The Risk Manager's Dilemma

The Salt Curve

2012-02-23T09:39:00.002-06:00

I recently stumbled across a soap crafting blog (link here) that reminded me of a dandruff shampoo that I used to work with years ago. The product had coal tar as the active ingredient. Cleaning shampoo out of a small process tank (this was R&D afterall) was a chore because it continuously created suds during rinsing. We used to add salt to create a gel so we could scrape it out. Additionally, I have seen salt added to older generation hand creams to adjust viscosity. Those systems ...
were typically triethanolamine / stearic acid (soap based) creams. Here's how it works.
The salt curve is adding salt (sodium chloride) - an electrolyte - to an anionic / nonionic surfactant containing system to create a gel. This works in the following way.
Anionic/nonionic surfactants provide a specific charge density on the their outer surface. Addition of sodium ion lowers the charge density of the micelle surface and increases the density of micelles. This makes them able to become closely packed together and creates a thicker solution. Too much ion causes a higher charge density and the micelles will repel each other. The result is a thinner solution.
Just like in our diet, too much salt is bad. The system viscosity falls off a cliff above 3.0%.

Another useful link here: Chemists Corner

Ergotamine...Not Just Your Mother's LSD

2012-02-11T08:26:00.000-06:00

Image via Wikipedia

Sandoz was an early pioneer in drug discovery. It made enormous strides working with galenical or plant derived drugs. A particular milestone in the history of drug discovery was the benefits of ergot. Ergot or ergot fungi refers to a group of fungi of the genus Claviceps. Most of the ergotamine work occured early in the twentieth century starting in 1906 with the discovery of the inhibition of the pressor effect of adrenaline. It's many derivative compounds have had an impact on blood pressure therapy and migraine headache relief. Some years later, one of the more interesting derivative ergot amine discoveries was the now recreational drug LSD.
LSD was a serendipitous discovery ...
by Albert Hoffman. Hofmann first synthesized LSD on November 16, 1938 but set it aside for five years. Hoffman revived research into the drug years later and fell victim to its effects quite by accident. He accidentally absorbed a minute amount through his fingertips and started tripping. He described his experience this way: " ... affected by a remarkable restlessness, combined with a slight dizziness. At home I lay down and sank into a not unpleasant intoxicated-like condition, characterized by an extremely stimulated imagination. In a dreamlike state, with eyes closed I found the daylight to be unpleasantly glaring, I perceived an uninterrupted stream of fantastic pictures, extraordinary shapes with intense, kaleidoscopic play of colors. After some two hours this condition faded away."

The drug was eventually mass produced and marketed by Sandoz under the trade name Delysid. However, by the 1960's the abuse of the drug by the counterculture created such a backlash that the substance was banned. Hoffman blamed Timothy Leary an American psychologist and a strong advocate for its non-medical use for derailing what he thought was a promising psycho-active drug although admitting its potential for abuse.

Small Pox Vaccine and Cow Pus

2012-02-02T21:41:00.000-06:00

Image via Wikipedia

I suppose that most people never give much thought as to how smallpox vaccine is made. It's starting material is cow pus. Pretty disgusting but it is true. Wikipedia gives a pretty good history regarding the development of the vaccine. Here it is: Small Pox article. A former colleague described to me the process. She said that the side of a cow was scraped and then infected with cowpox. The disease was allowed to spread through the scraped area. The pus is removed ...
from the infected animal and eventually makes its way into becoming smallpox vaccine. I assume the animal was euthanized afterward. Pretty sad for the cow.
Some female hormone medications have a similarly disgusting start: the urine of pregnant mares. The urine is spray-dried and added to tablet preparations. The mares have catheters throughout their pregnancy. Pretty sad for the mare.
It is pretty obvious that not all drug synthesis has a specifically chemical origin. Sometimes they are biologically derived. They are pretty disgusting too.

Related articles

The Inverse Rule of Dispensing Medicine: Oversupply Drives Up Prices

2012-01-16T18:42:00.001-06:00

Image via Wikipedia

I often tell my colleagues that pharmaceutical pricing and medical care in general do not follow the traditional rules of supply side economics. It is the threat of government regulation keeps prices in check. I am then accused of supporting socialized medicine and subscribing to a traditional liberal piety for subsidized healthcare. Hear my argument first before you pass judgement.
Let me compare Canada and Boston ...
as an example. There are far more MRI machines in Boston than there are in Canada. Boston has much higher healthcare costs compared to Canada. Canada has better quality health outcomes. Why is that? The oversupply of MRI machines in Boston leads to greater access to the machines. Health insurance insulates the customer from the true costs of the this diagnostic tool. Also, the Boston consumer tends to be highly educated and demands this level of care so it is prescribed more often. In this example the patient outcome is no different than Canada. This is because true need for MRI is always met in Canada. The MRI is often overused in Boston or the US in general for simple sprains or other injuries that could be easily tracked by a healthcare professional using traditional methods. Keep in mind that any diagnostic tool presents risk such as being exposed to radiation. Indiscriminate use of these machines may raise other risks in the patient's future.
Here is an example of my other point regarding pricing. Imagine that you have a life-threatening illness and a single tablet can cure you. What would pay for that tablet? Would you sell everything you own or otherwise risk dying? I suspect that there is little upper limit on pricing that could be charged unless companies were concerned governments would step in and control prices.

Which is exactly what happens in Canada.

Emulsions Formulation Overview

2012-01-15T22:03:00.000-06:00

Emulsions Formulation Overview

View more presentations from jimmmcelroy510.
Most of my working career has been focused on the development of liquids and semi-solid formulations. I have written this presentation as a primer for those folks who have a mostly solids background in pharmaceutical formulation.

