How we drive continuous improvement in our data quality - Truveta

How we drive continuous improvement in our data quality

In our previous blog covering our approach to data quality, we outlined four categories that we use to measure every record on our platform: Completeness, Validity, Timeliness, and Representativeness. In this blog, we’ll discuss a fifth category which is also important to the quality of our data: Conformance.  

When we measure conformance in a health record, we are checking to ensure that source data from our health system members meets baseline criteria needed to process those records in our pipeline. This criteria includes basic, row level data such as patient identifiers and birthdates as well as proper formatting of this data.  

All of our members agree to this baseline quality before joining our network. Once they begin sending data to us, we work closely with members to help them maintain adherence to that baseline. This includes a dashboard which we set up for each health system member.  

The dashboard provides members with real-time data quality metrics for all of their incoming health records. Members can drill down on conformance and timeliness metrics for their data as well as a separate set of metrics taken after our system has normalized records for consistency

We also provide financial incentives to members that meet and exceed their data quality requirements.   

 

Truveta monitors data quality before and after we normalize health records for consistency. These metrics are displayed on a real time dashboard for health system members to help them maintain and support data quality.    

 

This feedback and incentive loop is unique in our industry. Often, health data providers have little control over the quality of the source data they receive (self-reported patient data is a prime example). On that hand, Truveta forms close partnerships with our health system members that drive continuous data quality standards and improvements right at the source. The result is that we’re able to process more records, more efficiently while also maintaining high quality standards.  

And, of course, there’s always room for improvement. While our health system members work on their data quality, we’re driving our own continuous improvement initiatives on our side.  

For example, as the size and diversity of the data in our platform grows, the AI model that we use to drive consistency across that data evolves right along with it. Our growing team of medical-credentialed clinical informaticists are constantly fine tuning the model with new and updated data, including our own additions. The smarter our AI model becomes, the better the accuracy and quality of our data.     

At the same time, our de-identification and security systems are continually evolving so that we can stay ahead of ever-evolving security threats and meet the highest privacy standards. In this way, Truveta offers researchers a complete data package, delivering high-quality data that’s ready to support the requirements of peer-reviewed published research and regulatory filings.  

To learn more about how we’re setting new standards in quality, real-world data for medical research, check out our new whitepaper, Our Approach to Data Quality