The Use of a Single Center, Multi-Location Research Structure For Participation in Large Clinical Trials: The Cleveland Clinic Model

..

As funding for clinical research trials decreases and sponsors of clinical studies are under increasingly stringent timelines to obtain FDA approval and get new therapies to market, the need increases for well-managed, efficient, and effective clinical trial sites.  Recruitment and retention of large numbers of subjects in a relatively short period of time will soon become a standard expectation of all sponsored trials in order to meet these increased demands. A single center, multi-location (SCML) model has recently demonstrated recruitment and retention success in the Heart and Vascular Institute at the Cleveland Clinic.

To date, no other healthcare enterprise has conducted clinical trials in this manner. The regional network model LIONs (local identification and outreach networks) used similar methodology with the interest of recruitment and retention; however, despite using the same study staff and documentation being stored centrally, each location within the network was part of a unique health system and had a separate principal investigator (PI) and institutional review board (IRB) approval. The LIONs model had positive results on recruitment numbers as well as similar data quality when compared to individual sites in the same trial.¹ The LIONs model reflected increased per-patient costs initially when compared to single locations; however, these costs did decrease over time. The model discussed here achieves similar goals but further improves the efficiencies of running clinical trials.

Although there are many benefits to research network models, there are key differences between the SCML model and other more common network models, including those utilizing a single PI. In the SCML model, each location is part of the same institution, allowing for one contract and budget to be done, which eliminates the negotiation of multiple contracts. Since the institution is a single center from a trial management perspective, this also allows for a single-site initiation visit and site training to be conducted as well. Similarly, the storage and maintenance of regulatory documents are streamlined in the same manner. In a network model, these may be stored at each individual site. A network model would also warrant monitoring visits to be done at each site within the network, whereas, the SCML model could be reviewed in a single visit.

The SCML research model allows for rapid high-volume enrollment in large-scale clinical trials. Utilizing multiple locations within a unified health system increases research visibility and availability to subjects in local facilities while increasing the pool of potential participants from which to recruit. The advent and use of electronic medical records (EMR) plays a pivotal role in the creation of a single-center, multi-location research model, in that research and clinical data is collected in a uniform manner and immediately available electronically throughout the healthcare enterprise. This decreases variability between locations, as well as minimizes study maintenance. This model decreases trial startup timelines by using a single PI, site contract, and IRB application and approval. Each site within this model is given a unique site number by the sponsor.  Collectively, these locations comprise one unique center. The trial discussed in this article was successful in recruiting 80 subjects in a period of 3.5 months, with 68 randomized subjects and total subject retention to date of 98.5%. The trial is a long-term follow-up study of which is nearly two years complete; follow-up visits are conducted on site at six-month intervals. 

The purpose and application of this model is best suited for individual health systems with multiple operating locations participating in industry-sponsored clinical trials in efforts to continue to make viable contributions to the study with optimal efficiencies. Each location must be part of the same health system, with unified management and business systems, in order for this model to apply.

For the SCML model to be successful, a strong collaboration among involved locations is necessary. The collaboration assists not only with research recruitment, but also creates a stronger working relationship between locations at all levels of interaction. Often, community-based physician practices do not feel aligned with the academic aspect of medicine within a large health system. Providing the opportunity for these physicians to participate in research fosters professional development, satisfaction, and retention. Although not statistically evaluated in this study, we found that convenient and familiar facilities can also increase subject adherence to study procedures and retention.

The model

Personnel: In this trial, there were a total of three locations, with at least one co-investigator (Co-I) at each location. The co-investigators ensured investigator presence at each facility to address any immediate issues that arose during the visits. Otherwise, the PI addressed all concerns via regular meetings with the study team. The use of EMR throughout the Cleveland Clinic enterprise allowed the PI to actively review eligibility criteria for each potential subject regardless of the subject’s facility location. Additionally, the EMR allowed immediate PI review of adverse events. Research coordinators were fluid for all of the locations, traveling among sites to meet subjects’ needs. This was feasible, since each of the investigators and all of the study personnel were employed by the Cleveland Clinic.
  
PI oversight: The PI was instrumental in selecting and personally training all of the Co-I’s on the trial. In addition to reviewing each subject’s eligibility with the study coordinators, weekly meetings were held between the PI and at least one study coordinator. Monthly team meetings were held with the PI and entire coordinator staff.  The PI was also available via pager or email for unscheduled contact when needed.  Immediate concerns and adverse events were assessed electronically.

Regulatory structure: Studies have shown that repetitive ethical reviews have been at the root of many wasted resources as well as delays in conducting human subjects research.² In a letter to the editor in Palliative Medicine, it is indicated that in a multicenter trial that included six different institutions, each of the institutions required individual ethical reviews, which resulted in time delays as well as the need to seek increased funding.³ Although the network model discussed in the letter is different than the Cleveland Clinic model—in that each institution is separate—the idea that multiple ethical reviews will add to the study timeline and cost remains relevant.

The Cleveland Clinic IRB operates on an electronic system referred to as “WebKit” for all transactions. WebKit allowed enterprise-wide access to IRB documents. The trial was conducted under a single IRB application and approval, which included conduct at all of the individual locations. There was one contract and one payment schedule between the sponsor and the Cleveland Clinic. This was possible because all the locations used were owned by the Cleveland Clinic. In addition to streamlining the budgetary process, the ethical process was also streamlined, allowing the local IRB to review one application.