Who Has Cooties?

2012-01-09T20:04:00.000-06:00

Image by Jess Beemouse via Flickr

My daughter works as a hairdresser. The other day she had a small child as client. Who had just been successfully treated for head lice. The little girl’s parents wanted her hair cut in order to manage her hair better. This reminded me of a product that never made it to the market place.
Standard treatment for head lice (pediculosis capitis) includes application of pyrethrin or permethrin insecticides to the scalp. The new formulation took a completely different approach from using pesticides. It used heat in the form of an exothermic reaction. This required an interesting balance of creating enough heat to kill the parasite but not burn the patients scalp. I am not sure if they succeeded in this. I have not seen it marketed but I liked the out of the box thinking on that project. Ideally nobody likes applying a pesticids to one's scalp. Killing cooties with an exothermic foam seems like a great idea.

Tastes Like Pool Water

2011-12-26T09:05:00.000-06:00

Image via Wikipedia

As a process scientist you are occasionally asked to scale-up mediocre formulas and re-purpose equipment. I recall a story from a colleague from a former company. It was about a poorly preserved suspension and stainless steel tubing that was converted to chlorinating USP Water.

Here is the background.
He had a liquid suspension product that because of its very high pH it was difficult to preserve with traditional preservatives such as parabens. Parabens are not stable at high pH. The formula problem was solved by spiking the suspension with chloramine to ensure that the product was sterilized prior to bottling. There were two downsides: the product tasted like pool water and he ended up corroding a perfectly fine stainless steel tubing.
I was amazed that consumers purchased this terrible tasting product but I was even more amazed that the process scientist did investigate whether stainless steel was compatible with chlorine. It is not. To me this is an example of poor execution at several levels. First, there clearly is a failure to properly investigate and fix a formula problem before it was scaled up and commercialized. Second this led to another poor decision regarding the chloramine spiking of USP Water in stainless steel. Information regarding stainless steel and chlorine compatibility is readily available from many sources. This is a frustrating lack of due diligence.
The problems were eventually fixed by using an adequate and better tasting preservative thus eliminating the need for chloramine. This occurred after many product and equipment investigations and significant downtime. A little more homework up front could have saved a lot grief down the road not to mention that it would have saved money and provided a better tasting product.

Are You Sponge Worthy?

2011-12-21T12:30:00.000-06:00

Image via WikipediaAre You Sponge Worthy? I remember this episode (The Sponge) of Seinfeld very well. I worked briefly on the contraceptive sponge back in the early-mid 90's. It was owned by a different company then. That company took it off the market due to production issues. Hence the shortage and Elaine Benes concern about using up her sponge stash.
I thought this was a ...

remarkable product in two ways. Iit was one of the few contraceptive products available over the counter. The available options are largely spermicidal gels and condoms. The sponge has effectiveness rates of 77% to 91%.It tends to be more effective in women who have not had children previously. The sponge works by blocking the cervix and releasing a spermicide (nonoxynol-9). Women who had children tend to have a more dilated cervix and the effectiveness is reduced.
Secondly and more relevant to this blog entry is how the sponge is made. The contraceptive sponge is constructed of soft polyurethane foam containing 1000 mg of nonoxynol-9 a surfactant. I guess sperm is easy to kill.
The polyurethane foam is formed by reacting a diisocyanate (OCN-R-NCO) with a diol (HO-R-OH). The first step of this reaction results in the chemical linking of the two molecules leaving a reactive alcohol (OH) on one side and a reactive isocyanate (NCO) on the other. These groups react further with other monomers to form a larger, longer molecule.
Water (a blowing agent) reacts with the isocyanate to create carbon dioxide gas, which fills and expands cells created during the mixing process. The interesting thing to note here is that all of the active ingredients can be added to the water phase.
There are three steps to the reaction:

1. A water molecule reacts with an isocyanate group to form a carbamic acid. Since carbamic acids are unstable they decompose forming carbon dioxide and an amine.

2. The amine reacts with more isocyanate to give a substituted urea. Water has a very low molecular weight so even though the weight percent of water may be small, the molar proportion of water may be high and considerable amounts of urea produced.

3. The urea is not very soluble in the reaction mixture and tends to form separate "hard segment" phases consisting mostly of polyurea. The concentration and organization of these polyurea phases have an impact on the properties of the polyurethane foam.

It is an interesting product with an interesting technology.

Sipping The Cherry Kool-Aid

2011-12-11T09:11:00.001-06:00

Image via Wikipedia

Solving problems often occurs in the context of needing to get a solution that is inspired, inexpensive and gets you back into business quickly. So why is it that most solutions are uninspired, expensive and take too much time to solve? I think the problem is that most engineers find solutions to problems in very limited ways.
There are many problem solving tools that help engineers solve problems but the hardest thing for most anyone to do is to think outside of what they know and solve problems in a unique and insightful way. Most of us are tweakers rather than innovators.

