Lean Outsourcing Models for Clinical Trials

A number of recent trends within the industry, including reduced headcounts and an increased focus on “nimble” product development strategies, have prompted many companies to pursue “lean” models for the outsourcing of clinical development. Potential advantages of such models include resource and geographic adaptability, reduced fixed resource consumption, and use of “best in class” resources for each task, whereas perceived challenges include a lack of “tight control” over outsourcing partners, a tendency for sponsor staff to revert to traditional behaviors despite resource reductions, and difficulty finding personnel experienced in such models.

On behalf of Purdue Pharma, The Avoca Group gathered experiences from companies that have evaluated and employed lean models, for the benefit of those that are exploring them. In-depth interviews were conducted with management representatives from 14 sponsor companies and three CROs that are currently conducting clinical development using “lean” models. For the purpose of this project, “lean outsourcing models” included any designed by sponsor companies to reduce or minimize the level of resources used for CRO oversight compared to traditional, “standard” models (Figure 1). During interviews, respondents were asked to describe their models in detail, including the responsibilities held and value added by each party; the decision criteria that led to choice of the models; and their histories with the models, including both facets that have worked well, and challenges. Specific tools, technologies, and best practices used for oversight, and specific traits of sponsor personnel that work most effectively under the model, were also discussed.

Rationale for Lean Models

For some of the smaller sponsor companies interviewed, development of a full-capability internal clinical research staff was never part of their corporate strategy and would not have been practical; thus they had always employed “lean” outsourcing models. Other companies had originally performed clinical trials using in-house staff and/or more “traditional” outsourcing models, but had moved at least some portion of their trials to leaner models because of headcount reductions and/or other budget constraints, a strategic mandate to refocus internal staff into areas of higher value add, movement toward increased (or increasingly variable) volumes of clinical development activity, and/or dissatisfaction with the quality or efficiency obtained using traditional outsourcing models. Positive results were reported to include not only greater efficiency in use of both sponsor and CRO resources (due in part to greater operational control by CROs and hence greater freedom to innovate and willingness to contain pricing), but also improved/faster problem-solving with fewer escalations (by getting some cooks out of the kitchen); improved chemistry across sponsor-CRO teams; and improved ability to utilize lessons learned due to increased commitment to the relationship.

Success Factors for Lean Models

Through the interviews, it became clear that successful lean outsourcers had made carefully considered decisions in each of seven key areas, as follows.

#1: Core competencies to retain in-house

Sponsor executives reported that their companies had made carefully-considered decisions about the functional competencies that would be retained in-house. For regulatory reasons, all maintained at least project management/oversight in-house; the nature of any other competencies retained in-house was driven by corporate strategy regarding areas of internal expertise thought to provide competitive advantage and/or by historic factors, i.e., past difficulty outsourcing successfully in certain operational areas, and/or strong desire to retain current functions for which highly skilled and experienced staff were already in place.

Some examples of competencies to be retained in-house included: 

• Senior-level project management. All individuals interviewed expressed the importance of employing very experienced individuals, preferably functionally cross-trained and competent in advanced aspects of project management (risk assessment, proactive issue recognition, use of metrics, etc.), to perform project management tasks under a lean model.

• Protocol concept. Often sponsor companies felt themselves to be in a superior position of therapeutic area expertise compared to their providers, as well as being more intimately familiar with their business strategy for a product candidate. Thus most frequently, protocol concept documents were developed by sponsors in-house, even under lean models. However, there were exceptions where sponsors were branching into new therapeutic areas, or employing a business strategy that involved regular in-licensing of product candidates without any particular therapeutic area pattern. It should be noted, however, that it was quite popular to have CROs build the full protocols out from the sponsor-generated outlines. CROs were thought to have the operational expertise to do this well, as well as to feel greater ownership in execution when they had a significant role in protocol development.

• Internal competency in Quality Assurance was felt by some to be second in importance only to project management. Given the reduced level of hands-on oversight under a lean model, a well-conceived and efficiently executed strategy for quality oversight becomes increasingly important. Sponsors that did not possess this competency internally often outsourced it to a specialty vendor, independent of their CROs.

