Clinical Trial Design Automation

December 1, 2013
Srini Dagalur
Applied Clinical Trials
Volume 22, Issue 12

Incorporating CDISC-based libraries to store study components for study design and building eCRF pages.

There have been significant changes in the way clinical trials are executed today—with a move towards eClinical where technology-based solutions are being widely used to drive automation and overall efficiency into the clinical trial end-to-end process. The majority of these current approaches are more siloed around systems such as CTMS, EDC/CDMS, IVR and submissions management (i.e., trial master file, CSR, etc.). There still exists certain phases such as study design and study build within the clinical trial process that can have a significant impact based on reduction in overall time for getting trials ready for study execution across the clinical sites.

This article provides a framework called clinical trial design automation that helps establish a recommended approach to manage the study design and study build phase of the overall clinical trial process. The approach includes process and technology to aid in speeding up critical elements of a clinical trial lifecycle—elements covered are from study design (protocol design) to study build. This article highlights how this recommended approach will drive efficiencies by reducing overall timelines for building eCRF-based clinical trials including the drive towards reusing protocol elements and eCRF page elements using standards-based libraries. The conceptual architecture provided describes the components of the architecture and the process upon which this recommended approach for clinical trial design is founded on. Most of the components necessary for the proposed approach can be configured based on pre-existing solutions in the marketplace, while the core component termed clinical design interchange doesn't exist and will need to be designed and built. The use of industry data standards and CDISC-based messaging format for storing and communicating is proposed. The key mechanism in the proposed approach is the creation and maintenance of study components which are constantly being built in a clinical library resulting in reusable components that enables flexibility and adaptability to changing needs/regulatory requirements and support for global clinical trial execution.

This recommended approach has significant value for key clinical trial stakeholders—sponsors, CROs, and eClinical vendors, each with the potential for driving efficiencies in the study design and study build phases of clinical trials resulting in the overall cost reduction of clinical trial execution while enabling flexibility to meet new regulations and clinical trial trends.

Current state of clinical trial design

Life science organizations are facing challenges with escalating costs emanating from increased operational complexities and ever-changing regulatory requirements within the United States and across the globe. In many cases, primary increases in costs for bringing products to market are around planning and execution of clinical trials. Further, the clinical trial market place is experiencing the following key trends as a result of increased cost of drug development, reduced probability of drug success, and failure to meet trial deadlines due to trial operational complexities.

  • Maturity of CRO's, emergence of boutique CROs, and other support organizations. Studies show that CROs1 complete drug development faster than the drug companies themselves, without sacrificing data quality. Boutique CROs are focusing on specific therapeutic areas and provide expertise such as protocol design, patient/investigator recruitment, etc.

  • Globalization impact has increased clinical trials being conducted outside of United States and Europe. It has been driven by reduction in costs, increased patient recruitment across varying demographics, and less bureaucracy.

  • Drive toward adaptive trial design. Eighty-nine percent of all drug candidates from the initiation of Phase I through FDA approval fail in the clinic. Adaptive trials offer the potential to enable more levels of doses to be studied using maximum tolerated dosage over a select patient population.2 Adaptive trials improves the decision making in the identification of drugs with reduced probability of success.3

  • Implications of EHR/EMR. Aiding in patient recruitment4, 5 by identifying patients that meet clinical protocol criteria. Clinical data captured via these methods are helping answer questions about the safety, effectiveness, and costs of new treatments.

Clinical trial design automation

The proposed approach is based on building clinical studies centered on the following steps:

  • Developing study/protocol based on individual protocol components that are stored in a reusable CDISC-based repository.

  • Designing individual eCRF page elements that are stored in a eCRF library.

  • Leveraging study maps which define eCRF pages based on linking eCRF page elements to study/protocol components.

  • Building entire eClinical studies based on study maps.

The intent of the proposed approach is to enable the reduction in cycle times based on process improvements including reuse of clinical trial design content (study design and study build) while providing for a flexible framework that adapts to meet clinical protocol needs (Figure 1).

