Today contract resource organizations (CROs) continue to dominate clinical research because they can accelerate drug development and reduce trial costs for sponsoring, large/midsize pharma companies and biotech start-ups. Because of this trend CROs share a responsibility with their sponsors to establish trial efficiencies and to identify, correct or eliminate common compliance deficiencies.
The FDA cites the most commonplace compliance deficiencies as inadequate investigator oversight, protocol deviations, poor record keeping, insufficient investigational product accountability, and subject protection and consenting issues.
To address these deficiencies, risk-based monitoring (RBM) has emerged as a powerful, technology-based approach. RBM utilizes analytics to direct the clinical research team’s focus onto critical data and processes that have the greatest potential risk to impact patient safety and data quality. Technology is very important to RBM because it enable the implementation of analytics, maximizing the use of a company’s data and allowing for smarter decision-making. It also provides centralized monitoring to detect trial data trial anomalies, resulting in better data quality, sooner fast-fail actions, and more successful drug to market pathways. Conducting RBM in the cloud enables trials to be more flexible, adaptable, affordable, scalable and agile.
The RBM approach has been endorsed by the leading industry consortium, TransCelerate and by globally accepted industry authorities including the ICH, the FDA, and the EMA. It has begun to be recognized as the better alternative for creating cost effective and timely drug studies that keep the focus on the quality of critical data and provide a quick means of addressing patient safety issues.
Recently, Oracle Health Sciences and North Carolina Biotechnology Center, a state-supported, non-profit corporation, sponsored a panel discussion on RBM in the Cloud. The talk focused on how RBM can improve clinical research for CROs. The panel included James Streeter, Global Head of Life Sciences Product Strategy for Oracle Health Sciences, MaryAnne Rizk, Global Head, CRO Business Partnerships & Alliances, Oracle Health Sciences, Warren Pence, Director, Adaptive and Intelligent Monitoring (AIM), Clinical Development Services, PPD, DFS Pharma and Vivian Doehling Director, Business and Technology Development, NCBiotech.
Oracle on RBM
James Streeter, providing the panel keynote, discussed how Oracle works with the pharmaceutical industry to support its current and future trends, developing integratable solutions that are modular and scalable in the cloud today, as well as provide a solutions bridge to future-proof the research business.
“We take a holistic approach to RBM, including RBM capabilities in each piece of our end to end solution. RBM is included in each of our clinical research and development, regulatory, data analytics, business processing and cloud technology, remote monitoring and eMonitoring solutions, all of which are very important to CROs today.”
Mr. Streeter cited Oracle’s Siebel CTMS solution, “Our CTMS solution is cored to understanding what's going on with a trial. It includes the risk assessment categorization tool. The systems contained here can help CROs take advantage of what they need to do and how they need to define risk up front. The key risk indicators and clinical development analytics are tools used look at what's going on, monitor and identify the risks, understand what needs to happen, and how to address those risks. Those activities are all feed back into the Siebel CTM System. Then the partial SDV support, whether we're looking at the CTMS and the EDC, and even SDR, are supported with our CTMS and EDC systems and how they come together.”
He then addressed other pieces of Oracle’s end to end clinical trial solution, “Another important set of considerations is the training, planning, and tracking for the sites within the CTMS system. As we look at risk, we need to ask, How do we handle those risks? How do we address them? How do we record them back into the CTMS? It's important to record these. The regulatory agencies want to see at the beginning of the study that its risk was defined, how that risk would be addressed, and then at the trials end, that those risk mitigating steps were taken.”
“We also have our EDC and IVRS systems. As part of this, our Data Management Workbench cleans and organizes data, making it ready for analytics. It takes data from different systems and finds a use for it that will increase the quality of the study itself.”
“CTMS is where most companies collect and document what's going on, whether it's in the initiation, the start-up phases, or even at the close out phases. A lot of that information is captured now in the CTMS system and then in the middle of the study it's sent to a clinic in analytics. Our Clinical Development Analytics are an important piece of that.”
“Oracle also offers safety and surveillance systems. Aggregating the signals from the early detection process of Statistical Control Environments are another important piece when we're looking at analytics from monitoring the risk of a clinical trial.”
“Additionally, we integrate other patient data from third party systems. We're seeing a lot of companies now just looking at the patient data and can learn a lot from that. For example, we can see how sites are performing the data, whether there’s any fraud detected, when data was entered, and the quality of data from one patient compared to another. These are important pieces that can be looked at in analytics.”
Mr. Streeter concluded, “So, again, it's that holistic approach from planning all the way through disclosure. These are what we call the planning, design, execution, analysis and disclosure phases.”
Moving forward, Mr. Streeter says, Oracle sees a decreased focus on source data verification (SDV). He added, “CRAs will focus on source data review (SDR), rather than SDV. They’ll really look at the data and understand what's going on. This new demand is creating the emergence of a new role – that of the data scientist. Because there's so much data being collected that can tell us so many different things, we need this new data scientist role to look at data differently. He or she needs to look at the data patterns and identify what's important, what's not.”
