Lee S. Scheible

	As the biopharmaceutical industry continues to move down an outsourcing path, it is essential that logical sourcing strategies and operational processes exist and align with corporate strategy. This alignment will ensure desired outcomes are realized, such as shorter timelines, enhanced trial enrollment, and reduced overall costs while maintaining quality. Sponsors are often so focused on specific clinical deliverables that they often overlook the essential need for sourcing strategies, detailed deliverable specifications along with robust operational and oversight processes. In a report by Booz & Company based on interviews with top-20 pharmaceutical companies, it found that while sponsor companies have expanded their level of outsourcing and embraced deeper and more integrated third party organization (TPO) relationships, sponsors have done so without carefully thinking through the necessary support mechanisms and without aligning their outsourcing and corporate strategies.

 This lack of proper planning and oversight can often lead to operational efficiency, quality, and regulatory issues as regulatory bodies are now heavily focused on TPO oversight. This article discusses the essential steps for developing and implementing a logical sourcing strategy.

Step 1: Developing a logical sourcing strategy (what is needed and when)

	In order to have a truly logical and effective sourcing strategy, the sponsor must first have a clear understanding of its current and near future portfolio. This includes not only understanding therapeutic area needs, but also understanding the portfolio at the compound and trial levels, including trial volume and design. A logical sourcing strategy must be developed that provides flexibility in order to meet a variety of therapeutic study designs and compound-development needs while also ensuring consistency, quality, and timeliness of data and clinical processes, ultimately ensuring patient safety, data integrity, and the on-time delivery of the portfolio while controlling developmental costs. 

	A portfolio analysis is needed initially to develop the initial strategy, and then periodically in order to proactively plan and adjust the sourcing strategy as the portfolio changes. The analysis should be conducted over a suggested three-year time span in order to proactively set the sourcing plan and to ensure that the proper number and types of vendors are selected to meet the projected work volume. Contingency plans may easily be added to the overall strategy to buffer for any sudden portfolio shift. Critical trial criteria for successfully conducting the study need to be determined in order to fully understand the true clinical and sourcing needs for each trial type making up the portfolio. Trial types might include trial phase (IA, IB, IIA, III, etc.), trial size, complexity, specialty design needs (i.e., adaptive design) and registration needs. 

	The critical trial criteria might include clinical expertise, operational expertise, access to patients, investigator connectivity, site interface, recognized areas of therapeutic expertise, geographical presence, service capacity, timeliness, and strategic advantage. See Table 1 for an example of a critical trial criteria table.

Vendor and sourcing model analyses are now needed to align with the previously determined portfolio needs. Vendor models would include the larger, full-service contract research organization (CRO), the smaller specialized niche CRO, and the full-service clinical network. Sourcing models such as full (or project), functional (or process), and role-based sourcing also should be compared and aligned to the portfolio needs. The detailed analysis information needs to be collected using a request for information (RFI) process. It is crucial that a well-designed RFI document with specific questions measuring for critical trial criteria, cross-functional services, vendor capability, and capacity is utilized to ensure an accurate vendor and sourcing model analysis is performed.

	Questions seeking the same information should also be placed in multiple sections of the RFI document to serve as an accuracy check. The RFI document should also include standardized definitions and data collection templates to ensure data entry consistency. The RFI data collection process should be a completely web-based process with data aggregation, filtering, and reporting functionality. The RFI should be reviewed by a cross-functional analysis team to ensure that an accurate vendor model analysis occurs in order to ensure optimal trial-model alignment within the overall strategy summary. See Table 2 for an example of a vendor capability table.

Analysis of the detailed portfolio and critical trial criteria data compared to the vendor capability data will allow for the preparation of a sourcing strategy summary including optimal vendor model types and target vendor pool numbers used to finalize the vendor pool. The vendor pool numbers are extremely important, ensuring that there are not too many vendors not allowing for steady stream utilization, but not too few in order to allow for portfolio/trial-need flexibility.

	A Phase I/II adaptive design trial has very different needs compared to a large, global, Phase III registration trial. The summary should also be written in a user-friendly manner listing out the primary and secondary options or models by therapeutic area (if applicable) and trial type, along with a brief justification explaining the benefits and reasoning of the suggested options. To optimize use, the strategy summary needs to be converted into a very simple, user-friendly tool. Figures 1 and 2 below illustrate examples of a user-friendly, logical decision tree and a vendor selection consideration tree, respectively. 

Step 2: Implementing using logical operational processes (how to make it happen)

	In order to optimize the benefit of the sourcing strategy and ensure delivery of the portfolio, a logical operational process flow must be developed and implemented consistently across all study teams within the sponsor organization. The first key step in the process is the selection of an optimal TPO from the TPO pool based on the trial design or program needs. It is recommended that both a primary and secondary TPO choice be made in case the primary TPO choice has capacity issues or competing trials at that time. The selected TPO must be checked and verified for proper qualification dependent on which services are being outsourced. It is highly recommended that all TPOs or at least the core set of primary TPOs within the overall pool be periodically reviewed and qualified prior to being assigned a project in order to minimize trial startup delays. Periodic portfolio reviews are recommended and as the portfolio and sourcing strategy shifts, the TPO pool and qualifications may need to adjust as well.

	Next, a request for proposal (RFP) packet must be compiled for the project. The RFP packet must include a draft or final protocol, detailed specifications and operational plan(s) dependent on the services being outsourced. These detailed documents are critical to ensure optimal TPO comprehension of service expectations, compliance, and ultimately delivery of the key deliverables. This includes the development of an accurate service budget/proposal to avoid future trial delays, issues, and change orders. Once the final contract and budget are negotiated and executed, the project kick off meeting may occur where the operational plan is further discussed along with the development of mutually agreed upon communication and risk plans. Communication flow is extremely important to ensure that the trials are implemented and managed in an effective and streamlined manner. Optimal flow would need to flow from function to function, sponsor to vendor, and from vendor to vendor, especially if functional/process sourcing is being utilized to ensure compliance with functional process requirements and to avoid duplication and other issues. The flow must also be coordinated and facilitated by one primary function (i.e., project management or external sourcing), though all functional areas will have partial ownership.

