, which serves as a “hub”; and go live costs likely are substantially smaller. -An exciting HL7 standard known as FHIR (Fast Healthcare Interoperability Resources) appears to have a decent amount of traction in the industry.  Organizations such as Cerner, Epic, Meditech, and GE Healthcare have implemented this RESTful* Service based API on top of their recent EMR application releases.  Organizations are starting to use these APIs to build and deliver new applications and mobile apps to improve patient and staff healthcare activities.  In addition, the FHIR APIs allow support for interoperability among systems.   With this insight into one of the more exciting use cases for Real World Data & Real World Evidence.  EMR to EDC integration will continue to be explored with our customers.  The Oracle Health Sciences team plans to be there and working closely with them as the technology, policy, and business process challenges are resolved.  Ultimately this will become a mainstream approach increasing clinical research efficiency to deliver new therapies to patients more rapidly.   *Representational state transfer (REST) or RESTful Web services are one way of providing interoperability between computer systems on the Internet.   Greg Jones is responsible for Enterprise Architecture Strategy for Oracle's Health Sciences business. 

A new real world data/real world evidence (RWD/RWE) industry trend is emerging.  That is, electronic medical records (EMRs) to electronic data collection (EDC) integration.   Here, instead of trial...


Metadata Management in Clinical Trials

Metadata management in clinical R&D is centered on the concept that each piece of data collected for a clinical trial, as defined by that trial’s protocol, can be managed independently.  Each piece of metadata and logical groupings of many metadata items together can be governed and managed in the organization. This includes version control, data edit rules for that data item, and transformation rules for that data item as it changes to support the analysis process.  In addition, it also supports the ability to trace that data element through the entire clinical trial lifecycle. This trace-ability extends from the time it’s initially captured during the clinical trial through to that data element’s submission to regulators.  This trace covers all information on how that data element contributes to proving the efficacy and safety of a new therapy for regulatory approval.   As a quick example, blood pressure can be a metadata item for a clinical trial.  One can attach edit rules to that blood pressure item to insure accuracy. When the user inputs the data, the rules assure that it is edit-checked properly and that transformation logic is defined to change the representation of the blood pressure data element during data entry to a completely different representation for an analysis dataset that will be written in SAS code to prove efficacy and safety of the therapy.   I can do all that and maintain version control of that blood pressure data element as I change its representation over time.  Here’s another example. If I use that blood pressure data element consistently across all my clinical trials, then when I change that data element and produce a new version; I can query my metadata system on which clinical trials in my portfolio will be impacted by changing that blood pressure data element.   If I am running multiple clinical trials in which each has potentially hundreds of data items to be collected, then metadata management can help me manage those clinical trials operationally.  Metadata management also helps me to insure that I maintain full regulatory compliance and trace-ability as data goes through its lifecycle from capture to submission.  As mentioned at the beginning, this industry trend has been a long time coming.  The industry move over the last 10 years away from paper based clinical trials to electronic data capture based trials set the stage for this type of capability and for the operational savings from a successful metadata management solution deployment.  Unfortunately,expanding from the basic functionality described and scaling it up across a large scale clinical trial operation has proven to be very elusive to date for many organizations.  The metadata management highway is littered with several organizations that have failed to y deploy it successfully in their complex clinical trial environments.  The reasons for failure are complex, including the challenge of the activity and the resulting impact on the business processes of these large, complex, early- pioneering organizations.  Recently, there’s been a new wave of momentum!  Oracle Health Sciences’ (OHS) powerful partner ecosystem around its clinical R&D applications is kicking in to drive the next set of attempts at metadata management.  Specifically, OHS partner Accenture is leading the charge in close collaboration with two OHS top customers - GlaxoSmithKline (GSK) and Eli Lilly & Company – to tackle, once again, this very complex problem space.   Accenture is working to build a module called Metadata Registry (MDR).  The Accenture, GSK, and Lilly team is working to build this module with the above mentioned capabilities.  Progress is very promising to date!  They do have a large number of the above mentioned capabilities implemented successfully and are going through testing with GSK and Lilly.   In addition, the team will ultimately integrate the MDR with OHS Central Designer and Data Management Workbench applications.  This integration fulfills the promise of the enormous value of the MDR module.  Clinical trial metadata can be managed, version controlled, and more, within the module. Then, those same metadata data can be pushed into our Central Designer and Data Management Workbench applications at study startup and for the management of changes to study(s) in progress.   This will reduce the amount of time it takes to manage clinical trials operationally in each, respective, company’s portfolio and will increase the trace-ability and audit-ability quality each company needs for regulatory compliance.     Greg Jones is an Enterprise Strategy Architect with Oracle Health Sciences.

Metadata management in clinical R&D is centered on the concept that each piece of data collected for a clinical trial, as defined by that trial’s protocol, can be managed independently.  Each piece of...


