Abstract
Classical drug exposure: response studies in clinical pharmacology represent the quintessential prototype for Bench to Bedside‐Clinical Translational Research. A fundamental premise of this approach is for a multidisciplinary team of researchers to design and execute complex, in‐depth mechanistic studies conducted in relatively small groups of subjects. The infrastructure support for this genre of clinical research is not well‐handled by scaling down of infrastructure used for large Phase III clinical trials. We describe a novel, integrated strategy, whose focus is to support and manage a study using an Information Hub, Communication Hub, and Data Hub design. This design is illustrated by an application to a series of varied projects sponsored by Special Clinical Centers of Research in chronic obstructive pulmonary disease at the University of Pittsburgh. In contrast to classical informatics support, it is readily scalable to large studies. Our experience suggests the culture consequences of research group self‐empowerment is not only economically efficient but transformative to the research process.
Keywords: institutional management teams, organization and administration, information services, information storage and retrieval
Introduction
Clinical pharmacology is the quintessential discipline of Bench to Bedside‐Clinical Translational Research (BB‐CTR). Central to this discipline is the coming together of teams of researchers, who come from different branches of science, to create project teams that apply their expertise to understand mechanisms that influence drug efficacy and are of value in guiding drug development. The natural evolution of drug development starts with complex studies conducted in small numbers of subjects, evolving to studies in larger numbers of subjects, but with less complex design. The infrastructure support to this transition has evolved as an ad hoc process in which project management and data management have emerged as different disciplines. The logistic challenges of multicenter studies conduct merit the separation of these activities into clearly defined areas of expertise. Unfortunately for early phase studies, the scaling down of skill sets and tools used in large studies proves inefficient and cumbersome when applied to small, nimble, and rapidly evolving protocols. There is a need to meet the challenges of this early phase from a different perspective.
The approach we have developed and advocate to meet this challenge views the same fundamental requirements from the perspective of management of a small project team. We have bridged the gap between project management and data management with communication tools. Our underlying philosophy has been to design integration of low cost, easily available, robust tools to provide for ease of use by team members, with sophistication of information technology being kept in the background. The objective is to empower each member of the team and maximize cross‐disciplinary interaction. Somewhat to our surprise, this team building focus is easily scalable from the earliest start of a pilot study to large multicenter studies and even to programmatic research management. Even though no individual element is new, we have suggested this orientation is suf ciently novel and more importantly efficient that it merits a commentary in the Clinical Translational Science literature. The primary objective of this presentation is to make the case that BB‐CTR poses specific challenges in integrated systems management. We use a case model presentation to illustrate how the needs of the discipline of BB‐CTR have resulted in the evolution of electronic Management‐Clinical Translational Research (eM‐CTR) Systems (an electronic management system of BB‐CTR). This strategy provides efficient, easily acquired infrastructure to manage small, in‐depth, mechanistic studies. It has proved to be a scalable entity that is sufficiently flexible for application in multiple small networking multicenter studies, and large‐scale genotyping, phenotyping studies, and Community‐Based Participatory Research (CBPR).
From the perspective of a clinical translational scientist, application of eM‐CTR is proving to be transformative technology that is changing the culture of team approach to this discipline. The fundamental element in our strategy is that all the information required for all aspects of a study conduct are readily available at the desktop computer screen of each participating member. The variety of informational elements range from: protocol development, logistical conduct of the study, data acquisition and organization, and data curation to data analysis and presentation. The perspective of an information technologist has moved from a “give us the data, we will manage it and upon request give you back what you need” paradigm to a seamless integration of information flow under the control of the Principal Investigator (PI), but accessible on a need to know basis to each team member. This change in emphasis adopts the principle that powerful information technology (IT) tools are already readily available at low cost. The assembly of a functional entity is dependent and providing the most appropriate tool to meet the needs for each aspect of function in such a way as to minimize the education needs and communication gaps between team members. This approach allows for a high level of sophistication, in which the research team uses powerful technology that is maintained in the background. The increase in ef ciency, defined as the work productivity for each team member, is now enabling us to address more and more challenging studies with an increasingly limited funding base. This transforming protocol management, together with new technology capability (from imaging and in laboratory technology), is transforming the future of CTR.
