When we demonstrate Acuity to may users, some ask can they build their own visualization within the system? The thinking is often that this avoids back and forth changes with our data scientists developing the dashboards.

My response to them is, do you write your statistical analysis plan or do you have the study statistician write it? Building data analytics might at first glance seem as easy as creating charts in Excel, something many of us can do.

Acuity in its current release is not designed for self-service analysis. The challenge that Acuity is trying to solve is aggregate multiple different data sources and then present information derived from the data in an intuitive manner that solves the core business problem of identifying the critical data points for effectively managing a clinical trial. It also aims at providing a consistent and user centric design approach to interpreting and analyzing the key performance indicators with the right level of filters, hierarchies and baselines etc.

Let me try to explain the reasons why DIY analysis is not our focus.

Business Use case: The areas that acuity is being used are around well-defined business process where critical data points and process are pre-defined and not exploratory in nature.

  1. Portfolio planning and analysis
  2. Risk based monitoring
  3. Operational and executive report
  4. Data quality oversight.

Data formats – Data behind these reports are not always standardized; as much as we would love to work with data standards, clinical systems and data domains continue to present us with new variables and therapeutic area specific data points. That is the whole of point of research: if all the data are standard then there is no point in testing new hypotheses. When the meta data behind the reports keeps changing then it is not simple to keep exposing the different schema for generating reports.

Audience – the third most important factor to keep in mind is, the system is aimed at end users who are at in the majority not experienced in data analysis and presentation. Many have a totally different job focus where their primary responsibilities are to manage portfolio of trials, manage budgets, maintain relationships with investigative sites and so on. These users also need documented training on how to use dynamic drill down reports and understand patterns and trends seen in data visualizations.

Value chain – The last but not least point we need to understand is that Acuity is not merely a data visualization tool, it is meant to address the whole data value chain where data is converted into insights and then translated to decisions and those decisions are then tracked to actions. It is designed to comply with ICH E6 R2 requirements to keep track of central monitoring outcomes.

Ultimately the challenge isn’t in providing self-service dashboard design, it is understanding that how to aggregate data sources, derive the right information from data sources and present it effectively is a specialized role. The nature of clinical trials as individually designed experiments with unique facets to their data sets ensures that will likely always be the case.