Population Health and Oracle Healthcare Foundation’s Partner Ecosystem

As the Population Health Strategist for Oracle Health Science (OHS), I enjoy the ability to continue my educational quest for healthcare knowledge.  Between reading the Federal Register and House bills, catching up with CSPAN, and keeping upwith my friends from industry, I am amazed at the number and variety of population health applications currently available.   When I first started 20 years ago, the concept of “population health “ was something closer to evidence based medicine. Today, population health is synonymous with a range of US healthcare market subjects including: patient identification, cost analysis, clinical care gaps, precision medicine and early identification,outcomes measures,and EHR Implementations. We, who are on the OHS team, look at the idea of population health from the ground up. We aggregate a healthcare organization’s data and use it multiple times for any question posed, today, tomorrow, a year from now, or five years from now.   Our OHS team has invested in Oracle Healthcare Foundation (OHF) as the data aggregation and normalization engine that can fulfil population health discovery and care transformation, both clinically and financially. OHF offers a fit-for-purpose, analytics platform that provides a data acquisition, data integration, data warehousing, and data analytics solution.   The solution meets and exceeds current market conditions for organizations evaluating Value Based Care, Quality Measurement Performance, and Internal Cost and Care Team Effectiveness. It evaluates an organization’s information, turning data-driven insight into action.  In addition, OHS is active in recruiting top-quality population health partners to build out our partner ecosystem. These efforts leverage and extend OHF in the vast population health analytics space. OHS also invites its partners to join the Oracle Validated Integration Program. Prior to the invitation, the OHS strategy team takes into account several, fluid factors including: · The organization’s ranking according to IDC, KLAS, Gartner, Forrester, and other analyst organizations · A review of how the partner organization addresses the healthcare market’s business needs · A review of the landscape for emerging healthcare trends and initiatives at federal, state, and local levels  Oracle Validated Integration, available through the Oracle PartnerNetwork (OPN), gives customers confidence that the integration of a complementary partner software product with an Oracle application has been validated, and that the products work together as designed. This can help customers reduce risk, improve system implementation cycles, provide for smoother upgrades and ensure simpler maintenance. Oracle Validated Integration applies a rigorous technical process to review partner integrations. Partner companies that successfully complete the program are authorized to use the “Oracle Validated Integration” logo.  At this year’s HIMSS17 in Orlando, FL, OHS (Booth #3349) is pleased to join the population health conversation with our newest Oracle Validated Integration partners: ENLI Health Intelligence, SpectraMedix, and SCIO Health Analytics . These partners have developed and tested pre-built integrations between Oracle Healthcare Foundation and their population health analytics applications.   About Enli Health Intelligence Enli Health Intelligence™ is the market leader in population health management technology. Enli enables care teams to perform to their full potential by integrating healthcare data with evidence-based guidelines embedded in provider workflows across the population and at the point of care. @enlihealthintel,HIMSS17 Booth #2723 About SpectraMedix SpectraMedix empowers health systems, hospitals and other provider organizations to transition to fee-for-value and shared-risk programs using advanced quality measure, performance reporting and predictive modeling solutions in support of DSRIP, PRIME and operational activities.@SpectraMedix,HIMSS17 Booth #1889 About SCIO Health Analytics   SCIO identifies, risk stratifies and leverages predictive modeling (claims based) on patient populations based on actionable care gaps in order to design effective programs and meet value-based care delivery initiatives. @SCIOAnalytics Lesli Adams, MPA, is Director, Population Health Strategy for Oracle Health Sciences. Visit us in the Oracle Booth #3349 at HIMSS17 in Orlando, Feb 19 - Feb 23, 2017.

As the Population Health Strategist for Oracle Health Science (OHS), I enjoy the ability to continue my educational quest for healthcare knowledge.  Between reading the Federal Register and House...


Recap: Northern California HIMSS Innovation Conference and Showcase by Rahul Dwivedi

The Northern California HIMSS Innovation Conference and Showcase in Santa Clara this January was very well attended with industry think tank executives and high profile industry leaders. Sessions sponsored by industry leaders such as Intel, Oracle, Salesforce, HealthCatalyst, and others provided leading views on personalized healthcare, innovation, and the impact of artificial intelligence (AI) and machine learning (ML) on precision medicine and on healthcare industry, in general. The event featured discussions on innovation models for real-time analysis of key performance task or workflow indicators that could optimize processes and advance healthcare through more predictive, prescriptive information and insights. More specifically, panels addressed a range of issues including:   Innovative thinking from outside healthcare to transform personalized health and patient engagement through digital, mobile and the cloud. Genomic and IOT - behavioral, clinical, and genetic data capture/analysis - Intel’s “inside-out model.” How formal collaboration optimizes “inside-out” commercialization, research and tech transfer, and venture fund approaches to bringing ideas to fruition and fertile markets. Generating actionable insights and speeding innovation by continually looking at new and better ways to manage data. Understanding how machine learning can be woven into existing population health and personalized care business intelligence and analytics efforts. Agenda:The Northern California HIMSS Innovation Conference and Showcase In his keynote address, Bob Rogers, Intel’s Chief Data Scientist for Big Data Solutions, described the current state of healthcare and the future role innovation would play in access to information, patient engagement, logistics, and hand-off. He explained that these are areas in which innovation could have significant impact in creating value for all healthcare industry players, from patients, to providers, to HIT companies. Highlights from Panel Discussions The panel discussion on AI and ML for precision medicine and population health, led by Oracle Health Sciences Product Director, Prashant Natarajan, discussed healthcare specific innovations in ML field and the importance of going beyond existing models that worked successfully in other industries. Mr. Natarajan emphasized that healthcare innovations are created to augment the clinician's ability to serve individuals and tackle issues at population level, not to replace the clinician. Describing how AI and ML can help develop the right patient treatment at the right time, he also provided insight on the nature of data quality and how traditional amounts of data (small data) can work more effectively. From Left to right: Oracle’s Prashant Natarajan; Zeeshan Syed, MD, PhD, Clinical Assoc. Prof., Stanford School of Medicine; Mitesh Rao, MD, Prof., Stanford School of Medicine; Oscar E. Streeter, MD, FACRO, Center for Thermal Oncology; and  Bob Rogers, PhD, Chief Data Scientist for Big Data Solutions, Intel Corporation. After the discussion some audience members expressed opinions on how big data analytics, privacy, transparency, and data security were very important.  They agreed with various points put forward by Mr. Natarajan. There was also some additional discussion on how both clinical (population health) and omics (precision medicine) data were stored together as an integrated platform. Medigram CEO, Sherri Douville, offered a presentation on data democratization and data governance. The session provided great insights into what makes data governance important and how to achieve success in an organization when leading data governance efforts. The panel speakers provided great insight through their real life experiences and their comprehensive understanding of innovation and data governance. During the panel, Fail Fast, Succeed Fastor Fail Permanently - Emerging Models of Medical Device and Clinical HIT Innovation, entrepreneurs and finance industry speakers provided several,successful emerging models of innovation employed in several different companies. Companies from medical device production to healthcare information technology have been employing these models to propel innovation, both internal and external to their organizations. The session offered a great overview of identifying new ideas to match demand and building solutions to achieve scale that reaps benefits. Overall, this was a great gathering of academic think tank executives and industry leaders coming together to discuss the current state of innovation and AI-ML in the healthcare industry and what to expect in the future.   Rahul Dwivedi is Principal Applications Engineer, Oracle Health Sciences.