In this commentary, we identify the challenges faced by a BB‐CTR team, describe our strategy to meet these challenges, and suggest that the culture of a BB‐CTR team is best supported by an integrated systems management approach we have called eM‐CTR. We propose that use of a hub strategy ( Figure 1 ) greatly facilitates the efficiency and quality of research to such an extent that it can be considered a disruptive technology. We use a recent case model to illustrate how choice of relatively low cost information technology tools have been used to support an extensive genotype, phenotype study that is being conducted to redefine the classification of chronic obstructive pulmonary disease (COPD) and an innovative clinical trial in a defined patient subset.
Figure 1.
Integral components of eM‐CTR.
eM‐CTR—The Hub Concept
Irrespective of detailed study structure, the evolution of a research protocol follows a stepped process over time with well‐defined milestones ( Table 1 ). Each step has system wide requirements that need to be met. The traditional model of a PI writing a proposal, which the team implements, carries too much risk of lack of attention to detail to be tenable. From the outset, protocol planning, design, and execution is a team process that requires integrated management.
Table 1.
Life cycle of BB‐CTR research project.
| Steps in life cycle of project | Investigator is assisted in… | |
|---|---|---|
| Step 1 | Preplanning | Team building |
| Step 2 | Planning | Project design |
| Step 3 | Transcription | Project transcription to regulatory approved protocol |
| Step 4 | Implementation | Study conduct |
| Step 5 | Analysis | Presentation |
The contemporary challenge to a PI with a new project is to ensure that the generic infra‐structure needs for the type of research proposed are easily available from the outset, can be readily put in place, and are readily scalable.
From a pragmatic perspective, the hub concept illustrated in Figure 1 provides a flexible, efficient, infrastructure base to meet this challenge. The hub concept allows each hub to be implemented using different off‐the‐shelf technologies, while it also allows seamless integration into a coherent service.
The Information Hub is required at the first step in meeting the needs and transition from Steps 1, 2, and 5 in the project life cycle ( Table 1 ). In this paper, information management is defined as the management of documents (e.g., protocols), information on collaborators, bibliographies, and textual discussions between collaborators. The primary requirement is for a centralized, but shared idea management process. By definition, team assembly is a process whereby individuals with different skill sets, who often use different jargons (languages), but have complementary roles in the study, come together to plan the protocol from concept to implementation. These needs predicate infrastructure specifications.
Multiple customizable IT tools are available on the open market to meet these needs. Based on our research users, comfort level with Microsof Windows, we have selected SharePoint™ as a low cost, easy to incorporate system that has the further critical advantage, that the desktop user interface of this system is similar to the Microsof Office interface that most users are already familiar with, thus minimizing cultural acclimatization and need for education for new users. Additionally, the SharePoint™ interface provides basic support for MAC 1 and Linux users. 2 It meets our philosophical strategy in that this software package is readily interfaced with sophisticated technology applications to adapt the front end for user friendliness.
Efficient protocol development is a key feature of eM‐CTR. Using the Information Hub in eM‐CTR, a protocol‐specific site can be established within an hour, prior to the first protocol meeting, and be introduced at the meeting as the central Information Hub that is web accessible to all project members for the duration of the study. A further attractive feature of the eM‐CTR Information Hub is that access by team members can be on a need to know and need to modify basis. Despite the enormous variation in research context, the research team structure and individual member's responsibilities are remarkably similar. This permits design elements to address not only the PIs, perspective, but the individual perspectives of each team member, from clinicians, research coordinators, specialized skill laboratories or imaging services to lab technicians and financial administrators.
The eM‐CTR Information Hub allows us to def ne role‐based access to documents, as well as role‐based collaboration networks. The further opportunity is that by stratification, cross‐disciplinary team building can be encouraged (by defining a horizontal integration of specialized functions, in contrast to and to complement the vertical integrated team approach) in a broader management matrix for multicenter studies. Each subset within this stratification has specific needs in their viewing of the information and can be supported by function specific views of the information. There is also the added value that there is an opportunity for a shared ability to solve problems with other peer members at the same level within their discipline across teams.
