Glass particle contamination: is it here to stay?

We would like to bring attention to the problem of glass particle contamination of contents of single-dose glass ampoules that could occur upon opening of such ampoules, and which if injected parenterally, would be a hazard to the patient. This risk is particularly high for patients in the operating theatre and intensive care unit as they receive medications almost exclusively in parenteral form. Glass particle contamination of glass ampoules is a known hazard [1] but because glass particles used to be large enough to be easily visible, they could be avoided during aspiration of the drug. It is now rare to find grossly visible remnant glass particles in opened ampoules. We wonder if this is due to the absence of glass particles or that glass splinters are now too minute to be grossly visible. If it were due to the latter, it would create a false reassurance among medical practitioners and anaesthetists, leading to the inadvertent aspiration of glass particles while drawing up the ampoule contents. To investigate this problem, used single-dosed glass ampoules were collected from operating theatres at the end of each theatre list. A visual inspection for grossly visible glass particles was carried out. Large glass particles were seen in three of the sampled glass ampoules. One millilitre of sterile water was then injected into each glass ampoule. The glass ampoules were then gently swirled for 15 s, to ensure that all glass particles, if any, were washed off the wall of the ampoule. To derive the average number of glass particles per ampoule for each group of drug ampoules (for example, 1 ml, 10 mg morphine ampoule), only the contents of ampoules from the same group were aspirated using the same 3 ml syringe and the same 18G (1.2 mm × 40 mm) non-bevelled (B. Braun®) drawing-up needle. To ensure that all glass particles, if any, that had settled to the bottom of the ampoule were drawn out, the needle was pointed downwards, touching the base of the ampoule, with the ampoule slightly tilted during aspiration. With the needle still attached, the contents of the ampoule were then pushed out of the syringe and emptied onto a spread out, clean, green towel. Once the towel had absorbed the water, the towel was then visually examined closely for any shimmering of glass particles. The number of glass particles observed was noted. After counting, all glass particles were carefully removed and discarded into the sharps bin. Next, a vernier calliper (accuracy to nearest 0.02 mm, Mitutoyo®) was used to measure the thickness of the wall of the neck of the ampoule. Three readings were taken from each ampoule and an average was recorded. The results of our study showed that, out of 510 glass ampoules, a total of 113 glass particles were found, giving an average of 0.22 particle per ampoule. When sampled ampoules with matching wall thickness (0.70–0.80 mm) and type (metal-etched) were compared, we found that those with a larger size (larger than 2 ml capacity, n = 13) have a greater averaged number of glass particles per ampoule (0.46), compared to that of smaller ampoules (0.14), where n = 162. For sampled ampoules with matching wall thickness (0.80–0.90 mm) and size (larger than 2 ml capacity), we found that metal-etched ampoules (n = 55) have a greater averaged number of glass particles per ampoule (0.47), compared to that of colour-break ampoules (0.24), where n = 93. These results are in concordance with other studies [2,3]. Furthermore, of all the ampoules with comparable size (less than or equal to 2 ml capacity) and type (metal-etched), we found that those with thicker wall (more than 0.80 mm, n = 20) have a greater averaged number of glass particles per ampoule (0.35) compared to that of ampoules with wall thickness less than 0.80 mm (0.14), n = 162. We would also like to highlight the particularly high averaged number of glass particles per ampoule found in the nine sampled glass ampoules containing etomidate emulsion (0.56). Being a white emulsion, the presence of remnant glass particles is not easily seen, hence might lead to reduced vigilance of glass particle contamination during aspiration of the drug. Other injectable non-glass particles found include paint particles from the white ring marker of colour-break protamine sulphate ampoules. The data from our study clearly demonstrates that even with the use of an 18G needle for drawing up the contents of the ampoules, smaller glass particles can still pass through the needle into the syringe. Glass particles of this size can easily be injected into patients. This is particularly so if drugs are routinely administered via the injection port of the intravenous cannula, which is a safety measure designed to decrease sharps injury to the medical staff. As ampoules for medical use are made of type I borosilicate glass [2], the sequelae of glass particle contamination requires a considerable time to develop. Previous animal studies had demonstrated that the infusion of glass particles led to pulmonary silicotic changes and nodular fibrosis of the liver, spleen and small intestines as a result of foreign body reaction [2]. Even though our method of visual eyeballing for glass particles may not be very sensitive as some fragments may only be visible under light microscopy, those visible to the naked eye are probably over 75 µm in diameter [4] and would have lodged in the pulmonary capillaries if administered. Pulmonary capillaries have an average diameter of 10–12 µm. Besides being chemically inert and possibly recyclable, one advantage of packaging drugs in glass ampoules is the ability to vacuum-seal. However, as plastic is just as comparable in these aspects, and as the manufacturing processes still cannot prevent fragmentation of glass, there is an urgent need to call for drug manufacturers to move towards using plastic ampoules as an alternative packaging for drugs now available in glass ampoules. Not only might this overcome the problem associated with infusion of glass particles but this would also prevent problems of sharps injury to staff, such as laceration of fingers. In the meantime, to minimise administering glass particles to patients, we recommend either the use of in-line filter [5], or aspiration through a fine-bore needle [6] as well as avoiding aspiration from the most dependent area of the glass ampoule [5]. All that said, the practicality of using filters and fine-bore needles for aspirating large volumes might be a consideration, when implementing these recommendations. There is also a need to look into effects of paint particles contamination from the coloured ring marker of colour-break ampoules. We would like to hear from the rest of the anaesthetic fraternity.