ISSUE 1 MARCH 2013

Health Sciences
A resource dedicated to the convergence of the life sciences and healthcare industries

Journal

Clinical Data Warehousing

How does the theory translate into practice? An industry perspective
The Clinical Data Warehouse a New Mission-Critical Hub

Jonathan Palmer, Oracle Health Sciences

A Clinical Data Warehouse Solution to Improve Operational Efficiencies

Colin Burns, ICON Clinical Research

A Dynamic Platform for Data Integration, Standardization and Management

Brooks Fowler and Nareen Katta, AbbVie

Clinical Research Innovation through Shared Clinical Data Warehousing

Jerry Whaley, Pfizer

Health Sciences Journal

Issue 1 March 2013

Contents
Foreword 3 The Clinical Data Warehouse a New Mission-Critical Hub 5
Jonathan Palmer Senior Director for Clinical Warehousing and Analytics, Oracle Health Sciences

A Clinical Data Warehouse Solution to Improve Operational Efficiencies13

Colin Burns Senior Director of Global Data and Technologies, ICON Clinical Research

A Dynamic Platform for Data Integration, Standardization and Management18

Brooks Fowler Global Head of Data Sciences, AbbVie Nareen Katta Senior Manager of Data Sciences, AbbVie

Clinical Research Innovation through Shared Clinical Data Warehousing 24

Jerry Whaley Senior Director of Development Business Technology, Pfizer

p2

Health Sciences Journal

Issue 1 March 2013

Foreword
The life sciences industry is changing. The common driver for this change is the need to improve overall efficiency and cost-effectiveness. In response, organizations are rigorously seeking ways to extract maximum value from their assets and become more operationally and economically efficient. Driven by ever increasing cost constraints, regulations and competition, organizations look to create new business models through acquisitions, global outsourcing and collaborative partnerships. The need to better manage the life sciences industrys key asset, data, has become extremely important. Maximizing the value of data can only be achieved through improved data management, standardization, storage, and accessibility which ideally should be universally available to everyone across an organization and its partners. In response to these challenges, clinical data warehousing is evolving from a purely in-house solution to an essential tool for harnessing and maximizing potential from life sciences industry data. It is becoming a business-critical platform that supports decisions across the clinical trial portfolio, is central to collaboration, and fundamental to business survival. This issue of the Health Sciences Journal explores clinical data warehousing from various industry perspectives, with particular focus on meeting business needs, implementation, governance, challenges and solutions, and associated benefits. Jonathan Palmer, a senior director for clinical warehousing and analytics at Oracle Health Sciences, provides insights into this area by defining a clinical data warehouse and describing the drivers for implementation, importance of data standardization, necessary requirements for successful implementation, and how changing trends in the life sciences and technological industries have impacted clinical data warehousing. Increased outsourcing, partnering, and globalization across the life sciences industry has also created the need to improve overall working efficiency and communication between partners and service providers. Colin Burns, a senior director at ICON Clinical Research, provides his views on the role of clinical data warehouses in contract research organizations (CROs). Specifically, how they have been used to convert technical data into a usable format accessible to all users and how this is subsequently used to inform operational decisions. In contrast, Brooks Fowler, global head of data sciences, and Nareen Katta, senior manager data sciences, at AbbVie, provide a pharmaceutical view point on clinical data warehouses. They highlight some specific industry and business challenges that led AbbVie to consider a clinical data warehouse solution. The article explains how AbbVie has used its clinical data warehouse to integrate, standardize, and manage data effectively with a phased implementation, including key challenges and benefits to users.

p3

Health Sciences Journal

Issue 1 March 2013

Foreword (continued)

Pfizer, the worlds largest research-based pharmaceutical company, has responded to changing trends in the life sciences industry by re-assessing its clinical trials operational model, particularly with regard to IT infrastructure. Pfizer is paving the way for a radical approach to data management in the clinical trials space. Jerry Whaley, senior director of development business technology at Pfizer, describes the companys vision to build a new clinical data warehousing platform that can be shared by companies across the industry. The premise is to refocus company resources on scientific discovery for better healthcare, rather than just on the management of clinical trial data. Pfizers vision is to find other companies with similar mindsets with whom such a platform can be shared for mutual gain.

p4

Health Sciences Journal

Issue 1 March 2013

The Clinical Data Warehouse a New Mission-Critical Hub

Jonathan Palmer is senior director for clinical warehousing and analytics at Oracle Health Sciences. Jonathan first joined Oracle in 1997 where he has had various roles across business development, consulting, and product development. Since 2008, Jonathan has been involved in defining new product strategy targeted at developing innovative solutions for the life sciences industry.

Jonathan Palmer Oracle Health Sciences

Demands for clinical warehousing are increasing dramatically. From being viewed as a data gathering tool, a clinical data warehouse is now moving into a new phase of becoming a business-critical platform. Such a platform can support all clinical decisions across the clinical trial portfolio, be central to collaboration, and fundamental to the survival and agility of the business. To fully comprehend the meaning, applicability and relevance of a clinical data warehouse, and how it can potentially benefit a company, we must understand how it is different from a typical data warehouse, what has driven its need, and how its adoption can be maximized to drive clinical development. Traditional warehousing versus clinical data warehousing Traditional business data warehousing is a well-established IT discipline, the primary focus of which is often to deliver decision support capabilities to drive productivity and efficiency gains across a business. Typical examples can be horizontally focused, such as in inventory management, or in industries such as banking, telecommunications or retail. These warehouses are based on well-defined structures, data sources and goals. For example, a warehouse focused on inventory management allows a business to manage stock, assess sales, and answer well-defined questions to drive efficient stock management in response to sales activity.

p5

Health Sciences Journal

Issue 1 March 2013

The Clinical Data Warehouse a New Mission-Critical Hub (continued) In contrast, the clinical trials segment of the life sciences industry is unique in its needs based on the variability of its data structures driven by trial design. The primary focus of a clinical data warehouse is to facilitate extraction of value from clinical data. This can aid study design, prove efficacy and safety of new products, and support regulatory queries. Productivity and efficiency gains, whilst important, are rarely the key focus of a clinical data warehouse. Traditional warehouses are less focused on regulatory compliance, and the need for full traceability of data life cycles is often less relevant and compared with the high demand for such capabilities from the life sciences industry. As with all warehouses, a key feature of a clinical data warehouse is that it should allow a company to store and release value from data assets to drive better decisions. For example, it may enable a pharmaceutical company to realize value from the vast amount of clinical data generated from a single trial, all trials in a particular program, a therapy area or, indeed, all the trials in the company. The ability to standardize, pool, analyze, explore and mine data across large and disparate data sets has previously been challenging for the life sciences industry. Whilst a traditional warehouse uses only a finite set of data sources, there are potentially hundreds of data sources in the clinical space, each with a different structure, variability and frequency. Drivers for change Historically, data have been managed and stored in distinct silos in a function-centric model for example, data for the clinical data management, biostatistics or safety groups. Whilst synergies exist across these groups, often data in silos cannot be accessed by other teams. This often creates data lag, as data must be requested from one group to another, and manually handed over. Transparency and availability of data across the breadth of the organization has been a challenge. As the industry has evolved, there has been an increasing need to access, combine and share data across multiple information domains. Further, there is the need to effectively align clinical and administrative data to provide a complete picture of study conduct, from an operational, safety and regulatory perspective. There have been a number of key changes in the industry including: 1) Industry consolidation 2) Cost constraints 3) G  lobalization, outsourcing and virtual technology 4) Aggressive generic drug replacements 5) I ncreased scrutiny by governmental and regulatory bodies.

