Facilitating Data Integrity Compliance With a Rapid Micro Method

The phrase “Data Integrity” was unknown in the QC microbiology lab until recently. Findings with HPLC systems where stored data was reworked to get compliant product or tests were repeated until a “good run” was obtained opened the floodgates of observations and warning letters. The chart shows the trend in warning letters over the last few...

The phrase “Data Integrity” was unknown in the QC microbiology lab until recently. Findings with HPLC systems where stored data was reworked to get compliant product or tests were repeated until a “good run” was obtained opened the floodgates of observations and warning letters. The chart shows the trend in warning letters over the last few years.

Recently auditors have been applying the CFR requirement that all manually recorded information needs to be verified by a second signatory, the so called “4 eyes” observation. Within the micro testing process there are many steps that open the possibility of error or fraud. Sample taking (did the dish go in the air sampler?), did the sample get incubated for the correct time at the correct temperature, did the operator read the plate and record the correct result, did the recorded result get transferred to the batch record?

Most companies are reacting to the “4 eyes” requirements however there are many interpretations of the requirement. Implementations range from 2 analysts reading every plate (then dealing with enumeration differences between the analysts) to random checks to formal checks but at 3 to 6-month intervals. The variation in approaches may lead to closer scrutiny by an auditor as consistency of approach leads to confidence.

The most conservative approach would involve removal of the human variable. While this is very difficult for the sample collection the incubation, enumeration and data transfer to a LIMS database can easily be automated. Automation can verify whether samples have not been taken early so missing sample points can be re-taken rather than wait 5-7 days to find data is missing that is required for batch release.

The Growth Direct™ system facilitates data integrity through the enumeration and reporting phases. Sample worklists can be downloaded from a LIMS platform e.g. MODA for EM or a general system e.g. Labware to the Growth Direct system and collected samples validated against the list on loading for incubation. The automation of the incubation process ensures the correct incubation time and temperatures (including change over for serial Inc) are accurately controlled. Interim plate reading through the incubation and display allow faster reaction times in the event that a sample demonstrates an action or alert level excursion. On completion of the incubation the results can be approved on the Growth Direct prior to printout of reports (in the stand-alone mode) or dropped to the LIMS system for approval electronically on line. In this mode complete security is obtained from start of incubation to formal batch record.

Concerns for Data Integrity are negated by complete isolation of the data and processing from any human access. All data and meta data are in encrypted files with only the final cfu result with associated meta data and audit trail available for review. As the colony detection software has been fully validated as being accurate and reproducible there is no requirement for any human intervention to remove or reclassify any count. For those sites that require an enumeration of molds vs bacteria the Growth Direct software allows the addition of a field to enter the mold count from visual verification of the plate. This addition does not alter the original total count detected by the system.

To learn how Rapid Micro Biosystems can help your company, please contact us here.

To learn more about Growth Direct™, please visit here.

Source: www.rapidmicrobio.com