Menu Back to Poster-Presentations-Details

W-08: Using a Continuous Learning Risk Repository System to Drive Efficiencies in Identification of Clinical Protocol Risk Patterns

Poster Presenter

Thaddeus Urban

Sr. Clinical Data Manager
Syneos Health
United States

Objectives

Share patterns of study protocol risks identified while conducting 161 preliminary risk evaluations (PRE’s), including both pre- and post-award wins across all phases and therapeutic areas using a newly created internal cloud-based, digital, near real-time, risk repository software system

Method

Client-provided materials (RFPs, study outlines, protocols, etc.) were reviewed between Oct 2015 and Nov 2018 to support study team’s understanding of the risks associated with an opportunity and to communicate their awareness to Sponsors. The analysis of these reports from the system is provided.

Results

A total of 161 Preliminary Risk Evaluations (PRE) were completed between Oct 2015 - Nov 2018 across therapeutic areas, e.g., Oncology/Hematology (39.1%), CNS (21.1%), General Medicine (37.3%), and Medical Device and Diagnostics (2.5%). Of those, 57% of PRE’s were performed in support of an RFP/pre-award opportunity and 43% were conducted to assist the study team at the time of award win. Initially risks were identified, categorized and reported out manually. By early 2017, a cloud-based risk repository software system was developed to expedite the risk assessment process with greater efficiency. 6,897 individual protocol specific risks were identified with an average of 42.8 risks per study; 41% from oncology/hematology, 30.6% from general medicine, and 27.2% from CNS. Risks within the repository were systematically organized into three major categories: protocol, planning, and conduct. Within each category there were multiple processes and sub-processes. The Protocol process, including Protocol Strategy and Design, Statistical Analysis Plans, Regulatory Interface, Subjects, (i.e. inclusion/exclusion criteria) and Treatment represented 56.8% of the identified risks while 40.9% of the risks were during the Planning stage. The top three sub-processes in which risks were identified included protocol subjects and treatment (30.0%), protocol design and strategy (22.7%) and project management (12.0%). Based upon the data, the quality of a well-performed study risk assessment was dependent on the type of information and the maturity of the clinical study design as provided by the Sponsor. Specifically, those areas whereby more people (subject selection criteria) and expert-level science (protocol design and strategy) were involved seem to show more likelihood for variability and study risk. In contrast, the areas that were more routine and operational, with less variability, such as project management, recruitment and investigational product handling posed fewer risks.

Conclusion

Conducting a systematic risk assessment or PRE early on in the clinical study program should be the benchmark adopted across the industry. For a single study, the ability to identify potential or perceived risks and provide suggested preventions or controls early on will have a positive impact on subject safety and data quality and have downstream cost and time benefits. Based upon this review, individuals assigned with clinical study risk assessment roles, e.g. project managers, medical directors with design oversight, project leads, etc., should largely focus their risk mitigation efforts on subject selection criteria and study design where the majority of risks seem to occur. Performing individual clinical study risks at an expert-level would typically require a facilitated group-input process. In this review, a curated, digital risk repository system of clinical study risks and preventions from accumulation of expert knowledge and lessons learned was used to help identify risks and preventions within a 24 hour period that could save study teams time and resources. As the risk repository matured in content and users, risk assessment conduct time appeared to decrease as the study team logistics became more efficient and the database of risks and preventions began to accumulate new information with more and more lessons learned from each and every prior study, aka compounded effect of lessons learned. In short, these types of digital continuous learning systems, when applied to clinical studies, can have positive effects on study conduct and subject health and compliance when study risks can be identified earlier and mitigated to reduce downstream burdens. Authors: HESKE S, URBAN T, WIMMER J