home > epc > winter 2012 > sites unseen
European Pharmaceutical Contractor

Sites Unseen

Clinical trials sponsors are moving from 100 per cent site monitoring of research data, thereby reducing costs and saving time, and resulting in a more efficient and successful outcome. However, there are crucial strategies that should be adhered to in order to ensure a streamlined process.

Over the past year, the FDA has recommended that clinical trial sponsors move away from their reliance on full on-site monitoring of research data Ė an extremely time-consuming and costly process that typically involves thousands of hours of manpower per trial Ė and conduct more centralised and risk-based monitoring. Risk-based monitoring is both more effective and safer for patients because monitors can review smaller amounts of more focused data with greater accuracy. When implemented correctly, it can also lead to huge cost savings; sponsors can save 23 per cent of trial costs with a risk-based approach (1).

As a result, more and more sponsors are interested in implementing a targeted, risk-based monitoring strategy to increase safety and efficiency throughout the trial process. These new methodologies are still being explored in clinical trials, but some best practices are beginning to emerge. The key factor in successfully implementing risk-based monitoring is to have the right underlying tools and processes in place to efficiently alert monitors to potential issues without bogging them down in extraneous data. With the proper use of modern technologies and methods, sponsors can automate and streamline complex criteria and processes to increase efficacy and efficiency of the entire monitoring process. As a result, risk-based monitoring can drastically reduce the risks and overall costs associated with clinical trials.

The Goal of Monitoring

Monitoring is one of several crucial components utilised to ensure the quality and integrity of a clinical investigation. Monitoring uncovers potential problems, such as identifying data entry errors or missing data, provides assurances that study documentation exists, assesses the familiarity of the siteís staff with the protocol and required procedures, and provides a sense of the overall quality of a site (2).

More recently, the FDA has strongly encouraged the adoption of centralised rather than on-site monitoring where possible, because it allows for more effective analysis across all sites. Several publications suggest that data anomalies (such as fraud, including fabrication of data and other non-random data distributions) may be more readily detected by centralised monitoring techniques than by on-site monitoring (3). In part, centralised monitoring is possible due to the powerful insights into data that modern analytics tools and technology provide, in ways that simply arenít possible otherwise. The limits of onsite monitoring have become readily apparent. Firstly, trials are plagued by the costs as these can reach anywhere from 25 to 30 per cent of the entire cost of the trial (4). Secondly, there are problems with accuracy and effectiveness, as itís impossible to wade through all the data in a study with 100 per cent accuracy and alertness. This often results in data that is only skimmed, meaning that large problems can go unnoticed.

Centralised monitoring can solve many of these problems. With modern alerting systems, those responsible for detecting anomalies can create rules in their analytics systems to identify them. For example, they can be set up to signal whenever a patient has the same weight for more than four visits (a common type of fabricated data). When these alerts are triggered, processes can be implemented to look into the issue, potentially with on-site monitoring if necessary. One particular advantage of this method is that as the trial progresses, a library of alerts can be built up as more and more types of anomalies are found. Over time, the collection of alerts becomes quite sophisticated and effective at proactively identifying problems.

This leads to risk-based monitoring, which is an exciting and powerful way to increase patient safety and reduce costs. Risk-based monitoring starts with each site scheduled for the minimum number of on-site visits. Centralised monitoring is then used to target additional on-site monitoring visits toward troubled or high-risk sites. Additional on-site monitoring visits are targeted through triggers, which can be based on any aspect of the data; such as quality or enrolment. Over the course of the study, the risks and triggers can be reassessed as more data is collected. The key advantage of risk-based monitoring is that it uses available resources more efficiently based on where theyíre most needed. Triggered monitoring uses a combination of metrics, such as enrolment data, quality metrics and safety signals, to determine where monitoring resources would best be used.

Benefitting from Risk-Based Monitoring

There are several primary factors that must be considered when determining whether or not risk-based monitoring is appropriate for a risk-based study:

Study Complexity
As studies get more complex, more on-site visits are typically required to ensure compliance and correctness. In addition to this, if endpoints are not thoroughly clear and objective, it may be too difficult to rely on centralised monitoring to trigger on-site visits. However, even complex studies can heavily rely on risk-based monitoring if the infrastructure and technology are in place.

