Visual and subvisual evaluations

At each moment of the study, a visual evaluation was carried out on a white and a black background with a flashlight, with the unaided eye by two different technicians, before mixing, after mixing and after 1- and 4-h storage as recommended by the European Consensus Conference [1].

The subvisual aspect was analyzed by measuring the turbidity using a UV-visible spectrophotometer (SAFAS mc², Monaco). This method allows to evaluate the formation of microparticles in solution. The absorbance light was scanned at three wavelengths (350, 410 and 550 nm), also recommended by the European Consensus Conference [1]. A change in the initial absorbance of more than 0.1 between the measurement after mixing and after the 4-h of storage was considered as a turbidity and so incompatibility of the two drugs involved.

A second subvisual test was performed applying the Light Obscuration Particle Count Test (PAMAS SVSS). The particle counter is based on the principle of light blocking which allows the automatic determination of the number of particles according to their size. Requirements exist for the number of particles larger than 10 µm and larger than 25 µm. According to European Pharmacopoeia Reference Standards [10], for containers of ≤100 mL, the values to be respected were ≤6,000 particles of ≥10 µm and ≤600 particles of ≥25 µm per container. For containers of >100 mL, the values were ≤25 particles of ≥10 µm per milliliter and ≤3 particles of ≥25 µm per milliliter. Measurements were carried out in 10 mL glass tubes. Before each measurement, the outer surfaces of the glass tube were cleaned with a water jet from E-POD^® “Pure Water Remote Dispenser”. Before proceeding with particle counting, the contents of each sample were mixed by 20 successive slow inversions of the container, then the mixture was left to stand for 2 min to avoid the possible presence of bubbles which may be generated by agitation. A first portion of 1 mL of the mixture was removed and the number of particles was evaluated on the three next portions of 1 mL. The mean number of particles of these three portions (total volume of 3 mL) represented the number of particles in the mixture. Between each measurement, the particle counter tubing was flushed with two portions of 1 mL of water from the E-POD^® “Pure Water Remote Dispenser”. Measurements for each pair were repeated three times and the mean number represented the particle count.

The last test performed at any time was the pH measurement (Bioblock Scientific pH meter) of each solitary drug before it was combined, and then after having mixed the drugs in pairs. For each pair, the pH had to show no change of more than 0.5 pH unit within 4 h to be stable.

These compatibility tests were assessed at different times: after mixing the two drugs (T 0H), 1 h after preparation (T 1H), and after a 4-h storage (T 4H), as 60% of the compatibility studies published in the literature are carried out over a 4-h period [11].

Acetylsalicylic acid

Acetylsalicylic acid is a drug compatible in pairs with 80.0% (8/10) of the drugs tested presented in Table 2. The acetylsalicylic acid/potassium canreonate mixture was incompatible because of visible precipitation at the 9/1 ratio after preparation and at the 1/1 ratio after a 1-h storage. Practices also revealed the frequent combination of acetylsalicylic acid with Nutryelt^®, which is a nutritional intake of micro-nutrients: iron, copper, manganese, zinc, fluorine, iodine, selenium, chromium, molybdenum [12]. The acetylsalicylic acid/Nutryelt^® combination in pair was pink at 9/1 and 1/1 ratios after the preparation within 4 h, while the ordinary Nutryelt^® color is yellow. In the absence of chemical data and by applying the precautionary principle, this mixture was considered incompatible.

Amiodarone hydrochloride

Results for mixtures of amiodarone hydrochloride with other drugs tested were 84.6% compatible (11/13). The amiodarone hydrochloride/cefazolin mixture at 1/1 and 1/9 ratios demonstrated turbidity at all times of analysis. The absorbance variation between T 0H and T 4H was 0.48 at 350 nm, 1.29 at 410 nm and 1.10 at 550 nm at the 1/1 ratio and 0.37 at 350 nm, 0.23 at 410 nm and 0.12 at the 1/9 ratio. Figure 1 shows that the mixture was whitish at the 1/1 ratio and opalescent at the 1/9 ratio from T 0H and at each time of analysis. The amiodarone hydrochloride/furosemide mixture was whitish at the 1/9 ratio at all times studied and also exhibited turbidity. The absorbance variation between T 0H and T 4H was 1.42 at 410 nm and 1.71 at 550 nm to the 1/9 ratio. These mixtures were also considered incompatible and should be administered separately.