To illustrate this, I have an extreme example of world view bias. There are many people who believe in alien visitation to earth. I recall reading an article about scientists finding a pre-Columbian skull in a burial mound that was severely deformed. The author of the article posited that this skull was overwhelming proof that this was the remains of an alien from outer space. The author did not research (or perhaps care) to further discover that that the deformation of skulls was a fairly common practice among these ancient peoples. That would violate the world view of alien visitation to earth.
Industry problems are similar. Pharma rarely hire scientists from other industries. Most hiring managers look for all kinds of unneeded GMP knowledge or specific dosage form experience. Those things can be learned on the job. The end result of hiring within this small clique is group think or what I like to refer to as "sipping the cherry Kool-Aid." If everybody is drinking Cherry Kool-Aid how will we ever know if Lemon-Lime flavor might be better. Problem solving requires working on the right problem with the right mix of people and not trying to solve it by by limiting the context of the problem to exactly the stuff that is front of us.
This brings me to the Apollo 13 movie. There is a classic scene of trying to fit a square filter into a round filter hole. The engineer spills all of the stuff that is inside the capsule onto a conference table and says this what we have to fix this. The problem is very straightforward. There is limited set of objects because it is a very small spaceship. It is a very specific task. So they get to work and problem solved. In reality, there are still systemic flaws to the spaceship design. Why are there two different filtering systems?
I have found that industry treats problems similarly. Often they only fix the problem at hand and rarely move forward with tackling the systemic issues. It requires a dedicated management team to provide time and resources needed to solve the underlying issues.

Is That Broken Glass In My Topical Cream?

2011-12-06T22:05:00.000-06:00

Image via Wikipedia
Some history here. I used to manage all of the personal care products for the site where I was located. I remember a particular consumer complaint about finding broken glass in the product. I was at a complete loss since we did not have any glass anywhere near the processing areas. Paradoxically I was also getting low assays for one of the active ingredients: Pramoxine.Hcl a topical anaesthetic. When it rains it pours...

This is the first time we started making this product at the site and I was the main person who developed it. Needless to say I was committed to getting this fixed. When I examined the returned product it sure did look and feel like very small bits of glass. So I had the lab examine it and tell me what it was. It turns out it was Polaxamer 188 and Pramoxine.Hcl. Polaxamer 188 is a surfactant. The two were combining to form crystals that looked and felt like broken glass. This was a new complaint in a product with a long history. Clearly it was something I did when I transferred the process.
The product in question was a market leading antibiotic cream. It was what is known as a harvest brand that was bought and transferred, to the manufacturing site where I worked. I transferred the process exactly or so I thought. My investigation uncovered a subtle but important difference.
The original process combined the active and surfactant in a "spaghetti pot." It was mixed by hand using a spatula. I used an electric mixer with a high shear impeller. That extra energy formed the crystals. I quickly returned the process back to "hand" mixing.
The low assay for Pramoxine.Hcl was related. It turns out that the lab used a canula to draw product into a syringe prior to adding into the a flask. The canula opening was so small that the Pramoxine/Polaxamer co-crystals did not pass into the syringe thus resulting in a low assay result. Even small seemingly insignificant changes to the process can have interesting consequences.

Document Management and The Theory of Everything

2011-11-24T12:12:00.001-06:00

Image via Wikipedia

It is a vexing industry problem to develop an easy way to do document development and document management. A lot of us rely on brute force methods of writing, tagging, executing and storing documents. What I mean is that even with software the problem is still very difficult and labor intensive to manage.
In the pharma world we have to comply with the rules and regulations with regard to CFR 21 Part 11. The predicate rule definitely applies to GMP documents. That means the software has to be able to properly secure, audit trail and maintain document integrity. If there are digital signatures additional measures have to be taken. Additionally, the needs of document management require metadata and useful connections between documents and other enterprise systems such as laboratory management software. Right now most software would not be backward compatible (or at least it would not be easy) to existing document management systems. The systems rely on the user to make the connections and supply the correct metadata. All of them are inadequate. Here is why.

A validation document has a relationship to the plant floor as well as other documents. Let's explore what those relationships are so that I can highlight the problems. I will use a process validation as a case study. A good process document like the process validation is built upon a foundation of other activities and other documents. For example:

training
preventative maintenance/autonomous maintenance
calibration
procedures
equipment train
raw materials (includes packaging)
testing (raw and finished)
method validation
cleaning validation
filing (country to be sold)
packaging validation

This is not a comprehensive list but it demonstrates that myriad of relationships with regard to documents and lists. It is up to the developer of the document to connect the dots by either creating a electronic relationship, an attachment to a paper document or some hybrid of both.
An ideal system would use minimal inputs such as product name and immediately start to create the relationships and prompts needed to tie the document together.
Sharepoint 2010 is starting to lead the way. This product can be the front end of enterprise systems such as a LIMS, SAP, Documentum, Trackwise, etc. A well designed software can automatically make the relationship connections. Document development and storage must be simple and easily defined for the end user. It just requires a well thought out relationship plan with minimal manual meta data inputs. The developer needs to keep the end user in mind and to think big. There are lot of enterprise softwares out there that are scattered and standalone. It is time to connect the dots. When the relationships are defined and the meta data is entered the theory of everything with regard to documents is complete and comprehensive.