Technology Infrastructure
In order to be able to rely on triggers and a centralised monitoring system, the technology must be in place to provide the required answers. Questions sponsors should ask to determine whether or not a studyís technology infrastructure is well-suited for risk-based monitoring include:
  • Does the study implement electronic data collection systems that are conducive to centralised monitoring?
  • Are audit trails and metrics data easily accessible?
  • Are the analytics tools in place so that data can be analysed in a timely fashion? Not simply clinical data, but metrics data as well?
  • Are all systems, including patient enrolment, CTMS, safety and EDC systems easily accessible?

Risk Assessment
Depending on the overall risk profile of the study, it may or may not make sense to use risk-based monitoring. If itís expected that the sites are going to have extensive problems implementing the protocol, then it is important for the overall success of the study to rely heavily on the on-site monitoring.

Study Size and Geography
If the study is extremely large or geographically spread out, then it can drastically increase the cost of monitoring. The complexity and logistics of using on-site monitoring can become cumbersome and expensive. In these cases, risk-based monitoring can keep costs low without sacrificing patient safety; it even has the potential of improving it.

Setting Up Trigger Factors?
There are many important factors to consider when setting up risk-based monitoring. First of all are standard triggers, which are important for all studies. For example, trials should always consider flagging sites that continuously submit statistically improbable, fraudulent, fabricated or missing data. Triggers can be set up on all data dimensions that are commonly problematic. As an example, if too many patients have the same weight after each visit, then there is probably a problem with the site that would require an on-site visit. Another trigger could be if there is an unexpectedly high amount of missing data. There are many other types of potential triggers that can be set, and a library of common triggers can be developed and reused in subsequent studies.

Itís also imperative to identify studies where the protocol specifies the most important types of data relevant to that specific study. For example, it may be the case that certain types of baseline characteristics, such as age or concomitant treatments, are not particularly important for a specific study. However, there are other metrics that are particularly crucial for the success of the study, such as not following protocol-specified definitions. When areas of specific concern are known, then triggers can be targeted specifically toward them, increasing the likelihood of a successful study outcome. According to the FDA guidance, the types of procedures and data should include those which directly impact the studyís reliability or patient safety, so these should be incorporated into a risk-based monitoring approach as well. Itís important to remember that triggers can evolve over the course of the clinical investigation, as long as the monitoring procedures were well thought out and established in the studyís design. As new information is collected or new potential problems are found, sponsors can introduce new triggers.

Putting Methods and Processes in Place
Itís important to have well-documented procedures in place describing the studyís monitoring process. These procedures govern the entire trial and are crucial when defending the studyís results to a regulatory body. Regarding risk-based monitoring specifically, it is crucial to have sophisticated trigger tools in place that remove potential errors throughout the process, as improperly executed triggers will quickly ruin an entire study.

Instead of relying on antiquated tools such as Excel spreadsheets, there are analytics tools and other systems that can be put in place to make the most out of risk-based monitoring. For example, itís important to have an analysis system that accesses all available data, including: patient enrolment, the clinical trial management system, electronic data capture and safety systems. All systems containing data that will feed into triggers must be accessible from the analytics tools.

Often sponsors rely on systems that output Excel spreadsheets, and then manually sort and filter on a weekly basis. This instantly leads to a less effective and problematic risk-based monitoring system. In addition, it also drastically limits the number and complexity of the triggers that can be put in place. With an automated system, sponsors could include dozens or hundreds of triggers, set up to catch all types of problems that would warrant an on-site visit. With a manual process such as an Excel worksheet, it becomes tedious and prohibitively expensive to have more than a handful of triggers. Itís also important to have advanced tools in place that allow data to be explored from many dimensions simultaneously. If these tools arenít thorough, automated and powerful, then there is a much higher risk of problems with the risk-based monitoring process.

