With every passing day, a growing number of our experiences become more data-driven. From apps for personal exercise, to those that manage industry measurements or assess social attitudes, data plays an increasingly important role.
This role is nowhere more keenly understood than in the world of clinical trials, where getting the best therapies to market as fast as possible is a direct result of strong, high-quality data. Sending clean, high-quality, clinical trial data results to regulators enables biopharmas, contract research organizations (CROs), and medical device companies to prove the safety and efficacy of new, life-saving therapies and deliver them to market for the patients who need them.
However, the field of clinical data management is fraught with frustrating difficulties. Issues including weeks of waiting for data access, fragmented data, and siloed systems cause clinical teams to waste valuable time cleaning and reconciling data, instead of analyzing it. The time used to resolve these kinds of trial data issues is also costly, resulting in anything from small delays to catastrophic setbacks that can require repeating an entire trial. As the volume of trials continues to grow, and the variety of data and data sources continue to increase; these data management challenges can only multiply over time.
To gain greater insight into current clinical data management practices and their impact on trial results, Oracle Health Sciences and Pharma Intelligence conducted a new global study* that surveyed 155 biopharma, CRO,and med device clinical data management professionals on their clinical data management concerns and challenges.
The research results provided valuable insight into trial operations and data quality issues, as well as opinions from the survey group on critical trial data management priorities in the next five years.
The following are the four top insights from the survey.
Let’s take a closer look at each of these.
#1) Managing clinical trial data is manual.
A full 87% of respondents said that they used up to 10 different data sources (Chart 1) within a single clinical trial, citing top operational challenges as data completeness (51%), quality (45%) cleaning (43%), and inconsistency (39%) (Chart 2).
“The kind of clinical data quality issues such as those highlighted in this report can have significant negative impacts,” said Julie Barenholtz, Principal Clinical Data Manager, Cytel Inc. “As a data company, we are always looking for ways to improve the quality of the data and process it efficiently so that patients have access to treatments as quickly as possible.”
A total of 95% acknowledged that manual effort was involved in aggregating, cleaning, and transforming the data (Chart 3).
Also, two out of three respondents experience issues with these manual processes. (Chart 4).
#2) Real-time access to trial data is not the norm.
A majority of respondents (62%) do not have real-time access to their clinical data (Chart 5). This can cause a wide variety of problems that can go undetected (Chart 6). However, even for those who have real-time access to their clinical trial data, they still experience issues.The following were cited in the free text comments section:
Respondents named protocol issues, enrollment issues, and skewed trial results as the most common problems that can go undetected without timely access to clinical trial data (Chart 6).
When asked how long it takes to receive clinical trial data from internal and external partners once it has been requested, survey respondents revealed that Electronic Data Capture (EDC) and laboratory data had the fastest access times, while emerging and less standardized data sources, such as mHealth/loT and biomarker data, had the longest delivery times (Chart 7).
#3) Data governance is the #1 issue for regulatory compliance.
A full 81% of respondents cited data governance issues as the biggest challenge with clinical trial data in meeting regulatory compliance. Respondents indicated that duplicate data/inconsistent data, data quality, and data lineage/traceability were the top three challenges (Chart 8).
They also named inconsistent data, missing data, and patients missing visits as the top three most critical problems to catch when looking at clinical trial data (Chart 9).
“Data governance is our top concern because clinical data quality issues can hinder a trial’s completion,” said Melonie Longan, Director, Data Operations, Functional Services, Premier Research.
When asked about the biggest risks to a trial as a result of data governance issues, study respondents cited additional and large data reconciliation initiatives, not being able to prove therapeutic efficacy, and patient replacement during the trial as the leading risks (Chart 10).
In addition, respondents identified trial delays and failing to identify critical patient-risk issues as two of the most critical problems that can result from clinical trial data issues, with audit findings, increased costs, and submission rejections following closely behind (Chart 11).
“Wasting precious time reconciling clinical data issues can be detrimental and costly to our customers,” said Vicki Gashwiler, Associate Director, Strategic Development & Market Access, MedTech Division for Novella Clinical. “Our top concern is proactive data management and data monitoring to reduce the risk of clinical data quality issues slowing down a trial. These delays can have significant financial implications for our customers.”
#4) eSource data is the future.
While it is hard to predict the future, the majority of respondents agreed on where the future challenges lie.They cited the management of mHealth clinical trial data sources as the most important issue to be addressed in the next five years. In addition, finding resources that were capable of managing new clinical trial data and eSources (such as mHealth) was highlighted as the next biggest issue to address (Chart 12).
Combatting Clinical Data Challenges
The researched revealed that delays in accessing and analyzing clinical trial data can seriously affect data quality, create unnecessary difficulties for trial patients (due to missed, real-time signals), and jeopardize the overall value of the study data for regulatory submission.
Here are three ways to combat data management challenges:
Oracle Health Sciences Data Management Workbench (DMW) is an end-to-end, advanced technology solution for aggregating, cleaning, and transforming data. Providing automated workflows and reusable library templates, it accelerates timelines, improves compliance, and provides seamless integration with other trial solutions such as Oracle Health Sciences InForm, Oracle Health Sciences Thesaurus Management System, and external data sources. It supplies study teams with a complete picture of trial data in real-time so that they can make better decisions more quickly, effectively, and easily.
This kind of advanced technology can eliminate these clinical data management challenges. It can remove the need for manual data review and its associated errors, lower trial and patient risk, and provide more reliable, high-quality data results to regulators in order to speed life-saving therapies to the patients who need them.
The survey showed that many industry respondents lack confidence in the quality and completeness of their clinical trial data, from an audit and compliance perspective. The potential for small data issues to undermine expensive, long-term studies is a major problem for the Life Sciences industry. While improving data management may take time and resources, the survey showed the cost of inaction is higher.
Are you struggling with clinical data management challenges?
Contact us for a conversation.
* The survey was conducted by Pharma Intelligence and sponsored by Oracle Health Sciences and ran from the beginning of August 2018 into September 2018. The largest percentage of responses came from clinical researchers, data scientists and clinical programmers from biopharma organizations and a small percentage of medical device companies and Contract Research Organizations. Respondents were from around the globe with 61 percent from North America, 20 percent from Asia Pacific and 17 percent from Europe.
** According to Good Documentation Practices, reflected in ICH E6/R2 and more specifically called out in the recent MHRA GxP Data Integrity Guidance and Definitions data integrity can be defined by attributable, legible, contemporaneous, original, and accurate (ALCOA) actions. Recently, ALCOA added a plus for complete (including all change history of data and its metadata), consistent (data is chronological and protected from unauthorized changes with nothing missed in sequence), available (data is accessible in a usable format to the right people at the right time, including auditors), and enduring (data is available through the appropriate retention period) actions.