Written by Sarah Jamal
Business Intelligence (BI) has become an increasingly popular topic in clinical trials as clinical project managers are expected to make smarter decisions on intelligence derived from clinical trial data. Most systems have some type of analytics capabilities, but do they have the granularity needed to make the best use of clinical and operational data? What should you consider when looking for an analytics tool?
- Harmonization: a true single source of truth that does not require you to join multiple datasets and tie data from multiple studies together;
- Ease-of-use: not all companies can afford to have in-house expertise for mastering an analytics tool, therefore intuitive tools are important to enable faster team adoption and on-going usage;
- Flexibility: no study is the same, which means that static dashboards won’t allow you to explore the true potential of your data. A combination of ‘out-of-the-box’ functionality and flexibility in dashboard customization is optimal.
Data works better together
The industry has been talking about the concept of using a single source of truth for some time, but are you accessing harmonized data from all your studies? When assessing an analytics tool, you should ensure that data from all your studies and all capabilities (e.g., data collection, randomization, trial supplies, third-party data, etc.) resides in the same place, including clinical, operational, and metadata. Data works better together as you can switch from looking to perform cross-study analysis for your full portfolio to analysis of study-specific data, looking at current data to reporting on all historical data, and obtaining meaningful site metrics to evaluate performance across studies. Furthermore, harmonized data is better explored when using a single data model without the need to join multiple data sets, which is often a cumbersome task.
Ease of use, speeds adoption
Time-consuming enablement is increasingly a barrier for adopting a new tool, due to resourcing constraints and staff turnover, with teams needing to adapt quickly. Having an easy-to-use analytics tool is paramount – ‘drag-and-drop’ functionality, easy filtering, sorting and formatting, and intuitive datasets with robust supporting documentation to easily guide users are a must. Ideally, the tool should provide users with a starting point, such as template dashboards, for non-expert users to use and easily adapt to their particular needs. When complexity levels increase, such as the need to derive custom data fields, make sure this is simplified, for example, if there is programming required, basic knowledge should be sufficient. This is typically supported by smart editors that can alert you of syntax errors, the use of predictive text and prebuilt functions that can be easily learnt and deployed, as well as, documentation that provides use cases for real-life scenarios. Sharing data and dashboards must be straightforward, as simple as clicking a button to provide non-system users with a printout of your outputs.
Tailor made, customized to your needs
While it is important that your tool can be easily used by any user, it should also be flexible enough for more experienced users to go deeper. Relevant aspects you should consider include:
- Can outputs be customized, for instance, adding a title and a logo to reference a specific project or sponsor or even using a specific color scheme?
- Can the data be queried, for example to filter a specific datapoint?
- Can outputs be reused and easily modified for different purposes?
If you want to make the best use of your clinical data, consider the benefits of having data harmonized, selecting a user-friendly tool, yet powerful enough to allow for tailor made outputs.
Learn more about Clinical One Analytics.
For more information, visit Oracle Clinical One or contact me if you have additional questions.