p6

Health Sciences Journal

Issue 1 March 2013

The Clinical Data Warehouse a New Mission-Critical Hub (continued) These drivers for change have focused the attention of senior management on the need for greater data visibility, transparency and availability. Information silos have become barriers to clinical innovation, and organizations are increasingly seeing clinical data warehousing as a platform to support more agile clinical research, and as an essential base to maximize the clinical portfolio. Figure 1: A clinical data warehouse to enable collaboration through shared visibility Clinical data warehousing is becoming ever more important. Acting as a central hub for information storage, collation and archiving, a clinical warehouse is essential to delivering a consistent, single view of the data assets across an organization. It is essential that all team members, both internal and external, have access to a consistent view of the data to drive clinical research (Figure 1). Standardization and master data management Standardization is fundamental to combining and sharing data. Whilst most industries are based on well-defined standards, the clinical trial industry has struggled to agree on a universal standard due to the broad nature of trial data. Standards organizations such as the Clinical Data Interchange Standards Consortium (CDISC) and Health Level Seven International (HL7) have made significant contributions over recent years. Complete adoption of standards across the clinical portfolio is now recognized as being fundamental to enabling efficient process design, and extensive usage of standard software tools for analysis, exploring, and mining without significant reliance on trial-centric programming resources. Submitting standardized clinical trial data to regulatory authorities (for example, the FDA or EMA) allows analysis and review of submissions using standardized methodologies and tools, which streamlines processes and regulatory reviews, potentially resulting in faster drug approval timelines. As with all standards, they are constantly evolving and expanding, and must make provision for data which cannot be standardized. Variation is typically focused around efficacy data, and platforms are available to support this variability.

p7

Health Sciences Journal

Issue 1 March 2013

The Clinical Data Warehouse a New Mission-Critical Hub (continued) Another consideration is management of master data, which typically relates to process-centric entities (for example, address details of principal investigators or trial sites). If master data are not recorded and managed correctly, critical contextual information may be lost, leading to expensive manual data cleansing and sourcing to complete the formal regulatory submission. Master data management, although well-known in general data warehousing, is a relatively new issue in the clinical trials data space. Of course the problem is, as soon as you find one data inconsistency, then everything is gonethe whole structure is compromized, and the research is at risk. Evolving from clinical data management systems (CDMSs), to clinical data warehouses Historically CDMSs have been used to capture case report form (CRF) data and support cleaning cycles. Over the years the role of the CDMS has changed. Organizations typically sought to import as much data as needed into CDMSs to support new product submission. Laboratory data were often imported into CDMSs, which evolved into an idealistic single source or pseudo data warehouse. However, as clinical trials became more complex and data source variability became broader and more diverse, it became impractical to load all data into the CDMS. Conversely, the evolution of CDMS systems to electronic data capture (EDC) moved the focus from manual data processing to site-based data capture. Whilst succeeding in accelerating CRF data capture, the need to manage non-eCRF data sources remained unaddressed. Today many organizations pass the burden of
p8

data consolidation and integration onto data management programmers and/or biostatistics programmers, to manually combine a file-based data source with the eCRF-based data. This process is labor intensive, requires specialist programming skills, and can result in quality issues if data standards are not leveraged effectively. Using CDMS systems as a pseudo data warehouse, augmented by file-based approaches, inherently fragments the data. For a single drug trial this approach may be adequate as it is possible to navigate through the different files and obtain the required data for drug approval. However, to view all trials in a therapy area (for example, diabetes) and to explore across related trials becomes a complex undertaking as data are stored in separate files, requiring specialist programming skills to access and analyze. So has the industry survived? Sure. Companies have got new drugs to the market, but are often missing an opportunity to exploit huge inherent value embedded in their vast data stores. Why? Because in many cases the data are locked away, in a secret vault that only the biostatistician or statistical programmer has access to. As organizations have undergone this system evolution, it has become increasingly apparent that force-fitting all data into data capture systems (CDMS or EDC) is not a viable solution. Furthermore, file-based stores are inherently difficult to search and mine across. This reinforces the role of the clinical data warehouse as a purpose-built relational data store, which can flex, expand and scale to meet the varied needs of clinical trials.

Health Sciences Journal

Issue 1 March 2013

The Clinical Data Warehouse a New Mission-Critical Hub (continued) Globalization and outsourcing a new paradigm The life sciences industry is continually under increasing pressure to become leaner and reduce healthcare payers costs, whilst ultimately increasing end value to shareholders. Ongoing industry consolidation is partly driven by the need to be more efficient and the need to find ways to expand portfolios, and defend against aggressive competition. Twenty years ago the industry was incredibly cash rich and blockbuster-centric. Weve seen over recent years massive consolidation through mergers and acquisitions, huge patent expiries, and healthcare payer budget cuts. Clearly the industry needs to be a lot more agile and innovative. The world is becoming smaller through more integration, virtualization and collaboration. Enabling technologies, like the Internet, have opened up new, previously unavailable business models. Workforces are becoming more geographically distributed across different time and language zones so as to lower costs. To meet the demands that these changes bring, the role of the clinical warehouse becomes ever more important. The clinical warehouse is evolving from an internal knowledge base into a hub for leveraging new business models. For contract research organizations (CROs), clinical warehouses had previously been irrelevant due to the studycentric processing model. However, many CROs are now focusing on building warehouses that can act as a centralized and standardized platform on which they can add tools for extracting value from data. This allows CROs to differentiate themselves from their competitors, from commodity vendor to strategic partner. Similarly, as pharmaceutical companies look to reduce internal costs and optimize processes, they are increasingly leveraging global service providers and using clinical data warehouses as an integration and collaboration platform to enable full service, and hybrid (using both internal and outsourced resources), outsourcing. Challenges in implementing a clinical data warehouse The challenges facing clinical data warehouse implementation are largely determined by the managements view of the implementing company. Forward-thinking senior management view a clinical data warehouse as being fundamental to progression and essential for collaborative innovation i.e. the ability to adapt, be agile, acquire organizations, outsource, and generally be more efficient, cost-effective and competitive. Demonstrating a clear return on investment (ROI) is always challenging as clinical data warehouses are often multi-year programs with abstract, cross-functional concepts. To be successful, a clinical data warehouse requires continuous senior management commitment and sponsorship. It is important to build a broad, holistic picture for the organization rather than at a departmental level. Buy-in solely from one or two department heads is rarely sufficient. However, transition from batch-centric data preparation and programming to better solutions, more in keeping with the dynamic nature of the industry, requires a key company visionary, with the gravitas to communicate the overall benefits of a clinical data warehouse to the wider company.