	Recent publications have reported on a lot of dialogue and activity on many fronts on how to improve overall efficiencies in clinical trials and how sponsors may engage properly with their outsourcing partner(s) so that efficiencies and cost containment are a priority on both sides. Collaborative clinical/technical interactions between the select vendor pool and the sponsor need to occur. These interactions optimally need to occur at the portfolio, compound, and trial levels in order to ensure logical trial designs and trial placements occur for each compound that makes up the overall sponsor portfolio and to ensure a priority exists on both sides. Most sponsors only interact with their vendors at the trial level. In order to allow for proactive resource planning, the vendor-sponsor interactions need to occur at the portfolio and compound level. 

Step 3: Measuring and maintaining through efficient control processes (how to ensure it truly happens and continues to happen)

	Once the project services have been implemented, effective control processes must be put into place to ensure the deliverables are consistently delivered as specified in the contract.  The first control process is efficient TPO oversight implemented in a manner to maximize quality while also minimizing sponsor and TPO staff impact to help control costs. An oversight plan will be needed to ensure consistent implementation across the sponsor staff.

	Functional areas may need to develop separate oversight plans or a separate section within an overall plan to ensure their unique deliverable requirements are being met. Effective oversight needs to include both lagging and leading indicators. Lagging indicators would be used to measure historical results telling the sponsor how well the TPO has performed, while leading indicators would tell the sponsor if its sourcing strategy is effective and logical. Lagging indicators may include enrollment-to-plan and data-transfers-to-plan, while corresponding leading indicators might be number of initiated sites, countries with FPV, or outstanding monitoring visits (backlog).

	Clinical processes can start to go wrong well before the TPO performance turns the lagging indicator red. It is crucial that the right indicators be measured, paired, and analyzed or investigated to tell an accurate performance story-and then acted upon in a timely manner by the sponsor to ensure the TPO is performing at a level to deliver the trial and justify continued inclusion into the select vendor pool. In addition, positive trends need to be continued through shared learning and negative trends mitigated through corrective action preventative action (CAPA) development.  

	Indicators must be standardized across all TPOs governed by the strategy in order to have fair and accurate cross-TPO performance analyses. Indicator categories might include site selection, regulatory readiness, enrollment to plan, financial,

 dataflow, and compliance. In addition, it is recommended that an overall value model be developed and tracked along with the performance indicators looking at not only TPO performance, but TPO service fees and sponsor staff oversight that impact the overall value to the sponsor. The actual TPO service fees may be easily monitored by tracking the TPO invoices, while the actual sponsor staff oversight may be tracked using a time entry process (i.e., database, excel sheet) by service categories. The goal of the value model analysis is not to pick the least expensive TPOs, but rather the TPOs with the best overall value. Figure 3 above illustrates the typical vendor value model.

	Two additional areas important to the overall oversight plan are quality/compliance management and TPO governance. Quality /compliance management would include ensuring that all TPOs within the TPO pool have been properly assessed prior to use for proper procedures being in place, staff qualifications, system validations, security, privacy, and other key requirements.

	In order to ensure timely project initiations, the TPO pool must be qualified at all times and ready for use. To ensure this “state of readiness,” periodic reviews of the TPO pool, TPO services, service locations, and staffing must occur. These reviews could be incorporated into a periodic portfolio review, since the portfolio needs are directly correlated to the TPO pool needs. Thus, as the portfolio needs shift, the TPO pool would be reviewed, including TPO services, service locations, performance indicators, and need for additional assessments. Periodic checks for TPO service changes and needs for additional assessment could also be incorporated into the TPO audit cycle, which the sponsor’s quality department would periodically conduct usually based upon a risk-based approach. Quality indicators could also be added to the periodic performance reviews as an alert mechanism to the sponsor. 

	These alerts could be discussed at the TPO governance meetings, which are also part of the oversight plan. The key to having effective TPO governance is having the “right size” governance. Similar to performance indicators, excessive governance can be counter-productive, where the resource costs outweigh the benefit to the TPO and the sponsor. The key to determine the optimal level of governance is to have a risk management plan in place to determine the TPO’s relationship impact on the sponsor’s portfolio. Key risk factors to consider include work/project volume, service/project complexity and impact, safety and quality impact, budgetary impacts, and competitive advantages.

	Typical TPO governance levels start at the project team operational level, focused on a specific project, and then escalating to the enterprise operational level, focused on multiple projects, and, finally, to the executive operational level, focused on strategic development and efficiency. All of the levels focus to some degree on relationship development. The level and intensity of governance (i.e., meeting frequency, length, committees, communication channels, etc.) all depends on the risk determination so that the “right size” is applied.

	In order to have effective uptake and compliance to the oversight plan and control processes, it is crucial that a positive TPO relationship be maintained throughout the entire process. TPO relationship management is not easy, especially when managing multiple TPOs within the TPO pool. The key to a positive relationship is to create and maintain a collaborative environment between the sponsor and TPO staff. An effective communication plan and open channels are the first step to creating a collaborative environment, ensuring that both the TPO and sponsor staff is kept updated at all times, making them feel like an actual team member.

	A collaborative approach should be utilized for the development of the final, standardized indicators, so that the TPOs feel they contributed to the development of logical and fair indicators. These indicators will allow for cross-TPO performance analysis in a de-identified, confidential format, and all trial or project work with the sponsor will be measured and analyzed in one standardized manner. 

	Not all TPO specific indicator suggestions or requests will be met; this point needs to be stressed during the collaborative development process. The performance indicators must be reviewed on a periodic basis (i.e., annually, biannually) and user feedback collected for definition and analysis criteria enhancements based on the actual cross-TPO performance trends from the previous performance period. In addition, as business processes change, performance indicators may require additional modifications to remain applicable and effective. A final point to stress on performance expectations and indicator development is that a sponsor should never expect more from a TPO’s performance than it would expect from their own performance.