The Northern California HIMSS Innovation Conference and Showcase in Santa Clara this January was very well attended with industry think tank executives and high profile industry leaders. Sessions...


The Future of Value Based Care

The current healthcare delivery landscape is changing dramatically. Major regulatory reimbursement models are evolving from fee-for-service to fee-for outcomes and value. This transformation to value- based contracting involves multiple facets of the organization and an even greater demand for data. This shift requires health systems to leverage actionable patient outcome and cost analytics, as well as manage several other constraining challenges to address value based contracting, quality measure performance, internal costs, and care team effectiveness. Each health system must maintain market share in an environment of ever-increasing competition and ensure quality of care at reduced costs. Today, many organizations have an agnostic analytic strategy. They align their actionable intelligence to explore patient information, provider recommendations, resource utilization, and care outcomes.  All this information is evidence-based and demonstrates wise stewardship of resources. With this strategy, these organizations believe they’ll be prepared for the future. Over the past three decades, there has been a constant reformulation of acronyms for reporting mandates, a flurry of new scoring measures, and an expanding list of organizations that require reporting on utilization, treatment patterns, care outcomes, internal cost analysis, and reimbursement models.   To ensure that patient care at the point of care uses the right guidelines and the right utilization of service, with the right access to care without delay, deft organizations with agnostic analytic strategies should aggregate their enterprise data once and employ data governance to control its variability. These processes will ensure better clinical and financial decision making to optimize treatment planning.  In this way, these organizations will be more prepared for the inevitable changes in the future by employing nimble and flexible capabilities that can help meet the unpredictable legislation and reporting mandates ahead. The ability to deliver optimal care is dependent on data availability, actionable insight, and effective prioritization.  Join our conversation about the next generation of healthcare analytics that supports your organization’s population health, revenue cycle,care transformation, and clinical decision support activities. Register here. ###     Lesli Adams is Director, Healthcare Strategy for Oracle Health Sciences.

The current healthcare delivery landscape is changing dramatically. Major regulatory reimbursement models are evolving from fee-for-service to fee-for outcomes and value. This transformation to value-...

Life Sciences

Interpreting Big, Real World Data – the New Clinical Data Scientist Role

With cloud technologies becoming commonplace to store and manage big, real world amounts of clinical (genetic, BP, temp, etc.) and medical (EMRs, EHRs, outcomes, etc.)  data; and the growing popularity of wearable sensor devices to collect and transmit clinical trial patient data from remote locations on a continuous basis, the research world is brimming in terabytes of information*.  But what does one do with all this data? How can we sort through it to find those points that provide additional support for what is known about a drug’s effect on a disease? Better still, how can it be optimized to demonstrate breakthrough insights and new patterns in relation to the drug and the disease?  These questions pave the way for the introduction of a new research discipline -- data science. In her paper, Michaela Jahn, Global Clinical Data Manager at F. Hoffmann-La Roche, defined it as the application of a team's diverse informatics and analytical capabilities to retrieve and analyze data to support drug project decision making, drug and platform development. Data Science capabilities include bioinformatics, imaging informatics, biostatistics, data integration and visualization, text-mining, information science.   The Clinical Data Scientist This all sets the stage for the role of the clinical data scientist. Driven by the changes in data management, this role can be seen as an evolutionary step forward for the clinical data manager. Where the data manager was task driven and in reactive mode -- organizing and readying data for analysis -- the data scientist takes a more active discovery role looking for new anomalies and patterns in the clinical trial data that may suggest additional paths to insight. Additionally, the skills of the data scientist could prove critical for the clinical trial team. His/her discerning data discovery capabilities could not only help the team to  identify new data paths to explore, but also to guide them away from unproductive, cost/time wasting, data directions.  In her paper, Ms. Jahn defines the clinical data scientist as one who would require: comprehensive knowledge of all areas of data management and data delivery, an understanding of protocols, [the ability to interpret clinical study data, and knowledge of technologies needed for clinical studies from start-up to completion. She also recommends that the scientist be viewed as an equal partner in the study team and have the following attributes: • Have a clear understanding of protocols, their structure, primary and secondary endpoints (the accurate collection and extraction of data.) • Own oversight of study milestones and what is needed for data delivery • Conduct a data risk assessment (what data should be cleaned for a certain therapeutic area and what data can be left as is) • Understand the basics of statistics and programming • Support international standards • Understand the basics of the disease area • Oversee the external service providers • Adapt to new technologies • Help clinical scientist understand the data modeling and explore the data  The paper goes on to identify how the clinical data manager can become a clinical data scientist by evolve his/her skills in these areas.  Oracle Health Sciences solutions can help data managers grow into the data scientists by seamlessly eliminating the need to focus on the clinical trial software infrastructure, enabling the automation of daily intake tasks, and allowing budding data scientists to focus on using the solutions as tools for new and advanced data insights.  For instance, Oracle Big Data Discovery can turn massive amounts of raw, structured and unstructured data from systems like Hadoop into new insight in minutes. Oracle Health Sciences Data Management Workbench (DMW) provides an end to end, clinical data management solution for data cleansing, integration, and analysis. Oracle Health Sciences Clinical Development Analytics offers fast, fact-based insight into clinical programs for business decisions, increased R&D productivity and optimized, drug development efficiency. Oracle Health Sciences Cohort Explorer enables the clinical trial team to have self-sufficient capabilities to analyze and identify clinical cohorts. Oracle Health Sciences Empirica Study detects potential problems early in the pre-marketing clinical stage, enabling clinical R&D professionals to gain deeper insight into a drug in development’s safety profile. Oracle Healthcare Foundation offers actionable analytics for population health, precision medicine, and value based care.  Glassdoor lists the data scientist, as one of the 25 best jobs in America in 2016.Though, in a recent article Cio.com comments that currently, not only is there a lack of qualified talent for this emerging role, but also that companies hiring data scientists are still grappling with the most effective ways to utilize their skills. The amount of data collected in clinical trial will only grow larger.  In clinical trials, it will fall to the clinical data scientist to see new patterns and find new relationships in this data that eventually can save more lives and achieve better patient outcomes.  Finally, it would be interesting to combine the clinical data manager’s deep experience, as he/she evolves into a data scientist, and the new, creative, though less experienced perspectives of millennials coming into the clinical trial industry. Where experienced data professionals have deeper insight into data exploration, millennial scientists might provide fresh, unexpected attitudes on new data sources and combinations. Perhaps together these groups can optimize R&D data discovery even further.   ###   *The human genome typically includes a few gigabytes per person. Also, simply taking blood pressure (BP) three times/day in a two year, 500 subject, clinical trial is two million data points. (BP is two data points (Systolic, Diastolic). Therefore, 2 x 3 per day x 365 days x 2 years x 500 patients = about 2 million.) James Streeter is Global Vice President Life Sciences Product Strategy for Oracle Health Sciences.  