The Predicate Rule and CFR 21 Part 11

2011-11-15T21:58:00.000-06:00

Image via Wikipedia

I recall when CFR21 Part 11 was issued in the mid 90's. It caused considerable hand wringing in the pharmaceutical industry. It was immediately considered excessive and difficult to implement by just about every technical manager that I knew.
Part 11 defines the criteria under which electronic records and electronic signatures are considered equivalent to a paper record. The integrity of the record requires adequate security and audit trails. This is fundamental and not disputed by anyone. The problem lies in the fact...
that it was commonly believed that every computer system was subject to the mandate of Part 11. This was regardless of whether or not data was generated, used or stored. This interpretation slowed what could have been a very productive decade of software upgrades and control systems.
I am assuming the reader is already familiar with the rules regarding electronic records and I will not go into it in any detail here. If you are not here is the guidance: CFR 21 Part 11.
In 2007, the FDA recognized that pharma was lagging in implementing better controls and computerized systems that could have significant benefits to product quality. The FDA recognized the problem and has indicated that the predicate rule applies. Simply put the predicate rule is the underlying FDA regulations of GLPs, GMPs, and GCPs that govern activities in the life sciences or pharmaceutical industry. For example, the regulations require documentation is needed for maintaining records for product labeling for at least one year after expiry as per CFR 211.180 (b). If your company maintains these documents electronically then Part 11 applies.
As a contrary example I may ask "What if you have a computer system that keeps track of corrugate inserts used for shipping boxes of your product?" Let us assume that the information is for the company to keep track of inventory with regard to corrugate inserts. Does Part 11 apply? There is no regulatory requirement with regard to tracking the amount of corrugate inserts used in tertiary packaging. It is simply the company trying to manage inventory and not GMP obligations.
Like most things in the pharma industry; interpretation is key. It is important to fundamentally understand the intent of the law and implement the regulation mandates intelligently. Otherwise, time and money will be wasted on implementing unnecessary controls on your data.

Don't Panic! It's Only A Paper Tiger.

2011-11-14T22:28:00.000-06:00

Image by Pop!Tech via FlickrEarly in my career, I once had a friend remark that without government regulations industry "would pee in a jar and call it lemonade." I thought the statement was hilarious if not over the top. A few years later I ended up working for two companies that were under consent decree for poor quality control. They added hundreds of qaulity engineers and lab personnel to deal with the terms of the consent. So much for hyperbole.
The question is: how does this happen?
These companies have large quality control organizations, competent technical staffs and diligent business managers. I am sure they are all committed to their jobs and are passionate about doing good work. I am even sure that most of these folks are an ethical bunch. I know. I have worked with many of these people.
It is easy to think that mismanagement lead these companies over the precipice with regards to quality. Malcom Gladwell's article The Risk Pool might suggest otherwise. He posits that seemingly good decisions in the present can have devastating implications in the future. Companies choose to respond to a regulatory environment by creating a verifiable paper or electronic trail. Have you heard the following: "If it is not documented it did not happen ?" I have heard it quite a lot over the years. I think it is the stupidest comment ever made and to my mind has had dire consequences for industry. I agree that documentation is important in that it ensures a due diligence and at least some record of activities performed. A good document trail is a like a pilot's checklist. The cockpit of an airplane is very complicated and errors can be fatal. A checklist is essential that nothing gets missed.
The problem is that quality organizations and even the health authorities often treat documents as a means to an end. At times I feel like I work in two factories: one that creates products for consumers and one that creates paper to feed the quality organization and ultimately the health authorities.
Why insist on paper for even the most trivial of items? I have come across a single packaging shop order that had over 600 initials and signatures. I once asked the site quality director if he really thinks a shop order needs 600+ signatures and initials. The response was "..if that is what it takes to achieve quality then so be it."
Wow!
Lately, industry is recognizing that a lot of the paper we create is not value added. ISPE has issued an excellent white paper on the future of equipment and facility validation. This paper highlights a risk based approach to qualification and documentation. It makes the creation of paper a necessary outcome of activity. A future with good quality systems looks a little brighter if we are willing to change our mindset about paper and shift it to culture and quality related activities.

Bag Tricks For Toxic Compounds

2011-11-07T19:47:00.002-06:00

Several years back I was involved with a project that required me to work with very toxic powdered compounds. Inhalation of these toxic compounds could cause rapid death if handled improperly. I used to work in the paint industry in the mid 1980's but this was a whole other level of dangerous.

The process we were developing required the handling of these compounds from weighing to mixing to dispensing and final compounding into tablets. We had built temporary enclosures...
or isolators out of rolled sheets of high density polyethylene (HDPE) and added glove ports.

We hired a consulting firm to advise us on how to handle such toxic materials. The solution they provided was simple and interesting. They called it "bag tricks."

The handling process is simple. It works like this:

ensure that the bag is tied tightly with an electrical tie very close to the area containing the powder.
attach the open end to the fitting where the material will dispensed and seal it with a bag tie.
cut the tie that seals the powder side of the bag and dispense
when the dispensing is complete place two bag ties near the dispensing area. keep the ties about two inches apart
cut the bag between the two ties thus isolating the dispensing area and the remainder of the bag
tape the two florets where the cut was made

Appropriate PPE such as respirator and safety glasses should be worn during this process.
The interesting thing about using bag tricks is that it can accommodate any number of powder processes. I like it because it is low tech and easy to implement. It can keep dust to a minimum and thus reduce or eliminate cross contamination.

Suspensions Formulation Overview

2011-11-07T08:47:00.000-06:00

Image via CrunchBaseHere are some of my notes regarding suspensions. This presentation resides at Slideshare. Feel free to download and use it for your personal needs.

Suspensions Formulation Overview

View more presentations from jimmmcelroy510.

