Bika in Production labs
Hi L
> I have been making some test on BIKA. Please don't take as a critic, I am amazingly surprised by BIKA. It's hard to believe it's free and open source.
Always good to hear that!-) And all criticism welcome, users do not realise how much well documented bug reports are appreciated by devs who want to perfect their work soonest. OS matures quickly. Yours are frequently asked and I hope you don't mind me replying via the user group for everybody's benefit.
> Sampling
If you do not need to capture data about sampling, it can be disabled for simplified workflow
> we identify the product ONLY by the batch number.
Bika/Senaite uses Sample Batching to group Samples together where it makes logical sense, to manage them to publication in context, sans the system noise generated by other analysts. Client users may now create batches too.
Remember that for your inhouse production lab, Sample Type should be translated to Product. In production environments, Bika Batches normally contain all the samples and their results for a specific Production Lot, Work Order or Production Run. For the duration of the run, from raw material testing to the final packaged product.
Subgroups in a batch, can be used to track say the alcohol % in wine through Fermentation, Blending, Maturation, Bottling and Off-the-shelf subgroups
> I applied the following logic: The Analysis Specification defines the set of Analysis Services that will be applied, and it's limits.
Nope, when registering Samples, the Analyses, or Tests, requested for them are created from Analysis Services, ASs, the catalogue of testing services your lab offers. T
he Tests, Analyses, can be added 1-by-1 on the Sample registration form, or speeding it up by selecting a previously configured Analysis Profile (Panel or Mask) containing collections of ASs frequently ordered together.
Analysis Specifications provide, in your case, your product specifications, and their valid ranges for each of the test on the Sample. E.g. in government food regulations, there are for 100s of Products Sample Types, and Bika applies only the relevant specification for the Product and tests selected for testing.
> Once I have created the sample (I imagine this to be QC employee in charge of sampling the production lines), I receive it (I imagine this to be the QC technician in our lab receiving it from the other guy).
Labclerks are also authorised for Receiving Samples. You can label them at this point, or provide the production line workstations with label printers already.
> Also, in a pharmaceutical environment, the Analysis Profile should be linked with the product. Under BIKA behaviour, it's possible to accidentally choose the incorrect Analysis Profile, and therefore > analysis the product incorrectly, I mean with the unintended Analysis Profile
mmm, most lab users should pick that up. If you regularly sample the same sample points, you could also use Sampling Rounds, a further layer of automation, improved TAT and reduced human error.
> In a robust workflow the only input data DURING the day to day workflow should be which is the product you are willing to analyse, > the remaining data should come from setup data.
K, that is gap. Logging shortly
> Once I have completed I Save and Send the results (I imagine this to be QC technician loading the results)
Best to interface your instruments bidirectionally to completely eliminate human transcription errors. It will also reduce your TAT significantly, in our experience more than anything else in LIMS.
> I prepublish the results.
Prepublishing provisional results is optional.
You skipped important step, Verification. Authorised users, eg QC supervisors, review the captured results submitted by the Analysts before clients see them, in context of peer and QC outcomes.
Results out-of-range, OOR, are highlighted. The Verifier may then retract results for retesting or verify them. Once a Sample has some verified results, those may be prepublished.
> And then I finally published it. For this I need to go to the Clients –> Client-1 Samples Verified list (this can be the QC supervisor),
QC staff do the critical Verification step, there-after clients already get to see the results online. Publishing itself, posting of the COAs, is done thereafter, but is not as critical and you could assign it to clerks.
> I check the sample I press Publish. I understand workflow finishes here. Is it possible to edit the COA layout?
It is getting easier but still has a bit of a coding threshold. It is probably best to have a professional do it.
> Conclusion: I think the most compromising limitation to consider BIKA in a pharmaceutical environment is the apparent inability to link a > product with an analysis specification.
See above, Bika actually does that, often connected to Sample Types not Product, Here though, https://www.bikalims.org/manual/analysis-specifications
In other labs, you could have more than 1 spec per product, e.g import specifications per country, and only then the user has to pause and select the correct one. In Bika 3.2 we had a default spec feature but it did not make the cut to new Senaite code base
Bests, Lemoene