p9

Health Sciences Journal

Issue 1 March 2013

The Clinical Data Warehouse a New Mission-Critical Hub (continued) The implementation phase of a clinical data warehouse can be challenging. This phase requires essential specialist skills in areas such as process design, data standardization, modeling, and system integration. However, once the clinical data warehouse is in place, fewer specialist skills are required. Key emphasis must also be placed on management of process design and change, user training, adoption, and crossdepartmental co-ordination to ensuring that investment in the clinical data warehouse is maximized. A continuous program of monitoring and driving user adoption, streamlining data flow, and extending and enhancing use cases and tools-sets for data exploration, visualization and analysis are key to ongoing return. These projects do not stop when they go live but evolve with continuous improvement and adaptation (Figure 2). Figure 2: Schematic depicting the general capabilities of a clinical data warehouse
Drive Clinical Innovation Empower Clinical Teams Maximize Data Value Enable Collaboration Simplify Outsourcing Accelerate Regulatory Queries

Identifying use cases For a clinical warehouse project to be successful it is essential that clear and specific use cases are defined before project initiation. A use case can be analogized to breaking up a big problem into bite-sized pieces, targeted at delivering specific business benefit. These could be focused on reducing data handoffs, standardizing data, accelerating statistical analysis, or simplifying medical and safety review. Without use cases to drive project goals and ROI, such projects can become complex IT architecture programs with poorly defined endpoints delivering little business value at completion. Focus on specific use cases allows a clear understanding of the likely process changes and an understanding of the potential benefits of changing these processes. Once use cases have been identified, the right platform can be selected, along with associated technology and consulting services to assist companies in delivery. Phased delivery is one of the key things. Due to the many interdependencies of the components in these projects some companies adopt a Big Bang approach of trying to do everything at once, but these often fail to deliver as return on investment is too long. It is critical to implement use cases that give incremental return. A phased delivery, rather than trying to deliver universally on everything, is advisable. Defining specific use cases at the start, and phasing the implementation over several stages, allows the realization of tangible benefits, while successfully managing stakeholder expectations.

p 10

Health Sciences Journal

Issue 1 March 2013

The Clinical Data Warehouse a New Mission-Critical Hub (continued) Extracting full potential It is imperative to create a learning organization which can evolve, extend and expand, to extract ongoing value from a clinical data warehouse. There must be a conscious effort towards creating alignment between IT and business units of an organization. The clinical data warehouse platform should be an enabling platform from which the business can gain significant, repeatable value, for example improving response time to regulators, identifying potential new therapies, or moving towards individualized medicine. This is the ultimate vision for clinical data warehouses. Modeling and simulation is another area from which value can be extracted using a clinical data warehouse. For example, simulations may influence trial design in terms of identifying appropriate recruitment populations or helping to design efficacy parameters. Providing these groups with rich, collated, standardized data via a clinical data warehouse can enable them to effectively and precisely predict outcomes in various models. A good clinical warehouse can truly flip the 80:20 rule for a modeling/simulation analyst. Instead of spending 80% of the day searching and cleaning data for analysis, and only 20% on analysis, they can leverage the data warehouse, find their data quickly, and spend 80% of the day analyzing to accelerate research. The ability to easily mine and explore legacy data can also uncover hidden value, for example, a previously abandoned drug in one disease could be investigated for use in another disease. Likewise, mining and exploring data acquired through mergers and acquisitions could be used to augment current company data, thereby increasing data value (Figure 3). Figure 3: Optimizing collation and search leads to increase opportunity for innovation
Opportunity for Innovation

Value

Data Collation and Searching

Time
Clinical Data Warehouse Traditional Approach

p 11

Health Sciences Journal

Issue 1 March 2013

The Clinical Data Warehouse a New Mission-Critical Hub (continued) Emerging technologies and big data There is increasing recognition of a need for improved management of big data in the clinical space, and efficient aggregation and integration with core clinical data will be key to successful clinical warehouses in the future. Big data are unstructured compared with traditional (fully structured) or clinical (mostly structured) data formats. They can be obtained from sources or formats such as social media and include real world use of prescription and overthe-counter drugs. In this scenario, patients may use social networks to relate their drug experiences, for example, in terms of safety or adverse effects. It is critical for the drug manufacturer to be able to mine these data, recognize potential issues arising, and address or manage them effectively, thus providing pharmacovigilance insights on a marketed drug. Other key sources of big data will come from advances in patient genomic profiling, as well as wearable medical monitoring technologies. Combining such huge data sources with well curated clinical trial data will be essential to delivering individualized or personalized medicine. An approach taken by IT vendors to address big data is to design technologies that combine both software and hardware to support largescale data sources and data warehouses, for example, Oracle Exadata. As the life sciences and healthcare industries converge around delivery of individualized medicine, the need for such database machines will be critical to delivering targeted treatments. By exploiting such high performance data management platforms the industry will transition from drugdriven clinical trials to patient-driven. Final thoughts In addition to, and as a consequence of, the overall move of the life sciences and healthcare industry to become more cost-effective, efficient and competitive, it must improve the overall value realized from its key asset, data. A clinical data warehouse provides a solution for life sciences companies to better access, mine, explore and use data across their trials and portfolios. Furthermore, clinical data warehouses can be viewed as an essential tool for speed of access to data when considering globalization, outsourcing and merger and acquisition activity. Most importantly it can be viewed as a platform for accelerating clinical innovation. Clinical data warehouses represent an exciting area of current development and offer the potential to shape the way we utilize and manage clinical data. They will continue to evolve and, with the convergence of the life sciences and healthcare industry, will become a necessity in order to efficiently drive value from data and ultimately accelerate development of new therapies.

p 12

Health Sciences Journal

Issue 1 March 2013

A Clinical Data Warehouse Solution to Improve Operational Efficiencies

Colin Burns is senior director of global data and technologies at ICON. He leads the development of ICONs enterprise clinical data warehousing capability, the data arm of ICONIK Monitoring service, which optimizes the clinical trial execution strategy for each trial to help clients manage risk and improve efficiency. Colin has more than 15 years of life science and health science experience.