With cloud technologies becoming commonplace to store and manage big, real world amounts of clinical (genetic, BP, temp, etc.) and medical (EMRs, EHRs, outcomes, etc.)  data; and the growing...

Life Sciences

Real World Data vs.Real World Evidence

Now that advanced cloud technologies enable the collection, storage, and analysis of petabytes of information, pharmaceutical and biotech companies often use this information in the form of real world data (RWD) and real world evidence (RWE) for a wide variety of purposes including: health economics and outcomes research (HEOR), pricing, unmet needs, discovery/pre-clinical processes, and clinical R&D).  This post though, focuses on using RWD/RWE in Clinical R&D, and how these data can provide additional sources of proof for the safety, efficacy, and value of new drugs and therapies. Real World Data In Clinical R&D, though the terms real world data and real world evidence are used interchangeably, they are not the same. In a recent article, Accenture’s Jeff Elton explains that real world data is information gathered “…from myriad of sources -- the current standard of care, gaps and deficiencies in the care model, and patient reported outcomes -- that when linked together provide a view of a patient’s health history that can be acted upon using insights from advanced analytics.” Real World Evidence Conversely, real world evidence, which Dr. Elton describes as “a product of analyzed real-world data” and which can be generally recognized as “key conclusions that could be derived from [among other real world data sources] published studies in peer-reviewed journals,” can offer new, actionable insights. Dr. Elton goes on to say that “real world evidence involves using the growing wealth of real-world data, increasingly at the population level, to generate meaningful insights.”  For instance, groups within the American Society of Clinical Oncologists and the National Comprehensive Cancer Network use published studies to develop their recommendations for new treatment guidelines. Dr. Elton points out that advance analytics – using machine learning—can uncover new real world evidence (in the form of new data relationships and patterns) from real world data, and this can influence patient treatment planning. Real world data analytics output can also influence changes in clinical trial design – “how medical affairs experts may identify the ‘long responders’ to specific treatment approaches, and how commercial organizations evaluate the effectiveness of patient services programs.”  Pharmaceutical companies, too, can use real world evidence to uncover new treatment directions for specific conditions, establish follow-on research planning, develop value dossiers, and inform medical communications. The Patient-Centered Outcomes Research Institute is conducting innovative real world research studies testing therapeutic effectiveness “in a broad routine clinical practice.”  “Real-world evidence,” Elton states, “provides the validation between the results seen in regulatory clinical studies that initially supported approval, and the post-approval validation process, which showed that consistent or improving benefit was being realized.” Real world evidence can also supply health insurers with a means of assessing support of patient use and reimbursement charges.  Real World Data, Real World Evidence, and Oracle Oracle supports the evolving areas of real world data and real world evidence, not only for Clinical R&D, but also how they can impact the variety of health and research related areas mentioned above (in the first paragraph).  Keeping the focus on this post’s “RWD/RWE for Clinical R&D” angle, Oracle solutions can gather and transmit real-world data from a patient participating in a research study from the comfort of his or her own home. As a clinical trial participant, the patient wears a connected mHealth sensor with a unique identifier. The device remotely and continuously collects his/her real-world patient data, such as blood pressure and blood glucose levels, then sends the information, via Bluetooth, to the patient's mobile device. From there, the data is routed through Oracle IoT Cloud Service, an enterprise-class, highly secure, scalable, cloud repository that aggregates, summarizes, and disseminates the targeted data into Oracle Health Sciences InForm, or Oracle Health Sciences Data Management Workbench (DMW) to ready it for analysis. The data analytics are then combined with other patient data collected for the clinical trial and continue on to biostats efficacy and safety analysis. Ultimately, this data can become real world evidence submitted for regulatory approval of a new drug or therapy. Oracle’s Real World Data Analytics Platform for Life Sciences , which includes DMW, Oracle Life Sciences Hub and Oracle Health Foundation, and optionally , Oracle Big Data Discovery, Oracle Big Data SQL, and Oracle Big Data Appliance can analyze the data to provide the new insights into clinical trial progress, adverse events, and study outcomes.  Together they can add additional clinical trial results support for the efficacy safety, and value of new drugs and therapies, ultimately, saving more lives.  Oracle Health Sciences is always looking ahead to support the important data trends that drive innovation and advancement for today’s life sciences industry. Subsequent posts, will describe how Oracle supports RWD/RWE in additional areas for biotech and pharmaceutical organizations.  Want to learn more about Oracle’s RWD/RWE solutions? Contact: healthsciences_ww_grp@oracle.com ###   Greg Jones is an Enterprise Strategy Architect with Oracle Health Sciences.

Now that advanced cloud technologies enable the collection, storage, and analysis of petabytes of information, pharmaceutical and biotech companies often use this information in the form of real world...