Step 4: Replication using a logical document architecture (how to ensure it is all properly documented to allow for compliance and replication)

	A logical, streamlined third party management (TPM) document architecture is essential to ensure optimal process and deliverable specification comprehension and, ultimately, efficient service delivery. All too often the sponsor will create very detailed, function-specific operational plans and other TPM governing documents in silos (i.e., in a specific functional areas like data management or statistical services) without consideration of other functional area impacts.

	This cross-function oversight omission will often lead to conflicting process and specification conflicts. Excessive, duplicative details and content will often be included in the TPM document(s). Over time, as regulations, business processes, and requirements change-leading to needed document revisions-conflicting content will often arise as different functional areas update at different times and in different manners. These issues clearly illustrate the difference between having TPM documents versus a true TPM document architecture. A true architecture will consist of multiple standardized documents that are written in a manner to clearly define, govern, and manage both the work and relationship between the sponsor and the TPO. Each document that makes up the overall architecture must have a clear purpose and proper connectivity among the other documents. This will then allow for a logical hierarchy and standardized content landscape, helping to ensure optimal content comprehension and management.

	So, how many TPM documents make up a logical architecture? The first step to developing a logical architecture hierarchy is listing out the key content categories such as privacy, confidentiality, escalation, governance, onboarding, training, and records management, as well as several other categories that need to be covered within the hierarchy. The category listing needs to be developed by a cross-functional team ensuring all functional needs are considered and based upon all TPM regulations, requirements, and applicable business processes to ensure overall compliance. Once the list is reviewed and confirmed by the functions, the next step is to determine the logical combinations and overall placement of the content categories to ensure wide accessibility and optimal user comprehension.

	The overarching content landscape and hierarchy, including the logical number of documents, will then be finalized, always keeping accessibility, comprehension, and compliance in mind. Though architecture consistency across the TPOs is extremely important for optimal-user comprehension, not all TPO architectures may be exactly alike. Lower-volume TPOs may not require an extensive architecture, such as a high-volume TPO, and thus the sponsor needs to determine what approach is optimal for each TPO category, always keeping accessibility, comprehension, and compliance in mind.

	Using the content landscape as a guide or road map, the cross-functional team would then need to focus on the development of logical architecture templates. It is essential that key instructions are written in a manner that is user-friendly and allows TPO relationship managers and study teams to convert the templates to TPO-specific and trial-specific documents. The standardized templates would be developed by functional area subject matter experts (SME). Once approved, a template owner would be assigned to each template in order to ensure the templates are periodically reviewed and the content maintained as regulations, requirements, and business processes change. 

	To help ensure proper architecture utilization, and maintenance, a content maintenance and change-control process must be implemented as the architecture is developed. These processes are crucial to ensure the architecture content does not gradually drift back into silos, ultimately leading to duplicative, conflicting, and ambiguous content. In addition, a TPO conversion checklist tool or process flow mapping all of the conversion tasks is highly recommended to help ensure consistent and compliant implementation. Once TPO-specific and trial-specific documents have been created and finalized, the TPO relationship manager (enterprise level) and study team (operational level) would be responsible for document content maintenance and periodic reviews. It is highly recommend that all periodic reviews and document content decisions are recorded and filed in the trial master file showing a complete decision trail and oversight, in case of future audits or inspections.

	Centralized storage, maintenance, and accessibility are extremely important to ensure optimal architecture utilization and compliance across the sponsor and TPO staff. If staff is not able to easily access the key documents, then compliance will be negatively impacted, as well as inspection readiness. The end-user needs to know where to quickly obtain the most current version of the architecture documents and who to contact for assistance. Proper archival of previous document versions is also extremely important in order to have the ability to show the full document and oversight history. A centralized website with document management functionality, including proper metadata, which meets 21CFR Part11 requirements, is highly recommended. Proper account management and, thus, access to the site and site folders will also be important, so that sponsor staff may access all applicable documents while TPO staff may only access their TPO-specific documents for proper confidentiality. Site development and validation will be an important step to ensure all access and functionality is properly functioning.

	During these changing times as the biopharmaceutical industry is facing increasing healthcare reform challenges, sponsors are quickly focusing on the development of a more efficient clinical research model, where sourcing is an essential and strategic component. Use of a strategic, collaborative approach, including management of a pre-qualified TPO pool all based upon the true portfolio needs, is essential. Sponsors must conduct ongoing portfolio and performance indicator reviews to adjust the strategy and TPO pool accordingly, while also enforcing the utilization of the strategy across the study teams and functions. Development and maintenance of proper control and oversight processes clearly defined in a logical document architecture will help to ensure ongoing utilization of the overall strategy. 

		1. Getz, K. “Peril and Promise of Risk Imbalance,” INSIDE Outsourcing - Guide to Strategic Outsourcing - Trends, Surveys, Analysis (10-16) (Dec., 2012).      

 is an External Sourcing Relationship Manager (at the time of research); 

, PhD, is a Senior Director of External Sourcing; 

 is a Director of External Sourcing (at the time of research); all with the Development Center of Excellence, Eli Lilly and Co. 

Articles

The four essential steps to ensuring optimal clinical deliverable specifications and process controls.

Developing a Logical Sourcing Strategy

 As the biopharmaceutical industry continues to move down an outsourcing path, it is essential that logical sourcing strategies and operational processes exists and align with corporate strategy. This alignment will ensure desired outcomes are realized, such as shorter timelines, enhanced trial enrollment, and reduced overall costs while maintaining quality. Sponsors are often so focused on specific clinical deliverables that they often overlook the essential need for sourcing strategies, detailed deliverable specifications along with robust operational and oversight processes. In a recent report by Booz & Company based on interviews with top-20 pharmaceutical companies, it found that while sponsor companies have expanded their level of outsourcing and embraced deeper and more integrated third party organization (TPO) relationships, sponsors have done so without carefully thinking through the necessary support mechanisms and without aligning their outsourcing and corporate strategies.