Each of the key components of proposed solution architecture is described in detail below:

Structured protocol authoring platform. This component is based on document management solutions that provide structured authoring,6 wherein complex documents are broken down into smaller components and then assembled together via document maps to create final published documents in multiple formats. In the case of clinical trial protocols, it is based on decomposing clinical trial protocols (includes protocol amendments) into components and building protocol components independently and then assembling these components into protocol documents using pre-defined protocol document maps. All these components are stored in a document store called the protocol repository that serves as the primary content storage mechanism for managing protocol components. The protocol repository is based on the protocol representation model (PRM)—a CDISC standard that supports the planning and design of a clinical trial protocol including protocol amendments. Further, the submission data tabulation model (SDTM)—trial design model (TDM) datasets are used as a source of elements for PRM.

As this approach to protocol authoring takes effect, the protocol repository becomes richer in content and further drives speed of authoring protocols. This approach results in reduced overall time to approve studies/protocols based on:

  • Minimizing document hold-ups with different authoring groups.

  • Using a component library that brings speed to content creation by reusing elements and also developing standard components that do not change across protocols.

  • Enhancing support for regional/local/country specific protocol changes by targeting specific protocol components that are impacted due to regional regulatory needs, which also speeds up overall approval times compared to traditional approaches.

The implementation of this part of the proposed architecture can be based on configuration of pre-existing structured/component authoring solutions that meets this proposed approach.

EDC platform. The EDC platform7 manages the design and build of the eCRF pages including the underlying clinical database that support these eCRF pages, all during the study build phase. This architecture component provides this feature and also includes the eCRF design tool that enables the layout and creation of eCRF pages. The study definition interchange provides a standard-based CDISC integration layer that describes the protocol structure defined during the study design via a map. This map links pre-existing elements across protocol and eCRF page elements. During the study build phase (i.e., eCRF build) the EDC platform processes the map information to link each of all protocol component elements to its corresponding eCRF page elements based on these maps. The eCRF pages are then generated including the database to store eCRF results. Within this component as eCRF pages are generated, for those protocol components for which eCRF page elements do not still exist in the eCRF library, new eCRF pages including maps will be created and maintained.

Study definition interchange. This is the core component of the proposed architecture and functions as the connectivity layer between protocol authoring (study design) and eCRF build (study build). The study definition interchange component manages all mapping and any transformations that are necessary between protocol elements and eCRF page elements. It provides integration services that are based on CDISC standards linking protocol authoring platform to the EDC platform. The rules, transformations, and maps linking study/protocols to eCRF pages are also stored within this component. Maintaining all these elements requires a simple user-interface to create, update, and manage components throughout the lifecycle. On the study design (protocol authoring) side, all elements in the protocol repository are exposed in the form of PRM elements within the study definition interchange layer in order to be transformed into specific eCRF page elements based on pre-existing protocol maps and transformation rules. On the study build (eCRF page creation) side, the study definition interchange component takes each protocol component exposed as PRM components and maps it to eCRF page elements based on clinical data acquisition standards harmonization (CDASH) standard, a subset of CDISC.

Since the study definition Interchange component of the proposed solution architecture has to be designed and built, below are some key design considerations for this component:

  • Store all study definition metadata in this component.

  • Leverage dictionaries and CDISC standards to provide interoperability between clinical systems across study design and build (e.g., WHO Drug, PRM/SDTM/TDM/ODM/CDASH, etc.).

  • Align on standards naming/metadata across all areas (global, therapeutic area, study, etc.).

  • Maintain version control, data traceability, and audit trails for regulatory compliance.

  • Leverage an ETL engine that supports transformation and store maps for study metadata across standard libraries.

The study definition interchange is a new component that does not exist in the market and will need to be developed or extended based on some pre-existing solutions.

For the clinical trial design automation approach to be effective it is integral for the maintenance of both the standard libraries—protocol and eCRF—both of which manage and store protocol components across all clinical trials (Figure 3).

