Clinical Outsourcing in Europe

The outsourcing of clinical functions typically refers to the hiring of an outside entity to conduct some or all of the technical tasks associated with conducting a clinical trial. Outsourcing technical tasks such as writing, statistical analysis, study monitoring, database design, etc. to contract research organizations (CROs) was an increasing trend in the 1990s. As part of that trend, U.S. firms have been outsourcing site tasks to Europe for some time. Drawn by cheaper costs and faster start-up times, they almost always outsource investigative sites. That’s because very few sponsors employ licensed onsite caregivers.

In 1996, the Clinical Device Forum initiated a survey to better understand the clinical research practices of the medical device industry. The survey has been conducted annually since then, with questions focused on the management, regulatory, quality and financial aspects of device clinical studies; data had been collected from device firms from around the world.

This report includes data from 204 U.S. and European medical device firms from 1996-2003. Of the firms reporting, 156 are from the U.S., 48 are from Europe and 17 are from other countries.

One of the goals of the survey is to provide benchmarking information for outsourcing. The quality of data delivered by a site is a function of the quality of data demanded of it. Because European sponsors set the tone for their home countries, a look at their practices can help U.S. companies better understand what to expect when conducting clinical trials there.

Generally, employees who work for the sponsor and contribute to clinical research are called Clinical Research Associates (CRAs). They may find investigators, write protocols, monitor studies, analyze data, prepare reports or manage projects. The more CRAs a firm hires, the more sophisticated its in-house clinical research program becomes. A lower number of CRAs may mean a smaller program or a greater reliance on outsourcing.

We asked sponsors how many people worked on their clinical research projects. European firms responded that they have steadily reduced the number of employees they hire internally for this work, declining from 6.5 full-time equivalents in 1998 to 4.5 in 2003. U.S. firms, on the other hand, responded that they have increased the number steadily from 5.5 to 7.5 full-time equivalents in the same period.

Protocols are detailed plans for how the study will be conducted, covering everything from patient recruitment, device use, intervention procedures, data collection and personnel responsibilities. The quality, detail, completeness and organization of the protocol set the tone for the entire clinical study. As seen in Fig. 1, U.S. firms have consistently reported that they take more time to write new protocols than their EU counterparts. American firms averaged 25 days while European firms averaged 15 days. The decreased work/quality demand by European firms is reflected in either the lower number of CRAs employed or possibly a shift to more outsourced CROs.

The number of studies per year is a good indicator of the amount of work sponsors must undertake, either by CRAs or by outsourced CROs. In the late 1990s, U.S. and European firms averaged about 15 studies per year, with a relatively large number of subjects and investigative sites. In the current decade companies have settled at around 8-10 studies each year, with smaller sample sizes and fewer sites. U.S. and European firms averaged about the same number of studies per year. (See Fig. 2.)

The duration of a study influences its cost and complexity and is a function of subject recruitment ease and length of follow-up. The longer the studies, the more employee CRAs or outsourced CROs are required. The average duration of studies for both U.S. and European firms has trended shorter over the past six years, from 20 months in 1998 to 15 months in 2003.

If studies have grown shorter, they have also have become smaller (Fig. 3). Average European sample sizes have dropped from an average of 350 subjects per study in 1998 to around 170 subjects today. While the average U.S. study size has remained stable at around 200 subjects, European studies have declined to fewer than 50 subjects per study. So while European firms have maintained the same number of studies each year as their American counterparts, the studies they initiate are much smaller.

Another factor affecting the cost and complexity of a clinical study is the number of investigative sites (Fig. 4). In the late 1990s clinical studies were designed with an average of 15 sites per study. Large numbers of investigative sites are very difficult to manage, no matter how much human resource is thrown at them. It’s difficult to make sure all investigators do the same things and are informed of what the others are doing and to keep track of who has been paid for what. As the industry has matured, the number of sites per study has decreased. In the current decade studies have averaged around five sites each. European firms tend to use 2-3 fewer sites per study than U.S. companies.

