Rubber Coring of Injectable Medication Vial Stoppers: An Evaluation of Causal Factors

Introduction

Rubber stoppers allow easy access to drugs contained in glass medication vials, as they can be readily punctured through with a needle or a cannula. In hospitals in general and in intensive care units in particular, the great use of injectable drugs packaged in glass vials with rubber stoppers means that such a practice is very common. For example, in the intensive care units of our hospital, approximately 99,000 rubber stoppers of various medication vials were punctured in 2015. The presence of visible particles of various sizes in medication drug vials, caused by the coring of the vial’s rubber stopper by injection or preparation medical devices, has been regularly reported in scientific literature. Isolated cases have been described [1–4], as well as several studies on particle incidence in used medication vials [5–8]. The incidence of rubber stopper coring, assessed by the presence of visible particles, has been evaluated as ranging overall from 0 to 40 %, depending on the studies and the medical devices used for the piercing of the rubber stopper.

Clinical risks induced by intravenous administration of visible particulate matter are still incompletely characterised or documented, yet macroscopic particles have been incriminated as being able to cause phlebitis and venous inflammatory reactions [9], reduce tissue capillary perfusion [10], induce the formation of pulmonary embolisms, injection site reactions and granuloma [11], and have even been implicated in a deadly small bowel infarction [12].

The administration safety of injectable drugs is ensured notably by the absence of visible particles in the injected preparation. Regulatory texts worldwide concerning visible particle indicate that parenteral preparations are essentially free from visible particulates [13], clear and practically free from particles [14], or clear and free from readily detectable foreign insoluble matters [15]. However, direct intravenous injections, such as with propofol, remain at risk, as the medication withdrawn from the vial is directly administered to the patient.

Unfortunately, the studies mentioned previously are often lacking in detailed description of medications or medical devices implicated (for example by only referring to generic drug names of the medication vials, without mentioning the drug manufacturer, or without mentioning exact references and manufacturers of the medical devices used), which makes comparisons and possible cause-analysis difficult. Several hypothesises have nevertheless been considered, such as the nature or quality of the rubber constituting the stopper, the type and characteristics of the medical device used for puncturing [8, 16], or the puncturing method used by the nurse or physician [17, 18]. However, no study investigating the relative importance of those factors has been identified in the current literature, and the exact causes of particle creation through coring remain uncertain.

The primary aim of this study was to determine which combination of factors posed the least and the most likelihood of coring and particulate matter creation. Secondary aims were to determine particulate frequency and size, and whether elastomer quality influenced coring frequency or particle size.

Materials and methods

Results

Particle dimensions

Particles dimensions data is summarized in Table 3, and their distribution is presented in Figure 1.

Table 3:

Observed particle dimensions.

Particle generating risk situation	Number of particles	Length (mm)					Width (mm)
		Min	Max	Median	Mean	Standard deviation	Min	Max	Median	Mean	Standard deviation
2 mm bromobutyl stopper, short bevelled needle, 90° puncture angle	3	0.85	0.89	0.85	0.86	0.02	0.43	0.69	0.47	0.53	0.14
4 mm bromobutyl stopper, short bevelled needle, 90° puncture angle	2	0.92	0.92	0.92	0.92	0.00	0.57	0.83	0.70	0.70	0.18
4 mm chlorobutyl stopper, short bevelled needle, 90° puncture angle	30	0.59	2.82	0.85	0.99	0.49	0.48	1.01	0.72	0.73	0.13
4 mm chlorobutyl stopper, short bevelled needle, 45° puncture angle	11	0.85	1.20	0.99	1.01	0.10	0.72	0.99	0.81	0.81	0.09

Figure 1:

Observed particle dimension distribution.

The observed particle length was not found to be statistically different between the various situations (p=0.94), with an average length of 0.98±0.39 mm (mean±standard deviation). Overall particle width shows statistically significant differences between situations (p=0.015), with particle width being significantly different between particles from situation N°1 (2 mm bromobutyl stopper, blunt bevelled needle, 90° puncture angle) and situations N°4 and 12 (4 mm chlorobutyl stopper, blunt bevelled needle, 90° puncture angle) (p=0.019 and p=0.001, respectively).

