The FDA said today it inked a cooperative research and development agreement with statistical monitoring company CluePoints to identify clinical trial sites that don’t meet the agency’s standards.

The CluePoints platform is designed to use “statistical algorithms to determine the quality, accuracy and integrity of clinical trial data during and after studies are conducted,” the company said.

The federal safety watchdog said the agreement calls for CluePoints to mine the data for signs  of anomalies indicative of “fraud, misconduct or sloppiness.” The groups will develop and test improvements to CluePoints’ software, aiming to create a list of “anomalous sites” with data that are highly inconsistent with other trial sites.

The agreement will also support the exploration of the “moderators of treatment effect,” the FDA said, in which external factors, such as region or location, can affect the magnitude of a treatment’s effect. The groups said they’ll work to refine the scoring system the FDA uses to identify outlying centers and add exploratory tools to identify moderators of treatment effect.

“Anticipated benefits to the FDA of the CRADA’s data driven approach include the detection of anomalous sites which may have escaped detection previously, rapid turnaround of results, the ability to determine the nature and extent of data anomalies, and the ability to explore the interaction of various factors with data quality,” the FDA said.

The FDA said it hopes the project will improve data quality and inform clinical and statistical reviewers efforts.