Sample Prep definitions

[jennyweck]

Hi!
I am hoping someone can help me.
My company make various skin creams and lotions. Some are drugs so we analyse them using HPLC, GC, AA, etc. A question has come up that we are having some trouble answering.
During a recent FDA audit, it was noticed that we have consucted a couple investigations for out of spec results in which the conclusion was that the sample was not adequately dispersed into the solution (read - additional shaking gave acceptable results). The auditor asked what we did to change our method to ensure we had adequately shaken our samples. Honestly, we did nothing except to be sure we look at them more closely.
The investigations were not on the same product or same analytical method. Some products in their respective solvent yield a milky solution in which the insoluble materials settle out. Some fully disperse into the solvent.
Basically, what I am asking is what do other laboratories use as the definition for shaking and dispersing samples and how do you put into words when you know the shaking is complete?

[Consumer Products Guy]

Jenny - I've made my living assaying product types like yours.

What solvent are you using, and are you allowed to say which analytes you're assaying for?

Pretty much we use a solvent in which our active (API) is very soluble, and hopefully that solvent also dissolves or at least disperses our samples well.

We filter samples at various times of extraction to determine when extraction is complete. Are you manually shaking, using a volumetric flask, shaker, vortex-mix, or what?

[jennyweck]

Consumer Products Guy -
The products we make are mostly diaper rash type creams, hand lotions and shampoos.
One analyte in question is dimethicone in a dimethicone / ZnO cream (oil / water emulsion with various oils and petrolatum). We put the cream in a vol flask, add MIBK and heat to loosen the oils. Then shake by hand to disperse. We allow the ZnO to settle before analysing on AA (ZnO clogs up the nebulizer if we don't settle it out). We have had one lot of product with the issue I described.
For samples we analyse on HPLC, generally, we add sample to vol flask, add methanol, shake (some samples use wrist action shaker, some we shake by hand) filter through 0.45um syringe filter into vial and inject.
For the shampoos, it is pretty easy to see when they are dissolved. They fully dissolve into methanol. We analyse for Triclosan and preservatives.
For creams and lotions on HPLC, there are some ingredients that are not soluble in methanol. They are typically the oils that make up the emulsion. Our products are mostly water so generally not an issue. The one we have had the OOS issue with is a cream made largely of glycerin, water, dimethicone and a few other ingredients. Glycerin is not soluble in methanol so it basically floats in the solution. We analyse the preservatives in this product via HPLC.
All the analytes are readily soluble in the solvent for analysis.

[Miss_Calculates]

The thing with the FDA is they want some measurable outcome to your investigation done to the procedure so they know you successfully identified the root cause. Leaving the procedure the same when you "found" the original shaking instructions were inadequate doesnt fly with them. They want to know you completed a preventative action so more out of specification results dont happen later.

Best bet is to revise your procedure for a shakeing window or a minimum say "shake 5-10 minutes at XXX rpm on XXX equipment....or shake for a minimum of 5 minutes at blah blah blah" Then test a sample at appropriate shaking intervals and assess the difference between the results to be "statistically negligible" (I'm totally guessing at shake times here, do whatever you want). This can be some kind of notebook study and included in the method development report; or if its a GMP method, in the out of specification report.... then when you are training people, just let them know that if it looks bad at 5 minutes, they can add another 5 minutes.

A narrower window "looks better" if you go for a window, but as long as you test it out, you can do as wide of a window as you want. We had a shake step on some samples that lasted 1-72 hours... tested both ends on a few different samples (and some points between) and wrote it into a method.

Another out of specification preventative action can be "more training of individuals that did the testing". This can be documented by a group training session about the samples.

You just have to show something documented where you instituted a preventative action to the OOS results.

[jennyweck]

For those that have viewed the topic...
The Quality Engineer wants to use the term 'vertical shake by hand' to describe the shaking process for samples that we do not use a shaker or centrifuge on.
Has anyone ever heard of this term before? I do not want to add it to the procedure because I have never heard the term before and I think it would cause way more confusion that we don't need.

[paulw]

It is tricky to write a procedure that insures the sample is completely dispersed with parameters to insure so. Time is not good because some people may shake more vigorous than others etc.

I would write the procedure to say 'Shake until sample is completely dispersed', or some such. Then when an issue arose, and it was found that the sample was insufficiently dispersed, I would write in the investigation that the root cause would be sample prep, and that the corrective action would be to retrain the analyst on the sample prep. Document the training and be done with it.

Like Consumer Products Guy, I did this type of testing for a number of years. Lotions, shampoos, sunscreens and even a hateful diaper rash ointment. It was my experience that after a number of times assaying a sample, I could tell when it was appropriately dispersed, and from being second analyst for a couple of situations like you described, I could easily tell when the previous analyst had not taken the care needed just by looking at their sample prep.

Since you mentioned some oil/water emulsion type products, this is very similar to the formulation of sunscreens. For those we had typically used an 85%MeOH/15%H2O, with 1mL acetic acid per liter, and it seemed to do the dispersing and extraction just fine. I remember doing the typical gambit of sunscreen actives like that (octinoxate, octisalate, avobenzone, oxybenzone, homosalate, octocrylene) as well as parabens and benzyl alcohol. Perhaps just adding a fraction of water to bring in the glycerin would help you out on that one in particular.

[michaelbarnes42]

For those that have viewed the topic...
The Quality Engineer wants to use the term 'vertical shake by hand' to describe the shaking process for samples that we do not use a shaker or centrifuge on.
Has anyone ever heard of this term before? I do not want to add it to the procedure because I have never heard the term before and I think it would cause way more confusion that we don't need.

I don't like the term, "Vertical shaking by hand" becuase it doesn't accuratly describe shaking by hand and it can be too constraining, based on interpretation.

[krickos]

I don't like the term, "Vertical shaking by hand" becuase it doesn't accuratly describe shaking by hand and it can be too constraining, based on interpretation.

Neither do it. In this case it is better to put it plain and clear for instance:
dissolve by turning the flask upside down x times, if not visibly dissolved after x turns , add z ml more of Y, then turn updside another 10 times, when dissolved dilute to volume with Y.