Monday, July 20, 2015

The CAPA Detective

By Laurie Meehan

Every industry has its detectives.  Mechanics and journalists hunt for leaks.  Software testers and security experts look for bugs.  Meteorologists, defense strategists, and forensic accountants all look for trouble brewing off shore.  The level of intrigue may vary.

In the GMP world, it’s the CAPA investigator who serves as detective, identifying how and why manufacturing defects occur.  The root cause is seldom obvious or sensational, but usually a confluence of seemingly innocuous deviations that can compromise an otherwise high-quality manufacturing process.  That’s not terribly flashy, so you probably won’t find “The CAPA Detective” in among the other crime dramas making their debut in this year’s primetime lineup.

Or maybe you will.

The Victim: Blockbuster sleeping aid, Swell-Naps Gel-Caps
The “Crime”: Gel caps hopelessly bricked together in container
The Chief Investigator:  Hard-boiled, 20-year veteran, Sgt. CAPA

The Usual Suspect
Everyone’s initial suspicions leaned understandably toward Excessive Heat, the long-recognized nemesis of gelatin encapsulating agents. However, after a preliminary physical examination of the victim, veteran detective Sgt. CAPA was quickly able to exclude Mr. Heat from the suspect list.  It turned out that the gel caps had not melted, eyewitness testimony to the contrary.  Instead, some of the gel caps had become brittle.  So brittle they broke, leaking their liquid contents into the jar and cementing the remaining gel caps together.  Convenient as it would have been to blame the whole sordid affair on a consumer’s faulty AC unit, there’d be no rush to judgement on CAPA’s watch.

CSI: Warehouse
Looking for crimes with a similar M.O, Sgt. CAPA searched the warehouse and inspected the entire Swell-Naps Gel-Caps inventory, putting the following facts into evidence:
  • Roughly 1 in 8 were bricked
  • All had been subjected to the same acceptable temperature, humidity, and lighting conditions

The evidence seemed to exonerate the warehouse, so CAPA decided it was time to pay the manufacturing facility a little visit.

CSI: Manufacturing Facility

Hoping the paper trail would produce some answers, Sgt. CAPA pulled the batch records for the warehouse inventory.  He found out that the gelatin powder used in the manufacturing was sourced from two different vendors, which we’ll call Vendor A and Vendor B.  (The names have been changed to protect…well…you know why.)

He also discovered:

  • Vendor A’s gelatin powder had been used to make Swell-Naps Gel Caps for years; Vendor B was a relatively new supplier.
  • None of the bricked batches were made using Vendor A gelatin.
  • All of the bricked batches were made using Vendor B gelatin.
  • Not all the Vendor B batches were affected – a full 60% of the Vendor B batches sustained no apparent injury. 

Vendor of Interest
Clearly Vendor B warranted a closer look, so CAPA launched a thorough investigation into B’s background.
  • All C of A*s for B’s gelatin shipments were in order.
  • All validation testing produced a suitable and stable encapsulation.
  • The Vendor Qualification process had been conducted and documented according to Swell-Naps’ SOPs.
  • Gelatin samples had been analyzed and all parameters were found to have been within the specified ranges, though sometimes approaching the upper thresholds. 
If it’s not the product, Sgt. CAPA mused, then it’s gotta be the process.  But if it’s the process, why does the defect occur only with Vendor B gelatin?  And why in less than half the cases?  Muttering to himself, he returned to his office to scrutinize the evidence again over lunch.

Cracking the Case
Hours later, the greenish glow of the computer screen reflected off CAPA’s face, pulled tight in concentration.  A full cup of vending machine coffee sat cold and forgotten in its Styrofoam cup.  Next to it, the brown paper bag holding his liverwurst on rye remained unopened. 

Then suddenly, it hit him.  The answer had literally been staring him in the face.  He glanced from one side of his wireless solar-powered keyboard to the other, and smiled the smile of exquisite satisfaction.

All of the defects had taken place during the Second Shift!!!

One by one, Sgt. CAPA interviewed the second-shift manufacturing crew, and all the pieces began to fall neatly into place.  For years, the shift supervisor had worked with Vendor A’s product, a very consistent, fine-grained material.  Long ago he learned that it reliably yielded a uniform gelatin blend when mixed for 10 minutes -- the minimum amount of time prescribed by SOP.  He concluded that further mixing was a waste of time and money, and proudly passed on the benefit of his experience to his subordinates.

But the size of the particles in Vendor B’s powder was more variable, and so required more mixing, perhaps as much as the maximum SOP duration of 15 minutes.  Following what they thought was their boss’s preferred MO however, the second-shift operators had stopped mixing the Vendor B blend after 10 minutes, never checking for uniformity.  Their supervisor was experienced enough to know that new materials would have meant new rules.  He would have checked his Vendor A assumptions and coached his crew accordingly.  But by an unlucky coincidence, he was in the midst of an extended company leave when the gelatin powders were switched.  The temporary supervisor was unaware of his adopted subordinates’ routine practice.

So, because of a misapplication of advice, the second shift operators had produced a gelatin that was not uniformly blended, causing the Swell-Naps Gel-Caps to become brittle, break inside the bottle and brick the contents of the jar.

The validation team should have run batches using gelatin powder at the upper limit of the particle size range.  If they had, they might have caught that a long mixing time was crucial and then confirm production procedures and SOP adherence.

The shift supervisor should have made it clear that his mixing advice pertained to a specific set of ingredients.  The SOP should have made it clear that mix duration was a judgement that depended on visual inspection at 10 minutes.

The second-shift operators might have questioned whether their supervisor’s advice held for this new set of circumstances.  The temporary supervisor might have observed his team a little more closely.  But there were no egregious deviations, just a combination of minor lapses in communication and testing coverage.

It’s not an uncommon story, and just another day on the job for Sgt. CAPA.

*C of A = Certificate of Analysis

No comments:

Post a Comment