Mary Arnould, associate director at Bristol-Myers Squibb (B-MS), told attendees at the DIA annual conference last week in Washington, DC that B-MS, which has implemented RBM fully in about 15 studies beginning in November 2013, follows the TransCelerate recommendations and methodology closely as it “really looks at risks generalized through the programs, and then mitigations can be made.”
B-MS performs both program-level risk assessment and protocol-level risk assessment to enable the creation of risk-mitigation plans, which includes site monitoring plans and data review plans, and which also feed into the creation of quality risk indicators (QRIs) and other thresholds.
B-MS uses a group of central monitors to remotely assess quality risk indicators (QRIs) across sites and countries to try to identify outliers and trends on the country, study and site levels.
More specifically, B-MS has its remote monitors perform SDV (source document verification) of 100% of the critical data for 20% of the subjects, and SDR (source data review) for 40% of the subject records, Arnould added.
The company still has “one or two mandatory on-site visits,” anywhere from four to six weeks after the first patient is treated, she said, but they’re not at a fixed interval and they have to be triggered by assessments from the off-site monitors.
Off-site monitors also aggregate info received from central monitors to take a look at the issues and evaluate if they’re “real or noise,” which allows them to make a decision on whether to make a phone call to the site and set up a training teleconference, or to call and ask questions and provide some “coaching,” or to schedule an on-site visit.
“If there’s a fair amount of SDV and SDR required at a site, that’s enough of a trigger for an on-site monitoring visit,” Arnould said.
In terms of how long it takes to complete an off-site assessment, B-MS found that based on a survey of its sites that on average it takes about 90 minutes globally. The company also made the decision early on “not to invest in a tech platform to support risk-based monitoring and as a result we have a lot of frustrated site monitors, but we’re now looking to implement” a new technology platform, she said.
As far as the perceptions and myths around RBM – B-MS found that “most sites don’t like it if they haven’t participated in it, so we tell them what it is and what it’s not,” she said, noting the change in perception usually aids the adoption of RBM.
Based on this preliminary success of its RBM, Arnould said the company has mandated that as of Q2 of 2015, all full-development and most early development studies will be initiated with RBM.
“I think one of the primary successes or benefits is the earlier identification of issues and ability to find the errors that matter and to put in mitigation strategies to address them,” she added. “From a sponsor’s standpoint, it saves us a lot of time on the back end of not having to deal with errors repeated throughout the study, and I think sites appreciate having us identify mistakes early on so they don’t have to go back and correct them later on.
“For sites and academic centers, there should be less demand on study coordinators’ time. Another benefit, and one of the comments from sites, is that they appreciate the more purposeful communication,” she added.
Celgene, PPD Collaboration
In 2012, Celgene launched a partnership with CRO PPD and a year later the company piloted an RBM strategy with a Phase III study that utilized PPD’s adaptive and intelligent monitoring (AIM) methodology.
Peggy Zavala, senior manager of regional clinical operations at Celgene, told DIA attendees that because that pilot was successful, it created a team of experts from both PPD and Celgene in 2014 and since then has used PPD’s AIM across about 45 studies ranging from Phase II to IV, though the methodology is customized for each study.
“Probably the most important take home is to avoid monitor-driven decisions…make data-driven decisions that are adaptable based on the site performance. Don’t leave it up to the CRA to decide how to increase site monitoring – have your system do it, and have your system do it based on the site quality and the health of the site,” Warren Pence, associate director of AIM at PPD, also told attendees.
In terms of the site health assessments, PPD starts most sites off at a moderate-risk level unless the performance of the site is known. “All of the assessment and scoring is automated in CTMS (clinical trial management system) to ensure 100% compliance,” Pence added, noting that there’s also an override process that allows site risk to be upgraded or downgraded based on known risks.
PPD has now used its AIM RBM program across 127 studies for 33 clients, including 58 that are running on the current version of AIM, and two Phase II studies have been completed. Between November 2013 and May 2015, 39 studies have used site health assessments, and about 2% of those included site health assessment overrides, the majority of which were to increase site risk.
“It’s still an evolution – there’s a lot of people doing it but it’s evolving, and people who haven’t jumped into the RBM pool are thinking how they’re going to do it and it’s continuing to go through modifications and enhancements,” Pence said.