The Communication Hub is required by a project team as a resource to facilitate interactions between team members as component parts of the protocol development in Steps 1, 2, 3, and 5 of protocol evolution. Existing mechanisms of face‐to‐face group meetings and email chains are effective for very small groups when geographically co‐located, but rapidly become frustrating and rate limiting for larger dispersed groups. Desktop communication tools can play an important role in the linking of remote individuals especially for routine, well‐developed, and fact‐filling meetings, 3 such as protocol meetings. The infrastructure specifications of tools that could help this process are listed in Table 2 .
Table 2.
Requirements for electronic management support.
| Information Hub | Communication Hub | Data Hub |
|---|---|---|
| • Central easy to access location | • Distance video conferencing | • Central easy to access location |
| • Team member contact list access | • Distance teleconferencing | • Library of SOPs |
| • Version control of edit changes | • Written material desktop sharing | • Library of patient report forms |
| • Provides time management scheduling | • Integration into eM‐CTR | • Bar coded biorepository management |
| • Facilitates quality assurance during study | • Scheduling | • User friendly electronic data input access |
| • Management of bibliography | • Tracking | • Capability of handling multiple input |
| • Integration into eM‐CTR | • Security | • Multidimensional |
| • Tracking of users | • Backup | • Version control of changes |
| • Security | • Facilitates quality assurance during study | |
| • Backup | • Ease of data extraction for use | |
| • Integration into eM‐CTR | ||
| • Tracking of users | ||
| • Security | ||
| • Backup |
Once again, multiple IT tools are available at varying levels of sophistication. We have selected a low cost (<$150 per camera) and an intermediate cost of video‐conferencing systems that can be tailored to the investigators, budget. Both require high speed DSL or other types of broadband access. We have found teleconference, in conjunction with a convenient desktop‐to‐desktop intercommunication tool such as “Bridget™” (which we also route through our central servers), provides a useful solution.
In practice, provision of video conferencing linked to Bridget™ from the outset creates a further cultural change within the research team. The protocol group's self‐perception evolves from a hub started (at the face‐to‐face group meeting) with the PI in charge, to a distributed social network that carries the increased efficiency and flexibility of distributed processes over centralized processes found in other disciplines.
Thus, the combination of the Information Hub and Communication Hub are all that is required to the end of Phase II in having a project design ( Table 2 ).
The Data Hub is required in the translation of a concept to a detailed management of study logistics, study conduct, data acquisition, and organization ( Table 2 ).
In principle, implementation of a study moves the process from protocol development to transcribing the protocol into a Manual of Procedures (MOP). It therefore requires a different set of activities for the staff that actually conducts the research but can take advantage of the same Information Hub and Communication Hub IT tools, but now incorporates elements from the Data Hub. The proposal developed at the end of Step 2 is used to create the MOP. This consists of a series of protocol specific Standard Operating Procedures (SOPs) together with relevant electronic patient report forms, data forms, and a preplanned informatics study‐specific virtual electronic space in a multidimensional database to house study data within the Data Hub.
The Data Hub of the eM‐CTR has access to a library of SOPs and patient report forms from our prior experience in clinical pharmacology research, particularly in studies applying pharmacogenetic tools, to NIDDK funded studies in liver disease, NCI funded studies in bladder cancer, and NICHD funded studies in drug disposition during pregnancy. Located within the secure Data Hub is a SOP wiki, a collaboration tool with well‐known advantages such as version control and full content search. 4 The Information Hub also brings together prior experience from other collaborating investigators to target needs for individual protocols and identified the domain users. It is then a relatively trivial process to adopt SOPs and customize electronic case report forms efficiently, with a minimum of travel, and in a distributed way to be able to aggregate a highly individualized MOP for each of the protocols being simultaneously developed ( Figure 2 ). An analogous case report form wiki of already electronic formatted case report forms permits rapid assembly and adaptation of protocol specific electronic case report forms.