Colin Burns ICON Clinical Research

The changing trends in the life sciences industry in terms of outsourcing, partnering and globalization have created the need to improve overall working efficiency and communication between partners and service providers. Contract research organizations (CROs), such as ICON, understand the intense pressure to reduce cost and timelines for drug development while ensuring data quality. Accordingly, ICON is among the eCROs to have emerged that have the informatics capabilities to re-aggregate clinical trial data in a time and cost-effective manner. Increasingly, CROs now have a stake in clinical drug development through risk sharing with pharmaceutical companies and involvement at every step of the process. Clearly, this new model involves handling larger volumes of data from a vast number of different sponsors and in different formats. For ICON, the implementation of a clinical data warehouse was an obvious solution to manage these large volumes of disparate data. Through the identification of a specific business case and definition of precise project parameters, the company has succeeded in implementing a clinical data warehouse that delivers operational efficiency and, consequently, competitive advantage.

p 13

Health Sciences Journal

Issue 1 March 2013

A Clinical Data Warehouse Solution to Improve Operational Efficiencies (continued) Business challenges/unmet needs leading to a clinical data warehouse The CROs main business challenge, in terms of data, is the increasing volume and variety of sponsors/clients that it typically engages with. CROs typically handle large volumes of disparate data from different sources presented in various formats. One challenge is how to effectively and efficiently access and manage disparate data while maximizing value, increasing usability and maintaining data integrity. ICON realized the limitations of previous approaches to data handling which were unworkable and unsustainable over the long term. It identified an urgent need to move away from traditional and ad hoc practices, to a more harmonized approach to clinical trial data handling. ICON also realized the need to convert data, which may have been overly technically or scientifically focused, to a more accessible and usable form for operations teams. This would enable CRO clinical trial study teams and sponsors to more effectively access, analyze and use data on a real-time basis (compared to previous batch-centric approaches). Thus, a plan was defined to implement and establish a clinical data warehouse. We [ICON] knew what we needed to do and we went about doing that in a very focused and directed way. A CRO approach for implementing a clinical data warehouse ICONs approach to implementing a clinical data warehouse was specifically directed and purposed to meet the need as identified by its business case; to make large volumes of data usable and empower study teams to proactively manage their studies based on more insightful data. The solution had to be scalable, thus handle increasing volumes of data, and implementation and deployment had to be rapid in a specified time frame in keeping with typical CRO project turnarounds. In addition, to ensure these targets were met, the clinical data warehouse would include only new clinical trial study data. Legacy data and associated legacy data conversions were not incorporated in the clinical data warehouse an approach that is suited to CROs typically contracted scope of services within the trial execution phase. We [ICON] took a very purposeful and directed approach. We wanted to do something quick; we wanted to do something manageable. A general approach when implementing a clinical data warehouse is to deploy a range of technologies targeting various layers of data management from data input and governance to data output and reporting, publication, visualization and collaboration. ICON contracted with Oracle to supply key enabling products/technologies to form the basis of the clinical data warehouse platform for data custody, governance, and export to sponsors. In addition, third-party providers were contracted to supply supplementary tools for data access and visualization. Together all of the interlinked technologies ultimately allow clinical study teams to gain access to and insights on trial data. Finally, and most importantly, implementation of a clinical data warehouse requires support and buy-in from key stakeholders,

p 14

Health Sciences Journal

Issue 1 March 2013

A Clinical Data Warehouse Solution to Improve Operational Efficiencies (continued) particularly executive personnel, who can catalyze the speed of the overall project. For ICON, there was both executive and board support from the project outset as one of the companys strategic goals was to improve its information and informatics capability. A clinical data warehouse is directly associated with this goal, hence implementation was unanimously supported and deployment was successful. Implementation challenges, solutions and standardization ICONs clinical data warehouse implementation was divided into a number of manageable phases, to acquire, centralize, standardize and visualize data and subsequently, to address associated operational deployment and change management challenges. The challenges of rapid deployment were overcome by defining simplified, achievable use cases and having appropriate milestones by which to assess progress. There have been technical challenges, particularly during early implementation. ICON was one of the first to adopt this clinical data warehouse strategy and consequently experienced pain points associated with being pioneers. These have lessened over the years as other organizations have adopted the strategy and product solutions have been identified to address these early challenges. Data standardization has been a cornerstone to the success of ICONs clinical data warehouse and was an important consideration even before use case identification and project initiation. The CRO approach to standardization is different from a pharmaceutical company based on the nature of CRO activities. CROs typically engage with numerous, diverse sponsors, each with their pre-defined set of requirements in terms of a clinical trial, thus a CRO must set its own internal standards. ICON has achieved this through the development of a comprehensive clinical data standards hub, built off Clinical Data Interchange Standards Consortium (CDISC) standards and the study data tabulation model (SDTM), which drives operational data review, data visualization and data delivery. A clinical data warehouse as part of an overall integrated information platform ICONIK In July 2010, a year after signing the initial contract with their selected clinical data warehouse provider (Oracle), ICON launched its clinical data warehouse platform. The clinical data warehouse is a key part of ICONs overall integrated information platform ICONIK. For us [ICON] ICONIK is about trying to be very transparent with our client in terms of how we are running their studies, trying to drive operational efficiencies and improve quality, and being proactive in terms of what we do in the conduct of our studies. ICONIK is a powerful integrated information platform that consolidates, standardizes and visualizes both operational and clinical data, to provide a single holistic view of all study information to both sponsor and CRO teams. It offers near real-time access to clinical trial performance metrics, critical safety and efficacy data, and the ability to analyze these data in novel ways. The ICONIK information platform and associated operational processes improve data quality and

p 15

Health Sciences Journal

Issue 1 March 2013

A Clinical Data Warehouse Solution to Improve Operational Efficiencies (continued) subject safety while yielding significant operational efficiency gains (Figure 1). Oracle Life Sciences Data Hub powers the clinical data warehouse and alongside an operational metrics data warehouse, ICON has the key components in place to drive operational activities such as assessing study feasibility, trial start-up, subject enrollment, subject retention, and safety. ICONIKs integrated information platform provides four levels of knowledge to the sponsor and study teams: Operational efficiency automated processes to gather and surface information Transparency increased transparency to clients throughout the development process Visibility accurate and detailed information on site performance and risk Quality increased focus on data integrity and control of clinical data The ICONIK integrated information platform and associated operational processes has enabled the company to revolutionize management of clinical trials, such as improving study planning and design by access to historical and operational clinical data to guide protocol development and provide quicker evaluation of site feasibility, faster identification of delays or potential difficulties in site start-up, and the ability to obtain insights into patient eligibility and screening failures to improve retention rates, among others. Figure 1: ICONIKs integrated information platform provides four levels of knowledge to the sponsor and study teams

