A recent study conducted by a consortium of nine leading pharmaceutical companies has brought fresh insights into the challenges and complexities of detecting visible particles in parenteral drug products. This research, published in theJournal of Pharmaceutical Sciences, reveals that the probability of detecting these particles depends on various factors, such as particle type, size, and the number of particles in each vial.

This cross-industry effort involved major pharmaceutical players including Amgen, AstraZeneca, Bristol Myers Squibb, Boehringer Ingelheim, Fibrogen, Genentech, Janssen, Lonza, and Roche. The collaboration emphasized the importance of rigorous manual inspection standards and explored the intricacies of detecting different types of particles in vials using manual visual inspection techniques.

Detectability of Particle Types and Sizes

The study underscores the inconsistency in particle detectability based on their type and size. Polystyrene beads and rubber particles were found to have the highest probability of detection (PoD), with a robust PoD of over 70% for sizes larger than 250 µm when single particles were present. These results stood in contrast to other particle types such as glass, metal, and protein surrogates, which demonstrated significantly lower detection rates even at sizes of 400 µm. The study also confirmed that vials containing multiple particles had a higher detection probability than those containing just a single particle.

Figure 1 in the study illustrates this with nearly consistent blue horizontal lines, showing that multiple particles per vial increase the chances of detection, while single particles are often missed. For example, single glass particles required a size of 250 µm or larger to reach a detection probability of over 70%, while no other single particle type reached this threshold. Fibers, regardless of their color or size (up to 400 µm), were consistently the most difficult to detect, never reaching a detection probability higher than 50%.

The Limitations of Manual Visual Inspection

The study highlights the limitations of manual visual inspection, a process still widely used despite advances in automated inspection technologies. The human eye, unaided by magnification, often fails to consistently detect smaller or irregularly shaped particles, particularly fibers and smaller glass particles. The researchers point out that even with proper training, analysts struggle to classify and enumerate particles correctly, especially when multiple particles are present in a single vial.

Additionally, the researchers found that detectability does not increase in a linear fashion with particle size. In fact, certain particles, such as metal and fiber particles, exhibited unreliable detection rates even as particle sizes increased. These findings suggest that training for manual visual inspection must be tailored specifically to the types of particles most likely to be encountered in real-world scenarios.

Industry Implications

This study has important implications for the pharmaceutical industry, particularly in refining visual inspection processes and incorporating more advanced technologies. Automated inspection technologies offer a significant improvement in the detection of particles compared to manual inspection. These machines can outperform human inspectors by 120% to 200%, depending on the particle type and inspection conditions. The probability of detection (PoD) remains the core focus of both manual and automated methods. In the case of automated systems, the consistency in detecting particles—particularly those in challenging size ranges (50 µm to 400 µm)—is significantly higher than with human inspectors. Automated systems, devoid of fatigue and bias, can perform prolonged inspections with greater precision, making them ideal for high-volume production environments where human limitations become evident.

(Original Research: Mazaheri, M., Saggub, M., Wuchner, K., Koulov, A. V., Nikels, F., Chalus, P., Das, T. K., Cash, P. W., Finkler, C., Levitskaya-Seaman, S. V., Case, J., Parsons, J., & Gonzalez, K. (2024). Monitoring of visible particles in parenteral products by manual visual inspection—Reassessing size threshold and other particle characteristics that define particle visibility.Journal of Pharmaceutical Sciences, 113, 616–624. https://doi.org/10.1016/j.xphs.2023.10.002)