Article Text

Download PDFPDF
Classification of non-antineoplastic intravenously administered drugs according to their toxicity risk: the path towards safe drug administration
  1. Teresa Giménez Poderós1,
  2. Juan José Fernández Cabero2,
  3. Marta Valero Domínguez1
  1. 1 Pharmacy Department, Hospital Universitario Marques de Valdecilla, Santander, Cantabria, Spain
  2. 2 Faculty of Pharmacy, Universidad de Salamanca, Salamanca, Castilla y León, Spain
  1. Correspondence to Dr Teresa Giménez Poderós, Pharmacy Department, Hospital Universitario Marques de Valdecilla, Santander 39008, Cantabria, Spain; teresa.gimenez{at}scsalud.es

Abstract

Objective Extravasation is a potential complication resulting from parenteral administration of drugs. The purpose of this study was to characterise the physicochemical properties of non-antineoplastic parenterally administered drugs and determine their potential to cause a toxic effect on tissue.

Methods A list of drugs administered by intermittent or continuous intravenous (IV) infusion was prepared. A database was also established to collect information from the literature. Each active substance was classified according to its risk to cause tissue damage using the following criteria: (1) High risk: active substances presenting with any of the following characteristics: osmolarity of the IV solution form >500 mOsm/L; vasoconstriction; vesication; cellular toxicity; very common, common or uncommon adverse events such as phlebitis, necrosis or pain at the site of administration according to the Summary of Product Characteristics. (2) Moderate risk: active substances where the pH range was <3 or >11 or where adverse events at the site of administration occurred rarely, very rarely or with unknown frequency. (3) Low risk: active substances where the osmolarity of the IV solution was <500 mOsm/L and the pH ranged between 3 and 11. These active substances did not cause vasoconstriction, neither were they classified as vesicant or cytotoxic or presented with adverse events at the site of administration.

Results The risk classification list included 138 active substances, of which 86 were classified as ‘high risk’, 18 as ‘moderate risk’ and 34 as ‘low risk’.

Conclusion The classification of intravenously administered drugs according to their risk profile is useful to ensure their safe use, as it can be used to implement the necessary safety measures to prevent adverse events.

  • Drug Administration Routes
  • Drug Compounding
  • Administration, Intravenous
  • Safety
  • PHARMACY SERVICE, HOSPITAL

Data availability statement

Data are available in a public, open access repository. ‘Not applicable’.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What is already known on this topic

  • Extravasation is a potential complication resulting from parenteral administration of drugs.

  • Identifying and recognising irritant or vesicant medications is necessary to prevent the negative consequences of extravasation and to implement the safety measures.

What this study adds

  • A classification of intravenously administered drugs according to the risk of tissue toxicity.

How this study might affect research, practice or policy

  • Depending on the risk of each drug, measures must be implemented to prevent damage in case of extravasation.

  • The methodology used in this study can be applied systematically as a quality assurance measure with any drug.

Introduction

Extravasation is an adverse event related to the parenteral administration of intravenous (IV) therapy. It occurs when an irritant or vesicant injectable drug accidentally leaks into the perivascular or subcutaneous space instead of remaining in the vascular space.1 Extravasation may result in lesions ranging from slight local irritation and erythema to ulcerations and necrosis. It may also result in chronic pain, impaired mobility, irreversible neuropathy, loss of function and even amputation of a limb or death. Extravasation worsens patients’ health outcomes through the iatrogenic damage it causes, increasing their morbidity, mortality and the cost and duration of their stay in hospital.

Extravasation can be caused by mechanical factors (eg, site of catheter, small size, poor condition of veins, larger catheter size relative to vein size), physiological factors (eg, thrombus or fibrin sheath at the catheter tip, lymphoedema) and pharmacological factors (eg, cytotoxic drugs, direct vasoconstrictors or medicines exposing the cells to osmotic stress or to a non-physiological pH).2 Mechanical factors occurring either during initial catheter insertion or while the catheter is in place and physiological factors relating to pre-existing or emerging vein problems can be contributing factors. Regardless of the mechanism, specific management of infiltration and extravasation is usually determined by the pharmacological characteristics of the offending infusion.

There is no well-recognised standard definition of a ‘cytotoxic drug’.3 Cytotoxicity is primarily associated with antineoplastic drugs and they are subject to relatively well-defined management strategies in hospitals. For many antineoplastics, a drug is considered cytotoxic by its mechanism of action (direct disruption of DNA structure or mitotic function, interference with cell growth or proliferation, or with DNA synthesis) to cause cell death. Other non-antineoplastic intravenously administered drugs such as antibiotics have caused cellular toxicity by causing depletion of intracellular ADP and ATP levels and other enzymes that sustain cell function, leading to cell death.2 4 Irritant drugs can cause pain, local inflammation, burning and phlebitis but they never cause necrosis.5 This is what differentiates them from vesicant drugs that cause irritation, ulceration and necrosis. Any drug that has caused necrosis during extravasation at least once is classified as a vesicant agent.