Figure 1:

The amiodarone hydrochloride – cefazolin sodium pair was whitish at the 1/1 ratio (in the middle) and opalescent at the 1/9 ratio (right) at T 0H.

Some mixtures with amiodarone hydrochloride were visually and at spectrophotometer compatible, while subvisual particles were counted. In this situation, it concerned mixtures with bumetanide, cefotaxime sodium, potassium chloride, hydrocortisone sodium succinate, insulin, magnesium sulfate, nefopam, paracetamol, thiamine and vancomycine hydrochloride, detailed in Table 2. These poor particle counter results could be correlated to the production of gas bubbles in solution induced by polysorbate 80 in amiodarone hydrochloride which should not be considered as a particulate contamination. These results by the particle counter were also not interpretable, because this method was unsuitable to mixtures containing amiodarone hydrochloride [10]. Eur Ph suggests the use of method 2 (microscopic evaluation) for these preparations. This technic was not available in our laboratory, however, according to the absence of turbidity nor visible particle, these 10 mixtures were considered as physically compatible.

Atenolol

Regarding drugs mixed with atenolol, only 3/7 pairs (42.9%) were compatible. The atenolol/mycophenolate mofetil mixture showed a haze at the 1/1 ratio from T 1H. The atenolol/Nutryelt^® pair was incompatible because of the slightly yellow color of the mixture for the same reason as the acetylsalicylic acid Nutryelt^® pair. More than 6,000 particles of ≥10 µm were counted by the particle counter in the container of the atenolol/ciclosporine pair at all ratios at all times studied, except for the 1/9 ratio at T 4H. According to European Pharmacopoeia Recommendations, it caused the incompatibility of these two drugs in the same IV line. For the atenolol/metronidazole mixture, more than 6,000 particles of ≥10 µm per container were also counted at the 9/1 ratio at T 0H and T 1H. These three mixtures containing atenolol were also not considered to be physically compatible and should not be administered in the same infusion line.

Clonidine hydrochloride

Clonidine hydrochloride results were compatible for 50.0% of the mixtures tested (3/4). The clonidine hydrochloride/methylprednisolone pair was incompatible due to particle count at the 1/9 ratio outside of European Pharmacopeia Recommendations at each time of the analysis. The clonidine hydrochloride/mycophenolate mofetil mixture showed more than 6,000 subvisible particles of ≥10 µm and more than 600 subvisible particles of ≥25 µm per container at 1/1 and 1/9 ratios at each time. As amiodarone hydrochloride, mycophenolate mofetil contains polysorbate 80 in its excipients producing gas bubbles, which could distort results by the light obscuration particle test. Excluding the particle counter results which were not operable, this mixture was physically compatible by spectrophotometry and by the visual evaluation.

Furosemide

Among the mixtures tested with furosemide, only the furosemide/paracetamol pair was compatible, representing 25.0% of the mixtures (1/4). Opalescence and white mixtures were observed for the three other pairs. The furosemide/magnesium sulfate pair showed turbidity at the 1/1 ratio with an absorbance variation of 0.43 from T 0H to T 4H at 550 nm. For the furosemide/nefopam pair, the absorbance variation from T 0H to T 4H was 0.19 at 410 nm and 0.45 at 550 nm at the 1/9 ratio and 1.93 at 550 nm at the 1/1 ratio. This mixture showed whiteness at the 1/1 ratio and opalescence at the 1/9 ratio after the preparation, visible on Figure 2. The mixture furosemide/spiramycine was white and produced haze at the 1/1 ratio with an absorbance variation of 0.78 at 410 nm and of 0.56 at 550 nm from T 0H to T 4H. This pair was opalescent at the 9/1 ratio. The subvisual evaluation made it possible to classify the pairs of furosemide/magnesium sulfate, furosemide/nefopam and furosemide/spiramycine as incompatible mixtures for the concentrations studied.

Figure 2:

For the furosemide – nefopam mixture, the 1/9 ratio (right) was opalescent and the 1/1 ratio (in the middle) was whitish after its preparation (T 0H).

