Patients admitted to intensive care units (ICU) often require the IV administration of several drugs. Vasoactive drugs, analgesics, and sedatives are among the most widely used therapeutic groups and are usually administered in continuous infusion.

According to the systematic review conducted by Moyen et al. there is an average 1.7 errors/day associated with the process of drug administration in the ICU setting.1 On the other hand, the data reported by Merino et al. in a study conducted among Spanish hospital ICUs are a little better (1.13 medication errors for every 100 patients/day).2 Even so, medication errors are common in ICUs and require care from healthcare providers to minimize them.

Errors in the administration of drugs in ICUs are due to several factors: the use of high-risk drugs (vasoactive drugs, inotropes, sedatives, etc.) often administered in low doses due to their high drug strength, requiring dilution and a prior assessment to their administration. Another factor is the prescription of doses in different units of measurement or the high number of drugs used with each patient. Although it is an important advance with regard to safety, the use of intelligent infusion pumps has been associated with an important number or medication errors due to programming issues.3

The combination of these risk factors increases the chances of making mistakes in the most vulnerable patients due to their severity. Critically ill patients often have limited venous accesses. This means that different drugs are delivered using the same route of administration, which increases the risks involved when mixing incompatible drugs. The mix of incompatible drugs is a medication error that can have serious consequences for the patient such as therapeutic failures, micro-embolism or toxicity.4

The Y-site infusion of 2 drugs requires both drugs to be physically compatible.5 This coadministration occurs when mixing drugs in a 1:1 ratio and in the absence of visible signs of incompatibility like precipitation or change in color. On the other hand, for the safe coadministration of 2 drugs in the same diluent, the mix needs to be chemically stable. This means prior confirmation is needed that no significant change has occurred in the concentration of either one of the drugs present in the mix.6

Standardizing the concentration of infusion solutions is one of the most useful measures to prevent medication errors in the ICU setting, especially in high-risk drugs due to their potential to cause severe damage and because they have the highest incidence of medication errors.

Another highly recommended measure for the safe administration of drugs is having reliable information available on drug compatibility when administering common drugs in critically ill patients. However, information on drug compatibility is scarce and, on many occasions, difficult to interpret due to the different concentrations used, the lack of information on the assessment techniques used or the suspicious technical quality of the sources. The lack of information on the safe mix of 2 drugs creates problems in the daily work of ICU nursing teams. Added to the risk of complications associated to the administration of 2 incompatible molecules, this lack of information can make the nurse have to look for new venous accesses to administer the drugs separately whichincreases the risk of infectious or thromboembolic complications.

The goal of this review is to gather the information published on the physical and chemical compatibility of the most commonly used drugs at an ICU when infused through the same line via a Y-site. Also, to assess the quality of the information published and generate a compatibility chart with reliable and updated information to improve safety in the administration of drugs to critically ill patients.

Making sure that the use of drugs is safe is one of the main commitments made by healthcare providers with their patients. In the ICU setting and given the huge amount of IV drugs administered and the patients’ limited number of routes of administration, this safety is sometimes compromised due to the risks involved when co-administering incompatible drugs in especially vulnerable patients.

Online databases like Stabilis 4.0 are very useful to look for information on drug compatibility. However, the personnel administering the drugs finds charts much more useful because they can quickly look at the information they need at a given time. This is especially interesting in urgent situations when any delays caused by the healthcare providers can have consequences in the patient.

This review focused on analyzing the physical and chemical compatibility of the IV drugs most commonly used through Y-site infusion in the ICU setting and summarizing the information obtained in a double-entry chart. Physical compatibility studies are the most common of all because they are easy to conduct. Chemical stability studies, however, are not because they require more sophisticated analytical techniques to determine the initial and final concentration of drugs.

Despite this, the number of drug combinations studied is still insufficient. As Fig. 2 shows we could not find any information on the physical and chemical compatibility of all the combinations suggested; for instance, in the case of flumazenil and piperacillin-tazobactam we could only determine stability with 4 drugs and in both cases the 39 remaining combinations remained with no information.