Identifying and recognising irritant and vesicant drugs is essential to prevent the consequences of extravasation,6 enabling organisations to introduce policies and procedures to address the prevention and management of drug extravasation.

The purpose of this study was to characterise the physicochemical properties of parenterally administered non-antineoplastic drugs most commonly given to adult patients and to determine their potential to cause toxicity.

Methods

A literature search of MEDLINE/PubMed was performed to establish criteria for the classification of drugs on the basis of their potential to cause tissue damage. After reviewing the literature,2 7–9 each active substance was classified according to its potential to cause tissue damage as a result of its pH, vasoconstricting properties, cytotoxicity and theoretical osmolarity. Information about the incidence of adverse events such as phlebitis, necrosis or pain at the site of administration was also included.

For classification of the risk, we used the following consideration:

  • Drugs with extremely low or high pH levels are classified as moderate risk because pH is not considered an independent risk factor influencing vessel health, but together with other predisposing factors it may contribute to harmful outcomes.8 9

  • A drug is classified as cytotoxic if it is considered a hazardous drug in the Spanish National Institute for Occupational Safety and Health review and cytotoxicity has been proposed as a mechanism of tissue injury in the literature review.

The classification was as follows:

  • High risk: active substances were classified as high risk if they presented with any of the following characteristics: osmolarity of the IV solution form >500 mOsm/L; vasoconstriction; vesication; cellular toxicity; very common, common or uncommon adverse events such as phlebitis, necrosis or pain at the site of administration according to the Summary of Product Characteristics (SmPC).

  • Moderate risk: active substances whose pH range was <3 or >11 or which were associated with adverse events such as phlebitis, necrosis or pain at the site of administration occurring rarely, very rarely or with unknown frequency.

  • Low risk: active substances where the osmolarity of the IV solution was <500 mOsm/L and the pH ranged between 3 and 11. These active substances did not cause vasoconstriction, nor were they classified as vesicant or cytotoxic or presented with adverse events at the site of administration.

A list of drugs administered by intermittent or continuous IV infusion was prepared. Drugs requiring direct IV administration or belonging to specific therapeutic areas such as oncology, haematology or radiology were excluded.

A database was created to collect the information necessary to classify each active substance. Two authors (TGP and JJFC) independently collected the following information:

  • Active substance.

  • National code, trade name.

  • Excipients.

  • pH.10–13

  • Mechanism of action: vasoconstriction.10

  • Adverse events at site of administration, frequency: pain, phlebitis and/or necrosis following extravasation.10

  • Ability to cause vesication.5 13

  • Cytotoxicity.14 15

  • Theoretical calculation of osmolarity:

    • Drug (salt or non-ionised form): dose, volume (if the dosage form was an injectable solution) or volume following reconstitution with water (if the dosage form was a powder), molecular weight,16 number of particles formed in each solute by dissociation of the drug.10

    • Solvent: type (sodium chloride 0.9% or glucose 5%) and volume.

    • Total volume of the mixture.

    • Value of theoretical osmolarity.

For each active substance, one or more concentrations were proposed based on the recommendations of its SmPC and on the levels established in standardised mixtures or used regularly in the hospital setting. For drugs with weight or plasma level-dependent dosing schedules, the dose was calculated either for a patient weighing 70 kg or regular doses were established based on the standard dilution recommended by the compounding protocols of the hospital pharmacy department. For some drugs, alternative dose concentrations were suggested to cover a broad range of clinical and fluid loading scenarios, which is standard practice in intensive care units. All drug concentrations were established by diluting the active substance in sodium chloride 0.9% or glucose 5%, as those are the most common IV mixture solvents used in hospitals. Parenteral dosage forms supplied as ready-to-use solutions were also included; the solvent was considered to contain sodium chloride 0.9% or glucose 5%, depending on the excipients used.

Osmolarity is the total concentration of all the solute particles active in the solution.17 18 As osmolarity is additive, the total osmolarity of an IV mixture is the result of adding up the osmolarity value of each component. The formula used to calculate the theoretical osmolarity value was:

Embedded Image

where M is the molar concentration of the solute in the solution and n is the number of particles formed by the dissociation of one molecule from each solute.

A search was performed on PubChem16 of the active substance (or its salt form) to find out how many particles were formed following dilution (2D structure section).

The following formula was used to calculate the molar concentration of each component (M):

Embedded Image

where M is the molar concentration expressed in mol/L, concentration is grams of the active ingredient divided by the volume of the solution, expressed in g/L, mw is the molecular weight expressed in g/mol and mOsm=milliosmol.