Proposed maintenance process

Key groups for proposed process are described in detail below:

  • The governance committee is responsible for managing and approving overall content within study standard libraries: protocol and eCRF. They also approve QSD (quality system documents) that describe the process and policies to help manage study standard libraries.

  • The standard libraries librarian helps manage, track and get documentation ready to governance committees for approval. Based on approval status help notify users of standard libraries in terms of components that are ready to be published and made available for clinical studies within the organization.

  • The component proposer group is primary authors of components that exist in the standard libraries. This group is responsible for preparing documentation on new components including updates to pre-existing library components. Works with the librarian group in getting documentation ready for approvals via the governance committee.

These three groups in concert with the proposed process help maintain the study standard libraries and enable the proposed approach in driving efficiencies into study design and study build process.

Benefits of clinical trial design automation

The benefits for the proposed clinical trial design automation approach are as follows:

Study design—protocol development:

  • Structured protocol authoring drives: quality, efficiency,search, and collaborations.
  • Standards based development enables upstream anddownstream integrations.
  • Ensures compliance with SOPs and regulations.

Study design—clinical data capture build (eCRF):

  • Standardized protocol and eCRF libraries improve reusability,visibility, and maintenance.
  • Industry standards (CDISC) based metadata and datacapture.
  • Enhances management of eCRF while enabling enhanceddata collection methods.
  • Streamlined clinical development process, harmony betweenclinical operation and data management.

Trial document management:

  • Facilitates easier management of documents based on documentrepository with version controlled check-in checkoutcapability, approval workflow, and electronic signatures.
  • Improved productivity (notifications and task management)and compliance.

Access to data and information:

  • Data is exchanged electronically using the CDISC-basedXML engine instead of documents.
  • A single authoritative source of truth is maintained andconsumed by the protocol documents and trial specificeCRF pages.
  • Enabling data mining/analytics around clinical trial design.
  • Improve enforcement of business rules.
  • Increased agility for organizations to create and adapt tonew business processes and conditions such as adaptivetrials and translational medicine.

Conclusion

This proposed approach has significant value and benefits for the clinical trial stakeholders: sponsors, CROs, site users, eClinical vendors, and others (regulatory, IRBs) based on driving efficiencies and flexibility into clinical trials process, reduction in cost and time across study design and study build phases.

Additional charts for stakeholder steps and benefits of clinical trial design automation are available in this online article.

Srini Dagalur, PhD, is Specialist Leader, Life Sciences, Deloitte Consulting, LLP, 100 Kimball Drive, Parsippany, NJ, 07054, e-mail: [email protected] or [email protected]

References

1. M. D. Masri, et al., "Contract Research Organizations: An Industry Analysis," (2013), http://bit.ly/IcALFq/.

2. Tufts CSDD, "The Adoption and Impact of Adaptive Trial Designs," (2013), http://bit.ly/18bnGZg/.

3. A. Schafer, "Adaptive Trial Market Dynamics," Applied Clinical Trials Online, (2012), http://bit.ly/1edN4Nm/.

4. M. D. Hirsch, "EHR Alerts Help Recruit Patients for Clinical Trials," FierceEMR, (2012), http://bit.ly/JgRYZL/.

5. J. C. Dooren, "Making Clinical Trials Less of a Tribulation," The Wall Street Journal, (2011), http://on.wsj.com/17uQjuy/.

6. National Institutes of Health, Clinical Center Department of Clinical Research Informatics and Protocol Management Services, "A Web Based Protocol Writing System ProtoType," (2008), http://bit.ly/1byD4dZ/.

7. J. Shah, et al., "Electronic Data Capture for Registries and Clinical Trials in Orthopaedic Surgery: Open Source versus Commercial Systems," Clin Orthop Relat Res, 468 (10) 2664–2671 (2010).

download issueDownload Issue : Applied Clinical Trials-12-01-2013