Sponsors

Study Teams

Immediate Knowledge

Today, ICONs enhanced Sponsor Reporting Services offers a number of key benefits for the optimized visualization of data from the clinical data warehouse. The enhanced Sponsor Reporting Services provide a single source for study team members to access study information in a consistent manner across the lifecycle of a study. Using clinical data from electronic data capture (EDC), interactive voice response (IVR), eDiary or Central Laboratory, a study or a program of studies can be instantly evaluated from a scientific, safety and quality perspective. Any operation, from the detection of a safety signal to the data quality analysis of a solitary site, can be performed in a few clicks.

p 16

Health Sciences Journal

Issue 1 March 2013

A Clinical Data Warehouse Solution to Improve Operational Efficiencies (continued) Main users of the clinical data warehouse, benefits and impact The main users of the clinical data warehouse are the end consumers (i.e. the clinical trial study teams, clinical data teams, the medical monitors, study start-up teams, quality assurance teams, who are the main beneficiaries of the clinical data warehouse) rather than data programmers. Such endusers may not necessarily log on, but do consume all data, insights and analyses of study performance and status and use these to inform operational decisions. One of the best examples to illustrate this is the ICONIK Monitoring service, where the centralized monitoring team routinely uses holistic scientific data analysis, together with clinical research associate site knowledge to direct central and site monitoring activities. The centralized monitoring team has access to and continually reviews real-time investigator performance and risk metrics, all of which are predictive of overall investigator performance and compliance. Investigators with abnormal behavior patterns are tracked and analyzed centrally in order to evaluate the need for site-specific action and ensure a focused approach to monitoring. Study teams are managing monitoring resources in a flexible and intelligent way, employing resources as and when they are required based on the demands of the study. We [ICON] have 8,500 employees in the company and our view is that the clinical data warehouse is the foundation from which we get the data that all of the teams consume; without the clinical data warehouse and operational data warehouse they would not have access to this. Conclusion A clinical data warehouse offers a data-handling solution for CROs as it enables the centralization and governance of clinical data which ultimately facilitates the publication of data in a usable format. For ICON, the main focus of the clinical data warehouse was to convert previously technical data to a more usable and understandable form for clinical trial study teams and sponsors, which could then drive operational decisions. In other words, operational efficiency was at the heart of the decision-making process. ICON believes that its implementation of a clinical data warehouse has differentiated it from other CRO competitors. The companys pioneering efforts to implement ICONIK has facilitated access to useful, comprehensive real-time data for its study teams and sponsors, giving the company a competitive advantage over other CROs. As the company looks to the future, the aim is to continue to focus on operational activity and identify niche areas within this space on which to deliver added value to its sponsors/ clients. ICON has successfully implemented and deployed its clinical data warehousing solution and 3 years from initial deployment of ICONIK, the company has already realized its main aim, which was to ensure accessibility to operationally useful data. The clinical data warehouse will continue to evolve and deliver benefits and efficiencies.

p 17

Health Sciences Journal

Issue 1 March 2013

A Dynamic Platform for Data Integration, Standardization and Management

Brooks Fowler is the global head of data sciences at AbbVie. Brooks is specifically accountable for data management operations, clinical informatics and clinical sample management operations. Brooks began his career in pharma with G.D. Searle in 2000. He joined AbbVie in 2003 as a section manager of clinical data management. Over the course of the last nine years, Brooks and the AbbVie team have designed and implemented enterprise solutions for EDC, ePRO and IRT.

Brooks Fowler and Nareen Katta AbbVie

To effectively compete in the current economic climate and, in the face of changing trends in commerce, the life sciences industry has had to evolve and become more cost-effective, efficient and responsive. There is increased emphasis on optimizing the clinical trial process and enabling maximum use of data, the industrys key asset. To this effect, pharmaceutical companies such as AbbVie are continually searching for ways to maximize value from data. Better analyses of clinical trial data and optimization of operational aspects (for example, administrative and financial) of each trial can improve both cycle time and efficiency. It is about cultivating previously unused data whether its clinical or operational, and putting it to good use. In addition, AbbVie recognizes the need for better data management, achieved through better IT solutions. Indeed, although IT is not a core competency of the life sciences industry, there is a high demand for up-to-date IT infrastructure and solutions. Thus, an IT provider or specialist company builds and implements the IT infrastructure while the pharmaceutical company uses this infrastructure to consolidate, mine, and explore data, thereby informing clinical and operational decisions with reduced need for specialist skills internally.

Nareen Katta is the senior manager, data sciences at AbbVie. Nareen is specifically accountable for managing the companys EDC system and clinical databases, including design and definition, data integrations, standardization and ETL operations.

p 18

Health Sciences Journal

Issue 1 March 2013

A Dynamic Platform for Data Integration, Standardization and Management (continued) Addressing industry needs and business drivers AbbVie identified a number of unmet needs that led to a clinical data warehouse as a solution. These included the: 1)  Absence of an archiving solution in the companys current and legacy clinical data management systems (CDMSs) there was no functional system from which archived clinical data could be accessed on demand, as there was no archival facility in the previous CDMS for clinical data from the companys own trials, or data inherited from mergers and acquisitions 2)  Use of numerous and varied data entry systems, thus data were disparate rather than standardized making analysis challenging 3)  Use of data management systems with a fixed structure restricted data integration as data had to be in a certain format 4)  Inability to perform cross-study analyses the companys vision was to create a system that allowed all users, and not just specialists like statisticians, to conduct ad hoc analyses and be able to visualize data, thereby maximizing the value of data. AbbVie needed to address key business drivers, including minimizing the number of manual steps required to access data. It was imperative to identify the right IT tools for the right functions thereby allowing near real-time data access in a consolidated, accessible and effortless manner. In addition, AbbVie required a solution that was dynamic and allowed upgrading and switching of systems as new versions or peripheral applications became available. Thus, a solution that could readily evolve with minimal disruptions was needed. A flexible clinical data warehouse presented the most suitable solution based on the fact that it does not have a pre-defined data structure. As a result, it was possible to integrate data from any data structure, for example other clinical and operational systems, and subsequently make these data conform to AbbVies structure templates, that are source system agnostic within the warehouse, with minimal disruption. Furthermore, with a clinical data warehouse it is possible to build integrated data access layers. These enable non-specialist users to readily access data and perform cross-trial data analysis.

