X

The Health Sciences Blog covers the latest trends and advances in life sciences and healthcare.

BI the Way

Craig Morgan
Head of Marketing, Study Startup

The potential of eClinical technologies can be optimized through the use of cloud-based tools for analyzing business intelligence with actionable visual reports that help to identify study bottlenecks and improve performance.

The clinical trials sector is heavily invested in technologies that track how studies unfold. But, putting that information to good use requires turning real-time visibility into actionable data. It is not enough for a sponsor or contract research organization (CRO) to know that one site enrolls quicker or is speedier than others at turning around the study budget. Instead, it is more effective to understand what it takes for all sites to complete those tasks in a timely manner and where the bottlenecks are, so steps can be taken to turn more sites into high achievers. This requires access to critical information that, when quickly spotted in reports, is actionable and allows project managers and other stakeholders to be proactive in making decisions faster and better, based on fact. Putting this data in a useful format requires processes regularly referred to as business intelligence (BI).

Business Transformation

Until recently, use of BI in clinical trials has been far from commonplace. But that is beginning to change, driven largely by a need to revamp how studies are conducted in today’s ultra-competitive global marketplace. In an insightful piece, Kramer et al, acknowledged that while clinical trial technology has become routine, the supporting business model has not evolved alongside it. The continued use of obsolete methods to track study conduct reflects ties to processes shaped by previous generations of paper-based business models for clinical research. Business transformation is a must if the sector is to benefit from what new technologies have to offer. This is where BI comes in.

At a time when it takes an estimated eight months to move from pre-visit through site initiation—and the cost of initiating one site is in the range of $20,000 to $30,000—altering how study information is collected and used is a transformation long overdue. For starters, BI is best when gathered by robust reporting tools that offer visualizations. These visualizations make it easy for sponsors and CROs to hone in on which tasks are behind schedule, which countries are struggling, and which sites are yet to be activated. Not long ago, this reporting capability had been missing from many of the available electronic clinical trial technologies, but newer solutions are now addressing this gap.

Breaking-Down BI

A decidedly tongue-in-cheek article about BI describes a less than positive spin on today’s clinical trial operations, referring to the processes around study execution and analysis as a ‘slavish, box-ticking Sisyphean nightmare’.  And this is before factoring in the problems caused by older, so-called ‘lead-footed tools and technologies.’

Fortunately, newer technology enables BI, which is all about the flow of captured data from the patient to the investigator to management, and then onto analysis. In simple terms, BI is a widely accepted, technology-driven process for analyzing data and presenting actionable information to help make informed business decisions faster. BI—a term first coined in 1989—is meant to optimize processes that support users in identifying and addressing business problems. In its early days, this approach was used to tackle prevalent business problems, such as how to accelerate check-out lines or improve inventory control.

Dashboard Metrics

BI processes are developed from an array of tools, applications, and methodologies that allow organizations to collect data from internal systems and external sources, prepare them for analysis, and create reports, data visualizations, and dashboards to make the resulting analytics usable for stakeholders. In applying this general definition to the clinical trials market, it is important to ask, “With volumes of data being generated, which ones are needed for BI?”

For clinical trial operations, data that is to be collected and analyzed for BI are characterized by the protocol and statistical analysis plan. A BI solution can plug into one or more data sources, such as the clinical trial management system (CTMS) as well as an electronic data capture (EDC), safety, and/or a financial system. The solution then aggregates and summarizes data, then displays key performance indicators usually accessed through a dashboard.

Using this data, it is possible to lay out the roadmap for identifying bottlenecks and other risk factors that may throw the study off course. For trials, the promise offered by BI is in its capacity to foresee those risks, compare results to milestones and, ultimately, reduce cycle time. By comparison, traditional spreadsheets—still widely in use—lack the ability to link critical data, creating a void when it comes to the visualization of data and decision-making.

Start-Up Improvements

Study startup is widely acknowledged as an aspect of clinical trials in dire need of improvement. Bringing a higher level of predictability and quality to this multi-step process is of critical importance to industry stakeholders. In particular, study startup activities—for example, site identification and feasibility, contract and budget negotiations, patient recruitment activities, and managing regulatory documents and drug accountability—have traditionally been handled via laborious manual or siloed processes. Because these tasks have frequently been performed without an organized approach, resulting inefficiencies have led to missed timelines and cost overruns.

However, there have been significant improvements to study startup tied to electronic, real-time document collection and data reporting systems. Stakeholders are starting to align these technologies with critical path project management and resource centralization initiatives in a bid to reduce trial initiation times. Still, much work is needed to implement greater use of reporting tools that bring BI to study startup.

Reporting Tool

Oracle Health Sciences Study Startup solutions permit the collection of study startup data—including information about individual site performance, country performance, and submission activities—and are collected from existing cloud-based solutions such as study startup, CTMS, EDC, and the electronic trial master file (eTMF).

The ensuing reports contain a wealth of data and serve as the basis for analytics that help stakeholders determine study status, as well as identify and resolve bottlenecks. Reports can either be a standard part of the solution or ad hoc. Importantly, they can be shared quickly with members of the clinical team in a simple, secure method that generally entails clicking ‘share’ and providing an e-mail address that authorizes team members to see the dashboard and one of more reports. Information is available at the site, country, or regional level and can reflect individual studies or groups of studies across the portfolio.

Understanding how the reporting tool identifies a bottleneck may involve looking at documents on the critical path, such as site contracts or an informed consent form (ICF). As a study is unfolding, stakeholders can track how long it takes individual sites, as well as countries, to complete those contracts or ICFs. As new sites are initiated, it is also possible to monitor their progress. Is it taking them five weeks on average, or eight weeks? Is this longer than the established benchmark? If a report shows a trend toward longer completion time for contracts, stakeholders can take action to steer sites back on track. Another example illustrating the value of BI is how actionable data that speeds study startup can ultimately accelerate patient recruitment—a long-time bottleneck.

A high-level view of study progress by country and a predictive analytics report appear in Figure 1, showing how reports can be clearly visualized using pie charts, bar charts, or tables. Examples of items that can be tracked, analyzed, and reported might include: cycle times for various activities, length of time since cycle-start activities were completed, individuals responsible for completing tasks, and leading indicators.

Figure 1: Dashboards showing study startup visualizations

Above: This is a high-level executive view of study progress by country with projected work to be completed. The dashboard allows a functional manager and study manager to know the resource needs for the upcoming weeks and months based on work completed to date.

Below: Predictive analytics are key to site selection. This report enables teams to understand the study startup performance of sites and their investigators. It allows for risk management, especially when the contracted dates at one site are 71 days versus the average of 41 days for the remaining sites.

Meaningful Patterns

Improved study startup has the potential for overhauling how large quantities of study-related data are collected, handled, and parsed. If those data are consolidated and flow into a robust reporting tool that allows study team members to aggregate data, view standard reports or create ad hoc ones, and customize data visualizations, it becomes possible to uncover meaningful patterns within the data. This analytical approach defines BI and is a major step forward in ensuring visibility of real-time data sets that help stakeholders to be proactive in identifying bottlenecks and taking action based on facts. The intent is to use the technology to build a competitive edge and deliver promising therapies to patients sooner.

Revised and abridged version of article published in International Clinical Trials, November 2015.

Contact us for the details on strengthening and leveraging business intelligence for your trials.

 

.

Join the discussion

Comments ( 1 )
  • Steven Kenneth Tuesday, July 23, 2019
    Great article. Very informative.
Please enter your name.Please provide a valid email address.Please enter a comment.CAPTCHA challenge response provided was incorrect. Please try again.