 This lack of proper planning and oversight can often lead to operational efficiency, quality, and regulatory issues as regulatory bodies are now heavily focused on TPO oversight. This article discusses the essential steps for developing and implementing a logical sourcing strategy.

	Step 1: Developing a logical sourcing strategy (what is needed and when)

	In order to have a truly logical and effective sourcing strategy, the sponsor must first have a clear understanding of its current and near future portfolio. This includes not only understanding therapeutic area needs, but also understanding the portfolio at the compound and trial levels, including trial volume and design. A logical sourcing strategy must be developed that provides flexibility in order to meet a variety of therapeutic study designs and compound-development needs while also ensuring consistency, quality, and timeliness of data and clinical processes, ultimately ensuring patient safety, data integrity, and the on-time delivery of the portfolio while controlling developmental costs. A portfolio analysis is needed initially to develop the initial strategy, and then periodically in order to proactively plan and adjust the sourcing strategy as the portfolio changes. The analysis should be conducted over a suggested three-year time span in order to proactively set the sourcing plan and to ensure that the proper number and types of vendors are selected to meet the projected work volume. Contingency plans may easily be added to the overall strategy to buffer for any sudden portfolio shift. Critical trial criteria for successfully conducting the study need to be determined in order to fully understand the true clinical and sourcing needs for each trial type making up the portfolio. Trial types might include trial phase (Ia, Ib, IIa, III, etc.), trial size, complexity, specialty design needs (i.e., adaptive design) and registration needs. The critical trial criteria might include clinical expertise, operational expertise, access to patients, investigator connectivity, site interface, recognized areas of therapeutic expertise, geographical presence, service capacity, timeliness, and strategic advantage. See Example 1 for an example of a critical trial criteria table.

	Vendor and sourcing model analyses are now needed to align with the previously determined portfolio needs. Vendor models would include the larger, full-service clinical research organization (CRO), the smaller specialized niche CRO, and the full-service clinical network. Sourcing models such as full (or project), functional (or process), and role-based sourcing also should be compared and aligned to the portfolio needs. The detailed analysis information needs to be collected using a request for information (RFI) process. It is crucial that a well-designed RFI document with specific questions measuring for critical trial criteria, cross-functional services, vendor capability, and capacity is utilized to ensure an accurate vendor and sourcing model analysis is performed. Questions seeking the same information should also be placed in multiple sections of the RFI document to serve as an accuracy check. The RFI document should also include standardized definitions and data collection templates to ensure data entry consistency. The RFI data collection process should be a completely web-based process with data aggregation, filtering, and reporting functionality. The RFI should be reviewed by a cross-functional analysis team to ensure that an accurate vendor model analysis occurs in order to ensure optimal trial-model alignment within the overall strategy summary. See Figure 2 for an example of a vendor capability table.

	Analysis of the detailed portfolio and critical trial criteria data compared to the vendor capability data will allow for the preparation of a sourcing strategy summary including optimal vendor model types and target vendor pool numbers used to finalize the vendor pool. The vendor pool numbers are extremely important, ensuring that there are not too many vendors not allowing for steady stream utilization, but not too few in order to allow for portfolio/trial-need flexibility. A Phase I/II adaptive design trial has very different needs compared to a large, global, Phase III registration trial. The summary should also be written in a user-friendly manner listing out the primary and secondary options or models by therapeutic area (if applicable) and trial type, along with a brief justification explaining the benefits and reasoning of the suggested options. To optimize use, the strategy summary needs to be converted into a very simple, user-friendly tool. Figure 3 illustrates an example of a user-friendly, logical decision tree. 

	Step 2: Implementing using logical operational processes (how to make it happen)

	In order to optimize the benefit of the sourcing strategy and ensure delivery of the portfolio, a logical operational process flow must be developed and implemented consistently across all study teams within the sponsor organization. The first key step in the process is the selection of an optimal TPO from the TPO pool based on the trial design or program needs. It is recommended that both a primary and secondary TPO choice be made in case the primary TPO choice has capacity issues or competing trials at that time. The selected TPO must be checked and verified for proper qualification dependent on which services are being outsourced. It is highly recommended that all TPOs or at least the core set of primary TPOs within the overall pool be periodically reviewed and qualified prior to being assigned a project in order to minimize trial startup delays. Periodic portfolio reviews are recommended and as the portfolio and sourcing strategy shifts, the TPO pool and qualifications may need to adjust as well. Next, a request for proposal (RFP) packet must be compiled for the project. The RFP packet must include a draft or final protocol, detailed specifications and operational plan(s) dependent on the services being outsourced. These detailed documents are critical to ensure optimal TPO comprehension of service expectations, compliance, and ultimately delivery of the key deliverables. This includes the development of an accurate service budget/proposal to avoid future trial delays, issues, and change orders. Once the final contract and budget are negotiated and executed the project kick off meeting may occur where the operational plan is further discussed along with the development of mutually agreed upon communication and risk plans. Communication flow is extremely important to ensure that the trials are implemented and managed in an effective and streamlined manner.  Optimal flow would need to flow from function to function, sponsor to vendor, and from vendor to vendor, especially if functional/process sourcing is being utilized to ensure compliance with functional process requirements and to avoid duplication and other issues. The flow must also be coordinated and facilitated by one primary function (i.e., project management or external sourcing), though all functional areas will have partial ownership. 