p 19

Health Sciences Journal

Issue 1 March 2013

A Dynamic Platform for Data Integration, Standardization and Management (continued) A phased approach to clinical data warehouse implementation Once the unmet needs and business drivers had been defined, the company scored and ranked business issues via internal interviews to highlight possible approaches for finding a solution. The approaches were categorized as process, application or infrastructure changes. Through this screening, application emerged as the most commonly requested change. The company responded to this by replacing its existing data management system with a clinical data warehouse. AbbVie wanted its solution to serve as an end-to-end CDMS with both data warehousing and data management capabilities; the warehousing aspect for data aggregation, standardization and reporting and the CDMS for data cleaning, blinding and medical coding requirements. We [AbbVie] were looking for, not only a clinical data warehouse and repository, but a full blown clinical data management system. The clinical data warehouse was deployed over two phases. Phase 1 involved assessment and implementation of core functionality, as determined by a cross-functional group in workshop settings, while Phase 2 involved the addition of tools and further refinement. Prior to Phase 1, AbbVie performed a proof-of-concept test to assess the core functionality, process change and use cases, thereby reconfirming the suitability of a clinical data warehouse solution. To define how processes and the business, as a whole, were likely to change as a result of a clinical data warehouse, the company extrapolated and mapped the final project outcomes to the base requirements. This exercise defined the end-user interaction with the new framework and highlighted areas that would need further development in order to optimize functionality. There were various integration processes during Phase 1, for example with electronic data capture (EDC) and Laboratory Information Management (LIMS) systems, making it possible to amalgamate and consolidate data with the core system. To ensure continued accessibility of data to the statistics teams, the extraction methodology for pulling clinical data from the clinical database to the analysis database was redesigned to fit the clinical data warehouse. For example, a previous storage facility of metadata was repurposed for the clinical data warehouse and enabled the company to begin processing some studies through the warehouse. In addition, a metadata driven study setup utility, a parameterdriven edit check engine to enable discrepancy management and integration with coding solutions (for example, Oracle Thesaurus Management System) used to standardize medical encoding terminology across studies, were developed. With core functionality achieved, the company was able to deploy its clinical data warehouse at the end of Phase 1. In Phase 2, there was additional integration of tools to enable the users to more extensively use the system. For example, using metadata, AbbVie created a tool to allow users to identify a new study and search for similar studies from legacy data. In addition, there was integration with additional data

p 20

Health Sciences Journal

Issue 1 March 2013

A Dynamic Platform for Data Integration, Standardization and Management (continued) sources like AbbVies Phase 1 management system, as well as bi-directional integration with the EDC system to enable discrepancy management with sites, thereby enabling endto-end data flow. Reporting and data browsing tools were added to further simplify user interaction and access of the clinical data warehouse. Finally, Phase 2 also involved process automation. The clinical data warehouse is currently in the early stage of rollout and is therefore restricted to use by the global data management and statistics divisions; however, the company envisages further roll-out and expansion of the user community in the future as reporting and visualization tools are added to the platform. We [AbbVie] are utilizing the tool across our global data management and statistics sites. We are making sure that the use of the system and the implementation is geographically dispersed rather than focusing it here [Chicago] at our single headquarter office. AbbVie expects to process all of its studies through the clinical data warehouse once it is fully scaled up. Thus, the overall intention is that every global site will utilize the new system, increasing operational efficiencies and cost-effectiveness. Key challenges and data standardization AbbVies main challenge with implementing a clinical data warehouse has been the fact that the new clinical data warehouse framework is a complete paradigm shift. A clinical data warehouse is an entirely novel undertaking and completely different to the companys previous experience. It was challenging to fully comprehend the capabilities and select appropriate tools to be integrated onto the technology platform. However, the IT provider was instrumental in this endeavor and provided the guidance and expertise needed to manage the process change. To be able to translate this [a clinical data warehouse] into a future vision and be able to execute it was a challenge. It is a technology framework, not just a business process. Implementing a clinical data warehouse was a major IT initiative and to ensure its success it was important to improve both the IT and business infrastructures, including but not limited to process and resource development, which were likely to impact the project. The volume of technology and process integrations required also presented a challenge. One of the drivers for implementing a clinical data warehouse was to reduce the amount of manual activities. Automation of manual tasks required an assessment of both present systems/processes and future clinical data warehouse environment capabilities/processes. Based on these, tasks for automation were identified.

p 21

Health Sciences Journal

Issue 1 March 2013

A Dynamic Platform for Data Integration, Standardization and Management (continued) Stakeholder buy-in was less of a challenge. Through specific use cases and business cases there was clear demonstration of improved efficiencies, drug safety implications of data integration (in terms of having aggregated data for regulatory queries and analyses), and the cost justification which will become increasingly evident over time as more clinical trial data are collected and reduction in manual effort manifests more broadly. Due to these elements, unanimous support from project sponsors was won. The clinical data warehouse has enabled AbbVie to implement the Clinical Data Interchange Standards Consortium (CDISC) Study Data Tabulation Model (SDTM) standards. The company required flexibility to allow incorporation of data from various sources, but to also have industry-recognized standards. To do this, a number of tools were added to the platform to convert native standards to CDISC STDM formats, which were then accessible to users. In addition, a data governance team was built to manage these processes. Benefits to users AbbVie has realized a number of benefits from its clinical data warehouse solution including the ability to extract value from metadata and legacy data. For example, using legacy data, programmers are able to more efficiently design future trials and processes. In addition, a single information hub has allowed the use of one or two key reporting visualization systems which provide data in readily usable formats to end users. Previously there had been numerous, different reporting and visualization tools providing data in diverse formats. Sample management logistics is another potential benefit and may allow sample tracking from origination and collection to process end within a robust warehousing environment. A clinical data warehouse also provides a central repository for storing data inherited from mergers and acquisitions. It provides an open but secure framework onto which acquired data can be archived, mined and standardized, as required. Clinical data from three legacy CDMS applications inherited from mergers and acquisitions have been archived to date. An additional benefit, though not specifically identified by the company when defining the use cases, has been the ability to lock clinical trial databases more quickly. The ability to manage blinding and un-blinding of sensitive clinical data in the clinical data warehouse contributes to further reduction in the cycle time. For example, in previous systems, blinding data were added once the database had been locked and this was time consuming. With a clinical data warehouse, sensitive data can be uploaded and stored in a secure/non-accessible area long before database lock. Thus, on database lock un-blinding can potentially be performed instantaneously as all data are already on the system. With regard to being able to lock databases quicker and being able to change and increase the frequency of data refreshers into our [AbbVie] system we have realized efficiencies but I think it will be a while before we completely realize the efficiencies that are associated with the new system.

