| OHDSI Home | Forums | Wiki | Github |
Hello Everyone,
I know we have two tools such as Achilles and Data Quality Dashboard for assessing the data quality of the transformed data.
However, can I understand from you on what’s the difference between ACHILLES and Data Quality Dashboard?
Both assess the data quality. Am I right?
How is DQD different from ACHILLES?
I see that ACHILLES also has a set of rules similar to DQD.
Can someone help me understand this?
Hi Akshay.
Let me give a simple comparison. I leave it to others to provide a more in-detailed explanation of DQD, because there is much to say about the principles behind that framework.
Achilles is around for a while. It does characterisation of your OMOP dataset (i.e. counting occurrences of events) AND executes a small set of data validation checks (the Achilles Heel component). The characterisation component is the input for the ‘Data Sources’ component in Atlas.
The DataQualityDashboard was released during last years symposium. It only does data validation and does so very rigorously with a big set of checks. This framework is easily extensible with your own checks and the threshold for each check can be customised for your dataset.
At The Hyve, we are switching from using Achilles Heel to using DQD for data validation. We will still use Achilles for data characterisation.
Well said 
There is potential down the line for Heel to be retired in favor of DQD, but that is not imminent.
Removed the post as the issue is resolved
Did we reach a conclusion for Achilles Heel?
btw. very good talk on DQ is here
slides https://www.sentinelinitiative.org/sites/default/files/communications/publications-presentations/Sentinel_Marsolo_Curation.pdf
OHDSI has no “investigative” checks. We seem to be afraid of writing those. Our grade of notification had “mixed reaction”. (and mostly not moved to DQD).