	Recent publications have reported on a lot of dialogue and activity on many fronts on how to improve overall efficiencies in clinical trials and how sponsors may engage properly with their outsourcing partner(s) so that efficiencies and cost containment are a priority on both sides. Collaborative clinical/technical interactions between the select vendor pool and the sponsor need to occur. These interactions optimally need to occur at the portfolio, compound and trial levels in order to ensure logical trial designs and trial placements occur for each compound that makes up the overall sponsor portfolio and to ensure a priority exists on both sides. Most sponsors only interact with their vendors at the trial level. In order to allow for proactive resource planning, the vendor-sponsor interactions need to occur at the portfolio and compound level.  Figure 4 illustrates both the typical and proactive interaction models.

	Exhibit 4a, 4b: Examples of typical and proactive interaction models.

	Step 3: Measuring and maintaining through efficient control processes (how to ensure it truly happens and continues to happen)

	Once the project services have been implemented, effective control processes must be put into place to ensure the deliverables are consistently delivered as specified in the contract.  The first control process is efficient TPO oversight implemented in a manner to maximize quality while also minimizing sponsor and TPO staff impact to help control costs. An oversight plan will be needed to ensure consistent implementation across the sponsor staff. Functional areas may need to develop separate oversight plans or a separate section within an overall plan to ensure their unique deliverable requirements are being met. Effective oversight needs to include both lagging and leading indicators. Lagging indicators would be used to measure historical results telling the sponsor how well the TPO has performed while leading indicators would tell the sponsor if its sourcing strategy is effective and logical. Lagging indicators may include enrollment-to-plan and data-transfers-to-plan, while corresponding leading indicators might be number of initiated sites, countries with FPV, or outstanding monitoring visits (backlog). Clinical processes can start to go wrong well before the TPO performance turns the lagging indicator red. It is crucial that the right indicators be measured, paired, and analyzed or investigated to tell an accurate performance story-and then acted upon in a timely manner by the sponsor to ensure the TPO is performing at a level to deliver the trial and justify continued inclusion into the select vendor pool. In addition, positive trends need to be continued through shared learning and negative trends mitigated through corrective action preventative action (CAPA) development.  

	Indicators must be standardized across all TPOs governed by the strategy in order to have fair and accurate cross-TPO performance analyses. Indicator categories might include site selection, regulatory readiness, enrollment to plan, financial, dataflow, and compliance.  In addition, it is recommended that an overall value model be developed and tracked along with the performance indicators looking at not only TPO performance, but TPO service fees and sponsor staff oversight that impact the overall value to the sponsor. The actual TPO service fees may be easily monitored by tracking the TPO invoices while the actual sponsor staff oversight may be tracked using a time entry process (i.e., database, excel sheet) by service categories. The goal of the value model analysis is not to pick the least expensive TPOs, but rather the TPOs with the best overall value. Figure 5 illustrates the typical vendor value model.

	Two additional areas of oversight important to the overall oversight plan are quality /compliance management and TPO governance. Quality /compliance management would include ensuring that all TPOs within the TPO pool have been properly assessed prior to use for proper procedures being in place, staff qualifications, system validations, security, privacy, and other key requirements. In order to ensure timely project initiations, the TPO pool must be qualified at all times and ready for use. To ensure this “state of readiness,” periodic reviews of the TPO pool, TPO services, service locations, and staffing must occur. These reviews could be incorporated into a periodic portfolio review, since the portfolio needs are directly correlated to the TPO pool needs. Thus, as the portfolio needs shift, the TPO pool would be reviewed, including TPO services, service locations, performance indicators, and need for additional assessments. Periodic checks for TPO service changes and needs for additional assessment could also be incorporated into the TPO audit cycle, which the sponsor’s quality department would periodically conduct usually based upon a risk-based approach. Quality indicators could also be added to the periodic performance reviews as an alert mechanism to the sponsor. These alerts could be discussed at the TPO governance meetings, which are also part of the oversight plan. The key to having effective TPO governance is having the “right size” governance. Similar to performance indicators, excessive governance can be counter-productive, where the resource costs outweigh the benefit to the TPO and the sponsor. The key to determine the optimal level of governance is to have a risk management plan in place to determine the TPO’s relationship impact on the sponsor’s portfolio. Key risk factors to consider include work/project volume, service /project complexity and impact, safety and quality impact, budgetary impacts, and competitive advantages. Typical TPO governance levels start at the project team operational level, focused on a specific project, and then escalating to the enterprise operational level, focused on multiple projects, and, finally, to the executive operational level, focused on strategic development and efficiency. All of the levels focus to some degree on relationship development. The level and intensity of governance (i.e., meeting frequency, length, committees, communication channels, etc.) all depends on the risk determination so that the “right size” is applied.

	In order to have effective uptake and compliance to the oversight plan and control processes, it is crucial that a positive TPO relationship be maintained throughout the entire process. TPO relationship management is not easy, especially when managing multiple TPOs within the TPO pool. The key to a positive relationship is to create and maintain a collaborative environment between the sponsor and TPO staff. An effective communication plan and open channels are the first step to creating a collaborative environment, ensuring that both the TPO and sponsor staff is kept updated at all times, making them feel like an actual team member. A collaborative approach should be utilized for the development of the final, standardized indicators, so that the TPOs feel they contributed to the development of logical and fair indicators. These indicators will allow for cross-TPO performance analysis in a de-identified, confidential format, and all trial or project work with the sponsor will be measured and analyzed in one standardized manner. Not all TPO specific indicator suggestions or requests will be met; this point needs to be stressed during the collaborative development process. The performance indicators must be reviewed on a periodic basis (i.e., annually, biannually) and user feedback collected for definition and analysis criteria enhancements based on the actual cross-TPO performance trends from the previous performance period. In addition, as business processes change, performance indicators may require additional modifications to remain applicable and effective. A final point to stress on performance expectations and indicator development is that a sponsor should never expect more from a TPO’s performance than it would expect from their own performance.

