iSixSigma

Implementation in Pharmaceuticals Industry

Six Sigma – iSixSigma Forums General Forums Implementation Implementation in Pharmaceuticals Industry

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • #54331

    derya kucukaydin
    Participant

    what is the biggest obstacles to implementing Lean Six Sigma in Pharmaceutical manufacturing?
    Thanks.

    0
    #194850

    Katie Barry
    Keymaster

    @derya99 — Have you looked at iSixSigma’s Implementation and Pharmaceutical sections of content for help?

    https://www.isixsigma.com/implementation/

    https://www.isixsigma.com/industries/pharmaceuticals/

    0
    #194860

    Somewhat difficult question to answer, as there are several sides to the question.

    Generally, pharma raw materials, formulations, and processes are locked-down in early development, with this information driving production plant design and equipment selection (See QbD or Quality by Design). The challenge here is that the formulation labs are a long way from production, drug availability is often limited so batch sizes are constrained, and doing extensive processing investigations may amount to nothing because the ultimate scale changes going to production throw any developmental information into a cocked hat. That said, once the information comes down from Development, and are used in making the pivotal clinical lots, most of the easy options are off the table and will be blocked due to potential FDA filing issues.

    Then there are differences in what kind of production facility you’re talking about. Pharma production facilities have high regulatory constraints against making any changes to raw materials or processes. That said, basic Lean techniques like 5S work well in equipment storage, shift organization, and communications areas. Pfizer has issued papers on improvements in QC laboratory work flows using Lean Six Sigma concepts. (http://www.pharmamanufacturing.com/articles/2011/069.html) So, there are avenues where Lean Six Sigma can be implemented and help. However, if we’re talking production, you need to look at how your facility operates.

    For smaller, multi-product plants, you’re often forced into running relatively large batches due to the “economies of scale” concept. So you’re making fewer (6-12, or less) batches per year of some products, making collection of sufficient data for trending challenging. You’ve also got the regulatory aspect that says “don’t change anything without full validation”, which is significant time and money taken away from any other products using the same equipment. In most production facilities, people don’t want you to significantly change systems that are functioning, even if a little clunky. While you could do some experimentation in a pilot plant, many pharmaceutical production facilities don’t have full pilot-scale plants associated with them, scaling equipment is always tricky and often not reliable, and there’s still the cost issue. Certainly common processes (eg weighing) can be grouped together giving you replicates, but common process aren’t usually where the nasty problems are hiding. Often issues are hidden in compendial raw materials where critical, but unknown, non-compendial characteristics exist. Shifts in vendors, or vendor production methods, now throw your product from one side of the spec to the other and are the devil to track down because these are the materials you only buy 1-2 times per year and nobody can identify what the mystery characteristic is, let alone track it.

    With single product facilities you’ve got dedicated equipment and personnel churning out many lots a year, which is great for data collection and tracking. The challenge is often the facility is laid out to do that one thing in the one way and changing flows isn’t going to happen easily because the potential impact of a screw-up on sales is just too great. For most trending / tracking activities, this type of facility should provide more turn-overs and better data collection potential. Any recommended change that affects the process will still run into the need for extensive experimentation and validation prior to implementation, which means significant cost and risk to a product that may be running in an “acceptable” fashion.

    Well, this is much more of a rant than I was considering originally. Personally, I think the short answer to your question is “use of batch production methods”. In any scheme batch methods throw your ability to monitor, check, or change processes right out the window.

    0
    #200896

    MBBinWI
    Participant

    @katiebarry – curious. This isn’t the post that I clicked on (it was “When you can’t find a way to implement six sigma”), which showed only one post and 1 voice. Tried it 3 times and always had the same result.

    0
    #200897

    Katie Barry
    Keymaster

    @MBBinWI Weird. Same thing is happening to me. This might be above my pay grade tech-wise. I’ll alert the tech geniuses who keep me looking good.

    0
    #200903

    MBBinWI
    Participant

    @katiebarry – seems to be working correctly now.

    0
    #200917

    MBBinWI
    Participant

    @katiebarry – problem returns.

    0
Viewing 7 posts - 1 through 7 (of 7 total)

You must be logged in to reply to this topic.