Smart Data Quality (SDQ)

Transform Your Clinical Data Management Practices with AI

Accelerate data cleaning, medical coding, and time to query with a transformative AI engine from Saama, developed in collaboration with Pfizer and proven effective in accelerating the pharmaceutical company’s groundbreaking COVID-19 vaccine.

The Smart Data Quality (SDQ) solution “saved us an entire month,” says Demetris Zambas, Vice President and Head of Data Monitoring and Management at Pfizer. “It really has had a significant impact on the first-pass quality of our clinical data and the speed through which we can move things along and make decisions.”

Read more about the SDQ origin story.

How SDQ Changes the Game

Current, rules-driven approaches to clinical data management result in time-consuming reviews of vast amounts of data points, unmanageable query backlogs, inaccurate medical coding, and costly performance bottlenecks.

Saama solves all that with SDQ, by automating and accelerating your data management processes. With SDQ in place, you can instantly answer questions like these:

  • Is a concomitant medication consistent with an AE term?
  • Are duplicate medications given for the same condition?
  • Are related AEs, such as RECURRENT FEVER and UNKNOWN FEVER, of the same toxicity?
  • Are AE terms like DENTAL EXTRACTION linked to non-drug treatments?

Improve Medical Coding Accuracy

Despite the importance of medical coding during clinical trials, accuracy rates traditionally fall between 30-50% for adverse events (AE) and 50-60% for medications. Unclassified items must then go through another time-consuming manual query process for resolution.

SDQ’s medical coding module uses natural language processing (NLP) to auto-code AEs and medications. It can even auto-generate queries for items that can’t be properly coded.

The SCT workflow is extremely intuitive and, for the most part, automatic:

  • Terms from an EDC application, such as Oracle InForm or Medidata Rave, flow into SCT for auto-coding or querying
  • Deep learning models predict coding decisions for each term
  • A human user approves or rejects the proposed coding decisions
  • Approved terms flow back to the coder via import APIs or flat file downloads, creating a single source of all final coding decisions

Manage Data More Effectively with SDQ

Manage Data More Effectively with SDQ

To learn more about SDQ and arrange a demo, contact Saama today.