	Step 4: Replication using a logical document architecture (how to ensure it is all properly documented to allow for compliance and replication)

	A logical, streamlined third party management (TPM) document architecture is essential to ensure optimal process and deliverable specification comprehension and, ultimately, efficient service delivery. All too often the sponsor will create very detailed, function-specific, operational plans and other TPM governing documents in silos (i.e., in a specific functional areas like data management or statistical services) without consideration of other functional area impacts. This cross-function oversight omission will often lead to conflicting process and specification conflicts. Excessive, duplicative details and content will often be included in the TPM document(s). Over time, as regulations, business processes, and requirements change-leading to needed document revisions-conflicting content will often arise as different functional areas update at different times and in different manners. These issues clearly illustrate the difference between having TPM documents versus a true TPM document architecture. A true architecture will consist of multiple standardized documents that are written in a manner to clearly define, govern, and manage both the work and relationship between the sponsor and the TPO. Each document that makes up the overall architecture must have a clear purpose and proper connectivity among the other documents. This will then allow for a logical hierarchy and standardized content landscape, helping to ensure optimal content comprehension and management.

	So, how many TPM documents make up a logical architecture? The first step to developing a logical architecture hierarchy is listing out the key content categories such as privacy, confidentiality, escalation, governance, onboarding, training, and records management, as well as several other categories that need to be covered within the hierarchy. The category listing needs to be developed by a cross-functional team ensuring all functional needs are considered and based upon all TPM regulations, requirements, and applicable business processes to ensure overall compliance. Once the list is reviewed and confirmed by the functions, the next step is to determine the logical combinations and overall placement of the content categories to ensure wide accessibility and optimal user comprehension. The overarching content landscape and hierarchy, including the logical number of documents, will then be finalized, always keeping accessibility, comprehension, and compliance in mind. Though architecture consistency across the TPOs is extremely important for optimal-user comprehension, not all TPO architectures may be exactly alike. Lower-volume TPOs may not require an extensive architecture, such as a high-volume TPO, and thus the sponsor needs to determine what approach is optimal for each TPO category, always keeping accessibility, comprehension, and compliance in mind.

	Using the content landscape as a guide or road map, the cross-functional team would then need to focus on the development of logical architecture templates. It is essential that key instructions are written in a manner that is user friendly and allows TPO relationship managers and study teams to convert the templates to TPO-specific and trial-specific documents. The standardized templates would be developed by functional area subject matter experts (SME). Once approved, a template owner would be assigned to each template in order to ensure the templates are periodically reviewed and the content maintained as regulations, requirements, and business processes change. To help ensure proper architecture utilization and maintenance, a content maintenance and change-control process must be implemented as the architecture is developed. These processes are crucial to ensure the architecture content does not gradually drift back into silos, ultimately leading to duplicative, conflicting, and ambiguous content. In addition, a TPO conversion checklist tool or process flow mapping all of the conversion tasks is highly recommended to help ensure consistent and compliant implementation. Once TPO-specific and trial-specific documents have been created and finalized, the TPO relationship manager (enterprise level) and study team (operational level) would be responsible for document content maintenance and periodic reviews. It is highly recommend that all periodic reviews and document content decisions are recorded and filed in the trial master file showing a complete decision trail and oversight, in case of future audits or inspections.

	Centralized storage, maintenance, and accessibility are extremely important to ensure optimal architecture utilization and compliance across the sponsor and TPO staff. If staff is not able to easily access the key documents, then compliance will be negatively impacted, as well as inspection readiness. The end user needs to know where to quickly obtain the most current version of the architecture documents and who to contact for assistance. Proper archival of previous document versions is also extremely important in order to have the ability to show the full document and oversight history. A centralized website with document management functionality, including proper metadata which meets 21CFR Part11 requirements, is highly recommended. Proper account management and thus access to the site and site folders will also be important, so that sponsor staff may access all applicable documents while TPO staff may only access their TPO-specific documents for proper confidentiality. Site development and validation will be an important step to ensure all access and functionality is properly functioning.

	During these changing times as the biopharmaceutical industry is facing increasing healthcare reform challenges, sponsors are quickly focusing on the development of a more efficient clinical research model, where sourcing is an essential and strategic component. Utilization of a strategic, collaborative approach, including management of a pre-qualified TPO pool all based upon the true portfolio needs, is essential. Sponsors must conduct ongoing portfolio and performance indicator reviews to adjust the strategy and TPO pool accordingly, while also enforcing the utilization of the strategy across the study teams and functions. Development and maintenance of proper control and oversight processes clearly defined in a logical document architecture will help to ensure ongoing utilization of the overall strategy.       

, RPh, PMP, LSS Belt, CA-AM, is an External Sourcing Relationship Manager with the Development Center of Excellence, Eli Lilly and Co., email: 

, PhD, is a Senior Director of External Sourcing with the Development Center of Excellence, Eli Lilly and Co., email: 

, PMP, LSS Belt, CA-AM, is a Director of External Sourcing with the Development Center of Excellence, Eli Lilly and Co., email: 

	1. K. Getz, “Peril and Promise of Risk Imbalance,” 

	Gotts, I. (2009), Leading indicators vs. lagging indicators, available at: 

http://www.bptrends.com/publicationfiles/TWO%2003-09-ART-Leading%20vs%20Lagging-Gotts-final.doc.pdf

	1. The State of Clinical Outsourcing
	By Denise Calaprice Whitty
	

	2. Outsourcing’s Evolution from Commodity to Partner
	By Dave Scott
	

	3. An Improved Trial Model
	By Lee S. Scheible, Michael A. Pozsgai
	

	4. The Clinical Network: A Different Approach
	By Lee S. Scheible
	Community Oncology
	(November 2006)

	5. The Clinical Investigative Site Network
	By Lee S. Scheible, Donald R. Russell,
	and Karen J. Brinkman
	

As the biopharmaceutical industry continues to move down an outsourcing path, it is essential that logical sourcing strategies and operational processes exists and align with corporate strategy.