“In the future we see the industry moving from risk monitoring to towards risk management. That is, the smart interoperability among safety, medical monitoring, and clinical data management, using AI and machine learning to identify the data needed to consider for better outcomes.”
As Director of Adaptive and Intelligent Monitoring (AIM) Group at PPD, Warren Pence focuses on clinical trial management. Other groups making up PPD’s comprehensive RBM capability include those monitoring bio-stats, medical monitoring data management, data analytics, and remote site monitors.
Each group looks at data from different perspectives and identifies critical issues.
PPD’s AIM RBM Processes
Mr. Pence’s group takes a holistic approach to RBM by assessing the entire subject visit. “We do that for SDR and SDV. We know that every site has different levels of quality. So, we stratify them into three risk levels, low, medium, and high. We customize a specific plan based on each protocol for each client. Knowing that the quality of sites is not equal, for every interim monitoring visit, we perform a 16 factor site health assessment, or a mini-CAPA [corrective and preventive action] assessment. “
“Five of our questions are completed via Oracle’s Siebel CTMS solution. We pull identified issues from the CTMS system and compare them against four key risk indicators. This information is scored and characterizes the site into one of the three risk categories. We see that in most studies there are about 20% of high risk sites, 60% moderate risk, and 20% low risk.”
Mr. Pence also said that larger pharmas are using vital signs as a surrogate marker for analyzing SAEs and actually using that as a way to detect scientific misconduct and fraud early on.
He added, “Our AIM approach provides greater oversight than 100% SDV alone. To ensure quality, data integrity and subject safety, each monitoring plan is customized based upon the protocol, the risk assessment and the client. We focus on data and site processes that are critical to quality through a holistic review of subject visits. PPD has made a conscious effort to integrate our RBM into our existing systems and not buy other systems specifically to execute RBM. Our clinical cost savings were between 5% to 25 %, based upon the phase of the study, protocol complexity, indication, IP formulation, and aggressiveness in SDV reduction. This includes some clients with very complex oncology trials willing to do 40% SDR, 20% SDV on average. For one particular strategic partner, with over 12 clinical trials, we saved them about $15 million.”
PPD and Oracle
These AIM analytics are part of PPD’s existing systems. PPD currently uses on premise Siebel CTMS. It is shortly moving to the cloud version. Also, for its large pharma clients, PPD sets up teaching partnerships that standardized the way the CRO does all studies with that client. PPD has found that its smaller biotech clients don’t like to do that, as they don’t want to make the investment in using multiple vendors.
Also Mr. Pence has been given to understand that, from the pharma’s point of view, analytics tools and data collection tools are getting smarter. “The things that are being added to the tech lineage, analytics and data management tools are making it easier for the FDA to review trial results.”
He also anticipates what will be possible when PPD moves to Siebel Cloud CTMS. “When we move to the Siebel Cloud, we’ll be able to do sampling and grouping based on subject visit for some of our phase I and II studies. Instead of treating each visit during an oncology cycle as an individual visit, we’ll be able to approach all visits during that cycle as one, and the level of SDV or review would be the same. Monitoring visit response will be forced based upon completion of the report, and the report will be locked based upon submission. We’ll also be able to sample a set percentage of subjects.”
Many of PPD’s larger and midsize pharma clients are TransCelerate members, have embraced RBM principles, and take an aggressive approach to reducing SDV and SDR. PPD currently has a 172 studies using RBM. They range from 90% SDV to 5% SDV. The organization accommodates all those different methodologies.
Mr. Pence also added, “Another thing we’re seeing in large pharmas is the assimilation of data management, monitoring, and analytics groups into the same group.”
PPD’s smaller biotech clients are less aggressive about reducing SDV. PPD is only providing about 60% to 80% SDV for them. PPD’s smaller biotech clients are still of the mindset that 100% SDV (as well as interim analysis and data safety and monitoring board [DSMB] reviews) equal quality.
But, all that monitoring makes for errors. According to Mr. Pence, this is where centralized statistical monitoring can make the difference. The biggest challenge that Mr. Pence faces every day is convincing smaller biotechs that RBM is the way to go because it can identify outliers.
In summary, risk based monitoring has matured but is still gaining in efficiencies. CROs, with high industry maturity, are critical in providing RBM implementation for smaller companies today. As key supporters of RBM, regulatory agencies want to have higher data quality, and make it easy for regulators to approve drugs. Investments in risk based monitoring can help in these areas and will continue. Standardization for including RBM in planning, design, execution, analysis and disclosure is key.
Keywords: risk based monitoring, RBM, CROs, pharmas, biotechs, CRA's, data managers, clinical operations managers, clinical trials, ctms, edc