Supplier Relationship Management In Pharma

2011-10-27T20:23:00.000-05:00

Image via WikipediaHere is an interesting question: What does supplier relationship management (SRM) in the pharmaceutical world look like?
Answer: Not much.
Many pharma companies are struggling to implement lean manufacturing models in their production areas in order to streamline their processes and reduce costs. This is a necessary effort to maintain profitability but a very difficult one to implement in this industry. There are no industry wide standards and companies are often left on their own to solve supplier issues...
and establish mutually beneficial relationships.
In order to implement any lean initiative it is essential that the supply chain is managed extremely well. Supplier relationships are essential. The challenge is that there are many aspects to the pharmaceutical business enterprise do not exist in other industries. For example, the FDA and other global health authorities regulate drug approval processes that can take years, there is a lack of homogeneity across the globe with regards to regulatory filing needs, quota restrictions on ingredients (i.e. controlled substances), regulatory mandates on managing GMPs that may conflict with lean principles. I have often said that we manufacture two different types of products for two different customers: drugs for patients and documentation for the health authorities.
The supply chain can take many forms. They are:

raw materials such as APIs, excipients, packaging components
intermediates such as granulation that need further processing
finished goods for repackaging

The supply chain can also include many other aspects of the business as well. For example:

3rd party manufacturing
offsite stability testing
offsite testing laboratory
R&D
clinic

Industry wide standards would be very useful and beneficial. As I often stated pharma often has internal initiatives and shares little with its competitors. Toyota learned early that standards across the industry would benefit its own bottom line. Pharma would be well advised to do the same.

The Case Of The Missing API

2011-10-20T10:00:00.003-05:00

Image via WikipediaIn theoretical physics, if mass is missing from the equations one can create "dark matter." If there are not enough dimensions to solve a problem just add more. I think "super string" theory has eleven dimensions or so. In industry, missing mass can be a real problem. There are no options on dimensions other than the usual three. If the missing mass is a an active pharmaceutical ingredient (API) then that is a significant problem.
I remember working on a major brand anti-fungal aerosol spray...
and it always assayed low for API. The API was generally in specification but low. Then it failed assay. It was a very vexing problem.
I witnessed the production of the product and watched the operators weigh, dispense and mix the all of the ingredients. I watched the filling operation. No obvious problems. So where was the API? There was some speculation that the alcohol vehicle was flashing off. I could find no evidence of it. I concluded that it had to be there and the problem had to be the method assay.
The method was part of the USP monograph so I had a difficult time challenging it with my analytical lab colleagues. I argued that if the method is not the culprit we are violating the conservation of mass: a basic tenant of physics. I spent a great deal of time researching the release data. Years and years of data demonstrated that we had accuracy but not precision. It turns out that the laboratory investigated further and found the method to be at issue.
In my many years of process development I have always found that a classic struggle between the lab and the process guys as to who was at fault for out of specification data. In my experience it usually is the process and rarely the lab. Mass can get lost due to the process. However, if the process proves to be sound beyond a doubt then either physics is wrong or the assay method needs to fixed.

Shake Well

2011-10-10T10:30:00.001-05:00

Image via WikipediaI have indicated in my previous post that scaling up sunblock emulsions can be a challenge. Traditional surfactant systems are less than ideal because formulators learned that traditional emulsifiers allow the insoluble UV absorbers to be solubilized when wet. There goes the "waterproof" claim. Functional polymers such as Pemulen TR-1 and TR-2 are often used instead. In general, traditional formulas are sensitive to heat, microbiological contamination and transport.
We often joked...
that there were only three basic formulas that every marketed formula was based on. However, the basic science of the formulation can get lost when manipulating the base formulas. I have an excellent example of this.
One of the first projects I worked on when I joined a major sunblock manufacturer was a SPF 25 product intended for women. The package was purse sized and elegant looking. The formula had a previous life as a different brand for this company. Coloring, fragrance and waterproofing were refreshed for this formula. There were other small tweaks as well.
I initially made one pilot lab batch for stability. The formula was easy to make and there were no issues. It did use a traditional emulsifying system and not Pemulen. Two months went by and the formula began to separate on stability. I assumed that it was back to the drawing board and the launch would be scrapped. I was wrong. A decision was made to put "shake well" on the bottle and proceed. I could not believe it. This was not salad dressing after all. I was told to be quiet and to proceed.
I scaled this formula up from bench size to 8,000 gallons. We manufactured the product at a 3rd party site in Michigan. We were scheduled to make launch quantities of thirty batches. Unfortunately, the first batch of the validation sequence fell apart while transporting to the holding tanks. As I stated before that this was due to a poor surfactant system. I had to save the day by keeping this formula intact long enough to be filled in bottles. I succeeded in doing so and the launch quantities were made. We sold only 10,000 bottles nationwide. Millions of bottles had to be discarded.
Ironically, I was rewarded for meeting the launch timing.