Essentials for Developing and Optimizing a Logical Sourcing Strategy

For more information on this topic, please read:

Mission Possible: Rebranding Clinical Research by Kenneth A. Getz, Emma Sergeant and James Kremidas

The Clinical Investigative Site Network by Lee S. Scheible, Donald R. Russell,Karen J. Brinkman

Spotting the New Managed Site Networks by Kenneth A. Getz

As the marketplace continues to rapidly change, demanding better access to more cutting edge therapies at lower costs, the pharmaceutical industry is being forced to find more efficient and cost-effective approaches in conducting clinical trials while facing increasing external factors. Rising drug development costs, a limited amount of research participants, increased competition for research participants among pharmaceutical sponsors, and an increasing turnover of new investigators complicate the drug discovery process.

  Additional requirements from regulators (including more diverse subject populations; longer, larger, and more complex clinical trials; exhaustive safety studies; and increased postmarketing safety surveillance) add to the overall study length and increase development costs. 

The use of enrollment optimization processes, such as enrollment forecast modeling and trial simulations and the use of clinical investigative networks versus the typical decentralized single site model, should be considered to help improve the current trial model. The addition of well-designed e-clinical processes such as EDC, eSource documents, Web casting, and others should also be considered to realize optimal trial quality and efficiency. 

The study protocol is the foundation of the clinical trial recruitment rate. New study protocol designs, such as those for adaptive trials (which is an increasing trend among pharmaceutical sponsors), can have an effect on subject recruitment. 

Nearly one third of all clinical trial timelines risk extensions based upon under recruitment or under enrollment. Under enrollment can jeopardize the registration of a drug, while over enrollment risks trial budget overruns. 

Protocols that have not undergone a feasibility analysis can be problematic once implemented. Sections of the protocol may be utilized from similar historic studies. This can result in poor recruitment if key changes are not made in relation to the new study outcomes in key areas of the protocol, such as inclusion and exclusion criteria. Other factors affecting recruitment may be study implementation issues at the site or sponsor-related issues. Additionally, poor patient compliance may result due to ambitious study requirements. What solutions are available to this perplexing dilemma? 

Clinical trial simulation and modeling tools are emerging in early phase work as a means to identify appropriate subjects and study targets, streamlining a compound's overall clinical study plan by focusing resources on clinical studies having a higher probability of success. Simulations can range from very large-scale mechanistic models that require significant resources to utilize and maintain, to a reusable platform that allows for complex and repeatable in-silica (computer-based) studies. The use of these tools can lead to larger success rates in later stage trials. 

The availability of historic trial data is another invaluable tool to develop a data-based projection. It is ideal to have the study projection developed prior to study timeline development. Recruitment activities and acceleration techniques are more effective and less costly when implemented during the initiation of the study rather than as a rescue effort well into the study recruitment period. Key historic data that is essential in forecasting a study projection includes the percentage of sites recruiting zero subjects. Historic metrics, such as the time necessary for sites to become regulatory ready, can occur over weeks or months and need to be considered in study recruitment forecasts. Finally, the rate with which research volunteers randomize into the study can vary depending on the study's inclusion and exclusion criteria, regions of the world involved, therapeutic area, and indication. 

A common mistake is to develop a straight-line projection that incorrectly assumes all sites are able to begin recruitment on day one of the clinical trial. Another important consideration in managing the clinical trial recruitment period is to consider timeline knowns including study seasonality, such as the period between Thanksgiving and the new year. In Europe, similar seasonal considerations would include the summer holiday season of July and August. Finally, disease incidence and prevalence, certain treatment settings (i.e., maintenance or adjuvant regimens), and pathophysiology effect recruitment and study duration. 

Figure 1 is an example of a projection taking into consideration these important factors. The team projection is a simple straight-line projection used in early forecasting of the study. 

Registration studies are increasingly becoming globalized for both increased access to subject populations as well as for economic considerations. Global regions affect the execution of clinical trial recruitment. The Asia-Pacific, Europe, Latin America, and the United States have their own recruitment enrollment performances. Key differences may include disease incidence and prevalence, specific country regulatory requirements, unique Ethical Review Board review periods, and specific investigator performance. There are recent trends of executing clinical trials in developing and emerging markets such as China, India, Latin America, and Russia. 

Historic site study execution is another invaluable tool in the acceleration of studies. In the field, there is often anecdotal and reputation-based evaluation of site performance. However, a more empirical method of measuring site performance at the therapeutic, compound, and study level is the use of invaluable tools to select high-performing sites as a means of accelerating not only the subject recruitment but complete study execution. Additionally, historic site study quality is a key factor in site selection. 

The methodology used in Figure 2 looks at the total number of subjects recruited at the clinical sites and the mean time between subject randomization at the sites. Using a clustering model, sites are compared to each other in the specific study. The ideal site would recruit a large number of subjects with low mean time between subjects being randomized. 

High performing sites are illustrated as Cluster 4 performers in Figure 2. Sites having a lower number of total subjects recruited and a longer mean time between patients fall into Cluster 3. Finally, Cluster 2 sites recruited the smaller number of total subjects with the longest mean time between subjects. Sites recruiting one or zero subjects are not represented in the figure. Historical data indicates sites will perform in the same cluster grouping 70% of the time. Selecting historically high performing sites along with new sites can provide a level of assurance in recruitment reliability and study execution. 

Another means to accelerate clinical trials is to increase the number of subjects that would consider participating in a clinical trial. While health fairs and disease awareness campaigns can provide a temporary boost to enrollment in targeted areas around clinical investigator sites, the issue remains: The number of people willing to volunteer for a trial is very low. 

The Center for Information and Study on Clinical Research Participation (CISCRP) is addressing these issues. CISCRP's recent "Everyday Hero" campaign seeks to address the low rate at which people volunteer for clinical research, as well as to educate the public and enhance its perception of clinical trial participation.

  By improving the public's perception of clinical trial volunteerism, there is significant potential for increasing the rate at which clinical trials enroll, thereby enhancing public health. 