• Study start up/site interfaces were occasionally maintained in-house for relationship-building reasons, and to expedite the study start-up process.

• Biostatistics activities were sometimes maintained in-house, in some cases because analysis and interpretation of clinical trial data was thought to be a “core” strategic activity, and in other cases because of historical difficulty with outsourcing such activities satisfactorily. In contrast, routine data management and safety case management activities were nearly always outsourced fully.

#2: Functional vs. full-service

Some sponsor companies were satisfied with their CROs and found the full-service model to be simple (seamless, coordinated) and relatively resource efficient (internally), whereas others preferred to outsource functionally to what they believe to be best in class. Companies reported success with both strategies; considerations included the volume of work that could be committed to a provider (i.e., desire not to dilute volume if it’s already low), and process-related considerations (i.e., if highly specialized expertise, training, requirements, and/or processes are needed across programs, it can be easiest to outsource functions to the same provider across programs).

The sponsors interviewed did not typically move into full-service lean outsourcing without having had considerable prior positive experiences with a given CRO. Those that did not have this opportunity sometimes built a “qualification” stage for lean sourcing into the initial work with the provider. This could involve using lean oversight models for only a subset of trials. This stage would assess performance, or performance-based qualification criteria within a trial. If the CRO met that criteria, it would trigger a reduction in the level of direct oversight.

Sponsors that preferred to outsource functionally recognized that this approach was more resource-intensive for internal project management staff, but felt that the ability to engage the best available provider for each functional area was worth the additional effort, particularly given that the highly-qualified internal project management staff had the level of experience and competence to handle this additional burden.

#3: Number and types/sizes of CROs to engage

Sponsors took a variety of approaches to determine the numbers and types/sizes of CROs to work with under their lean models. Some, particularly those with narrow therapeutic foci and those transitioning from intensive oversight models, chose to work with only one. The rationale being that that lean outsourcing models require a level of trust and understanding of sponsor expectations that are difficult to achieve with more than one CRO at a time, and that lessons learned from a pilot could be applied to others in the future. An additional rationale, especially for smaller companies, was the need to concentrate business as much as possible with one CRO in order to command the volume discounts, dedicated resources, A-team members, process integration investment, and team stability required for success under a lean model.

Other sponsors, particularly larger ones, chose to utilize multiple CROs out of reluctance to place “too many eggs in one basket,” particularly given the reduction in intensity of oversight. Anticipated development pipelines with therapeutic area or geographic diversity too great to be appropriately serviced by one provider, and a desire for healthy competition among providers from both a pricing and a performance perspective, also factored into the decision.

#4: Use of independent contractors and/or specialty shops

The effective use of independent contractors was a nearly ubiquitous theme among successful lean outsourcers. As a means of augmenting sponsor capabilities when necessary, independent contractors tend to be stable and reliable resources, have a relatively high average experience level per dollar, and be role-flexible (i.e., contract CRAs can also be used, in some cases, for auditing tasks). These traits can help to offset turnover, and the use of mixed teams can be a very effective means of raising the bar for all staff, especially when accompanied by a program of performance management that looks at individual-level performance.

#5: Types of trials to outsource under the lean model

Although some sponsors outsourced exclusively using lean models, others used lean models only for a subset of their clinical trials. Decisions about which trials to outsource using lean models were generally made based on considerations surrounding clinical trial phase, therapeutic area, and region, as well as internal resource availability. Some companies used a cost-based approach, finding it more cost effective to manage different types of trials under different models. Others used a risk-based approach, for example using lean models for regionally limited vs. global trials; less complex protocols, indications, or populations; lower priority portfolio assets or phases; trials requiring great flexibility in resource requirements (e.g., adaptive); and/or only when they had considerable positive past experience with provider(s).

#6: Contractual provisions

Sponsor companies generally reported positive experiences with carefully conceived risk-sharing models, and those that had not incorporated such contractual provisions generally expressed regret. Risk-sharing contractual provisions can work best with lean oversight models since these reduce the probability that CRO performance will depend upon sponsor involvement at a large number of touch points.

