Anyone know for sure if the FDA really REQUIRES HPLC units to have a yearly check-out by a trained engineer, such as the manufacturer's service engineer, to be considered acceptable for drug products assay and stability assays? Or is this something that most do because of paranoia and fear? Obviously, FDA must realize that standards and samples are run at the same time so small variations in flow, mobile phase concentration, injection size, etc., cancel out.

As far as I know, the FDA does not require yearly checkout of HPLC's by a manufacturer's engineer. You do need to be able to prove that your system is working properly (i.e., system suitability tests) and should do routine (yearly is the suggested interval) OQ/PV tests. These can, and I believe should, be completed by the individuals who are using the instrument on a daily basis (of course, I guess they could be considered trained engineers). An SOP is certainly appropriate to document your company's approach to OQ/PV testing and the relevant specifications.

Regards,

Michael Hinsberg
http://www.hinsbarlabs.com

There is nothing which explicitly states that an HPLC must be calibrated, but from reviews of GMP trends (lists "483" observations received by companies from the FDA), investigators are citing more companies for not "calibrating" their HPLCs. This can vary from just particular components, most commonly the detector, to the entire system. In respect of the "paranoia and fear" then you can look at this in a proactive or reactive light - you put the procedure in place under the watchful eye of the FDA after they've given you a 483, or you put the procedures in place before they cite you.

The writer also mentions about getting a "trained engineer" to do the tests. We took the view that we could be more economical in the longer term by getting a few people on site trained up by the OEM to do maintenance and calibration tasks (many OEMs offer this service). We can prove to the FDA that anyone doing maintenance has training traceable back to the OEM. We make up our own calibration solutions/reagents and since we are a cGMP compliant lab, we can trace this preparation back to recognised sources (NAMAS/NIST, etc.) if required. We have used 3rd party companies to do the work (some OEMs will point out that they haven't been trained by them, but I do not know of anyone being cited because of this), which worked well for a year. However, the following year, the number of HPLCs needing the work doubled which greatly increased the cost, so we reverted to on site procedures (one of the other labs on our site do still use the same company, but have paid for about 100 days of engineers time, for C&M tasks and breakdown callouts, which suits them better).