Burns & McDonnell

Automating Pharmaceutical Inspections for Efficient Delivery

Written by Payton Fraley | May 9, 2022

In an effort to efficiently examine pharmaceutical vials and increase domestic production, many life science facilities are turning to automation as a way to quickly and safely inspect vials for defects.

Regulatory agencies, including the U.S. Food and Drug Administration (FDA), take the safety of detecting defects in pharmaceutical products seriously to prevent compromising sterility and stability. As part of updated FDA regulations, the process of container closure integrity testing (CCIT) evaluates the effectiveness of vials in maintaining a sterile barrier against potential contaminants.

To maintain patient safety, it is vitally important to prevent contaminants from being able to enter the vial and to conduct inspections to see that the product is essentially free of visible particulates. While the risk of contamination is low, the stakes are high: Protecting customers and maintaining the safety of the pharmaceutical supply chain is critical. The FDA conducts facility inspections to confirm that regulations are followed.

A Cost-Efficient Solution

Traditionally, pharmaceutical vials are checked manually by facility inspectors. Workers physically pick up each vial to check for glass or other particles, identify that a stopper is placed in the vial, examine to see if the vial is sealed with a cap, and inspect for a variety of other defects. An automated process instead uses cameras and robotics — first to identify a certain defect, then to examine every vial to see whether that same defect is present, greatly reducing the need for manual inspections.

The automated process machinery is capable of detecting defects that are impossible to detect with the naked eye. This often leads the automated process to reject more vials and accurately identify more defects than would occur during manual inspection. Typically, three to four vials are inspected per minute with human inspection and 20-40 vials per minute for semiautomated solutions. Depending on the size, a fully automated process can be expected to examine over 400 vials per minute. While the initial capital cost of automated process machinery can be significant — ranging from $1 million to $3 million — the savings from automating the inspection process can significantly cut costs for a facility in the long run.

Safety Remains Paramount

A crack in a vial’s sidewall, a weak seal on a cartridge, or an ill-fitting stopper or cap on a vial are all defects that could lead to potential contamination. In recent years, the FDA has shown an increased concern around particles in injectables and have renewed focus on maintaining sterility. In addition to CCIT, a manual Knapp study should be completed to determine if the proposed automated solution is equal to or better at accurately detecting defects within the product compared to manual inspection.

Most challenges revolve around an incorrect design or application of equipment, which is why it’s important to have a life science team that has the design and engineering experience needed to thoroughly plan and implement process solutions. For example, inaccurate sealing parameters or leaks occurring from a broken seal can lead to contaminated products throughout a pharmaceutical process line. Such a contamination can lead to unexpected side effects and serious repercussions for patient health and can result in costly facility cleanup, halts on production and damage to brand reputation, among other repercussions.

Automating pharmaceutical inspection processes can save both time and money for the facility — allowing patients to receive pharmaceutical products in the most efficient manner possible. This is key to maintaining increased production levels, while simultaneously keeping checks in place to keep safety at the forefront.

 

Global insecurities have highlighted gaps in production, leading to a movement toward onshoring pharmaceutical manufacturing. Learn more about where the U.S. manufacturing industry is headed.