Most people are impacted by event-driven analytics every day, even though they might be unaware of the term. The phrase “event-driven analytics” is just a fancy way of describing what happens when we get a notification, triggered by a specific event that requires our attention.
I get notifications all the time on my smartphone, and they generally improve my experience with whatever company or entity is sending them out. Here are just a few examples of the event-driven analytics that get me through the day:
Figure 1: Event-driven notifications help me take
These notifications give me valuable insights that impact my home, my style, and my health. And because they’re based on events I’m interested in, they allow me to take specific actions that can enhance my life, from getting my packages inside and cleaning my Roomba® to managing my investments.
Study Teams Are Missing Opportunities to Prevent Problems from Escalating
Getting notifications, or alerts, that are tied to specific events is especially important when the stakes are high. And frankly, the stakes couldn’t be higher than they are in the clinical trial industry. Without timely notifications about issues and anomalies, clinical trials can experience costly delays and patients can be put at greater risk.
Imagine how different things would be if you could get notified as soon as enrollment starts going off track at a particular site, when queries start backing up, or when a site is taking longer than usual to enter data.
In a webinar I conducted recently about using event-driven analytics to improve the performance of clinical trials, I asked attendees how quickly they typically find out about anomalies. Only 11% of respondents said they’d know within a single day. Twenty-two percent said they had to wait two weeks for this information, and another 28% admitted that problems would start to fester for two weeks before they found out about them. Imagine having a package in front of your house for more than two weeks without your knowing it!
Figure 2: Most study teams take too long to find out about anomalies,
giving issues time to escalate into major problems.
At Comprehend, we understand how applying the principles of event-driven analytics can transform the speed and performance of clinical trials. Our Clinical Intelligence platform is designed to automatically review behaviors (at the site, study, region, country, and other levels) and alert users when established thresholds are breached. The platform knows what events to look for, and can create tasks so the right users can take the right actions at the right time.
The whole process reduces manual labor and enables shorter decision cycles, improved collaboration, and faster resolution of issues before they get out of hand.
Figure 3: Advanced data technologies make
effective event-driven analytics possible.
Determining Relevant Events of Interest
What’s truly remarkable about the technology is that you can apply a risk-based, contextual approach to determining which metrics to follow and what level of importance to give them as the study progresses.
We use a risk formula that assigns weights to key risk indicators throughout the course of a study. These weights influence the overall risk calculation and make certain indicators more or less risk-averse in comparison to others, as appropriate. For example, enrollment may have higher weights towards the beginning of the study and queries later on. The model supports any number of indicators, and can be tailored to different levels or deployed as a standard enterprise model.
But how do you determine the right indicators to follow? In another poll we conducted during the webinar, the overwhelming majority of respondents said they wanted alerts related to milestone delays and site performance.
Figure 4: Most sponsors say they want timely information
about milestone delays and site performance.
Fortunately, industry standards can be leveraged to identify what to monitor and at what thresholds. Many of our customers use the Metrics Champion Consortium as well as TransCelerate. At the same time, the platform provides ample room for customization, so you can continuously fine-tune your analytics as insights are revealed over time.
The goal of event-driven analytics is to bring insights directly to users so they can take the right action as quickly as possible. When notifications become an integral part of your study teams’ workflows, you’ll be amazed you were able to get anything done without them.
For more information on improving performance and reducing risk with event-driven analytics, I encourage you to watch the webinar now. If you like what you see and need help putting these solutions in place, call Comprehend at 650-521-5449 or request a demo.