The Rise of FSP Outsourcing in Drug Development

The changing realities of clinical research today include more technologically sophisticated agents such as gene and CAR T-cell therapies and more complex trials that use master protocols and adaptive designs. Sponsors also face soaring R&D costs and the continuing challenge of recruiting and retaining patients for clinical trials. Simultaneously, advances in mobile technology, data aggregation and analysis tools, and risk management methodologies are changing how clinical trial data are collected and monitored, influencing how companies outsource core clinical research tasks.

In this increasingly complex environment, biopharmaceutical companies need access to more flexible approaches and resource models, standardized systems, and remote capabilities to deliver on their outsourcing requirements. This is especially important during the COVID-19 pandemic, which has exacerbated patient recruitment and retention challenges.

The established full-service model, in which biopharmaceutical companies retain a preferred contract research organization (CRO) to perform the full range of clinical trial activities, has been increasingly complemented by an functional service provider (FSP) approach, which allows companies to engage discrete expert services—such as medical writing, biostatistics, or clinical monitoring—when and where they need them. Research shows that large biopharma companies’ use of FSP is growing at more than 13% annually.

Adrian Otte, principal, Clinical Center of Excellence, at YourEncore, Inc., and former leader of global development operations at Amgen and Pfizer; and Chris Baker, senior vice president and global head of FSP, Parexel, sat down recently to answer questions about the FSP outsourcing model and their thinking on future trends.

What is it about the current environment that’s encouraging the growth of the FSP model?

Adrian Otte

Adrian Otte: There are several factors. The first is cost. FSP can lower developer costs by reducing redundant activities. That can happen when you have project management both at the pharma company and at the CRO. These are expensive resources. Oversight is also simpler and cheaper if the FSP partner uses the pharma company’s systems and processes.

The second driver of FSP uptake is engagement. Historically, the FSP model has tended to last a long time, and I think that’s because there’s a strong relationship between the two organizations and a clear definition of each organization’s responsibilities. The FSP model is an ongoing partnership with a shared commitment to overcome challenges—such as slow recruitment or flaws in the protocol—collaboratively, leading to less finger-pointing.

Finally, companies want greater flexibility. If you have a group of, say, 150 clinical research associates (CRAs) in a country, and some of your studies are going faster, and some slower, you can move the CRAs across studies, depending on where you need them the most. Flexible resourcing boosts efficiency and accelerates timelines.

Chris Baker: I agree completely. Biopharmaceutical companies are under intense pressure from their shareholders to improve productivity and effiency while managing costs. It’s vital to be able to flex up or down in a given geography or within a therapeutic area program based on corporate priorities.

FSP also expands the talent pool in a targeted fashion, providing companies with access to the skills and expertise they may not have in-house. It brings in relevant experts where and when you need them. Lastly, it cultivates continuity because FSP relationships tend to be long-term. Some of our partnerships with pharma companies have endured for more than a decade, allowing for high-quality delivery, enhanced trust, and well-established pathways of communication.

What has been the COVID-19 pandemic’s impact on FSP use?

Chris Baker

Baker: We’ve seen increased utilization of FSP in response to the crisis by three types of companies. First, there are companies that want to start trials of new COVID-19 diagnostics, vaccines, or treatments, or to repurpose existing drugs for COVID-19, but they don’t want to stop or slow down programs in other therapeutic areas. They’re scaling up via FSPs because the capability and capacity provided, in turn, enables them to work on COVID-19 candidates without impacting other ongoing studies.

Second, companies that need to bolster a specific type of expertise or fill a geographic gap are using FSP to do so in a highly targeted fashion. And, third, companies that have accelerated their use and acceptance of remote monitoring, virtual informed consent, and electronic medical records (EMRs) have turned to the FSP model. Before the pandemic, acceptance of these innovations was slower in some areas, but now the adoption rate has accelerated.

Otte: I think the inherent flexibility of the FSP model has successfully stood up to the test of COVID-19. It has performed well in upsizing and downsizing resources.

That said, no outsourcing model can solve the problem of accurately predicting patient recruitment, or speeding it up, completely. Unpredictable patient enrollment has been greatly exacerbated by the pandemic. We see studies stalled unexpectedly due to travel restrictions and social distancing mandates. Throw in the uncertainty of how the pandemic will play out in the coming months and you have an environment that demands flexibly deployable resources.

The features that make the FSP model effective during this current public health crisis will also make it relevant in a post-COVID-19 world. Regulators and pharma have shown a willingness to adopt remote monitoring and the use of wearables. They’re allowing remote access to EMRs and implementing hybrid/decentralized studies in which patients visit sites infrequently or not at all. The question is whether these changes will stick. I think they will. I think we’ll see much faster adoption of digital tools and the FSP model in the future.

How have advances in data aggregation tools, artificial intelligence (AI), and machine learning affected FSP?

Otte: Historically we are a conservative industry, but the jolt from the coronavirus crisis has accelerated the adoption and use of new technologies. We’ve seen promising, positive outcomes with these tools, so companies will be excited about using them going forward. Actually, data aggregation, AI, and machine learning were steadily gaining traction pre-pandemic.

I’m not sure I see these tools as providing a particular advantage for the FSP model, except that it will probably be easier to apply them in a more standardized environment in which all studies are done with a single set of systems and processes. That’s a hallmark of the FSP model.

