Perfect Clinical Data is a Myth: 4 Tips for Improved Data Collection

Anna Hrovat-Staedter
May 15th, 2018

Many clinical research organizations believe in order to leverage operational analytics solutions and reports, they have to have “perfect data.” This implies that every piece of the data they collect throughout their operations should be complete, accurate, timely and organized. At Forte, we frequently hear concerns from our community about the ‘state’ of data collected within their CTMS and how that data influences their reports and other integrated systems.

Their concerns are valid, as complete and accurate data not only allows you to report more confidently on your current operations, but it also ensures accuracy across systems when data is shared via integrated solutions. When you are reporting performance and operational metrics to leadership, you want to present the most honest reflection of your current state.  

Though we acknowledge these concerns, the truth is, there’s no such thing as perfect data. While complete, accurate, timely and organized data is a standard to strive for, it’s often very difficult to achieve at large research organizations. Duplicate data entry, user error, miscommunication and lack of collection requirements can all impact data quality and scope and lead to varied levels of data integrity.

However, while “perfect data” may not be achievable, that shouldn’t stop you from evaluating your current data collection processes and working to improve them. With the right tools, processes and workflows, better data is still within reach. Consider these tips for cleaning up existing data and ensure more consistent organizational data practices.

1. Evaluate your current data “footprint”

Before you can address cleanup activities or introduce changes to standard collection procedures across an organization, it is important to first evaluate your current state. What do you currently collect on a per protocol, per management group or per study basis? And what do you do with the data you collect? For example, you may be curious about evaluating your study activation timelines, but don’t consistently track PRMC submission, or IRB submission across all studies. Without this data, you may not be able to identify a bottleneck that is influencing your time to study activation.

The process of cleaning data requires you to not only dive deep into data collection, but also understand what you want to do with the data you collect and why. Once you have identified what you want to do with your data, you can pinpoint what you need to collect in order to achieve that objective, evaluate your current collection process and make the necessary changes either to historic data or future collections to support this goal. 

Learn more about the necessary steps to clean up existing data and ensure more consistent organizational data practices during our free upcoming webinar. Register today! 

2. Identify the scope of your cleanup

Data clean up can feel overwhelming, especially when the problem seems too large to tackle. Narrowing down the scope of the project can make it easier to begin and help prioritize the most critical areas in need of improvement. When determining where to start, ask yourself:

  • What do you want to learn from your data?
  • What do you need to collect to better understand your operations? 
  • How consistently and accurately are collecting that data now?
  • What do you do need to revise your collection, or start new collection processes?

3. Map out the details

Once you’ve identified your scope, the next step is to pinpoint what information you need to begin data cleanup and who can help you get that information. In this step, determine the:

  1. Relevant timeframe: a start and end date for your data collection.
  2. Data fields that need to be completed: where there are gaps in your data.
  3. Key stakeholders to involve: who should be a part of the project.

With this information, you can build the framework for your project and focus on what is needed to meet the necessary data requirements. For example, you may want to evaluate accrual of minority populations across different management groups. In order to ensure tracking across all management groups, you need to involve a leader of each department to ensure their coordinators are up to speed on tracking the information you need. It is also important to inform those who are responsible for tracking and data cleanup about the purpose behind your procedures and requests. If coordinators are more aware of what goal their data entry is serving, they will be more likely to enter complete and accurate data.

4. Don’t wait to implement a solution

When we hear hesitations from research organizations who don’t think their data is ready for a business intelligence solution, we tell them it’s likely their data won’t be ready until they implement a solution. Systems like Forte Insights, our business intelligence system, can help organizations take the first steps toward improving their data integrity by highlighting gaps where cleanup is necessary, giving direction on which cleanup areas will be most impactful and providing purpose for mandated data entry across an institution.

Forte Insights includes data readiness dashboards that give you an overall score for how much you can trust your records and shows where and what to fix ranked by the overall impact that fix will have on your reports. Using these dashboards, you can more easily determine the scope of your data cleanup and help prioritize your efforts.

To learn more about how Forte Insights can help your organization take control of your operational data, and maximize it to guide strategic decision-making, download the Forte Insights Overview today.

And join us for our free upcoming webinar to hear expert presenters address common concerns, provide best practices for building a successful data integrity framework and share their journey to establishing organizational data integrity. Register today! 

No Comments

Leave a Reply

Your email address will not be published. Required fields are marked *