Sponsors are required to monitor investigative sites to verify the study plan is followed. A poorly monitored site often results in low-quality data. The time between monitoring visits (monitoring frequency) is one measure of the quality, but some sites dislike being monitored because it disrupts the workday and takes time away from patients. We wanted to know if European sites would welcome the same frequency of monitoring as U.S. sites. We found that the time between monitoring visits were about the same for firms on both sides of the Atlantic, trending upward over the past six years. Today most sites are monitored every 2-3 months.

Clearly, fewer, shorter and smaller studies (fewer subjects, fewer sites) indicate less clinical research work is being performed by both U.S. and European firms. Domestically there are more employee CRAs to perform the work, while in Europe the opposite is true. So, either the quality of European studies has fallen because the workload per person has increased or the work is being outsourced.

For benchmarking, the cost of a study is often measured by the costs per subject paid to investigative sites. These costs include payment for medical procedures, purchasing investigative and control devices, administrative and staff costs, IRB fees, overhead charges and other study-related payments made to the site. Total costs divided by the number of subjects enrolled at the site yields the per-subject cost.

In 2002 we began collecting data on per-subject costs. As shown in Fig. 5, the amount reportedly paid by U.S. sponsors in 2003 was four times the average paid by European sponsors, which grew dramatically from the year before. This is a major factor in motivating American firms to outsource their clinical sites to Europe. While the quality of data does not always meet American standards, it is far less expensive and may ultimately reduce the amount of U.S. data required by the FDA.

Where are these studies outsourced to? We asked respondents to indicate the countries in which they conducted clinical research last year. The seven most popular areas in which American firms conducted clinical studies are listed in order: the U.S., UK, Germany, Canada, France, Italy and The Netherlands. When we plotted the popularity of these same countries in 2003 we noticed interesting shifts in host country popularity. A lower percentage of studies were conducted in foreign nations, with the popularity of home-grown research increasing.

The most dramatic decrease in popularity occurred in the UK. This is likely due to increased regulatory activity. The popularity of Germany, Canada, France and Italy also decreased, probably reflecting a decline in the number, length and size of studies.

We’ve looked at several factors that affect the outsourcing of technical tasks and concluded that the number and scope of studies have fallen. We also know that the number of employee CRAs have increased in the U.S. while falling in Europe. But we were left with the question of whether the European workload has been shifted to outsourced CROs.

First, in the U.S. the number of firms reporting that they did not use a CRO last year increased to 42% from 27% in 1999. It was a bad year for device CROs, and many of them have privately reported decreased income. It was a good year, however, if you wanted to work directly for a sponsor. U.S. firms are shifting from outsourcing technical tasks to hiring their own employees, indicating that they believe conducting a study in-house will result in better quality control of the data.

Europe, though, is another story. The number of firms reporting that they did not use a CRO last year rose to 75% in 2003 from 67% in 1999. European device firms have never embraced the outsourcing of clinical research technical tasks, but because they have also cut the number of employee CRAs, these firms are apparently spending less human resources on clinical research. Other reports indicate their lack of effort is resulting in a decrease of data quality.

Finally, we asked the straightforward question, “How much did you spend on CROs last year?” U.S. firms in 2003 reported spending an average of $100,000 per company while European firms reported an average of $20,000 per company in 2003. Even though the small European sample may skew the data, Europe clearly is not the best market for CROs. In fact, most European CROs survive by performing overseas studies for U.S. clients.

While our survey shows some movement away from outsourcing, it’s important to note that there are two parts to the outsourcing of medical device clinical research: the investigational sites, which are almost always outsourced, and technical tasks. Several overriding issues influence clinical outsourcing, including the economy, research and development activity, demands of regulatory agencies and others. These issues influence specific requirements such as the number of studies, duration of study, number of subjects and number of sites. These requirements also determine whether in-house resources can handle the workload. In the U.S., the trend appears to be moving away from outsourcing technical tasks domestically, but if the investigative site is in Europe, there is a good possibility that a European CRO is involved. In Europe, the outsourcing of technical tasks continues to lag that of the U.S., but European companies will use a CRO if the study site is in the U.S.