Chromatography Forum

Hi ,
What is the best steps to eliminate reprocessing in empower ?
Since you cannot unsign and then creates replicate data/another result set ID although with unique ID number what is the best practice and check to do before clicking process

I have seen 4-6 result sets of same sample set with error in processing , incorrect quantification etc,,until final process is correct.. so when it comes to checking and sign offs it is 4-6 times the work

ie ,,,how to get it right first time (checklist to do before clicking process?)

Thanks

A few tips:

1. Don't run too many samples together than are low concentration or can be mistaken/misidentified as noise or baseline peaks. This can lead to exhaustive reviews of sample sets and ensuring every single peak is picked up correctly.
2. Have a second trained person thoroughly check sample sets for correct labels/bracketing standards, amounts, dilutions, sample references etc to minimize mistakes when the run is acquired.
3. Split runs up over several sample sets- if you have 50 samples to run, consider 2 sample sets of 25 samples each.
4. Check project tablespace regularly to avoid running out mid-processing.

If you have that many result sets, then you have a big problem with either analyst training or people not managing their work properly, typical of high pressure throughputs where efficiency is sacrificed for the latest deadline. A written SOP of typical checks before processing should be implemented.

...Also, force commenting on processing. Consider only the highest numbered result set ID when reviewing and be certain that reviewers and data authors are aware that excessive reprocessing is a negative indicator of performance. People who do lots of reprocessing probably do need more training. Avail yourselves of the analytics functions available in your CDS.

I agree with DR. When I started using Empower 10 years ago, the processing method used to give me sleepless nights! I couldn't get my head around the multitude of functions in one method that had so much influence over everything else plus the relational database was a lot to take in i.e. this method and that version and the ID numbers etc etc. It was only through reading up on the literature and googling articles about it that I became more proficient. Im sure Waters have some great training sessions but if you are looking for info on processing methods there is plenty online.

Hi ,

Thanks

Any thoughts or anyone carrying out checks to avoid reprocessing

ie
what should be don't process or report , labelling , naming of standards ,functions correct etc

Is One sample set and one result set 90-95% of the time a realistic target and how to get there ?

Is this a new lab? Surely some internal procedures exist regarding chromatography? Or is it on the fly development work?

As for what to check, id say aim for 90% right first time and set up a new SOP which clarifies your bracketing standard technique (start and end, overlap etc), the wizard will assign the labels and process functions if you set up samples sets there, Then just use this as a template for future runs.

Get sample sets checked before hitting run is critical- are component editor amounts correct? Are dilutions and sampleweights correct? Check that there is a washdown at the end of run etc.

Hi ,

New GMP with limited experience of Empower
My background is many years using Chromeleon, a few years using Chemstation, Ezchrom/Openlab and Empower

All above have pros and cons but Chromeleon head and shoulders above the rest (Ie processing , audit trial and ERES without replication of data results id etc)

So to aim for right first time consider :
Clarify bracket std techniques
Check before processing by second person
Split runs out
Checklist before clicking process
Further training
Sample set templates
Generic labelling

Yes, that checklist is a good starting point and you can then see what works or doesn't work and build on that.
Another thing to consider is the use of custom fields. If your staff are limited in experience, a training course with Waters would be a massive investment in knowledge for the lab and would prevent setting up fiddly and complicated excel spreadsheets when many calculations for assay, related substances etc can be done by coding in a simple formula and validating it before release.