Heparin sodium

The heparin sodium/paracetamol pair was the only pair with heparin sodium that was physically compatible (25.0%, 1/4). Other results for heparin sodium mixtures were wrong. The heparin sodium/levetiracetam mixture suggested the presence of subvisual incompatibility at the 1/9 ratio at T 0H and T 1H with more than 6,000 subvisible particles of ≥10 µm per container counted, same for the heparin sodium/midazolam pair at 1/1 and 1/9 ratios at T 0H and T 1H, and for the heparin sodium/nefopam pair at the 1/9 ratio at T 0H and T 1H. Based on European Pharmacopeia Recommendations, these three pairs could not be considered as compatible even if the results of the particle counter at T 4H were in conformity, because in practice, the mixture of drugs is supposed to be infused to the patient after its preparation.

Insulin

Regarding the mixtures made with insulin, 71.4% of them were compatible (5/7). The insulin/pantoprazole sodium pair was incompatible because of haze at the 1/9 ratio showing an absorbance variation of 0.46 at 350 nm from T 0H to T 4H. The insulin/remifentanil pair was also incompatible because of more than 6,000 subvisible particles of ≥10 µm per container counted at T 0H at all ratios, T 1H at 1/1 and 1/9 ratios, and at T 4H at the 1/9 ratio.

Nefopam

Of the 20 mixtures made with nefopam, 95.0% were physically compatible (19/20). Only the nefopam/cefotaxime sodium pair was incompatible due to a count of more than 6,000 subvisible particles of ≥10 µm per container at the 1/9 ratio at each time of the analysis.

Pantoprazole sodium

None of the mixtures obtained with pantoprazole sodium were compatible (0/6). The pantoprazole sodium/paracetamol pair became yellow at 1/1 and 1/9 ratios at T 4H and presented turbidity at the 1/9 ratio with an absorbance variation of 0.24 at 350 nm and of 0.13 at 410 nm from T 0H to T 4H. The pantoprazole sodium/pyridoxine mixture changed from white to orange after a 1-h storage at all ratios, and precipitated at the 9/1 ratio, visible on Figure 3. This pair also was incompatible at the 1/1 ratio with an absorbance variation of 0.87 at 350 nm, 1.94 at 410 nm and 0.83 at 550 nm, and at the 1/9 ratio with a variation of 0.98 unit at 350 nm, 0.92 at 410 nm and 0.27 at 550 nm. The pantoprazole sodium/sodium selenite pair also showed turbidity at 9/1 and 1/1 ratios with an absorbance variation at the 9/1 ratio of 0.15 at 350 nm and at the 1/1 ratio of 0.13 at 350 nm, while no change was visible to the naked eye. The pantoprazole sodium/thiamine mixture became orange at the 1/9 ratio at T 4H and the absorbance variation at this ratio was 1.85 at 350 nm, 0.68 at 410 nm and 0.37 at 550 nm. The pantoprazole sodium/phloroglucinol pair precipitated at all ratios after a 1-h storage. The pantoprazole sodium/phytomenadione mixture produced more than 6,000 particles of ≥10 µm per container at all ratios and at each time of the analysis leading to the incompatibility of this mixture according to European Pharmacopeia Recommendations.

Figure 3:

Pantoprazole sodium – pyridoxine turned from white to orange after a 1-h storage and precipitated at the 9/1 ratio (left: T 0H, middle: T 1H, right: T 4H).

Table 3 summarizes the results of combinations with these nine main drugs presented above and allows the conclusion of an overall compatibility of 68.0% for all the mixtures included in this study (51/75).

Use of main drugs in ICUs

Systemic anticoagulation such as acetylsalicylic acid reduces mortality in patients on mechanical ventilation, decreasing cardiovascular events and preventing organ failure [13].

Amiodarone hydrochloride is widely used in ICUs for the treatment of arrhythmias, especially since its use in combination is usually effective and safe [14].

Atenolol is a commonly used β-blocker in the treatment of cardiovascular diseases and to reduce the risk of re-infarction and the related mortality after myocardial infarction. Its cardioselectivity reduces the risk of inducing a respiratory deficiency problem, making atenolol a drug of first choice in ICUs [15].

Clonidine hydrochloride is usually used in ICUs for its sedative, analgesic and anxiolytic effects, outside of its approved indications [16]. It also gets combined with other drugs in the same IV route.