It's 3 In The Morning And There Goes My Career

2011-09-29T21:25:00.001-05:00

Image via Wikipedia
Sunscreens or sunblock products are an interesting subset of the over the counter pharmaceutical dosage form. They are part of the USP monograph but really do not provide a therapeutic dose. They function to mitigate the effects of ultraviolet radiation. Wikipedia has a great entry on this.
Some background on the sun care business. As you may have guessed sun care cosmetics are a seasonal business and sun block certainly falls into this category. The sun care business has great deal of competition and women are almost exclusively the purchasers of these products. Our marketing group was constantly trying to create "new news" in this category. This resulted in several new product launches a year. This can be quite a challenge for a process scientist.
I was working on a major brand sunscreen...
that had an interesting problem. The emulsion flipped during the cool down portion of the processing cycle. The emulsion I was working on had several interesting characteristics in that it could reflect, deflect and absorb ultraviolet radiation. More importantly as a development scientist was that it was an "oil in water" emulsion. Interestingly it had a greater proportion of oil relative to water at 55:45. A lot research went into scale-up of this product and we had used a relatively new instrument that Lasentec makes to watch the emulsion process in real time. We discovered that the emulsion flipped constantly until final cool down.
Originally this product was scaled by another scientist. He left the company and it fell on me to complete the project and support the launch. The product was validated and released to the trade. We were making several batches in a row to fill the distribution pipeline. It was around batch 20 or so that the flipping issue occurred during cool down. Once the product flipped during cool down there was no way to recover. So not only was there a huge expense because of lost batches but the "new news" associated with the launch in jeopardy. At this point there was 30 million dollars committed to marketing and advertising. I'm sure there was millions more tied up in product already in the pipeline. It was up to me to save the product or risk my career and not succeed in coming up with a fix.
When I started evaluating when batches flipped I noticed that all the failures occurred on third shift. My first thoughts were that the problem was either utility or operator specific. I remember flying out to the site on a Sunday in order to work through the night to witness the problem firsthand. Typically third shift operators tend to be young and the least experienced. Seniority often gets you first and second shift. The operators that greeted me that night were in fact young and quite energetic.
Remarkably I was able to solve the problem that night. The solution was kind of interesting and related to youth in an odd way. The first thing that I noticed was how rapidly the heating of the batch occurred. The validated process did not have rate of heating specified. The controls were all manual and the guys ramped this thing up to 55-60 as quickly as possible. The batch record then called for them to hold the batch at this temperature for five minutes and start cooling the batch.
The folks on first shift often heated the batch very slowly. They were not in nearly the same youthful rush. The difference was that the batch had a dwell time at 55-60C for much longer on first shift than it did on third shift. That was the difference. Five minutes did not provide enough energy to form the emulsion. My predecessor did not scale this properly in spite of the Lasentec. I was able to keep my job and the launch went smoothly.

So True...

2011-09-05T07:07:00.000-05:00

Image via WikipediaDouglas Adams came up with the following rules about the way we react to new technologies:

1. Anything that is in the world when you're born is normal and ordinary and is just a natural part of the way the world works.
2. Anything that's invented between when you're fifteen and thirty-five is new and exciting and revolutionary and you can probably get a career in it.
3. Anything invented after you're thirty-five is against the natural order of things.

Equipment and Facility Qualification Overview

2011-09-03T08:14:00.000-05:00

Image by Getty Images via @daylife
A good qualification plan is essential to the success of any validation. Qualification is required to meet Code of Federal Regulations (CFR) regarding current Good Manufacturing Processes (cGMPs). System Qualification has 4 major components and can be defined as follows:

Specification Development
Planning
System testing and documentation
Equipment and document lifecycle management

Specification development is an essential part of the qualification process. It is impossible...
to develop an adequate validation testing protocol without system specifications. Specifications define in detail what the system does and more importantly what the end user needs to perform a job function.
Planning requires the coordination of several disciplines. Systems are often very complex and require transport and placement into the facility. Testing may need to be performed at the site of manufacture. System vendors often supply documentation that can be used as part of the qualification process. This documentation may differ from standard formats used at the site. A good plan will anticipate challenges and define the qualification strategy needed to bring the system online and into compliance.
System testing and documentation are important not only to meet regulatory requirements but important to the business. Testing should be sufficiently adequate so that the system turnover is successful and ready to use when required.
All equipment and its related documentation need to be maintained throughout its useful life. It is important that the site provides adequate tools to manage the lifecycle. For example, a calibration and preventative maintenance program is necessary so that systems function in a robust and consistent manner. The same is true for documentation.
Documentation must be maintained and reviewed periodically. The regulatory environment changes over time. Systems must be managed through these changes and kept up to date.
A master validation plan for an equipment and facility installation or modification summarizes the validation strategy for the project. A plan is necessary when a project has many variables and a lengthy timeline. It is important to confer with the Quality Unit with regards to establishing a plan and its contents.
The project deliverable must be clearly defined in the plan. A validation plan must detail the Good Manufacturing Practices (GMP) and Good Engineering Practices (GEP) and clearly delineate the differences between the two.
The ideal master plan should have the following components. They include:

user requirement specification (URS) documents that will be created.

vendor generated scripts will be used or if the test scripts will be written by the system engineer
system commissioning that will be performed.
specify exactly who approves the documents and which components of the project are designated as such.
project and validation timelines
specify that a trace matrix of all activities be traced to the plan and specification documents and be described in the final report

Once the plan is approved it can be modified only with Quality agreement. The plan should require that a final report will be issued when the qualification is complete and/or the project has stopped.
A user requirement specification are essential to describing end user needs that is resident in system functionality. There are many acronyms for system design. The proper distinction is as follows:

a user requirement specification (URS) is a high level system document that describes system functionality that meets the needs of end users. In the traditional V-Model it correlates to the system Performance Qualification.

a functional requirement specification (FRS) is a specification that describe the system functionality in great detail. In the traditional V-Model it corresponds to the Operational Qualification.

a design specification (DS) describes the hardware and physical characteristics of the system. In the traditional V-Model it corresponds to the Installation Qualification.

I will only go into the URS in detail here.
Writing the URS requires a detailed knowledge of the end user’s needs and the system functionality. It is important to discuss differences in functionality with the end user’s expectations. It is not always possible to meet all of the end user’s expectations because of budget constraints or existing technology. This document is the agreement upon which both the business needs and the technically possible merge.
The following system operation should be described in the requirement document:

modes of operation (e.g., start-up, shutdown, test, fallback)
actions required in case of system failure
safety
interface with other systems, operators, etc
system relaibility

Each requirement must be testable. Avoid trivial requirements that do not add value to the system description.