One main physical location allowed research staff to have a dedicated work area with secured access in which to store regulatory documents and subject records. The regulatory binder and signed consent forms for all locations were stored in this area. The center had one regulatory binder with designated tabs for items that pertained to one specific location. These items included FDA form 1572, monitoring reports, and shipping/delivery notifications. During the time of study visits, research staff would obtain the study records of patients scheduled to be seen and transport them in a locked portfolio to the location conducting the visit. Following the research visit, the record was returned to the main secured location for storage in the same manner. The use of one dedicated work area minimized research non-compliance, increased the ease of monitoring visits, and ensured the study was conducted the same at every location.

Each of the three locations was assigned a unique site number by the study sponsor due to drug delivery and interactive voice response system (IVRS) programming. As a result, each location required a separate 1572 with the location address of drug delivery. The PI and all Co-I’s were identical on each of the forms. During subjects’ enrollment visits, they were assigned a study number that corresponded to the specific location of study visits. For this reason, once randomized at a particular location, it was imperative that the visits continued at that same location.

On-site monitoring and audit visits were completed and encompassed all three of the locations. Regulatory and source verification were done from the main location, along with a physical visit to review pharmacy sites. Utilizing a risk-based monitoring (RBM) plan, the majority of monitoring visits were done via phone conference. Each call covered all of the locations, so duplicative calls did not need to be made. 

Study visits were conducted in exam rooms within the clinical area of each location. Coordination between study staff and clinical staff was essential to ensure that adequate space and equipment would be available during the time of the study visits. This was done while maintaining regular patient flow in all clinical areas.

Randomization and study drug distribution: Study drug storage and distribution was done through investigational drug services (IDS) and retail pharmacies (for the locations that did not have an IDS on site). Uniform procedures were used at all pharmacy locations, which were developed and implemented by IDS. IDS was responsible for oversight of retail pharmacies performing quarterly process audits to ensure uniformity and compliance with all research-specific FDA guidelines. These quarterly audits took place for the first year of the trial and then on a consult/as-needed basis after that time. 

Copies of each report were stored in the site regulatory binder. In addition to the quarterly audits, regularly scheduled meetings were held with the research manager, project manager, IDS manager, and the manager of retail pharmacies. This was done to coordinate study startup and all processes involved in conducting the trial. Much attention was paid to developing processes that allowed research visits to be completed per protocol, but not disrupt the workflow of retail pharmacy customers. The investigational study drug was shipped and returned from each individual location. Subjects were told at the time of the informed consent visit that once they were randomized at a certain location, each subsequent visit would take place at that same location. Once kit ID numbers were given through IVRS, the standard requisition was filled out (signed by PI or Co-I) and sent to the pharmacy to be pulled from the supply and labeled for the subject. The investigational study drug was stored separate from retail pharmacy supplies and dispensing logs were rigorously maintained.

Laboratory specimens: In a similar fashion to drug delivery, laboratory services were also needed for the trial. The Cleveland Clinic has a research special studies laboratory that process and ships (as needed) study specimens per protocol. The research special studies laboratory is located at the main campus of the Cleveland Clinic. All specimens were sent to the research special studies laboratory regardless of location. If the specimens were collected outside of the main location, they were sent through the hospital courier system and delivered to the lab for processing.

Opportunities for further development: Further development of this model could include the utilization of one site number. As this model becomes increasingly more common, there will be a need for a more intuitive IVRS system that allows for this structure in randomization and drug delivery under one site number. This would have a variety of benefits, including decreasing regulatory burden, preservation of available sites for sponsor use, and avoiding duplicative documentation.

Additionally, education will need to be conducted surrounding the use of the model for trial sponsors, as well as contract research organizations (CROs), to promote understanding and usability of this model.

Discussion/conclusions

The current changes in the U.S. healthcare system as well as FDA requirements to bring new therapies to market have created a challenging atmosphere in which to conduct clinical research. Each layer of new therapy development is under increased pressures to operate efficiently, which has a trickle-down effect to the next layer of development. It is essential that at all levels these expectations can be met, while bearing in mind the integrity of the institution and the trial data. By using a sound SCML model, these goals can be achieved at the clinical site level to satisfy the mission of the organization as well as the demands placed on them to perform quality clinical research in a highly efficient manner.

In order to maintain successful participation in clinical trials, institutions must consider adapting current management methods of clinical research trials. Our SCML model has proven to be a successful and minimally disruptive way to maintain recruitment and retention. Changes in medical practice, such as the use of EMR, makes the use of the Cleveland Clinic supersite model a viable option at other institutions. 

References

1. W. Kernan, C. Viscoli, L. Brass, et al.  2011.  “Boosting enrollment in clinical trials: validation of a regional network model.” Clinical Trials 8: 645-653.
2. J. Menikoff. 2010. “The paradoxical problem with multiple-IRB review.”  New England Journal of Medicine 363: 1591-1593.
3. S. Cadell, G. Ho, L. Jacques, et al 2009. “Considerations for ethics in multisite research in pediatric palliative care.” Palliative Medicine 23 (3); 274-275.