Taking Stock of Today’s Global Site Landscape

The investigative site landscape is poised to change dramatically during the next decade. Many forces are at play: the convergence of clinical research into clinical practice; consolidation; more demanding site selection and site management practices; and ongoing efforts to improve investigator competence and credentials. Tufts Center for the Study of Drug Development (CSDD) recently conducted an assessment of the current state of the investigative site landscape to reset baseline data with which to monitor this

Ken Getz

 anticipated change in the landscape structure. 

The results of our study show a global community of unique FDA-regulated principal investigators (PIs) that has remained highly fragmented, nascent and unstable during the period 2008 to 2015. The largest segment is comprised of PIs who conduct only one or two clinical trials each year. During this period, one-third of all PIs have been first-time filers.   

But there are early signs that the landscape is scaling and maturing, particularly among PIs in North America. The rate of globalization is slowing and shifting where recent growth in the number of active principal investigators in Europe has outpaced that in North America and in the rest of world.  

Taking measure of the landscape

Consistent with past studies and to assess changes among the global community of individual PIs, Tufts CSDD analyzed data from FDA’s Bioresearch Monitoring Information System (BMIS). This database captures FDA form 1572s for each investigator administering an experimental drug under an investigational new drug (IND) designation. The full database containing approximately 525,000 records was analyzed after duplicate records had been removed.  

This year, Tufts CSDD considered using ClinicalTrials.gov as its primary data source. As a federally mandated registry, ClinicalTrials.gov is becoming more valuable but it has a number of limitations, one being that it lacks data on individual clinical investigators, instead providing a less granular view at the investigative site or facility level. As such, ClinicalTrials.gov cannot detect the specific location of individual remotely-based PIs affiliated with a site network or when a facility loses a specific investigator. Data from ClinicalTrials.gov has also not been validated to test for accuracy and completeness. The BMIS data set has a high level of completion and accuracy with the vast majority of sponsors reporting that they 

submit their 1572 forms periodically and routinely.

Tufts CSDD researchers gathered descriptive data on unique investigators filing form 1572 between 2008 and 2015. In segmenting the landscape, investigator maturity was defined as the number of consecutive years that a given PI had filed at least one form 1572, with first-time filers considered the least mature. PI productivity was defined as the volume of distinct clinical trials for which forms 1572 per investigator were filed. Turnover rates were defined as the percentage of unique principal investigators who filed at least one 1572 form in a given year but had not filed another 1572 form within the subsequent four-year period. Past research conducted by Tufts CSDD and others has demonstrated that if an investigator fails to file a new 1572 form within four years, it is highly likely that he or she will not file again.

Fragmentation but early signs of scaling

Since 2008, the number of unique investigators has grown 4.6% annually. During the past three years, the number of unique PIs has been growing at an even faster annual rate of 5.5%. At the end of 2015, there were 33,920 unique FDA-regulated PIs globally.

In 2015, there were 10,496 unique PIs (31% of the total) filing at least one form 1572 for the first time. Indeed, each year since 2008, approximately one-third of all unique FDA-regulated PIs are first-time filers.

Since 2008, the proportion of unique PIs in North America has declined from 63% of the total to 55%. The proportion of PIs based in Northern and Western Europe, however, has increased from 17% to 22% of the total-representing an 8.4% annual growth rate. The proportion of PIs elsewhere in the world rose from 20% to 23%. The number of investigators in these parts of the world grew at a 6.7% annual rate; well below the growth rate for this segment observed in the 2000 to 2007 period. 

Between 2008 and 2015, the number of unique investigators conducting a small number of clinical trials annually has been growing at a much slower rate than those conducting a high relative number of trials each year. Unique investigators, filing only one 1572 each year, grew by 3.2% annually, below the overall landscape growth rate and five times lower than the highest filing volume group. 

Early signs of scaling are most notable in North America. The proportion of investigators that are single-filers there has declined one to two percentage points each year between 2008 and 2015. At the same time, the proportion of higher volume filers (four or more clinical trials annually) has been rising steadily. Northern and Western European countries and those in the rest of the world are experiencing a similar shift, though it is not as pronounced.

Rising productivity and turnover

Unique PIs in North America have increased their productivity by 26% overall since 2008. Each PI managed an average of 1.9 distinct clinical trials in 2008 compared with 2.3 in 2015. Investigators in Northern and Western Europe and in the rest of the world saw more modest increases in productivity from approximately 1.2 distinct clinical trials in 2008 to 1.4 in 2015, representing an overall growth rate of 15.7% (Europe) and 12.9%  (rest of world) over the seven-year time horizon.

Turnover rates among unique FDA-regulated PIs remain very high, particularly among those investigators conducting a small number of clinical trials each year. Turnover rates also vary by geographic area, with the highest levels observed outside Northern and Western Europe and North America.

Approximately four out of 10 unique FDA-regulated PIs worldwide-or 10,838 individuals-who filed at least one form 1572 in 2011 have yet to file again.

Half of all PIs filing a single 1572 form each year choose not to file again in subsequent years. One in five unique PIs filing two or three 1572 forms in a given year choose not to conduct another industry-funded clinical trial. And approximately 5% and 1% of unique global PIs filing four to six 1572 forms and seven or more forms each year, respectively, choose not to continue participation.  North America–based PIs have the lowest relative turnover rates. 

Anticipating structural change

The results of this recent Tufts CSDD study capture a global landscape of unique FDA-regulated PIs that remains very fragmented, with high levels of inexperience and turnover. But there are signs that the landscape is beginning to scale and mature.

Sponsor companies-and to a greater extent contract research organizations (CROs) responsible for study conduct performance-are major facilitators of landscape change. Organizations have been establishing preferred relationships with site networks and large institutions. And several CROs have been consolidating the landscape since 2014 (e.g., Bioclinica, Icon, and PPD).  

Sponsors and CROs are demanding greater speed and efficiency from their investigative site partners through a number of primary initiatives. They are looking to use more robust and comprehensive site identification and selection solutions. They are vying to integrate site management infrastructure and systems to improve contracting, budgeting, recruitment and performance oversight.

Sponsors and CROs are looking for greater study conduct control to increase patient recruitment and retention effectiveness, particularly in light of the growing number of clinical trials targeting smaller and more stratified patient subpopulations. And sponsors and CROs are establishing more effective investigative site relationships to support higher levels of patient engagement.

In response, we can expect the traditional site landscape to continue to scale and mature at the same time that a variety of new study conduct models-including those integrated into healthcare settings and those accommodating remote study volunteer participation (e.g., telemedicine and home nursing networks) become more commonplace.

Ken Getz, MBA, is the Director of Sponsored Research at the Tufts CSDD and Chairman of CISCRP, both based in Boston, MA. email: kenneth.getz@tufts.edu

* The author wishes to thank and acknowledge Cerdi Beltre, Carrie Brown and Stella Stergiopoulos for their collaboration on this study.