Most particles have dimensions inferior to 1 mm, but particles with lengths superior to 1 mm were observed only in situations implicating 4 mm chlorobutyl stoppers. Particle shapes are mostly homogenous and roundish, and look like “cores”, but some other shapes have been noted (Figure 2).

Figure 2:

16 particles observed from situation N°4 (short needle bevel, puncture angle of 90°, through a 4 mm chlorobutyl vial stopper), observed through optical microscopy, magnification times 40.

Infrared spectrophotometric analysis

Obtained infrared spectra are presented Figure 3 (bromobutyl stoppers) and Figure 4 (chlorobutyl stoppers).

Figure 3:

Infrared spectra of the used bromobutyl vial stoppers, (A) from the 2 mm bromobutyl vial stopper and (B) from the 4 mm bromobutyl vial stopper.

Figure 4:

Infrared spectra of the used chlorobutyl vial stoppers, (A) from a 4 mm chlorobutyl vial stopper and (B) from a 2 mm chlorobutyl vial stopper. The arrows indicate differences between (A) and (B).

Infrared spectra show significant differences between chlorobutyl and bromobutyl samples, for wavelengths between 1,200–1,100 cm⁻¹ and around 900–700 cm⁻¹.

Bromobutyl samples present identical spectra, however the chlorobutyl samples’ spectra differ between themselves for wavelengths between 1,100 and 1,000 cm⁻¹ and at 700 cm⁻¹, indicating a clear chemical difference between samples.

Discussion

In our study, we have shown that visible particle formation can be an alarmingly common phenomenon, as visible particle formation was observed in four analysed situations out of 16 using commercially available medication vials, with particle incidence ranging from 0 % to 75 %, depending on the type and thickness of the elastomer, the type of needle, and the needle insertion angle.

Out of the four causal factor studied, the type of needle seems to be of primordial importance, as all of the present particles were detected after use of a standard short bevelled metal needle, and the use of a blunt metal needle totally eliminated particle formation, for all tested situations. Recent data published by Wani et al. also reported the profound impact of the type of medical preparation or administration device used [8]. They compared particle formation after the stopper had been punctured using two different preparation devices, an 18 gauge sharp hypodermic needle and a blunt plastic safety needle, and found an overall frequency of coring incidence of 40.8 % when using blunt plastic safety needles and of 4.2 % when using sharp metal needles. The comparison between their study and our data indicates that the blunt property does not seem in itself sufficient to guaranty minimal coring risks, as two different blunt devices gave opposite coring results. Such a difference could be explained by other intrinsic properties like for example the type of material and quality of the manufacturing (sharpness of the edges, thickness of the needle wall). Data published in 2015 by Eskander et al. [19], who studied the incidence of coring with 3 different access devices, including a blunt plastic safety needle and a metal 18 gauge sharp bevelled needle seems to confirm such a hypothesis. In their trial, the use of a blunt plastic canula cored up to 22 % of propofol medication vial stoppers, whilst the use of the blunt bevelled metal needle reduced the coring incidence ten-fold, without however totally eliminating it.

Also, both Wani et al. and Eskander et al. found that coring rates differed greatly between medications vials, indicating that the quality of the stoppers is probably uneven and manufacturer dependant. However, as batches, brand names and medication manufacturers were not disclosed, no conclusions can be made concerning the type of quality of the rubber vial stopper. In the same way, there is no indication about the puncture method used.

The needle insertion angle also seems to have an important impact on particle incidence, as a 45° initial puncture angle reduced particle formation to un-detectability if the initial particle formation was of less than 7.5 %, and reduced it by more than 50 % when the initial particle formation was of 75 %. The puncture angle is therefore an important factor to help reduce particle formation risk, but in practice it will not be enough to totally eliminate it. In 2001, Shiroyama reported an 18 % incidence rate of coring by 18 gauge injection needles during the use of 100 propofol vials [7], and subsequently proposed what was described as being an advanced puncture technique to reduce coring incidence, by inclining the puncture angle to 45° [18]. Such a technique was also reported by Roth et al. as reducing the risk of coring [17]. In 2013, Campagna et al. studied the occurrence of rubber coring of prednisone acetate vial stoppers. Despite using a needle insertion angle of 45° to 60°, they found an incidence of 10.5 % when using a metal 18 gauge sharp bevelled needle [6], thus confirming our data that a 45° needle insertion angle does help reduce coring incidence but without totally eliminating it. Also, the more elaborate a puncture method, the less likely it is to be fully applied in sometimes urgent clinical situations, thus reducing even further its effectiveness. Therefore, it could be said that such puncture methods are an incomplete safety measure where it comes to reducing visible particle formation.