Factory Acceptance Testing (FAT) is operational testing performed at the vendors site. The intent of FAT testing is twofold:

test the system operation
learn system functions with the system expert (i.e. the vendor)
identifiy and fix deficiencies

FAT is often an essential part of system(s) testing. The FAT is performed at the vendor facility. The FAT provides the project engineer and business unit personnel their first look at the new system. Both the project engineer and business unit personnel can use the FAT to understand and train for future use of the system. In addition, the FAT is an excellent way to ensure vendor accountability for contracted system(s) you are purchasing.
Test scripts for testing the system can be written by the system engineer. Sometimes vendor generated scripts are used. The scripts should be traceable to the user requirement specifications (URS). Commissioning is typically an engineering function Quality Assurance reserves the right to accept or reject the test scripts based on test integrity. The burden is on the system engineer to provide test scripts that meet best practices with regards to system testing.
FAT test scripts often do not have a defined form other than that they be traceable to the URS.
Operational testing must be performed with the the system engineer present to either participate or witness testing. If there are system deficiencies, a list must be created that describes the deficiencies in detail. The list must be agreed upon by all parties. The parties must agree in writing to remediate any or all deficiencies.
The vendor must commit to remediate the deficiencies prior to shipping and/or during Site Acceptance Testing (SAT). Some deficiencies may be accepted “as is.” This is acceptable as long as all parties agree in writing.
An official report is useful and probably necessary to summarize all work performed.
Site Acceptance Testing (SAT) is an essential part of system(s) testing. The SAT provides the system engineer and operation personnel their first look at the new system on the factory floor. Both the system engineer and operation personnel can use the SAT to understand and train for future use of the system.
Test scripts for testing the system can be written by the system engineer. The scripts should be traceable to the URS.
Operational testing must be performed with the system engineer and operations personnel present to either participate or witness testing. This is great opportunity for training for engineers and end users.If there are system deficiencies, a list must be created that describes the deficiencies in detail. The list must be agreed upon by all parties. This includes vendors, operations and engineers. Deficiencies that were discovered in FAT must be reviewed and their remediation assessed. New deficiencies may be discovered during SAT. This may be the result of shipping or other reasons related to installation. The vendor or system engineer must commit to remediate the deficiencies. Trivial deficiencies may be accepted “as is.” SAT test scripts do not have a defined form other than that they be traceable to the URS. Ideally, SAT test scripts should have a final report.

It is important to test the integrated train and document that testing in the PQ. Typically, commissioning (FAT and/or SAT) testing does not completely test system integration but simply tests the equipment as standalone systems.If time and resources allow integration testing during the system start-up is the ideal. The validation deliverable should leverage increased IPA testing. It is important to ensure training, PMs, and calibrations are complete and up to date.
Release to Manufacture indicates system readiness. This means commissioning is complete. This includes training, preventative maintenance schedules were updated, calibration performed.
Release to Commercialization indicates system capability has been verified. The equipment and facility systems are complete and approved by Quality for use.
A final report must be issued in order to release the system for commercial use. Product is ready for sale.

USP Water: An Overview

2011-08-20T17:14:00.001-05:00

Image via WikipediaUSP water is the lifeblood of the pharmaceutical industry. I am often surprised by how little process engineers know about the process of purifying water. Let me me take you through the steps.
Recall from chemistry that osmosis is a natural process. If fluids of two different solute concentrations are separated by a semipermeable membrane then the low solute fluid moves to the high concentration fluid. The movement of solvent across the membranes to equalize the concentration is called "osmotic pressure." Reverse osmosis applies a pressure on the high solute side that is greater than the osmotic pressure. This forces the high solute fluid to move to the low solute fluid and concentrates the salts on the high solute side.
An industrial scale reverse osmosis system can take many forms. Here is an example...
of how they are typically built in the pharmaceutical industry. All industrial systems pre-treat the incoming water. Typically, a sediment filter in the form of green sand (magnesium oxide) is used in conjunction with membrane filters. This removes rust and debris from the water.
Chlorine must be removed prior to the main membranes because chlorine will rapidly degrade the membrane usefulness. This is done by injecting sodium metabisulfite solution. Newer RO skids use activated carbon bed filters. This is a preferred option in my opinion. NaSO4 can be quite irritating to work with. Next, the pH of the water needs to be lowered in order to precipitate out the calcium salts and dissolved carbon dioxide. This is done by a NaOH injection. This keeps the membranes from fouling with calcium salts and ensures that the total organic carbon (TOC) meters will not reject water due to dissolved carbon.
Now the real action begins. The water purification process is performed by proprietary membranes.The membranes are sheet material that are wound around a permeable core. The membranes can filter to the molecular level and leave behind salts and pesticides. There are a number of companies that make them as well as the skids. US Filter and Siemens are two reputable companies that design and sell excellent equipment.
Most RO systems are single pass. I have used a dual pass system to make USP Water but it requires a lot of maintenance and wastes a lot of water. The ideal is to use deionizers to bring the water to acceptable conductivity levels.
Finally, the water system needs to sanitize the final pass water before it is released into the loop. There are several ways to do this. They can be alone or in combination the following: ozone, 0.2 micron filtration, heat and ultraviolet light. Remember that USP Water requires three things: low TOC, low conductivity and be free a objectionable organisms.
The recirculation loop is the USP Water delivery system. It is usually constructed of stainless steel. Stainless steel must be 316L and electro-polished. Orbital welding must be carried out by the Tungsten Inert Gas (TIG) technique using non-consumable electrodes, with additional cold-wire feed where necessary. Stainless steel is excellent for a heated loop (> 65C) and is rigid enough that it requires a minimal number of hangers. There should be no dead legs in the system. A rule of thumb is that a drop or pipe protrusion that is over 5 times the pipe diameter can constitute a dead leg. A dead leg will substantially increase the possibility of a biofilm.
This is a very high level description of a UPS water system.

