Updates

---

There, I’ve said it, and I’ve worked as a chemist my whole career. For me, it started in college, where my professors were horribly messy and disorganized. Unmarked flasks with bubbling fluids, mysterious white powders on balance pans, sinks full of dirty and chipped glassware, and countertops strewn with out of service equipment were the norm in the college labs. Early in my professional career I carried this attitude into the industry labs I worked in, and I wasn’t alone.

Walk through most chemistry labs and you will see hoods stuffed with last month’s abandoned experiments, desiccators crammed with last week’s samples, and reagents stored randomly around the benchtops. It is not unusual to find septa, vials, Kimwipes, Pasteur pipettes, and even chromatography vials on the floor. Open the drawers and look in the back of cabinets for a walk down memory lane, with ancient equipment, random obsolete parts, and the occasional abandoned sample prep - a data integrity nightmare.

But why? When I look at investigations, I see contamination, mix-ups, and erroneous reagents/equipment as primary causes of laboratory error. Many of these costly investigations could be avoided by keeping a clean and orderly laboratory in the first place. A clean lab leads to efficient workflows, fewer investigations, a more inspection-ready facility, and is a safer work environment. With all of these benefits why wouldn’t we choose to work in a clean and orderly laboratory?

After escaping from my professors’ slovenly laboratory habits, I joined a startup pharmaceutical manufacturer as the second chemist hired. We kept this new facility spotless and took great pride in its appearance. As a startup, we had a lot of venture capitalists and inspectors coming through so it wasn’t optional, we always had to look great. One of the many benefits of this was obvious at our first FDA inspection. I was nervous, but the senior chemist said “let them come inspect us all they want, we work in a clean and compliant fashion every day and don’t have to change a thing or worry one minute”. I smiled and let the stress fall away, taking this attitude with me into the rest of my career.

Most labs are shared spaces. A lack of ownership of the space can leads to messiness. Why should I clean up after that slob? A solution to this is zone assignment. Although it is not always possible to assign chemists to personal workspaces, chemists can be assigned a discreet area to keep clean. Publish a map of the lab with the names of the people responsible for a weekly deep cleaning alongside their assigned zone. Rotate the cleaning assignments regularly so nobody can complain that they are stuck with a tough zone. Finally, make sure the lab leadership all have zones assigned as well. We lead by example, so when chemists see the lab director sweeping and scrubbing his zone they get the message that it’s not OK to be a slob in this lab!

When I first moved into laboratory management, I had a mentor tell me "Eric, don't worry about the analyst that has a few laboratory investigations, worry about the analyst that never has a laboratory investigation". This piece of advice chilled my blood, but stuck with me through years of management and was a valuable wake-up call.

We tend to believe that analysts are honest, have little incentive to cheat, and "It wouldn't happen in our organization". This is likely true, but an incredibly dangerous attitude. The damage a single ethics breach can cause is enormous, so in the spirit of appropriate risk assessment, it is crucial that these issues are spoken about openly.

In 2005, Able Laboratories Inc. had their entire organization destroyed by laboratory fraud. “The damage from the fraud at Able Labs was devastatingly complete. Consumers were put at risk, a company that employed 500 people was destroyed, and shareholders were left with nothing in the end. This is the legacy of the fraud perpetrated at Able Labs by these defendants.”*

There is no shortage of other examples of laboratory fraud damaging company profits and reputations. As leaders in the laboratory industry, we have an obligation to view our own organizations as potential sources of fraud. To do otherwise is simply naive.

The laboratory is always under pressure to provide fast, passing results. The laboratory is praised when passing results are provided, and often criticized when failing results are obtained. Human nature dictates that ethics breaches will occur in some analysts if enough pressure is applied. Be very careful how much pressure you put on your staff to generate rush testing results. Given enough pressure, some analysts will eventually cheat rather than miss a deadline.

How to avoid these problems? Make it very clear that it is OK to fail a sample, it is OK to miss the occasional deadline, and that complete honesty is a requirement of employment. This culture can be developed with an ethics training program provided to all new employees with a one-year refresher for existing staff. The program should be given by senior management to make it clear this is a top priority for the company. You should have a third-party ethics hotline to facilitate complaint investigations, and robust quality review systems that can detect fraud.

These ethics controls will be a crucial aspect of your overall data integrity program. No amount of passwords, audit trails, or electronic systems will stop an analyst who is determined to cheat. Show your inspectors that your quality systems address the topic of fraud openly and directly.

Although ethics training can be stressful, analysts are grateful to know their organization prioritizes ethical behavior. Many analysts would come to me saying they appreciated working in a culture where complete honesty was clearly identified as a top priority. I always left the training sessions with a joke: "We never want you to cheat to meet a deadline... We may ask you to compromise your weekend, but never your ethics."

* March 8, 2007: Four Managers of Now-Defunct N.J. Generic Drug Maker Admit Fraud

Having worked in two contract labs and multiple manufacturing and research organizations, I can speak with confidence when I say there are flaws with many contract laboratory arrangements.

The primary problem for most organizations who hire a contract lab is that your work becomes just another sample to be tested in a large queue. The contract lab queue probably contains many more lucrative projects with more established customers. When resources are scarce, your sample can fall to the bottom of the contract lab's queue both in scheduling time and talent.

The second problem is the fact that some contract labs have little incentive to admit to laboratory error when unexpected results are obtained. Your project is just another $500 test. It matters nothing that your multi-million dollar product is affected. Having worked with several contract laboratories on out-of-specification (OOS) result investigations, a common theme is that a checklist was performed, the equipment calibration was checked and the analyst was interviewed - and guess what? No unusual events were identified! The contract lab then usually suggests a retest. Retesting without first finding root cause can put you in a terrible position, so you must resist this temptation!

How to overcome these problems? Try to become their most important customer. Send the contract lab regular work that they can count on. Send them forecasts of expected future work so that they can plan. Let them know your annual budget for their services. Go visit their facility, and not just for audits, but to grow the business relationship. Insist that they follow your SOP on OOS investigations, and pay them to do a thorough investigation even when laboratory error is found. It is a small price to pay when you consider the overall value of your product compared to the price of the investigation.

The best contract relationship I was ever in was one based on an annualized full-time-equivalent (FTE) arrangement. The contract lab was paid per FTE-day of work regardless of how many samples were tested. In this arrangement, the contract lab became a true partner and was able to see the value of the drug product. The contract lab was paid to search hard for the root cause of unexpected results, and was very flexible when rush testing was needed.

After supervising over 3000 laboratory investigations I have seen it all.  Bad instruments, methods, analysts, reagents, filters, columns, standards, glassware, and environmental conditions. Often I am approached with a statement from the analyst like this:  "I am sure this sample is truly OOS, I ran it right alongside a sample that gave perfectly normal results." Although sometimes the sample is truly OOS, we search tirelessly for laboratory error, and more than half the time are able to determine the OOS is a mistake somewhere in the lab.

Some of my most memorable:

• 20-mL volumetric flask was really 25 mL!
• Prednisone is not the same as prednisolone!
• Dissolution autosampler malfunctioned and diluted the samples
• Dishwasher had a broken coupling, leading to contaminated glassware
• Analyst decided to cut the stuck stopper off of a flask and transfer the sample to a new flask