Identical infrared spectra were obtained with both bromobutyl samples, indicating that there is no fundamental chemical difference between bromobutyl samples. As bromobutyl elastomer was implicated in particle formation when using a short bevelled needle with a 90° puncture angle, with a incidence irrespective of thickness (7.5 % and 5 %, for 2 mm and 4 mm respectively), it seems plausible that the stopper thickness does not significantly affect visible particle incidence. The huge difference in particle incidence observed between situations N°3 and N°4 (0 % versus 75 %, when using a short bevelled needle, 90° puncture angle, with respectively a 2 mm or 4 mm thick chlorobutyl stopper) could be explained by the chlorobutyl thickness (2 mm versus 4 mm). However, as infrared spectrophotometric analysis of the chlorobutyl rubber showed a clear chemical difference between chlorobutyl samples, and with regard to the absence of influence of elastomer thickness for bromobutyl samples, it also seems plausible that the observed difference be linked to the chemical quality of the elastomer, and not to the thickness of the stopper. However, as it is possible to have variations of the chemical composition of the stopper from different suppliers (who employ various additives such as vulcanisers, accelerators, stabilizers, pigments), it is difficult to extrapolate the results of the present study to all stoppers.

The particles observed had on average dimensions of around a millimetre, yet some particles were more than 50 % bigger, with lengths greater than 1.5 mm. Qualitatively the particles look mostly like cores, but the variation in size as well as the presence of other shapes suggests that there may be multiples ways in which the particles are formed (tearing, shearing or shredding of the vial stopper). The dimensions we observed are consistent with those reported in other studies. Riess and Strong reported a particle having dimensions of 0.4 mm by 1.5 mm [2], and the particles observed from prednisone acetate vial stopper coring had sizes ranging from 0.6 mm to 1.1 mm [6]. When it comes to risk assessment, larger particles, despite plausibly being able to cause more clinical damage, are more easily noticed once created, due to their larger size. Smaller particles, on the other hand, can be much harder to detect, and may be more easily aspirated through the needle into the syringe used for parenteral administration. In our study, 18 gauge needles were used, with an interior diameter of 0.927 mm, and as such many of the particles created could have been aspirated up through it (although not the bigger ones). Particle colour can also aid or hamper detection, as it is generally easier to detect an orange particle than a grey one. In any case, vigilance and systematic visual inspection of the solution to be administered is of great importance.

This work has investigated potential risk factors impacting visible particle formation. The use of puncture-free medication vials as the source of the rubber vial stoppers is a guarantee that the results were not affected by a previous puncture of the elastomer, and is also close to real clinical situations. However, the analysis of particles formed was realised before aspirating the contents into a syringe (which is the next step in clinical contexts), in order to avoid particles adhering to the syringe’s piston, rendering their analysis difficult. The formation of visible particles is clearly a not uncommon situation, with incidence varying with the quality of the products available (injection or preparation medical devices, medication rubber stoppers), and puncture technique. The number of samples tested for each situation (n=40) is high enough for correct statistical analysis between situations; however it is not high enough to identify systematically very low particle incidence (less than 2.5 %).

In our establishment, adult and neonate intensive care units are especially aware of problem, and in those units, blunt bevelled metal needles are used for direct intravenous administrations. However, visible particles are probably only the tip of the particle iceberg, as the clinical impact of sub-visible particles has also been increasingly investigated [3, 20, 21], as well as their potential relevance to latex allergies [22, 23], and the only way to simultaneously reduce risks linked to both visible and subvisible particles would be to promote the systematic use of appropriate medical devices, such as terminal online filters for all parenteral administrations, and not just for infusions, and/or the use of filter needles.

In conclusion to this work, we have shown that rubber coring can be partially reduced by the use of an adapted puncture technique, and totally eliminated by using blunt metal needles.