Figure 2.
Flow of SOPs to create study specific manual operating procedures (MOPs).
Conventionally, case report forms are completed in pen and paper at the patient's bedside and subsequently the data is transferred by dual entry by secretarial staff to create an electronic format. Our experience is that electronic forms, which take advantage of branching logic and drop down boxes, made available on a laptop computer at the patient's bedside, permits enormous efficiencies in time management and improved quality and completeness of data entry. We advocate the use of laptop or notebook initial data entry that is FDA compliant and meets HIPAA compliance for secure systems for collection, organization, and transfer of data.
The Data Hub also supports management of a virtual biorepository that manages, tracks, and organizes biological samples from the patient bedside, through laboratory analysis to integrating the acquired data in the data mart.
We utilize HIPAA‐compliant Freezerworks® inventory management software (Dataworks Development, Inc., Mountlake Terrace, WA, USA), which we have adapted to a web‐based environment. This system uses barcode‐based sample labeling to “capture” each biospecimen in the inventory management database at the moment it is drawn from a patient bedside, by swiping the prelabeled barcode affixed to each collection tube (or other receptacle) with a handheld barcode scanner. Bar coding technology is ubiquitous and efficient and requires low cost, portable tools that are easily available to carry out tasks which took many man‐hours in the past. The analogous example of grocery inventory management via barcode scanning at the supermarket is obvious.
The barcode inventory management system allows real‐time tracking of each and every sample as it travels from the collection site (bedside), to freezer storage, to transit to a laboratory for processing or analysis. Within the biorepository database, it records freeze‐thaw cycles (due to experimental manipulation or freezer power losses); it provides tracking information after analysis, allowing an investigator to trace a derived variable result from the patient to the laboratory and onto the database. If batches of samples are being tracked rather than each sample separately, the annotation is adapted to permit batch entry of results to the database.
The interactive component of the Communication Hub also permits the development of protocol specific data ontology ( Figure 3 ) in which there are clear definitions for semantics, and the interrelationships between each identified term. In addition, detailed planning can be accomplished for advertising for study subjects, patient recruitment, study logistics, identification of involved personnel, their education on study conduct, and plan to have appropriate resources available to conduct the study.
Figure 3.
Overall flow of protocol acquired information in the eMCTR hub.
Once the design, protocol, MOP, case report forms, and design of the virtual biorepository are completed, the activity of the research team switches gears to implementation ( Table 1 , Step 4). In this phase, clinicians and nursing staff have the logistical challenge of subject identification, patient accrual and managing the study collection of domains of interest. The logistic flow of patients through the research facility requires a scheduling component already available with the eM‐CTR site, where different members of staf have access to the same forum to ensure coordination at each site.
The data requires organization based on ontologies, using structured linked hierarchies, quality control curation, and then storage in a data mart ( Figure 3 ).
Once in the Oracle database, we have used its Online Analytical Processing System (OLAP) to support quality control and data extraction. OLAP is a technology that allows the user to quickly analyze information and can be used in the management of the protocol as it evolves. The information is summarized into multidimensional views and hierarchies. Instead of the typical tabular format seen in traditional databases, OLAP data are structured to have a multidimensional framework that can be represented as a data cube. The OLAP strategy for multidimensional modeling is an old technology that has been used heavily in the business sector, such as retail business and drug stores, long before it became popular in other industries 5 , 6 and is only recently being introduced to biomedical science. In business, measurable facts are mostly additive numeric values (such as sales, stock, etc.) and can be aggregated easily using sum and average. In biomedical research, however, many measurements are relatively complex. This requires a more sophisticated approach, with which we have experience and expertise. We have conducted previous extensive work on clinical data warehousing 7 , 8 that provides a framework for us to use multidimensional design for biomedical research applications. These design adaptations have potential for use in many important situations in biomedical research and are particularly valuable for assisting quality control and statistical analysis.