Baker: People have been talking about big data, data lakes, AI, and blockchain for years. Data sources are growing exponentially, but the key to benefiting from this innovation is the ability to connect disparate data sets longitudinally. For example, if one set of raw data from a study is based on wearables, another on EMR records, a third on handwritten notes from the principal investigator, and a fourth on MRIs or lab values, then the data must be integrated to provide meaningful and actionable insights. Within each study, the interconnectedness of data impacts how well we can analyze the outcomes and, critically, extrapolate the findings to understand the benefits and risks of a new treatment for other patients with the same condition.

What this means specifically for the FSP model is that CRAs need to be more technically savvy than in the past. They need experience and a high level of competence with electronic data capture, clinical trial management systems (CTMS), electronic trial management system (eTMS) software, and data-driven monitoring. And, of course, they need to be able to validate that the data are correct.

How have changes in clinical trials affected the CRA role in FSP?

Baker: As the complexity of clinical trials has increased, and the need to understand and use new technologies has exploded, the CRA role has become more important. The needed skill sets are more sophisticated than they were, say, 10 years ago. The bar has been raised globally for more complicated trials, especially in oncology and orphan drugs. Many trials now require a strong science, nursing, or physician’s assistant background because many advanced therapy medicinal products—such as CAR T-cell and gene therapies—require a deeper knowledge of the disease state, physiologic processes, and drug handling and administration.

We see more experienced Level 2 and 3 CRAs staying in the industry. Because the value of a CRA rises rapidly with experience and a solid track record—and they move up quickly—we are facing a potential shortage of Level 1 or entry-level CRAs. Therefore, Parexel enhanced its recruitment and training efforts in parts of the world where supply and demand for CRAs is not as tight. Additionally, intern programs help train and expand the pool of Level 1 CRAs.

Otte: Remote monitoring, risk-based monitoring (RBM), wearables, sensors, AI, new ways to get informed consent—these new platforms change everything for CRAs. The monitoring tasks they perform in the future will be completely different than what they’re doing today.

However, despite the development of RBM methodologies and centralized monitoring, we’re still highly reliant on the individual CRA to identify study conduct issues that could impact patient safety or compromise the scientific integrity of the study. Current trends in protocol complexity are creating a huge challenge for the CRA. There’s considerable work that still needs to be done to ensure that CRAs will be able to meet these new demands and be successful in these new jobs. In the FSP model, where the CRA typically works in a standardized environment of systems and processes focused on one sponsor’s portfolio, at least one set of uncertainties is taken out of the equation.

Training is key, but realistically we need to start evaluating the impact of that training by measuring the actual real-world competencies of CRAs beyond the occasional accompanied site visit.

How important is FSP staff retention to successful outcomes in clinical trials and in drug development?

Otte: Staff retention and continuity is a critical factor in collecting high-quality data from a clinical trial. If an investigative site is visited by three different data monitors in a year, then the data are handed off three times. This disrupts a trial on multiple levels and could compromise data quality. After each hand-off, new CRAs must be trained on relevant aspects of the trial and the patient journey—a steep learning curve—and trust and communication between sites and monitors have to be rebuilt. At best, all that hurts productivity.

Sites are more satisfied and willing to participate in trials when they work with CRAs who understand the sponsor’s product, portfolio, and way of working. When an urgent problem arises on site, staff need quick answers. It may seem trivial, but having a name and phone number and immediate access to a competent, knowledgeable response makes a big difference.

As I’ve noted, a CRA in a FSP model is usually fully allocated to one client and works exclusively with that client’s systems and processes. Maybe they’re working on multiple protocols, but they know that client. I am sure that helps CRAs feel more self-confident, and I believe that leads to reduced turnover.

Baker: One of the main reasons clients switch to a different FSP provider is their dissatisfaction with a current provider due to high turnover rates. Training FSP staff to understand a client’s culture, mission, and values and helping them integrate into the host organization is essential. If you find the right people, retention rates are high, and turnover decreases dramatically. Additionally, for safety or pharmacovigilance monitoring, we seek geographic hubs that have historically had very high retention rates, versus regions where the CRA market is hotly competitive and overly fluid.

What is the future of the full-service model in the face of FSP trends?

Otte: Full service is still a large part of the outsourcing business. It has worked in the past, and it can still work well. Where the full-service model falls short is when companies engage too many CRO vendors, each for a different study or program. Typically, full-service programs are put in place by the procurement department, and they end up contracting with two, three, or even more suppliers. This creates the appearance of competition but leads to confusion, inconsistencies, and difficulties for companies as they try to oversee the work of multiple vendors. One often ends up with a revolving door of new suppliers. This is not the fault of the individual suppliers, but rather the model and the way it is managed by both parties. If you choose the full-service model, it’s better to rely on one CRO with one set of systems and processes superimposed across your studies. Of course, there’s an assumed risk in depending upon a single supplier, but the quality and efficiency gains provided by standard systems and processes are generally worth it.

In my view, companies that combine full service, internal teams, and FSPs just haven’t settled on their strategic approach. Running all three models simultaneously may seem safer, but it’s inefficient and does not promote consistency or quality. Oversight is much harder and more complex when companies pursue multiple outsourcing models simultaneously.

Baker: There will always be value in the full-service model. For smaller companies, the full-service model is more feasible because they often don’t have the systems, processes and technology to oversee an FSP model.

For large pharmaceutical companies that use FSP, retaining too many niche FSP providers can be costly and inefficient. We see a lot of companies that pare down FSP providers from 10 to 15 globally to a more manageable three to four. That way they remain diversified, but increase efficiency.

As clinical trial programs grow increasingly complex, timelines shorten, and R&D costs soar, FSP will continue to prove its worth by providing cost savings, scalable expertise, and resourcing flexibility.