Taking into account these considerations, the following procedure was applied to each of the IV mixtures:

  1. The volume (in mL) of each component was multiplied by the mOsm/mL value of the component. It was considered that each mL of saline solution 0.9% contributed 0.31 mOsm/mL and that each mL of glucose 5% contributed 0.278 mOsm/mL. The mOsm/mL value was not calculated for excipients.

  2. The mOsm values obtained for the different components were added up to determine the total number of mOsm in the mixture (total mOsm).

  3. The volumes of the components were added up to find the total volume of the mixture.

  4. The total number of mOsm values obtained at step 2 was divided by the total number obtained at step 3 and the result was multiplied by 1000 to obtain an estimation of the osmolarity of the mixture in mOsm/L.

It is a known fact that the formula used to determine drug-solution osmolarity is not accurate, as osmolarity is best determined by direct measurement via osmometry. Indeed, excipients used to preserve the efficacy and/or stability of the drug may significantly alter the theoretical osmolarity provided by the formula. It is known, for example, that some excipients such as polyethylene glycol are hyperosmotic.19 To identify other potentially hyperosmotic excipients, a comparison was made between the osmolarity obtained by application of the formula and either the osmometry measurements provided by Manrique-Rodríguez et al or those contained in the drug SmPC, taking into consideration the osmolarity of the excipients used in the different drugs studied. The result of this comparison was used to make a final assessment of the extent to which the formula above can provide valuable information to healthcare providers and carry out a final classification of the mixtures into those above and below 500 mOsm/L.

Results

A list was obtained of the risks associated with a total of 138 active substances (see online supplemental table 1). Overall, 86 of the 138 substances were classified as high risk, 18 as moderate risk and 34 as low risk.

Supplemental material

Following the review of SmPCs for information on extravasation-related side effects (phlebitis, necrosis or pain at the site of administration), it was found that 58 active substances were classified as high risk. The drugs with the highest incidence of adverse events were octreotid, pentamidine and propofol.

The pH value assigned to each active substance (except ready-to-use medicines) corresponded with the value indicated in the SmPC of the drug used to compound the IV solution. Based on the pH value, it was found that nine active substances had extreme pH values of <3 and >11 (moderate risk).

According to the literature reviewed, 28 active substances were classified as vesicant, 11 were vasoconstrictors and 24 were cytotoxics.

For each active substance, theoretical osmolarity was calculated for at least one IV mixture. Overall, 272 IV mixtures were reviewed, of which 27 were ready-to-use mixtures or mixtures perfused intravenously in undiluted form. Following application of the formula, the theoretical osmolarity values obtained were compared with those produced by osmometry. For ready-to-use mixtures or those that are not diluted prior to administration (see online supplemental table 2), the osmolarity values obtained with our methodology were the same as those contained in the drug SmPCs and/or the value reported by Manrique-Rodríguez et al, with the exception of four drugs (clevidipine, lipids, propofol 2% and nimodipine). Osmolarity values could not be determined for these four drugs as an analysis of the excipients was not able to establish the presence of glucose 5% or saline 0.9% as solvents. Online supplemental table 2 shows the calculation of the theoretical osmolarity of nimodipine, considering that saline 0.9% was used as solvent. The calculation yielded a value <500 mOsmol/L, which contrasts with Manrique-Rodríguez et al who reported a much higher value. This indicates the importance of excipients (in this case, ethanol). For the rest of the drugs (online supplemental table 3), theoretical osmolarity with our methodology was significantly lower than that measured by osmometry in the case of nitroglycerine, urapidil, phenytoin, levosimendan, ciclosporin and digoxin. Excipients used in these drugs are ethanol, propylene glycol and macrogolglycerol ricinoleate/polyoxyl 35 castor oil. On the other hand, the theoretical osmolarity values calculated were higher than those measured by osmometry for some magnesium sulphate and dipotassium phosphate mixtures.

Supplemental material

Supplemental material

Online supplemental table 1 shows the authors' final conclusion and classification specifying whether the osmolarity of the mixtures was above or below 500mOsmol/L, after analysis of the excipients and comparison of the osmolarities. In cases where it is not known to what extent excipients affect the theoretical osmolarity calculated for a drug, the final conclusion was marked as inconclusive. For example, no information was available on the osmometry-determined osmolarity of the IV solutions of alprostadil, argatrovan, itraconazole, posaconazole and tacrolimus, although it is known that they contain excipients that could affect their osmolarity. In other cases, such as levosimendan, a comparison was made between the theoretical osmolarity calculated with the formula and the value provided by osmometry, but at a concentration different from the one used in our hospital. As levosimendan contains excipients that could affect its calculated osmolarity, the value was classified an inconclusive.