Itís also important to have an automated alert system in place. With an alert system, any number of complex triggers can be put in place during study startup. The system will then automatically notify relevant parties when any of the trigger conditions are met, for any site in the study. There are off-the-shelf software systems that allow this Ďsetup once, and waití methodology and make the entire risk-based monitoring process run more smoothly and efficiently. These systems will automatically notify the study coordinators and monitors as soon as the trigger criteria are met for any site, expediting the entire monitoring process. The trigger criteria can then be modified later as needed, and as specified by the protocol.

Upcoming Innovations in Risk-Based Monitoring

Risk-based monitoring is an area that will benefit heavily from technology as innovation continues to affect the industry. In addition to off-the-shelf packages described above, large libraries of pre-made triggers developed and curated by sponsors, CROs and software vendors are available. Examples include trial metric-based libraries or libraries specific to different therapeutic areas. When a new trial starts, the study organisers and planners will be able to use one or more of these trigger libraries to seed the initial triggers that will be used, and then modify them as necessary. This will allow a trigger best practices to emerge, make risk-based monitoring more effective, and reduce the time to getting started.

Eventually, other bodies will be able to hook into these trigger systems, such as the data safety monitoring boards, the Institutional Review Board (IRB) or even regulatory bodies. The end goal is to engineer a push-based trigger network, as opposed to a pull-based one. Pull-based networks require users to go out and specifically look for whatever data or action theyíre interested in, similar to when consumers go to a website looking for news. Push-based networks automatically notify the user, similar to how a cell phone beeps when an email is received. If used properly, push-based triggers and notifications can lead to huge efficiency gains and cost savings

Risk-based monitoring is the obvious progression in making the best use of study resources. Technology is facilitating this trend to make studies as safe and cost-effective as possible. Risk-based monitoring increases the success rate of the study while decreasing the costs, and is set to grow in popularity in clinical trials over the coming years.

  1. Visit: NEWPDF/Centerwatch%20monthly_ February2011_Article.pdf 2. Visit: GuidanceComplianceRegulatoryInformation/ Guidances/UCM269919.pdf 3. Visit: CTTI%20Monitoring%20Survey.pdf 4. Visit: Perspectives-Files/2012/Commentaries/ HSP-Drugs-Aligning-Cultural-and- Financial.pdf

Read full article from PDF >>

Rate this article You must be a member of the site to make a vote.  
Average rating:

There are no comments in regards to this article.

Rick Morrison is the Co-Founder and Chief Executive Officer of Comprehend Systems. Prior to founding Comprehend Systems, Rick served as the Chief Technology Officer of an internet-based data aggregator, where he was responsible for product development and operations. Rick has over 10 years of experience in writing software for clinical trials, including tools that are now used by the FDA and Big Pharma. Rick holds a BSc in Computer Science from Carnegie Mellon University, PA.
Rick Morrison
Print this page
Send to a friend
Privacy statement
News and Press Releases

Turkish Cargo maintains its steady growth.

According to the data obtained by WACD, the international air cargo information provider, in May; Turkish Cargo, the global air cargo brand providing service to 124 countries around the world, increased its tonnage rate by 7.1 percent, and grew substantially across the industry which shrank by 5.1 percent in the global air cargo market.
More info >>

White Papers

Delivering Fit For Purpose Biomanufacturing CHO Cell Lines

Fujifilm Diosynth Biotechnologies

The challenge during mammalian cell line development is to identify and isolate stable, high expressing cell lines producing product with the appropriate critical product quality attributes rapidly, reproducibly and with relative ease. Obtaining a host cell line that inherently exhibits desirable biomanufacturing attributes can therefore have a significantly positive effect on the identification of recombinant cell lines with desired traits during cell line development screens. In this study, we demonstrate that it is possible to exploit intrinsic heterogeneity within host cell populations and identify host cell lines which are more ďfit for purposeĒ.
More info >>

Industry Events

2020 Avoca Quality and Innovation Summit

3-4 June 2020, Amsterdam, The Netherlands

The 2020 Avoca Quality and Innovation Summit will take place 3-4 June 2020, in Amsterdam, The Netherlands.
More info >>



©2000-2011 Samedan Ltd.
Add to favourites

Print this page

Send to a friend
Privacy statement