Figure 4 illustrates the principle of the OLAP Cube developed for a phenotyping: genotyping study of COPD case study. Among its relevant dimensions are: demographics, clinical history, drug use, pulmonary ontology, concomitant diagnosis, pulmonary function studies, quantitative indices from CT scans, plasma cytokine levels, genotyping, and time. Each dimension has a hierarchy.
Figure 4.
OLAP cube for COPD study.
To assist with rapid reporting and analysis, OLAP is designed to support intensive, complex and ad hoc queries. An important feature of OLAP is the presentation of information at different levels of detail through aggregation (roll‐up) and disaggregation (drill‐down) of data over any two dimensions. Using OLAP technology, we can extract and present information using the “slice and dice” and “drill down” formats that make OLAP multidimensional data organization such a powerful tool. Using the cube illustrated in Figure 4 , users are able to view possible associations between any two dimensions, i.e., numbers of subjects studied to completeness of data entry during the study or pulmonary function studies and plasma cytokine levels, or haplotype to quantitative CT metric for interim or final analyses.
The underlying OLAP application permits all team members to flexibly query information entered into the system, be responsible for the data elements they are supposed to add, and create interim report cards that can be customized to each team member. This permits quality control, rapid reporting, data analysis, and data extraction for statistical analysis.
The secure and trackable “drill down” function can be used to export data into Excel, MS Access, or into any format used by statistical software. The statistician can upload the reports of their analysis, the statistical application code, data set, results, and analysis into the Information Hub. The secure hub has version control, so a complete record of statistical code, their associated data sets, and their revisions can be maintained.
COPD as a case example of eM‐CTR
COPD is a common progressive chronic debilitating disease that conveys major morbidity and is the forth most common cause of death in the United States. However, there is currently no evidence of a therapeutic intervention that influences its natural history.
The NHLBI has identif ed COPD as a critical unmet need in its prioritization of its recent portfolio and recently funded four Special Clinical Centers of Research (SCCOR) to address this challenge. The SCCOR in COPD at the University of Pittsburgh has used the eM‐CTR, built on our prior research experience, from the outset of planning. This experience illustrates the added value to our current level of research activity.
Our clinical research group had acquired preliminary data to suggest the hypothesis that COPD exhibits a spectrum of entities ranging from pure parenchymal destruction to pure bronchial immune‐mediated disease, with different pathophysiological mechanisms involved. Importantly the relative contribution of each process influences the disease outcome. An implication of this hypothesis is that if a mechanism is defined, which is only relevant for a subpopulation, then therapeutic modification of this mechanism would only be expected to alter disease pathogenesis in that patient subset.
This hypothesis was built on recent application of new technologies to discrete different study populations; the challenge was to integrate the ideas in prospective studies.
The new advances included: (a) breakthrough work in quantitative objective measurements derived from spiral CT chest scans. These measures provide independent quantitation of alveoli destruction in emphysema, and bronchial thickening due to inflammation; (b) quantitative histopathology of lung resected specimens in patients who had prior CT‐scans of the resected area; and (c) the identification of activation of T‐cell subpopulations exhibiting differential cytokine expression in both peripheral blood monocytes, inflammatory cells in the lung in progressive COPD; and (d) ability to measure multiple genetic, genomic, and proteomic endpoint measures.
Preplanning—team building
In order to prospectively address this hypothesis, the first step was to create a project team and build a purpose built “COPD Research” site for the Information Hub.
The team assembled covered a broad range of disciplines ( Figure 5 ):
Figure 5.
Prototype example: technological application to investigation of COPD nomenclature.
-
1
Clinical pulmonologist as PI.
-
2
Pulmonary physiologist.
-
3
Epidemiologist studying risk of lung cancer from smoking.
-
4
An expert in CT imaging quantitation.
-
5
Pulmonary pathologist.
-
6
Signal transduction molecular biologist.
-
7
Geneticist.
-
8
Basic science clinical immunologist.
-
9
Clinical pharmacologist.
-
10
Mathematical computational expert (modelers).
-
11
Bioinformatic specialist.
-
12
Statistician.
-
13
Nurse coordinator.