p 22

Health Sciences Journal

Issue 1 March 2013

A Dynamic Platform for Data Integration, Standardization and Management (continued) Furthermore, the system allows direct linkage to AbbVies EDC tool, thereby making it easier to access data with less administrative burden. The overall response from both stakeholders and users has been positive thus far. However, further time is required to fully appreciate the efficiencies and benefits of the new system, particularly with regard to future enhancements. Some of the expected benefits include system portability, reduced manual effort related to data cleaning and data loading, and for aggregating data for cross-study analysis. Looking towards the future As the company looks towards the future, the intention is to replace current processes (for example, various reporting tools) with the new system. Improved data archiving will allow storing of original data in a well-controlled environment with subsequent standardization and reporting in a readily usable format. This will provide an ad hoc analysis capability within the drug development process making it possible to assess, for example, whether anything was missed in the initial analysis, whether there were any safety implications of note, and whether a drug mechanism currently under investigation had been previously tested; all of which will be used to drive future decisions. AbbVie is also looking to capture large-volume data, particularly from its post-marketing registry trials which typically involve a large number of patients, into the clinical data warehouse. Future plans include partnering with health outcomes organizations and utilizing electronic medical record data and claims data from these organizations to maneuver the structure of these data into a more usable format for both internal personnel and health outcomes teams. In summary, AbbVies implementation of a new clinical data warehouse (integrated with clinical data management capabilities) has provided a platform that enables data integration, standardization and management. The company has focused on automating its data flow from collection to analysis to minimize manual steps, thereby decreasing sources of error and increasing operational efficiency. Clinical data warehouse implementation has been successful using a two-phased approach. The future user community is predicted to increase as data stored in the clinical data warehouse becomes more integrated and accessible. The clinical data warehouse is able to store both production and legacy data, allowing standardization, exploration, mining and analyses of these data to inform future decisions. The company views its clinical data warehouse as a dynamic, evolving platform that will eventually replace most of its current technologies and systems.

p 23

Health Sciences Journal

Issue 1 March 2013

Clinical Research Innovation through Shared Clinical Data Warehousing

Jerry Whaley is senior director of development business technology at Pfizer and is involved in the implementation of Pfizers clinical data warehousing solution. Jerry began his Pfizer career in 2001, as director, development informatics Ann Arbor site head. Before joining Pfizer, Jerry was an SAS programmer and supervisor at the Upjohn Company and a systems analyst, developer, project leader and manager at Parke-Davis. Prior to returning to Pfizer Jerry was vice president at Advanced Systems Development with responsibilities for business development and client implementation management.

Jerry Whaley Pfizer

Increasing partnerships between pharmaceutical and biotechnology companies and/or service providers is a key emerging trend in the life sciences industry. Specifically in clinical trial management, there has been an overall re-assessment of what constitutes competitive advantage with regard to data capture and management. The questions being asked include: does a custom-developed electronic data management platform really provide competitive advantage? Could using a standard platform help minimize the many issues caused by variability of data? Where is it best to focus pharmaceutical company resources? Unsurprisingly, focus is all on data. If, as an industry, we can holistically understand our data better and more in-depth, so not just as Pfizer-specific data but, for example, healthcare as a whole, then thats advantageous as it allows us to better analyze it. Pfizers clinical trials operational model has evolved over the years. The initial model progressed from conducting inhouse trials to outsourcing trials to 17 functional service providers. This has now evolved to the companys current position of having two alliance partners (ICON and Parexel), or contract research organizations (CROs). This change has been driven by a business need to be more cost-effective and manage resources more efficiently. This model enables the company to leverage the CROs expertise in execution of clinical trials

p 24

Health Sciences Journal

Issue 1 March 2013

Clinical Research Innovation through Shared Clinical Data Warehousing (continued) and allows Pfizers role to evolve, into a more oversight role. Pfizers oversight input on clinical trials also requires expertise and skills that need to be acquired over time. The consequence of this model is that it frees company resources to focus on analysis of trial data, rather than preparation of data. Such an undertaking requires a standardized data warehousing solution for data receipt, aggregation, access, and analysis. Pfizers intention is to create a road map to define and standardize processes for data integration and data sharing based on a communal data warehousing solution. In addition to enabling interactions with CROs, the visionary view for this type of clinical data warehousing is that it could also facilitate future interactions with multiple partners including other pharmaceutical or biotechnology companies, regulatory authorities, and companies absorbed through mergers and acquisitions. Pfizers Clinical Aggregation Layer (CAL) solution Pfizers vision is to create a cloud technology platform which facilitates efficient clinical trial operation for industry peers, to minimize duplication of effort in tool development, and drive process efficiency to accelerate new drug research. Company owned data handling tools and applications may not necessarily provide competitive advantage but do increase costs. In Pfizers view, as long as individual company data are secure and protected and there is appropriate legal and regulatory approval, data can be stored and processed from a central platform that is located externally to Pfizer, providing an opportunity for sharing technology across the industry. Pfizers data warehousing solution, known as the Clinical Aggregation Layer (CAL), consists of three core components (Figure 1A and 1B): 1) A clinical and scientific data warehouse (CSDW) to manage, aggregate, and analyze clinical trial data 2) An operational data warehouse (ODW) for trial performance metrics 3) A custom-developed trial master file (TMF) tool to keep a comprehensive record of all clinical trial activities. Data are loaded into CAL through various mechanisms, depending on data type and source (for example, data exchange adapters, secure file transfer protocols, etc.) such mechanisms being based on industry standards. Data stored in a metadata repository, are also uploaded into CAL Pfizer captures and maintains these metadata. It is imperative that input data are correctly referenced and indexed for such a solution to be effective. We [Pfizer] see this as sowing the seed of an industry infrastructure, thats our vision. This is not just a Pfizer solution; we are trying to seed this solution with partners such as Oracle and Accenture.

p 25

Health Sciences Journal

Issue 1 March 2013

Clinical Research Innovation through Shared Clinical Data Warehousing (continued) Figure 1A: Pfizers long-term technology vision
Parexel Systems Pfizer Run Studies ICON Systems
Future Acquisitions/ Future Partners

Information eXchange Hub (IXH)

Contextual Reference Data Module (RDM)

(PXL/ICON/FSP) On-going Studies: [OC/RDC/TMS]

This model is possible due to sufficient evolution of data standards, services, technology and IT infrastructure. The convergence of technology and business needs, tighter business models (with regard to efficiency and cost), and stringent regulatory processes and requirements have further reinforced the premise of such a solution. Figure 1B: Components needed for the new Pfizer platform

Pfizer owned and operated systems

Safety

Operational Data Warehouse (ODW) Study Operational Reports Programme/portfolio milestones Partnership Scorecard

Clinical Scientific Data Warehouse (CSDW) Aggregate reporting Visualization Data mining

Trial Master File (TMF)

Two Guiding Principles

Clinical Supply

One version of the truth

Exchanged data based on de ned data standards

Accenture/Oracle

Project Mgmt

Analysis and Review Tools Contextual Data

Reference Data Management (RDM)
Operational Data Warehouse Clinical Scienti c Data Warehouse Trial Master File

Clinical Aggregation Layer

Others

The CAL solution can be potentially both cost-effective and innovative and could drive the development of new tools available for a majority of users. Collective innovation may also result from being able to conduct in-depth analysis using shared data that are readily accessible from a standardized IT infrastructure. A shared platform may also be used to leverage trial data more broadly for example, companies conducting trials in a single disease area could, in theory, share placebo data if patient recruitment criteria were similar thereby reducing costs for the placebo arm of a trial.

