Tech, ExplainedHealthtech and life sciences
Part

Chromatography data system (CDS)

Every result with an audit trail regulators accept.

INSIDE: CHROMATOGRAPHY AND MASS SPECTROMETRYHOW IT WORKS · A 12-INJECTION TEST WITH ONE REPROCESSED PEAKThe analyst logsin with her ownuser IDThe CDS runs 12injections andstores the rawdata of eachOn injection 7,she moves apeak's end pointby handThe CDS savesthe change onlyafter she typesa reasonIt keeps boththe original andthe new resultA reviewerchecks theaudit trail andthe raw data,then approvesINSIDE: CHROMATOGRAPHY AND MASS SPECTROMETRYHOW IT WORKS · A 12-INJECTION TEST WITH ONEREPROCESSED PEAKThe analyst logs in withher own user IDThe CDS runs 12injections and stores theraw data of eachOn injection 7,she moves a peak'send point by handThe CDS saves thechange only aftershe types a reasonIt keeps boththe original andthe new resultA reviewer checks theaudit trail and the rawdata, then approvesINSIDE: CHROMATOGRAPHY AND MASS SPECTROMETRY ·A 12-INJECTION TEST WITH ONE REPROCESSED PEAKThe analyst logs in with her own user ID↓The CDS runs 12 injections and stores theraw data of each↓On injection 7, she moves a peak's endpoint by hand↓The CDS saves the change only after shetypes a reason↓It keeps both the originaland the new result↓A reviewer checks the audit trail and theraw data, then approves
Inside chromatography and mass spectrometry, the CDS runs 12 injections, records an analyst's reprocessing of injection 7 with her reason, and lets a reviewer check the audit trail before the result is approved.
opens a page inside this one a page in the same scene a page in another scene a concept page a label only, with no page

A chromatography data system (CDS) is the software that runs a laboratory's chromatographs and records every result, together with an audit trail that a drug regulator will accept. An audit trail is a record of every action taken on the data: who did it, when, and why. The CDS is a part of the setup described in chromatography and mass spectrometry.

The CDS does three jobs. First, it controls the instrument. It sets the run's method, such as the flow of liquid and the temperature. Second, it stores the raw data, the detector's signal over time, exactly as the instrument produced it. Third, it processes that data. It finds each peak and measures its size. This step is called integration.

Integration is where problems can happen. Where exactly does a peak start and end? An analyst can move those points by hand. A small change can make an impure batch look pure. For this reason, regulators read the audit trails of CDS software closely. Inspectors from the US Food and Drug Administration (FDA), who inspect Indian plants that export to the United States, have cited plants for deleted runs and unexplained changes. These problems are called data integrity failures.

Take, for example, a quality-control analyst at a tablet plant in Baddi, Himachal Pradesh, who runs a test of 12 injections. She logs in with her own user ID; shared logins are not allowed. The CDS runs the 12 injections and stores the raw data for each. On injection 7, the software draws the end of a peak in the wrong place. She moves the end point by hand. The CDS makes her type a reason before it saves the change: "baseline drawn into next peak". It keeps the original result and the new one. A reviewer from the quality team then opens the audit trail. He sees the change, the reason and the time, and checks the raw data himself before he approves the result.

The CDS cannot delete a run. A run that was started and stopped is still kept, and the reviewer can see it. These rules come from GxP rules, and the CDS software itself must be proven to follow them through computer system validation.

The approved result goes to the laboratory's LIMS, which decides whether the batch passes. Regulators' findings on data integrity are tracked by regulator inspection intelligence services.

Tech, Explained is a part of ingrid.fyi. Every page explains one piece of technology in plain words.