Even if we took all the possible combinations suggested into consideration and added the new data found, we would still have zero information on the physical and chemical compatibility of 470 combinations. This means that we only have data available for 50.3% of all the possible combinations suggested.

The most problematic combinations regarding incompatibility are drugs whose stability is closely linked to the pH interval; this is the case with sodium bicarbonate, furosemide or pantoprazole. Furosemide, for example, requires a basic pH to guarantee the stability of the molecule in solution, which is why the mix with acid drugs (pH<4) causes turbidity and precipitation.12

The presence of adjuvants in the pharmaceutical formulation, the concentration and exposure to extreme temperatures or luminosity are other factors associated with drug incompatibility.13 There are times when a given drug combination can be stable in a certain diluent and incompatible in another; for instance, dopamine is only compatible with amiodarone when both are dissolved in glycosylated serum at 5% because the latter in unstable in saline solutions at 0.9%. Thus, if this allegedly compatible mix is performed in physiological serum, a loss of concentration of amiodarone can occur with the corresponding risk of lack of therapeutic response.

On the other hand, in many cases, the quality of the studies published so far can be better. It would be good to have greater uniformity in the quality standards of this type of studies. For example, even though the pH is a critical factor in the stability of drugs in solution, it was only verified in 12 of the 27 papers. Similarly, turbidimetry or microscopy—more accurate techniques than visual observation for the detection of particles and changes in color—are underused. Over the last few years, several experts have published guidelines for the design of drug stability studies.8–11 We can only hope that this will improve the overall quality of this type of studies in the future.

Former authors have published reviews of these characteristics. For instance, Flamein et al.14 studied this problem in neonatal ICUs; Knudsen et al.15 shed light on the compatibility of analgesics and sedatives. Our review is based on the previous work done by Kanji et al.5 in Canada and López-Cabezas.7 in Spain. It has been completed with the new information available on drugs in our setting and data on the most widely used concentrations of drugs.

Overall, we found information on 82 new drug combinations from 27 different references including combinations of 3 beta-lactam antibiotics (ceftazidime, meropenem, and piperacillin-tazobactam) widely used at the ICU setting. Over the last few years the pharmacokinetic advantages of a prolonged perfusion route of administration of these 3 antibiotics have been confirmed.16–19

Perfusions at drug concentrations that exceed the usual ones are often used in the critically ill patient. In this sense, we could not find data on all drug combinations regarding the high concentrations used in the ICU setting (Table 1); however, in some cases, we did obtain information on lower concentrations than the ones reported in this review. These cases are shown on the compatibility chart (Fig. 2) as conditioned compatibility (I/C), that is, that the combination had been studied at a concentration different from the standard one.

The stability data reported in this review cannot be generalized to other drug combinations or concentrations different from the ones described. Also, the information provided is in regard to 2 drug combinations, and incompatibilities may be present with>2 drug combinations at a time, which is highly not advisable. Nevertheless, the drugs and concentrations selected are the most widely used in the adult ICUs of most hospitals.

Until we have new and better compatibility studies that shed some light on this issue, this review can be an easy-to-read update on the evidence available on the compatibility of the drugs most commonly used at the ICU setting. Its goal is to contribute to the safe administration of drugs to patients who can face the consequences of greater severity due to their frailty.

Genís Castells Lao: study design and idea, data mining, analysis and interpretation of data; paper draft or critical review of the intellectual material; and final approval of this version.

Montse Rodríguez Reyes: study design and idea, data mining, analysis and interpretation of data; paper draft or critical review of the intellectual material; and final approval of this version.

Judit Roura Turet: data mining, analysis and interpretation of data; paper draft or critical review of the intellectual material; and final approval of this version.

Marta Prat Dot: data mining, analysis and interpretation of data; paper draft or critical review of the intellectual material; and final approval of this version.

Dolors Soy Muner: study design and idea; paper draft or critical review of the intellectual material; and final approval of this version.

Carmen López Cabezas: study design and idea, data mining, analysis and interpretation of data; paper draft or critical review of the intellectual material; and final approval of this version.

The authors declared no conflicts of interest whatsoever.