Furosemide is used in ICUs to reduce short-term patient mortality. It helps limiting excessive fluid gain, while excessive fluid administration in post-operative and post-traumatic patients is a common occurrence [17].

Injectable anticoagulation such as heparin sodium is one of the most commonly prescribed therapies in ICUs. A low dose of heparin sodium is often needed to prevent a venous disease, which can contribute to the deterioration of organ functions. Higher doses treat a venous disease or an acute coronary syndrome [18].

Critically ill patients frequently develop high levels of sugar in the blood. An intensive insulin therapy with a tight blood sugar control reduces patient mortality in ICUs [19].

Nefopam is a centrally acting non-opioid analgesic agent commonly used for patients admitted in ICUs getting mechanical ventilation and receiving sedative and analgesic medications. They are integral parts of the complex management of these patients, to minimize their discomfort and to reduce the risk of agitation and accidental self-extubation [20]. Nefopam could be a safe and effective pain reliever in regards with the high risk of organ vulnerability of patients [21].

Critically ill patients in ICUs are at risk of developing gastrointestinal bleeding due to mechanical ventilation and coagulopathy. A low dose of heparin sodium is often required to prevent a venous disease, which can contribute to the deterioration of organ functions and liver or kidney failure. Proton-pump inhibitors, such as pantoprazole sodium are the most commonly used prophylactic acid suppressants in these circumstances [22].

Observational study

In this study, only mixtures which were visually and subvisually compatible by UV-visible spectrophotometer were tested using the particle counter, otherwise they were physically incompatible and the latter was not performed.

Sometimes, Isofundine^® was employed as a drug in combination with another drug, e.g., acetylsalicylic acid/Isofundine^®, sometimes it was used as a solvent, e.g., for the dilution of nefopam. Tests were not repeated on different batches of medications.

For combinations with amiodarone hydrochloride and mycophenolate mofetil, the presence of polysorbate 80, as an excipient contained in the formulation of these two drugs, led to the production of gas bubbles which could have been wrongly detected as particles by the particle counter. The production of bubbles was not due to a particulate contamination. For this type of mixture producing bubbles, a more suitable particle counting method would be a Microscopic Particle Count Test [10].

Large differences in pH between two drugs studied in Y-site were noticed, shown in Table 2. Pantoprazole sodium (4 mg/mL in 0.9% NaCl) pH was 9.70, while the pH of other drugs mixed with pantoprazole sodium was between 2.84 and 5.80. All seven mixtures containing pantoprazole sodium were incompatible. Furosemide had also an alkaline pH of 8.96 (in pure solution at 10 mg/mL resulting in large differences of pH with other mixed drugs. Four of five mixtures containing furosemide were incompatible. The combination of two drugs with opposite pH can also be a cause of drug incompatibility which should dissuade us from mixing two drugs with distant pH unless their Y-site compatibility is certain. This hypothesis is currently under study in our laboratory.

Data available in the literature

An incompatibility with the Light Obscuration Particle Test could equally represent an evolution of size, shape or count of particles, while the UV-visible spectrophotometer is useful for assessing subvisible aggregation and correlates to a visual incompatibility such as opalescence [1].

Care should be taken regarding intravenous administration of furosemide, as many drugs can interact with furosemide ranging from opalescence to precipitation, especially in combination with acid drugs, which can lead to vessel obstructions [23].

According to the summary of product characteristics, pantoprazole sodium powder should not be mixed with other drugs except for its dilution in 0.9% sodium chloride [24]. It was already known from other studies that pantoprazole sodium was often incompatible with other drugs [1]. No administration of mixture with pantoprazole sodium should be observed unless compatibility data can ensure this.

Many other combinations with heparin sodium and pantoprazole sodium are known to be incompatible in the literature. Heparin sodium is also a drug which must be administered alone when compatibility data are not available [25].

Limits

Only the physical compatibility has been tested. It cannot conclude on the chemical stability of these binary combinations, nor can it be extrapolated to associations of more than two drugs, nor to mixtures achieved in the same vial.

Physical compatibility studies should be performed for other drugs mixtures for which data are lacking. New requests must continue to be collected in ICUs and other units to achieve new Y-site compatibility assessments. In all cases, drug combination should always be administered with caution taking into account qualified data from the literature.