Life Sciences

Data Quality: A Critical Factor in Risk Assessment

Risk is an important factor to consider when designing and conducting a clinical trial. But, in the clinical trial, what is risk, really? It’s all about the data and the quality of that data. What is being measured? Are data sources validated? How many sites are involved? Are they all measuring the same things? How accurate is the data? On what was it measured? How often was it measured? Today, collecting and sharing clinical trial data is easier than ever with the aid of cloud-based technologies. But, there still is no guarantee that the resulting data metrics from all systems can identify any risk in a given trial or help researchers make decisions about protocol changes or monitoring of a conditional event. To do so, a recent Applied Clinical Trials article contends, there needs to be cross-industry, standardized, metrics tools that track data quality performance and identify risk factors. Over 15 years ago, the Tufts Center for the Study of Drug Development (CSDD) first emphasized the importance of using standardized performance metrics for clinical data quality (and therefore risk). In 2015, the Metrics Champion Consortium (MCC) defined some of the benefits that can be derived from these kinds of tools: Establishing clear, consistent performance expectations for internal and external operations. Facilitating adoption of best practices across sponsors and services providers. Ensuring consistent measures [which] reduces the garbage in-garbage out problem. Avoiding the cost of custom programming. Supporting comparison of performance across all studies within an organization, including across multiple vendors. Decreasing time spent trying to understand what is being measured and focusing on achieving meaningful process improvement. Oracle Health Sciences, Data Quality, and Risk Assessment Additionally, Oracle uses its CTMS system for risk based monitoring centralized issue management - tracking. As risks are identified within a clinical trial via review of patient or site data in any of the Oracle Health Sciences products (including Oracle Argus Safety, Oracle InForm, Oracle CDA), or by a partner solution (such as Cluepoints CSM ); the Oracle CTMS solution provides a centralized location to track and record the risk, the mitigation action for the risk, and the result. This data is then available in CDA to provide a consolidated report of risk, actions, and resolutions for submission with the study to regulatory agencies. The MCC also created standard definitions for common data elements from site activation to database lock. Additionally, specific to its Risk Based Monitoring initiative, TransCelerate has defined standardized tools for assessing and monitoring risks. The Oracle Health Sciences team, understanding the importance of data quality, endorses standardized tools and metrics, and supports standardization by building accepted standardized, industry metric tools into its clinical trial solutions. These tools help clinical researchers identify risks and pinpoint trial events that require further tracking or intervention. Oracle has moved TransCelerate’s* Risk Assessment Categorization Tool (RACT) into Oracle’s Siebel Clinical Trial Management System (CTMS) . Taking a holistic view of how risk can affect the entire trial lifecycle in any functional area of the study, it helps study planners identify potential study data factors that will require risk management from the trial planning through analytics. Through the data results, Oracle’s CTMS solution can identify risk areas for a drug and adjust for it, based on the results per trial. It can then roll-up those results and provide deeper insights on the full trial risk at the program level. Oracle has also incorporated TransCelerate’s Key Risk Indicators into out of the box dashboards in its Oracle Health Sciences Clinical Data Analytics (CDA) solution. With this tool in CDA, study teams can set up thresholds to gauge the status of each indicator at the study or site level and determine appropriate actions to be taken based on the Key Risk Indicator, as defined in their monitoring or quality plans. Standardization of data and reporting capabilities will become even more critical as more clinical trials are implemented with a risk based monitoring approach. Ensuring there are standard processes and data reporting capabilities will allow companies to assess their risk based monitoring approaches and actions to ensure effectiveness. Read more on Oracle’s view of source data validation and risk assessment in the white paper Beyond SDV: Enabling Holistic, Strategic Risk-Based Monitoring More questions? Contact: healthsciences_ww_grp@oracle.com *Transcelerate Biopharma Inc. is a nonprofit organization with a mission to collaborate across the biopharmaceutical research and development community to identify, prioritize, design and facilitate the implementation of solutions to drive efficient, effective and high-quality delivery.

Risk is an important factor to consider when designing and conducting a clinical trial. But, in the clinical trial, what is risk, really? It’s all about the data and the quality of that data. What is...

Life Sciences

Application of Artificial Intelligence for Clinical Development

The application of artificial intelligence (AI) is becoming ubiquitous in our daily lives.  Throughout the course of a typical day,  one uses a variety of applications and devices that automatically understand what is spoken and provide near real-time feedback to support decision-making at an unprecedented scale.  The question being asked now is how can Machine Learning (ML), one of a number of AI techniques, be used in the clinical development space, and, does it hold any value to accelerate development timelines and/or reduce development costs?   About  Machine Learning and Natural Language Processing ML encompasses a variety of algorithmic techniques that can be used to identify and infer patterns to support enhanced/automated decision making.  Natural Language Processing (NLP) is an algorithmic technique used in the ML space.  Using NLP, an application can be used to ‘read’ scientific text and infer the semantic context of the text, so that a human can search and find information more easily.  See the Supervised ML Figure 1 below.   In the clinical context this could include: Given a patient’s historical health data, predict propensity of a disease Given historical data about clinical trial locations, predict their risk profile Given a set of documents, build document clusters so that documents that are about the same topic are in the same cluster Find all the groups of patients that are similar to each other It is clear with the examples above, that the promise of machine learning has the potential to enhance decision making significantly throughout clinical development and beyond.  The principal benefit with ML and NLP is that they can be used to replace the analysis work performed by a human and can be infinitely scaled up as the volume and variety of data grows. What’s Behind the Magic? ML uses an algorithmic approach taking both structured and/or unstructured historical data through a mathematically driven process to generate a model that can recognize patterns and contextual meaning.   This process takes as its inputs:     Training data sets, used to train the ML algorithm and much larger input datasets used for subsequent analysis Standard medical dictionaries that provide reference word or term definitions Ontologies or textual annotations that describe relationships between terms A probability driven mathematical model that can be “trained” to address the types of input datasets to be processed (ex. scientific literature vs. Twitter feeds, each of which have very different contextual, semantic and linguistic characteristics)  Using these inputs, the ML algorithm can be trained to read the training datasets and extract relationships between the terms.  Additionally, a human can verify the accuracy of the relationships and “fine train” the algorithm, so that it reaches a level of comparable accuracy.  At the end of the training process, the larger input datasets are processed by the ML algorithm to extract new relationships to build a bigger picture and better understanding of the area of interest.  A significant advantage with ML is the ability to scale.  Once the algorithm has been trained, it can be used across ‘big data’ data sets. How Is Oracle Health Sciences Using Machine Learning? Oracle Health Sciences (OHS) has made ML part of its clinical platform technology to provide new and innovative business capabilities to industry.  Within Oracle there are world class ML experts, who, between them, have decades of expertise implementing AI based solutions across a variety of industries.  OHS is actively collaborating with Oracle Labs, in particular the Information Retrieval and Machine Learning Group (IRMLG).  As part of this collaboration and to illustrate the utility of AI, Oracle has developed a safety solution that uses ML to identify new Adverse Drug Events from a variety of data streams.  This solution uses OHS solutions and the expertise of the IRMLG to provide an intelligent process for automated case intake processing.     Fast Forward to the Future So what does the future hold for AI applications supporting drug development?  For the first time there is a nexus among the widespread availability/ accessibility of data, availability of AI tools, and very low cost of computation, all driving the development of AI techniques to augment and support what are currently highly human centric activities.  Only within the last few years, it’s apparent that there’s been an acceleration from typing to dictation on our mobile devices without a second thought.   This is primarily driven by ML techniques applied to big data.  Looking forward, the explosive growth in the area of Data Science, which capitalizes on AI technology, is set to deliver new capabilities that will transform how to use and learn from data.  For example, it is quite plausible that clinical development will benefit from AI by identifying new drug candidates using scientific literature, by optimizing clinical study execution via predicting study performance and execution and by assisting with post study statistical analysis. Srinivas Karri is Clinical Warehousing Cloud Strategy Director for Oracle Health Sciences.  Special thanks to Pallika (pallika.kanani@oracle.com) and John (john.k.mclaughlin@oracle.com) for their contributions to this article.