The clinical network model has been around for more than 15 years; thus, the use of clinical networks is not a novel approach. But in recent years, the clinical network model has shifted from a primarily decentralized structure often lacking streamlined, consistent processes across the network to a more centralized structure governed by core standard operating procedures, enabling the network sites to function in a more consistent and efficient manner. 

The old network issue of slower, more complicated processes adding more costs has become an issue of the past. Network processes that were often labor intensive, paper driven, and not properly enforced with network quality assurance staff have all been resolved, as networks have implemented internal quality assurance departments to ensure network core procedures are followed, and implementation of e-clinical processes offer a better approach. 

Networks are groups of independent clinical sites managed operationally by a central administrative core group. The central administrative group will often include various departments, such as project management, regulatory, safety, pharmacy, information technology, data management, finance, legal, IRB, and quality assurance. Each of the central departments, as well as the network sites, are governed by a set of standard network operating procedures maintained and enforced by the central quality assurance group. These procedures help to ensure that the network sites function in a consistent, compliant, and efficient manner. 

From a scientific or medical perspective, networks usually consist of central disease state committees made up of regional research leaders or disease state experts representing their areas or regions of the network. The disease state committees are led or facilitated by a research advisor or chief medical officer. The central disease state committee conducts scientific reviews to ensure that the best trial designs are run at the right time and with the right regions or sites within the network for optimal trial efficiencies. The disease state committee is assisted by the central network administrative group, which provides historical subject population and network enrollment data so that the best decisions are made for optimal trial placement within the network. 

Networks will sometimes differ in some characteristics, such as size, disease state focus (single vs. multiple), community based vs. academic vs. hybrid, and most importantly in their level of centralization (partial vs. complete). Regardless of the characteristics, it is extremely important that the sponsor accurately characterize the network and thus fully understand their organizational structure, capacity, strengths, limitations, level of centralization, and clinical services offered. Several tools and standardized approaches have been successfully used and previously published to assist with properly characterizing clinical networks in a timely manner. 

Networks have always offered various centralized Site Management Organizational (SMO) services, such as site selection, trial placement, trial feasibility surveys, enrollment forecasting, budget and contract development, site reimbursement, and project management along with various clinical site services such as subject recruitment, ICD administration, data collection, safety reporting, and pharmacy and clinical laboratory services. Recently, however, networks have begun to offer various CRO services, such as full data management services and reporting, medical writing, statistical support, regulatory document services, and trial initiation and startup along with quality assurance site assessments and audit services both within and outside of their network. 

This expansion of clinical services allows for true, one-stop shopping and thus a more efficient and consistent clinical model. But in order to ensure proper compliance, all networks and services should be assessed prior to any service outsourcing for proper procedures, training, system validations, capacity, privacy compliance, and overall regulatory compliance dependent upon what services are offered. 

There are several key elements to look for when analyzing a clinical network. Some of the most important include central standard operating procedures to ensure consistent and efficient network processes that help guarantee proper trial oversight, subject safety, and quality. Central quality assurance is important in order to maintain and enforce the procedures throughout the network. Additional key elements include central data management services that allow for more consistent and efficient data collection and centralized monitoring. Central regulatory document services allow for more consistent and efficient document collection for multiple sites to initiate under one regulatory process vs. multiple site processes, which can be very time consuming and delay trial initiations. 

Several additional key network elements to look for when analyzing and selecting networks include central pharmacy services, which help to minimize study drug shipments to multiple single sites and study drug waste, since many single sites may never enroll a subject. 

The network model offers several efficiencies that benefit not only the sponsor but also the investigator and the subject. The centralized administrative departments take care of most operational tasks, including contract and budget negotiations, trial initiations, regulatory document collection, IRB submissions, central pharmacy, and site payments, and also allow the investigators and their site staff to focus on conducting the trial, enrolling and treating their subjects (which helps with enrollment), data collection, and overall quality. The subjects benefit since the investigators and site staff spend less time on administrative tasks and more on treating and assisting them. 

Smaller, more remote clinical sites, which may lack the proper staff to perform all of the administrative tasks, can be considered for clinical trial participation as part of the larger clinical network. This increased opportunity allows for more cutting edge therapies to be offered to more investigators and their subjects while minimizing lengthy trips and costly travel expenses for subjects. In addition, the centralized or partially centralized network model allows for large efficiency and consistency gains since one regulatory document and IRB submission process is conducted to initiate multiple network sites. 

Central pharmacy services allow for one central drug shipment and less study drug waste. Central contracts and budgets allow for one budget and contract negotiation and development for multiple sites. Additional efficiency gains have been documented with central data management, allowing for centralized monitoring and more consistent data collection, as well as others which have been discussed in recent clinical network publications.

E-clinical processes are important because they maximize trial process efficiencies while also ensuring enrollment optimization processes, and network model efficiencies are optimized, as documented in recent e-clinical publications.

  These processes include virtual training (i.e., Web cast, Web sites, CD, DVD, Web conferences), which offer different options for training the clinical network sites and staff in a thorough, cost-effective, and timely manner. 

Document repositories (secure Web sites) are used for secure document delivery and enhanced communication among the network disease state and central administrative committees and at the satellite sites depending on the document type. The enhanced communication can help ensure that proper trial oversight and patient safety occur in the network trials. Secure email systems are helpful for rapid communication among the network and sponsor staff, while e-source documents and EDC systems can have a positive  impact on trials by allowing for centralized monitoring and more rapid data access, review, and query resolution. 

All of these e-clinical systems and processes also help to ensure that all of the network sites function in a consistent manner and that their records are maintained in a similar fashion. A key element is having a centralized network IT staff that will help to ensure that end-user functionality requirements are met, that system support services are offered, and that the systems are always regulatory compliant. Technical and help desk support and contingency plan development for prompt system and user issue resolution, often referred to as ancillary services, are also crucial for system and process acceptance and success. 

Steady enrollment and optimal trial metrics can become reality with the right processes and tools.