Options reported to work well included both fixed-price models, and risk-sharing models whereby CROs had access to bonuses in exchange for meeting performance targets and/or suffered penalties for not doing so. In this context, sponsors also reported positive experiences with contractually binding Quality Agreements, in order to ensure that quality requirements were clear and not compromised in attempt to meet time and budget targets. The utility of risk-sharing approaches was said to be maximized with early engagement of, and substantial decision-making power by, the CROs. For example, heavy involvement of CROs in the composition of protocols and timelines helps to ensure full commitment to milestone goals.

#7: Supporting tools and best practices

The effective use of tools and best practices was found by most of the participants in this research to be critical when operating under a lean model. Such tools/practices included:

• A playbook or manual that describes, in detail, the roles of CRO and sponsor team members (in particular placing restrictions on activities of sponsor functional area staff); standards and/or expectations for specific functional tasks, deliverables, and staff qualifications; communication and escalation plans; risk assessment and management activities; etc. Such documents are often long and detailed, and development from scratch can require significant levels of investment on the parts of both sponsor and CRO.

• A concise set of key performance, risk, and quality metrics, together with associated targets, thresholds for action, reports, and review plans. In addition to their obvious role in performance, risk, and quality management, negotiations surrounding metrics, thresholds, targets, and review plans are also often powerful aids in clarifying performance and role expectations, and in focusing oversight at the correct level.

• A fit-for-purpose governance model with well-articulated mandates, primarily including performance and relationship management, procedures for aligning on higher level issues with significant financial implications, and handling of escalated issues, along with a commitment to meeting at logical frequency.

• Ample face-to-face meeting time, to promote and maintain trust, openness, honestly, and transparency.

Although technology solutions that permit status transparency were mentioned by several of those interviewed as of potential benefit, the use of dedicated clinical data integration and dashboarding technologies for clinical operations oversight was not widespread. Reasons for not investing in such solutions included the financial investment, lack of resources or reticence to invest required resources to support system implementation, a feeling that available technologies offer much more than is required, and, for smaller companies, a low volume of active trials. Instead, the surveyed companies generally relied on a blend of reports from existing clinical systems owned by both the sponsor and the partner CRO(s), despite the common challenges of insufficient report frequency and insufficient accompanying information beyond the reported metrics.

Transition to lean models

Those companies that had transitioned from more traditional to leaner outsourcing models shared a number of lessons learned regarding the process. The importance of strong top-down support was found to be critical; senior management must not only articulate the model to others, but must hold staff accountable for living by it, for example by strictly upholding limitations on internal resource use. As part of demonstrating this support, the development of a joint (sponsor-CRO) transition plan was recommended, involving, for example, high visibility joint communications from leadership of both companies. If senior management support is not sufficiently strong and visible, divisiveness can remain, with not everyone on board with lean models, thus leading to a failure to gain the expected efficiency due to undermining of the model in certain areas.

In general, at least six months of considerable effort was said to be necessary to revise SOPs, develop joint metrics, develop charters and oversight and governance plans, compose operational manuals, refine early engagement and risk-sharing models, etc. Some companies found it useful for operating manuals and/or other documents to depict the transition in oversight over time, e.g., performance criteria for qualifying for lower level of oversight, as well as what each oversight level will look like.

Conclusions

In the shifting landscape of product development, the pharma and biotech arenas increasingly emphasize the need to use less internal R&D resources, increase efficiency, and reduce costs. One product development strategy is the utilization of lean outsourcing models for clinical studies. Our research shows that companies who successfully use lean models for clinical development have carefully evaluated several key areas in their decision-making, including what core competencies to retain in-house, the types of trials to be outsourced under the lean model, the number and type of CROs to engage, contractual provisions such as risk-sharing, and extensive use of supporting tools and best practices. These are among the key areas of critical importance in finding success with lean outsourcing models. 

Denise Calaprice, PhD, Senior Consultant, The Avoca Group

Mitchell Katz, PhD, Head of Clinical Research & Drug Safety Operations, Purdue Pharma