The application of artificial intelligence (AI) is becoming ubiquitous in our daily lives.  Throughout the course of a typical day,  one uses a variety of applications and devices that automatically...

Health Sciences

Clinical Research as a Care Option

Today the trend in healthcare is to provide the patient with as much value as possible. Though value comes in many forms, an important one that is often given less priority than others is patient engagement.  That is, enabling the patient to take a more active role in the treatment of his or her condition.  One study by the UK’s Centre for Health Policy focusing on patient engagement sought to shift the clinical paradigm from determining “what is the matter?” to “what matters to the patient?” Researchers found that increased patient engagement improved health outcomes and reduced costs, while it also aided insights into the data around the condition in question. One avenue for increased patient engagement is giving the patient the information that will enable him/her to participate in a clinical drug or therapy trial as a care option targeting his/her condition. A recent survey conducted by Eli Lily , Wilmington Health, PMG Research, Quintiles, Pfizer, and Harvard asked patients about their participation in an ongoing, four year, clinical trial on diabetes.  “The survey results consistently demonstrated a high and increasing level of patient satisfaction and engagement with clinical research, including satisfaction with access to care, efficiencies in care delivery, and the quality of care provided by the research staff,” said Katherine Vandebelt Global Head of Clinical Innovation, Eli Lilly and Company in a recent bylined article. The survey reported additional patient comments including:  “It [clinical trial participation] made me much more motivated to work on my diabetes.” and “Study participation has allowed me to manage my diabetes better than I ever have before.” Other survey findings were just as glowing. One hundred percent (100%) of participants thought that participating in clinical tresearch reducedtheir overall cost of heatlhcare and improved their overall interest and invoivelment in their overall helathcare. Ninely five percent (95 (% )thought that their participation had improved their overall quality of care.   Quintiles: Clinical research participation as a care option. 2015. [Whitepaper] The irony of these results is that less than one percent (1%) of Americans participates in clinical trials. Yet, a whopping 72 percent (72%) say they would if their physician allowed it. However, it is difficult for a patients to find a trials that fit his/her personal condition, let alone have a discussion with his/her physician on what is best for his/her specific case. Vandebelt advocates considering clinical trials as a medical care option because they have the potential to improve overall health outcomes. She goes on to explain that, unfortunately, today, clinical trials are treated as something separate from care. She says that needs to change She continues “ …This paradigm shift [to regard the clinical trial as a commonplace care option] will arise only from collaboration among patients, healthcare providers, health care systems, drug developers, and policy makers to realign clinical trials to center around the needs of the patient, not just around the collection of clinical data. When the model is structured to reward patient-centricity, new opportunities emerge to reinvent the value proposition for clinical trials, and clinical research can be more readily integrated into the overall continuum of care.” Oracle endorses patient engagement and centricity – including acknowledgement of the patient as the very important source of real world data in clinical trials and post-trial marketing research—via its mHelath initiative. This effort enables patients to participate in clinical trials and supply their real world data from the comfort of their homes, via wearable sensors sending data to remote devices, which upload the information to the Oracle cloud and then into Oracle Health Science solutions. According to the Quintiles white paper on the aforementioned survey, patients are more engaged when they can participate in managing their personal health, which not only leads to lower total cost of care and better outcomes, but also to improved patient satisfaction. James Streeter is Global Vice President Life Sciences Product Strategy for Oracle Health Sciences.

Today the trend in healthcare is to provide the patient with as much value as possible. Though value comes in many forms, an important one that is often given less priority than others is patient...

