In recent years we have seen a huge increase in the veracity, volume and variety of data in clinical trials.
 

In 2021, Phase III trials were already generating an average of 3.6 million data points – three times the data collected by late-stage trials a decade before¹.
 

This surge means we have to embrace advancements in the way we analyze data if we are to save study teams time and provide more meaningful, actionable insights.
 

Vanishing are the days of building pivot tables and graphs in excel and reviewing lengthy listings for data trends and anomalies. 
 

Instead, wider adoption of data analytics is allowing visualization of data across sources, near real-time monitoring and improved efficiency.
 

The Case for Visual Analytics

The current multi-provider, decentralized, digitally-enabled environment is leading to an increased number and variety of data sources. In addition, emerging, scalable technologies, such as telemedicine solutions and wearables, have changed the requirements for data curation, storage, and analysis tools. 
 

A comparison of data visualizations against traditional frequency table reporting in two placebo-controlled trials found visualizations better supported investigators to assimilate large volumes of data and enabled improved informal between-arm comparisons².
 

In addition, data analytics and visualization tools empower clinical trial teams to focus on specific risk areas and spend less time trying to analyze data in sub-optimal formats.
 

This risk-based approach is supported by ICH-E6-R2 regulation.
 

An example of where data visualization can support risk-based approaches is in electronic Patient-Reported Outcomes (ePROs). ePROs generate a high volume of data and have the additional challenge for site teams of monitoring subject burden. Clinical data visualizations, focused on adverse events, can identify under- and over-reporting. 
 

Different data types can also be brought together. For example, clinical events with respect to data entry can help identify problematic site behaviors. 
 

Designing Visual Analytics

The technological transformation currently taking place in the life sciences means data managers are increasingly involved in the design, build and review of visual analytics within clinical trials. 
 

Choosing the right visualization is key. A 2022 study found visualizations provide a powerful tool to communicate harms in clinical trials³. However, the same study emphasized the need to choose the appropriate visualization for the data type.
 

Data managers need to identify and define the requirements for visual analytics, and ongoing interpretation of results. There are numerous stakeholders involved in clinical trials and this can be a time-consuming task.
 

If done incorrectly, it can lead to too many, too few or unsuitable visual analytics being designed.  Identifying clear priorities at the start of data visualization design can help reduce this risk.
 

For example, visualizations must be intuitive, accurate and based on near real-time data to allow large amounts of sensitive data to be translated into meaningful information.
 

The type of data utilized is also key. Clinical, metric and audit data are essential for monitoring a clinical trial and enabling teams to monitor different aspects of the data collection process.
 

Interpreting Visual Analytics

Visualizations must often meet the needs of a broad audience.
 

However, they must always be designed with interpretation in mind to ensure patient safety and improve the data quality and efficiency of the clinical trial. 
 

If the data cannot be correctly interpreted, leading to actionable insight and improved trial outcomes, the visualization has not served its purpose.
 

Visualizations should allow interactivity and have drill down capabilities. These can be extremely valuable as the amount of data generated increases.
 

Historically, study teams had to follow up on large data tables to locate the points of interest and understand the context.
 

Interactive data visualization allows users to zoom in on data points and select specific points in the plot, highlighting specific issues and allowing the user to identify anomalies or patterns.
 

This can give greater insight and clearer interpretation of the risk being monitored.
 

An Opportunity for Safer, More Efficient Trials

In an increasingly data-rich environment it is vital study teams have the correct tools for analysis and interpretation.
 

We need to move beyond manual, time-intensive processes and embrace new technologies.
 

When used correctly, data visualization has the ability to improve risk monitoring, provide greater insight for study teams and increase efficiency.
 

To fully seize this opportunity, visualization should be interactive, intuitive and easy to interpret.

References:

1. https://www.globenewswire.com/news-release/2021/01/12/2157143/0/en/Rising-Protocol-Design-Complexity-Is-Driving-Rapid-Growth-in-Clinical-Trial-Data-Volume-According-to-Tufts-Center-for-the-Study-of-Drug-Development.html 

  

2. https://trialsjournal.biomedcentral.com/articles/10.1186/s13063-020-04903-0 

 

3. https://www.bmj.com/content/377/bmj-2021-068983 

 

---

Danielle Grindley, Senior Digital Health Analyst at PHASTAR, will be speaking at ACDM conference on Monday, March 13, from 1:35pm to 2:05pm, at Hilton Barcelona. The interactive session, Visual Data Analytics: Making sense of it all, will focus on the design and interpretation of visual analytics for clinical, audit and metric data in clinical trials using real life use cases.