In this instance, each of these team members had a positive contribution to add to the design as well as the execution of a series of protocols to study the COPD. The basic starting point for discussion was agreement in the challenge in studying multiple levels of a disease process to improve understanding of the relevance of many discrete biological processes, especially those diseases linked to gene variation or biochemical aberrations. The conventional approach does so by requiring large studies obtaining a small number of key variables to be able to discern an effect, due to either a marginal contribution of a variable to all cases or dilution of a relevant subset with cases where this effect is noncontributory. 9 As the heterogeneity and complexity of disease processes is becoming better understood, the limitations of confining enquiries to two or three domains of information are becoming apparent. Our contemporary focus changed to study multiple variables within COPD as a disease process with the intent of identifying subgroups of patients whose disease process can be more clearly defined to better characterize not one phenotype, but multiple phenotypes and better understand the complexity present in an individual patient. 10 This strategy requires substantial amounts of additional information. These multiple domains of information converge on the individual study subject at the time of the observation and recognize the central role of the volunteer subject as the most valuable asset within a study ( Figure 6 ). The contribution of these dimensions set the stage for predicting outcome. The more clearly the outcome can be predicted in the absence of an intervention, the easier it will be to identify the influence of an intervention.
Figure 6.
Example: inhaled cyclosporine clinical trial for COPD.
Based on this initial premise, studies were designed to redefine the classification of COPD based on molecular biology, further probe the role of immune mechanisms, and evaluate the innovative use of inhaled cyclosporine in only those patients with evidence of T‐cell abnormalities. Thus, studies encompassed different groups from as large as 800 patients for a long‐term follow‐up, to 60 patients for an in‐depth pharmacology intervention Phase II study.
Planning—protocol design
The task of coordinating group meetings from such diverse disciplines was not trivial. The role of the eM‐CTR for COPD was to provide the Information Hub in assembling the project team under the leadership of the PI to help in scheduling, create secured and backed up communication links via intranet project‐specific website between investigators, provide management help (in regulatory requirements, budget preparation, etc.), assist in version control of written material from the outset, and assist in the overall process management of project development. Immediate availability of this resource with minimal investment can be pivotal in success or failure in getting complex, diverse groups together, and to help share a common language and set of objectives. This support is of great value to both the young clinical investigator (still in training) and the established investigator.
Once the COPD research intranet website was established, access for read‐only or full editing capacity for each element included was allocated on an “as needed” basis for each individual. The centerpiece for the agenda was a planning time line agenda to assist in coordination. This COPD research site was the central forum for protocol development with built‐in version control for each of the elements for grant, protocol, and consent forms that are each posted. This resource not only centralized activities as they developed, but permitted information exchange of bulky packages of information without overloading email systems and provided ease of scheduling group meetings.
Group meetings were supplemented by the availability of the Communication Hub PC to PC videoconferencing desktop sharing communication between investigators that was particularly efficient in time management and minimizing travel needs. In practice, the protocol strategic development remained critically dependent on a small number of face‐to‐face group meetings to create a dynamic interpersonal interaction. These were then supplemented by distance communication after the identification of immediate action items for designated team members to be conducted after the meeting where the advantages of a distributed work team addressing different tasks at the same time become apparent.
At the conclusion of this process, the defined product from the collaborative interaction was a complete, written research proposal that was fortunate enough to receive NIH funding. A series of Institutional Review Board (IRB) approved protocols with associated consent forms, each with a MOP, the case report forms required for each component, and protocol specific data dictionaries. This process maximized the efficient use of the limited resources available to be ready for implementation when the funding was provided.
As important as the tangible products of these activities was the cultural change that took place in team members during this educational process. This observation is consistent with the observation of Schwier and Balbar, who reported that synchronous communication (i.e., videoconferencing) can lead to a “sense of community.” This sense of community was clearest in the coordinator team members. 11 Each team member had a sense of empowerment and involvement for their individual role, a series of interpersonal links had been forged and each member had access to the same set of well‐organized information on their own personal desktop. This continues to provide an important ingredient of the groups “Espirit de Corps” and group membership during the implementation of these studies.