Discussion

This study provides a broad overview of drugs used in hospitals classified according to their risk of toxicity. The information provided in this paper could serve as a useful tool to revise the current guidelines used by hospitals to administer the different drugs, including drug dilution and infusion protocols, and to implement the changes required for safe drug administration and prevention of adverse events. The methodology used here can be systematically implemented as a quality assurance measure when introducing new drugs into the hospital’s formulary.

This review, as well as the classification provided, is associated with significant limitations. The first is that, as the amount of information provided by drug SmPCs in Spain is rather limited, SmPCs issued in other countries were used although the drugs in some cases were manufactured with different excipients and identified under different trade names. Another limitation is related to the fact that the pH values reported here correspond to the active substance used as a basis for the IV mixture administered to the patient but not to the mixture itself, which could differ as a result of the dilution process; pH differences have also been observed across different manufacturers. Nevertheless, Manrique-Rodríguez et al 13 showed that there were no significant pH differences between mixtures containing different concentrations of active substances or between different trade names (based on an analysis of five drugs). They did, however, point out that an IV solution was slightly more acid when glucose 5% is used as a solvent. A comparison between the pH assigned to the drug on the basis of the SmPCs and those obtained for the IV mixtures by Manrique-Rodríguez et al 13 did not show significant differences, except for those indicated in online supplemental table 4, which do not affect our risk classification. Although not the subject of this study, it should be mentioned that the choice of solvent is critical as it affects the stability of the mixture and the activity of the drug. For example, terlipressin is stable only at a narrow pH range (3–4) but, if the final pH of the mixture is significantly higher than 4, the activity of the drug may be severely impaired.20

Supplemental material

Although our calculation of osmolarity did not provide an accurate value, it did allow easy calculation of an approximate (theoretical) value, which is most welcome considering that osmometers are not always available in regular clinical practice. Excipients were one of the main limitations to calculating the overall osmolarity of a drug. Pereira et al reported that some drugs with osmolarities much higher than 1000 mOsm/kg are extremely hypertonic (eg, digoxin, phenytoin, phenobarbital) as a result of using propylene glycol as an excipient.19 This study reviewed other excipients such as ethanol which may alter the final osmolarity of the mixture. The osmolarity of excipients should have been established to obtain a more accurate theoretical osmolarity value of the mixtures. However, to do so it would have been necessary to know their concentration and, unfortunately, this information is not normally available in the SmPC.

Hypertonic solutions may cause discomfort/pain resulting from their hyperosmolarity, which may be caused by the active substance of a drug but also by its particular excipient(s). Adding this criterion to the classification could alert healthcare providers to the significant toxicity inherent in certain excipients. Such is the case for verteporfin, a drug for which no risk factors have been described apart from a high incidence of adverse events.

The literature on the impact of direct cellular toxicity from non-antineoplastic drugs is also scarce.8 Studies on in vivo validation of the cytotoxic potential of drugs, tolerated drug concentrations and others factors in infusion therapy such as duration of exposure are needed. The only information on extravasations is contained in case reports and reviews. This emphasises the need to report all of them with as much information as possible. An example of direct cellular toxicity is amphotericin B which, after the appearance of thrombophlebitis and chemical irritation following extravasation, was classified as a vesicant although it is not associated with high pH levels and is not hyperosmotic or vasoconstricting.1 Vancomycin has been classified as a vesicant or irritant depending on the concentration of the active substance in the IV solution,5 which shows that cytotoxicity appears to depend on drug concentration.2

After a careful and separate review of each drug in the list, practical advice can be provided to healthcare providers on the measures that could be taken to prevent the damage they may cause in case of extravasation.

Conclusions

The classification of intravenously administered drugs according to their risk profile is useful to ensure their safe use, as it can be used to implement the necessary safety measures to prevent adverse events. Measures may include the design of decision-making algorithms for the selection of the most appropriate vascular access, standardising IV administration of drugs (standard solutions, increased dilution, change of solvent, change in administration rate) and development of guidelines on extravasation of drugs.

The methodology used in this study can be systematically applied as a quality assurance measure whenever a new drug is introduced into a hospital formulary.

Data availability statement

Data are available in a public, open access repository. ‘Not applicable’.

Ethics statements

Patient consent for publication

Ethics approval

Not applicable.

References

Supplementary materials

Footnotes

  • EAHP Statement 5: Patient Safety and Quality Assurance.

  • Twitter @TeresaHUMV

  • Contributors TGP accepts full responsibility for the work and/or the conduct of the study, had access to the data, and controlled the decision to publish. JJFC was involved in the data collection and MVD was the supervisor of this research and contributed to the paper.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; internally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.