Optimising the Site Selection Process

According to some estimates, up to 30 per cent of Investigators participating in a trial may recruit no patients. When considering this statistic, it is important to remember that every site—also those that have not recruited to target—has been through a process of selection. Whilst this process may determine whether a site can provide suitable staff and facilities in accordance with regulatory guidelines, it is evident that the site selection process is not foolproof in assessing whether or not a site will in fact recruit patients as anticipated, or at all.

This article evaluates a number of pragmatic approaches to more effectively identify suitable sites by proactively weeding out the types of surprises that—with the 20 / 20 vision of hindsight—were not so very surprising. We move beyond the basics of the site selection visit and consider the process as both a strategic and tactical challenge focussing on three key factors:

• The impact of company processes in setting the tone for success or failure
• Identifying the less obvious but crucial characteristics of successful sites
• The importance of investigator motivation.

The golden site profile
Looking to the investigators, prior experience and personal relationships often form the basis for selection. Given the increasing volume of competing trials, experienced investigators are in great demand. However, they may not offer the certainty of delivery they once did. Data also suggests that the number of investigators active for only one year in clinical trials is increasing. Taken together, the necessity to work with new investigators becomes apparent. In such circumstances, it is important to extrapolate from previous experience to identify the indirect and less obvious characteristics associated with good sites—characteristics that may not be apparent at the individual site selection visit. For instance, an analysis of stroke studies conducted at over 400 sites globally for over a five-year period showed that the top decile of investigators (in terms of recruitment) displayed the following characteristics:

• Located in countries / cities with an older population
• Practicing in a large or general hospital
• Located in more densely populated cities / countries
• Specialising in neurology with an emphasis on geriatrics.

Whilst perhaps not surprising, these findings allow the conclusion that the successful recruitment of patients in such studies is linked to the ability to identify investigators who fulfill this specific set of criteria.

Assessing investigator motivation
There is, of course, only so much that can be predicted from an investigator survey process and analysis of past data. Assuming the investigators are located in the same country, have access to similar patient populations, similar facilities and staff, why do we continue to see a broad range in recruitment performance across sites that appear similar or at least equivalent? To help explain this, we turn to two of the less tangible aspects of successful sites—investigator interest and motivation. For most investigators clinical research is only a small part of their practice, compared to the dominant obligation of patient care. Moreover, the investigators continually face time constraints and have to prioritise their time. In this context the importance of investigator motivation is evident. When considering motivation we should first address why investigators make the decision to take part in clinical trials? Reasons vary and are specific to each individual investigator, but generally include potential benefits to patient, scientific or medical interest and financial benefit. In today’s world of health economics, trials may also be a conduit not only to the test drug, but also the comparator drug that may already be on the market but cost-prohibitive to many patients. Regardless of what drives their motivation, it is necessary to confirm that investigators participating in a trial possess this key ingredient if the current cycle of delayed studies is to be broken. With experience, it is possible for the site selection staff to get a general impression of how motivated an investigator is. To more accurately assess investigator motivation, consideration should be given to the following practical approaches:

Strategic partnering – Depending on the healthcare system, institutions may have a Research & Development office that has worked with investigators over a number of years. By meeting only with the individual investigator and maintaining a transactional relationship, instead of collaborating also with centralised R&D functions, we may continue to select investigators in a relatively blinded fashion. However, developing shared objectives and working collaboratively with such central functions can add great value and prevent inappropriate investigators from being approached.

Asking the right people – It’s no surprise that the research nurse or other members of the team running trials day-to-day will be able to give a more accurate view of the practical implementation of a trial than the investigators themselves. Similarly, an investigator may not be aware whether support functions, such as the radiography department, can provide the volume and frequency of scans required. Such logistical challenges often are the hidden reason for study delays and should always be assessed directly and proactively with the individuals responsible.

Asking the right questions – Site selection visits that start with a protocol presentation followed by an invitation for the investigator to ask questions in fact veil how much time the investigator spent reading the protocol in advance of the visit. Instead, starting the visit by simply asking the investigator their opinion of the protocol is a good indication of their level of interest and motivation.

In summary, a site selection visit that ensures facilities and qualified staff are in place is a relatively straightforward process. However, determining if a site will actually recruit patients into the trial is another matter entirely. In order to make the process more effective and successful, we need to start by creating a culture and company processes that engender ownership and encourage the selection of the right sites. On a day-to-day basis, application of the assessment approaches explored in this article provides a greater probability of selecting investigators that will recruit as anticipated.

Do we as an industry set the scene for success or failure?
Before considering the investigators themselves, we should challenge ourselves as to our own role, as an industry, in building an environment that supports successful site selection.

• How often does a site selection visit result in the investigator not participating in the study? With up to 30 per cent of site not recruiting any patients, should we not also expect 30 per cent of the sites we visit to not go forward? Naturally, it is not this simple, but with relatively few sites dropping out at this stage it should be considered whether the balance is in the right place.
• How often are investigators re-selected who have previously taken part in trials but not recruited any patients? Again, this is not straightforward given the various reasons why an investigator may not have recruited patients to a previous trial, or why the investigator may be selected for continued trial participation. However, we should at least have a realistic view of the investigator’s likely contribution and account for this in study planning.
• Are the individuals who select investigators accountable for the subsequent performance of the investigators? One team of people may be responsible solely to select a set number of investigators from a limited pool, and may be facing considerable time pressure to do so. As such, their chief concern will be to meet the immediate commitment in terms of the number of sites. This can create a disconnect between site selection and the actual trial conduct, which may be exacerbated when the team managing the sites throughout the study receives such a list of investigators they did not select themselves. If not appropriately and proactively managed, this process allows for a lack of ownership and responsibility for site performance.
• Do the individuals responsible to select the investigators have the required experience to assess some of the more complex factors, for instance investigator motivation? CRAs may be trained to select investigators in line with ICH and the protocol. However, is the site selection staff trained to effectively assess the less tangible aspects of what makes for a successful site? Moreover, is the staff in a position to reject investigators they do not consider suitable?
• There simply may not be sufficient sites that meet all the criteria. This is an increasing problem in many therapeutic areas, notably oncology. Consider whether we are being too conservative in terms of investing in new investigators in emerging markets and facilitating their trial participation. Are we actually taking greater risk by not being more inclusive in this regard?

Selecting investigators who truly deliver is an issue that goes deep into the operations of a company—and this is ultimately where it is within our own control to set the tone for success or failure.

References

For additional information on methods to identify the golden site profile, please refer to the article ?Data Based Predictions? by Michelle Jones, MSc, CStat and Stephen Jones, MSc, CStat, published in Applied Clinical Trials magazine, March 2008, vol 17, no 3.