Life Sciences

From Big Data to Smart Data

Over the last few years we have all been inundated with the concepts of Big Data. There are 6,000 tweets from Twitter every second. Facebook stores 30 petabytes of user data. A Boeing Dreamliner generates one terabyte per flight. A Formula One car generates three terabytes in one race. Now we know why it’s called “big data”. In Health Sciences, big data is seen in a number of forms, such as the human genome (typically a few gigabytes per person, sequenced today at a cost of $1K vs. $10M ten years ago), vast amounts of epidemiology data (claims, safety, post-marketing data) and sensor/device data. Simply taking blood pressure three times per day in a two year, 500 subject clinical trial is two million data points. This is a lot of data. It could really change the way research is  perform research, how new therapies are discovered, and how patients are treated. However, it’s also a lot of stuff. You know, that ‘lot of stuff’. That’s what’s in your cupboards, garages, and houses, that you never, ever look at, and typically, have no idea what it is. You just know it’s taking up space and gathering dust. So, it’s better to use the term Smart Data. While this is not a new term, it is becoming more widely used. It also helps users move away from the over hyped “Big Data” and understand the opportunity of big data. The key to using this data is to understand how to store, explore, aggregate, analyze, present and act on it. To have meaning, and hence to be “smart,” data needs to have context. It needs to be correlated with something. There is no point in collecting heart rate every second of a two year clinical trial, if the heart rate data cannot be aligned to the exact date and time that the investigational drug was taken. When that analysis is done, it may very quickly show that Caucasian male subjects, aged of 35-45, show a 30%-60% increase in heart rate (HR) four to eight minutes after a dose, as compared to females of same profile. By further combining this physiological data with genomic data, it may become apparent that a further subset (cohort) of subjects with a particular genetic marker show a 70%+ HR increase, and hence, a safety issue for one cohort may be found.  We are now sailing, full speed ahead, into a new world of combining clinical, genomic, and healthcare data in ways that we could not have previously imagined. These new combining capabilities allow us to extract meaning from big data and turn it into smart data that can drive new therapies and advance clinical care. This is getting exciting!

Over the last few years we have all been inundated with the concepts of Big Data. There are 6,000 tweets from Twitter every second. Facebook stores 30 petabytes of user data. A Boeing Dreamliner...

Data Management

When Good Data Goes Bad by Srinivas Karri

Recently, the FDA published an article describing a notification it had issued to a contract research organization (CRO). The notification required certain bioequivalence studies to be repeated based on improper data collection and analysis processes. Bioequivalence studies, often conducted to bring generic drugs to market, establish that the generic drug has the same ‘effect’ as the original drug.  Obviously, these are critical trials, as they also establish the safety profile of the generic compound. The consequence of notification has been quite substantial.  It has been recommended that a number of generic drugs be removed from the market until accurate data is collected by repeating these trials, incurring significant, additional expense and impacting sales. According to an article in PulseToday.com, “European Medicines Agency (EMA) advisors said bioequivalence studies carried out on the drugs at Semler Research Centre in Bangalore were ‘flawed’ and ‘cannot be relied on...The EMA advisors concluded ‘the studies conducted at Semler cannot be accepted in marketing authorization applications in the European Union’ and, therefore, ‘no medicines can be approved on the basis of these studies’.” So what went wrong? Over the last decade organizations have had to become increasingly more stringent on their ability to collect and manage clinical trial data.  This has primarily been driven by the 21 CFR Part 11, GCP and GAMP regulations which describe methods and processes that need to be put in place to ensure that only authorized individuals have access to trial data and that adequate controls exist to prevent modification to data.  Clearly, at one CRO using spreadsheets to manage laboratory data, this method was insufficient.  There was clear evidence to show that by using spreadsheets to store and manage lab data, the CRO in question had manipulated data to create false results.   Fixing the problem Having looked at the announcements and recommendation, CRO should have deployed a data collection and management platform that would ensure that data could not have been improperly manipulated.  More specifically, the collected data should have been securely collected while allowing for downstream analysis and processing.  In this regard, Oracle Health Sciences Data Management Workbench (DMW) would have been ideal to support the CRO, and perhaps, should have been considered when building out its data management strategy. If you’d like a further description of our capabilities to support CROs and sponsors with building a data management platform see the attached for a set of capabilities and value statements for Oracle DMW. Srinivas Karri is Clinical Warehousing Cloud Strategy Director for Oracle Health Sciences. 

Recently, the FDA published an articledescribing a notification it had issued to a contract research organization (CRO). The notification required certain bioequivalence studies to be repeated based...

Life Sciences

Oh, That Looks Cool! – Riding the mHealth Wave

Innovation comes in many shapes and sizes from a unique user experience to game changing, disruptive, enterprise process changes. Innovation is something that's great to experience and deliver. It’s often detected by an initial emotional response typically triggered by some visual stimulus (or claim), which is somehow instantaneously converted to a real or perceived benefit. Conversely, innovation can become quickly blurred by hype, distracting us from the actual, realizable benefit. mHealth is a classic case of disruptive innovation in action. Thousands of vendors are active in this space from dot.com startups to mega-corps, such as Apple, IBM and Google. This is a clear indicator that‘something is happening’. But such huge market fragmentation is indicative of market confusion, ongoing pilot-itis, and a pre-cursor for mass vendor consolidation in the future. The promise of mHealth offers us all huge potential. The ability to capture real world data, direct from patients will be transformative. The ability to measure patient heart activity continuously via a sensor, analyze in stream, and enable a physician to consult remotely, is a significant change from today’s periodic, point in time model. Applying this to clinical trials, capturing this data can lead to reduced site visits through remote consultations, increased compliance to the drug studied, and most importantly, the collation of real world evidence to prove real outcomes.                     Geoffrey Moore’s marketing classic Crossing the Chasm illustrates the life cycle of new products. Once you have crossed the chasm you have a stable market that knows what it wants. Right now mHealth is certainly in the Early Adopter phase, and is likely to be there for the next couple of years, at least. But its potential “post chasm” is significant. From a strategy perspective it is clear that the pharmaceutical market is starting to emerge from a few, limited scope mHealth pilots to challenge regulatory barriers. Typically, pharmas, having tried a few devices and mobile apps, have seen both clear and fuzzy benefits. Now they are looking to scale. Oracle’ Health Sciences’ strategy is to build an enabling platform to Just Choose and Use. Pharmas should not be dependent on a plethora of disconnected vendors, such as managemydiabetes.com or neverfeelwheeszyagain.org, with associated proprietary apps and devices and multi-million dollar price points. For a Clinical Trial Program Lead it is critical simply to identify a desired device, plug it into a platform, and start a clinical trial with the comfort that data can be acquired,that it can be explored, and that it can be exploited downstream. Our strategy brings together the best of Oracle’s latest core tech PAAS services (such as Oracle Internet of Things Cloud Service) combined our best apps (Oracle Health Sciences InForm and Oracle Health Sciences Data Management Workbench), as well as our Big Data stack, to support a complete, end-to-end process from patient device acquisition.                                                                                              

Innovation comes in many shapes and sizes from a unique user experience to game changing, disruptive, enterprise process changes. Innovation is something that's great to experience and deliver. It’s...