Our experience with research teams has shown that the teams were composed of a combination of technology neutral, technology receptive, and technology adverse members. 9 , 10 For success, it is imperative that the eM‐CTR serves the entire research team. Our strategy of using the highest level of technology on the back end of the informatics system at our central location, and being receptive and adaptive to the technology level of the individual research team members on the front end at each information source domain has been very promising and effective in meeting project goals. We propose that eM‐CTR or integrated programs built around the same premises are a valuable support to CTR that can be provided throughout the CTR community including the consortia of institutions that hold Clinical Translational Science Awards (CTSA).
Summary
The potential for eM‐CTR to contribute to CTR that probes the pathophysiology of disease processes and develop interventions that favorably modify disease outcomes is enormous. It is increasingly becoming apparent that recent advances in technology linked to bioinformatics and management support systems will open up whole new avenues of science that until recently have been unapproachable. Intriguingly, the changes we are observing are more fundamental than transforming; they are disruptive in the sense that their application is creating a new research ethos. The availability, low cost and reliability of the eM‐CTR is feasible, practical, and adds cultural dimension to the processes and activities of translational research teams.
Conflict of Interest
The authors declared no conflict of interest.
Acknowledgments
We gratefully acknowledge the contributions of all of our authors for their efforts in completing this study. This research was supported by Grants: 1P50 HL084948‐01 SCCOR in Chronic Obstructive Pulmonary Disease (COPD); 1U54RR023506‐01 University of Pittsburgh, Clinical and Translational Science Institute; 5R01 DK059519‐05 Drug Metabolism and Chronic Liver Disease; 5R01 CA059834‐11 Drug Metabolizing Enzymes—Risk Factors in Bladder Cancer; 1 U10 HD047905‐03 Pregnancy and Drug Metabolizing Enzymes and Transporters.
References
- 1. Johnson B. SharePoint 2003 and Macs: How to make the most of the experience. MACTECH. 2006; 22(6): 22–28. [Google Scholar]
- 2. System Requirements for Windows SharePoint Services 3.0. Windows SharePoint Services: Microsoft TechNet 2008.
- 3. Gillam C, Oppenheim C. Reviewing the impact of virtual teams in the information age. J Inform Sci. 2006; 32(2): 160–175. [Google Scholar]
- 4. Aitken AM. Managing unstructured and semi‐structured information in organisations. Computer and Information Science 6th IEEE /ACIS International Conference ; 2007: 712–717.
- 5. Kimball R, Ross M. The data warehouse toolkit: the complete guide to dimensional modeling. 2nd ed, New York : Wiley, 2002. [Google Scholar]
- 6. Shoshani A. OLAP and statistical databases: Similarities and differences. In: Proceedings of the ACM SIGACT SIGMOD SIGART Symposium on Principles of Database Systems PODS ; New York , NY , 1997.
- 7. Parmanto B, Scotch M, Ahmad S. A framework for designing a healthcare outcome data warehouse. Perspect Health Inf Manag. 2006; 2(3): 1–8. [PMC free article] [PubMed] [Google Scholar]
- 8. Scotch M, Parmanto B. Development of a numerical‐spatial decision support system for community health assessment research: spatial OLAP visualization and analysis tools (SOVAT). Int J Med Inform. 2005; 75(10–11): 771–784. [DOI] [PubMed] [Google Scholar]
- 9. Seyal A, Rahman M, Rahim M. Determinants of academic use of the Internet: a structural equation model. Behav Inform Technol. 2002; 21(1): 71–86. [Google Scholar]
- 10. Parker H, Chignell M, Ruppenthal L. Measurement of technology mediated communication style: the communications preferences inventory. In: Proceedings of the 2002 conference of the Centre for Advanced Studies on Collaborative research; Toronto , Canada , 2002: 1–14.
- 11. Schwier RA, Balbar S. The interplay of content and community in synchronous and asynchronous communications: virtual communication in a graduate seminar. Can J Learn Technol. 2002; 28(2): 21–30. [Google Scholar]