It's A 2D Universe And Why I'm Really A Hologram

2011-08-09T21:16:00.006-05:00

Image by Denis Collette...!!! via FlickrI have just learned that noise in the data from the GEO600 Project seems to indicate that the universe is granular. This suggests that the universe is a 3 dimensional projection of a 2 dimensional surface. Think about that. Our everyday experience is a holographic projection of a distant 2D surface. So here I am typing this article based on something that is happening at the edge of our universe. Pretty cool.
A great article describes the potential discovery and its implications. Here it is.
If this noise in the data is based on some holographic principle...
it would have interesting implications for metaphysics. I am thinking Joseph Campbell and Carl Jung: two men's writings that I have spent many hours of devoted reading.
Joseph Campbell burst into my consciousness in 1988. The now famous television series on PBS featured Campbell being interviewed by Bill Moyers. He talked for several hours about how the world's multitude of myths are really different manifestations of the same thing. The many myths are disguised by culture and storytelling. More importantly to Campbell is that myth tells us who we are as humans and our place in the universe.
Carl Jung was a protege of Freud. He wrote extensively about a shared collective unconscious. He believed that all humans were hardwired for certain myth motifs. His most readable prose is his long essay "Answer to Job." He claims that Job is a landmark literary work in the "divine drama" in that it is the first criticism of God. The narrative is that YHWH does not consult his omniscience and Job suffers the wrath of God because of Satan's slanders. Ultimately, Jung claims that Jesus Christ is a mediator between man and god: he protects humanity against God and soothes the fear that we feel towards the being of God. The story of Job has a powerful resonance for me and this particular interpretation of the biblical text is the most mature in my opinion.
I cite these two men because I think that they understood something deeper about who we are and our connection to myth and our relationship to nature. Jung takes a religious theme known to Christians and Jews alike and reinterprets the story along his cosmic thesis of the collective unconscious. Campbell's many books and interviews display his vast knowledge of world religion and the struggles of the hero. Perhaps our unconscious has the imprint of a great cosmic struggle between man and his creator. A holographic universe in which reality is a hologram may have interesting implications. Perhaps it can be said that a better understanding of the physical sciences may allow myth may take a more important place in our modern life.

Hacks and Sisyphus

2011-07-31T08:15:00.009-05:00

Image by vintagedept via FlickrI recall a psychology professor in college who once claimed that 90% of people in a selected discipline are hacks. I was intrigued by this observation. He included pilots and doctors in his declaration which is frightening to realize. It really does put a perspective on every profession.
This is true for organizations as well. Every industry has its pacesetter Think of Apple. It is clearly an innovator in the truest sense. Most other companies are creating similar products but only after Apple raises the bar.
Every good organization has its "best guy" or "go to person." Perhaps some minimum level of non-hacks...
are needed to be successful. Maybe it is 10%. My guess is that the 90% hack rate claim is an arbitrary number. Nonetheless, it does state succinctly what I have witnessed over the years. In my chosen line of work I have witnessed many years of hackery by colleagues and contributed myself to the cause on occasion. Yet I am amazed by how most of my fellow hacks with their poverty of skills often inflate their abilities far beyond reality. The absurdity makes me think of Camus and Sisyphus.
The final chapter of Camus' work the "The Myth of Sisyphus" speaks of the condemned Sisyphus rolling a boulder up a hill to only have it roll back down again. This meaningless task points out that "the struggle itself... is enough to fill a man's heart. One must imagine Sisyphus happy." So it is.
If my professor's reasoning is correct then 10% of practitioners in a given discipline are good at what they do. These individuals create the tools that the hacks need to do an adequate job. Seeking solutions to problems is often reward enough. Even hacks given the right tools are able to reach acceptable solutions to unexceptionable problems.

Neanderthal: The New Human Ancestor

2011-07-27T14:15:00.009-05:00

Image via WikipediaThere has been some news coverage regarding humans and their interbreeding with Neanderthals. It has been speculated for some time that humans and Neanderthals may have interbred but no hard evidence existed. Recently, three different Neanderthal bone fossils were sequenced. The fossils were approximately 40,000 years old. Close examination shows that Europeans contain 1-4% of the Neanderthal DNA signal. Humans from sub-Sahara Africa do not contain the any parts of the Neanderthal genome. More interestingly the Neanderthal signal shows up not only in Europeans but also in the peoples of East Asia and Papua New Guinea. Neanderthals never lived in these areas.This means that humans and Neanderthals probably intermingled in the middle-east before migrating outward.
Some scientists are speculating that 1-4% of the Neanderthal genome is a minimum amount. Better sequencing data may raise the percentage to 10-20%. It is very difficult to sequence very old fossils. DNA sequences are never complete over long time spans. It may be some time before the higher number can be confirmed or debunked.
I have written earlier about evolution and how I believe that DNA sequencing will change the way we think about specie progression and change. I think this is a great example. The old model was that human species progressed from one form to another while our less fortunate cousins died on the evolutionary vine. I have always suspected that the human evolutionary picture is far more complicated.
Models of human progression are always fraught with the cultural, ethnic and racial baggage. It is difficult if not impossible to remove the bias that can go into these discussions. DNA evidence is difficult to dispute by rational people and may help humans to understand who they truly are.