Balanced Incentives and Healthcare Reimbursement

Healthcare payment reform is shifting again. At the rate of tectonic plates over the last 60 years, there have been several milestones that have utterly transformed the face of reimbursement. Diagnosis related Group (DRG) reimbursement, Managed Care capitation, and now the Medicare Access & CHIP Reauthorization Act (MACRA). MACRA will be a tsunami for Medicare reimbursement. With the Alternative Payment Model (APM) or the Merit based Incentive Program System (MIPS), reimbursement will be a synthesis of quality, technology, practice improvement, and cost reduction. What will MACRA accomplish? As with other historical reimbursement transformations, as Centers for Medicare & Medicaid (CMS) changes, commercial payers follow. For the MIPS alone, CMS will tie 30 percent (%) of their payments to performance outcomes by 2016, 50percdent (% ) by 2018, and commercial payers are targeting 75percent (%) of their payments by 2020. “Outcomes” involve a combination of clinical quality, resource use “cost”, health IT meaningful use, and clinical practice improvement activities (CPIA). MIMPS is the balanced approach. Currently 50 percent (%) of the performance weighted is based on quality and PQRS reporting. By 2021, the balance will be 30 percent (%) quality, 30 percent (%) cost reduction, 25 percent (%) technology usage, and 15% practice improvement. The goal is to ensure that participating providers are working collaboratively to transform the delivery of healthcare. CMS is using payment disincentives for those providers not able to meet these outcomes. With the macro level, CMS plan to have a net neutral impact for all Medicare providers on incentives; some will receive bonuses, some will not. Technology should always serve the healthcare provider, at the analytical level. MACRA is an organizational and multi-domain approach to transforming the reimbursement of care provided to those eligible for Medicare. This transcends to commercial payers and all patients due to process changes. How Technology and MACRA Will Transform the Industry Successful delivery of the MIPS activities above will be the synthesis of people, process, and technology: People to deliver the care, analyze patterns and establish quality improvement, care coordination and open access for those in the community, with technology as the critical infrastructure below both people and process. Actionable analytics and precise measurement to support reimbursement and margin analytics can only be attained if the full 360 degree view of the patient is available, consumed, de-duplicated, aggregated and delivered with the minimum amount of latency to the administrators, executives, and the care provider. Technology is necessary to support the business, with the right balance of infrastructure and “at the glass” costing and margin analytics, intersected with clinical and quality outcomes. Oracle and its partners are recognized for the right blend of technology and service offerings that deliver world class data aggregation and top in KLAS population health solutions for MACRA, delivery reform, and value based purchasing. Oracle Healthcare’s Offering Oracle’s Health Sciences has a staff of over 2500 professionals including physicians, PhDs, nurses, and clinical informaticists who guide the technical development and product offering for the healthcare market. Its Oracle Healthcare Foundation (OHF) is a unified healthcare analytics platform for data integration and warehousing providing clinical, financial, administrative, and omics modules. Building upon OHF, healthcare organizations can deploy pre-built business intelligence, analytic, data mining, and performance management applications from Oracle and its partners. These organizations can also leverage OHF’s out-of–the-box, self-service, analytics tools to build customized analytics applications.Currently OHF is implemented in such prestigious health systems as Los Angeles County Health Department, Penn Medicine, and Adventist Health System with 43 locations, and hundreds of individual hospitals participating. Oracle Healthcare Partners Oracle’s HSGBU has an active and expanding partner ecosystem that encourages pre-integration with top in KLAS population health vendors and Big 4 Healthcare advisory service providers. The partner ecosystem is active and mirrors the population health market to fulfill current and emerging business needs and include the full range of population health, evidence-based medicine, risk modeling, ACO, alternative reimbursement models, and value base contracting models (including MACRA). The following are two of the HSGBU’s partners that offer knowledge leadership and successful accomplishment in the MACRA and delivery reform space: Deloitte Deloitte is a Diamond-level partner and has been awarded Oracle’s highest honors for the past six years in a row. Additionally, Deloitte is the largest health care consultancy in the marketplace and a national and global leader when it comes to serving complex, multifaceted organizations. Analysts, including IDC, Gartner and Forrester, have both ranked Deloitte as a leader in EPM and Finance Transformation. Oracle and Deloitte have developed a “Triple Aim in the Box” solution in response to MACRA reform. Triple Aim in a Box combines Oracle’s packaged software applications providing a flexible, scalable platform and leading practice accelerators, with Deloitte’s implementation and integration services. Together, these apps and services enable clients to customize and configure a healthcare solution for their specific needs. This solution has a “start anywhere” approach that gives healthcare organizations the flexibility to advance their analytic capabilities across all three Triple Aim pillars, while prioritizing components that are critical to them. SCIO Health Analytics – Medicare Claim Analytics SCIO Health Analytics is one of the fastest growing technology companies in the US, having featured for four consecutive years on the INC5000 list and twice on the Deloitte Technology Fast500 list. In 2015, SCIO Health Analytics® Ranked as Major Player in IDC Health Insights’ 2015 Payer Analytics Marketscape Report. SCIO has been recognized by CMS with several CMS innovation awards focused on Medicaid members. Recently, SCIO was awarded the CMS Centennial Award for its relationship with the State of New Mexico. Learn more via our on demand webcast, MACRA, Analytics, and the Move from Volume to Value.

Healthcare payment reform is shifting again. At the rate of tectonic plates over the last 60 years, there have been several milestones that have utterly transformed the face of reimbursement....