Integrated repository for data required to track and manage study and development program execution

Integrated repository of analysis-ready data from our partners, P zer, and relevant external sources

Authoritative source of Essential Documents

Secure Data Exchange

p 26

Health Sciences Journal

Issue 1 March 2013

Clinical Research Innovation through Shared Clinical Data Warehousing (continued) The willingness of industry peers to be participants in such a warehousing solution is yet to be realized. To this effect, Pfizer is actively engaging in discussion with peer companies to gauge interest. Initial indications appear positive. Approach to implementing a shared clinical data warehouse The most evident differentiator of Pfizers clinical data warehousing solution is its accessibility of technology across the breadth of the life sciences industry. Pfizer is actively avoiding customization of its data warehousing solution, and is making every effort to maintain it as an off-the-shelf solution to allow for broad applicability. We [Pfizer] are trying to stay rigid to the fact that these are commercially available, off-the-shelf solutions Do not Pfizerize them; Do not customize them avoid this as much as possible. This approach allows reusability, ease of implementation and ease of support long-term. To further ensure this, Pfizer frequently engages in discussions with its IT provider to ensure that tools and applications remain generic, enabling easy upgrades, processing and most of all, seamless partnering with external parties. Although outsourcing operational functions (for example, execution of clinical trials) and using shared data warehousing platforms and technologies is not a new concept, it was unprecedented for large companies like Pfizer. The long-term aspiration is to externalize most processes and maximize the use of external expertise to drive Pfizers healthcare goals. Implementation of the Pfizer solution necessitates a phasedapproach to make this complex and challenging undertaking a more manageable and viable prospect. Future releases include additional functionality within the CSDW, across other therapy area and clinical trial teams, followed by ODW-related operational data. Stakeholders have readily championed and supported this project from its genesis. The CAL solutions goal is to be instrumental in increasing Pfizers operational efficiency and re-focusing resources towards accelerating clinical research. Use cases supporting a clinical data warehousing solution and potential users A number of identified use cases supported the need for the CAL solution. The most immediate was associated with the new operational model of increased clinical trial externalization. This required consolidation of data to a single location to allow Pfizer easy access to trial-related information and data. In addition, the need to consolidate data gained through mergers and acquisitions. Pfizer identified the need to improve the ability to explore, analyze and mine both clinical trial and operational data so as to maximize its value. From a regulatory and compliance perspective, the availability of all trial data in a single location could facilitate a quicker response to queries.

p 27

Health Sciences Journal

Issue 1 March 2013

Clinical Research Innovation through Shared Clinical Data Warehousing (continued) Presently, the two critical primary users of the CSDW component of the CAL solution are clinicians and statisticians. Users can explore, analyze, and mine clinical trial data on a single platform and, as a result of data standardization, can leverage a broad range of tools to drive research. Other users are the companys partners and service providers who upload trial-related data which Pfizer can use for data analysis, as well as monitoring trial progress. The ODW provides operational metrics which can be used to drive decisions (for example, what geographical region may be suited to a trial in a particular disease area). In order to remain compliant with regulatory requirements, the TMF solution provides definitive proof and record of all clinical trial activities. Such factors contribute to cost-benefits, better time efficiency and management (through standardization) and, increased data value realization. In addition, the CAL solution could simplify the role of industry regulators and auditors. For example, where previously there were different processes/systems for each company, with the CAL solution there is a single system to understand. Thus, processes such as auditing/inspecting could become more efficient based on the reduction of industry systems an auditor would need to be familiar with. This model can facilitate a more progressive and efficient life sciences industry in that regulators have only one infrastructure and set of applications that they need to understand, audit and ensure compliance. Planned trial throughput via the CAL solution Moving ahead, Pfizers aim is to route as many clinical trials through the new operational model as possible. Typically the company runs approximately 800 trials in a given year the intention is to transfer a proportion of trials (approximately 100) to the new model by the end of year one and to accelerate throughput to approximately 300 in year two. The longterm vision is to decommission all legacy processes and applications. Although there is a company-wide effort to transfer clinical trials to the new system, there is recognition that this must be done in a controlled manner to maintain data integrity. To this effect, the Clinical Data Interchange Standards Consortium (CDISC) and Study Data Tabulation Model (SDTM) have played important roles. In addition, standardization facilitates amalgamation of data following mergers and acquisitions an activity frequently associated with Pfizer.

p 28

Health Sciences Journal

Issue 1 March 2013

Clinical Research Innovation through Shared Clinical Data Warehousing (continued) Standardization is an important piece of our [Pfizer] strategy because without it, aggregating these data would have a lot less impact and a lot less value. The past, the present, and looking to the future In the past, Pfizer attempted to build and implement its own in-house data warehousing solution but this proved challenging. Nonetheless, many lessons were learned from early efforts which have influenced the CAL solution including the value in using a generic, robust, commercial off-the-shelf technology tool-set acceptable to other industry peers as a shared data warehousing platform. We [Pfizer] felt we chose our tools wisely; it was important to choose a scalable and industry leading tool-set that others would embrace. The present implementation of the CAL solution has not been without challenges. Some of these include converting ingrained legacy business processes to new processes and systems, availability of required stakeholders/personnel for making implementation-related decisions and meeting stringent deadlines and, as with all ambitious projects, managing budgets effectively. Looking to the future, Pfizers view is that this is merely the beginning of the journey. For the company, the CAL solution provides an innovation bed for managing, analyzing, accessing, exploring and extracting maximum value from data, whether legacy or newly generated. In addition, it provides a simple and standardized means of collaborating with multiple partners. Overall, the vision is that this joint data warehousing solution, which is a novel concept in the clinical trial space, will enhance innovation both from a scientific and technical perspective. To achieve this vision, Pfizer has leveraged the expertise of global IT service providers, including Oracle, to provide it with a clinical data warehousing solution that acts as an integration and collaboration platform to enable full-service, hybrid outsourcing, as well as to support internal processes.

p 29

STAY CONNECTED TO ORACLE HEALTH SCIENCES.

Join the Health Sciences Social Media Hub.
Follow Oracle Health Sciences thought leaders through our social media channels. Whether you join us on Facebook, follow us on Twitter, or become part of our new Health Sciences Forum on LinkedIn, youll see how Oracle Health Sciences is leading the conversation.

Join our LinkedIn group: Health Sciences Forum

Like us on Facebook: facebook.com/OracleHealthSciences

Follow us on Twitter: twitter.com/OracleHealthSci

p 30

This publication was produced by Oracle Editorial support was provided by ApotheCom Copyright 2013, Oracle and/or its affiliates. All rights reserved. Oracle and Java are registered trademarks of Oracle and/or its affiliates. Other names may be trademarks of their respective owners.

Oracle Corporation World Headquarters 500 Oracle Parkway Redwood Shores, CA 94065 U.S.A Worldwide Inquiries: Phone: +1.800.633.0643 oracle.com/healthsciences Printed on recycled material.