Understanding of the formulations of intravenous medicines and of the procedures involved in the care of children and neonates is essential for any pharmacist managing paediatric patients
Evelina Children’s Hospital
Intravenous (IV) administration of medicines presents special problems in children for a variety of reasons, including the small size of the patients’ veins, the small circulating volume and altered pharmacokinetics. Things that adult patients could easily tolerate, such as a little extra fluid or sodium can seriously compromise a pre-term neonate.
Moreover, there are some problems in day-to-day management of neonates that simply do not occur with adults. Simple procedures such as weighing or identifying the patient are not always straightforward. For example, if a 0.5 kg neonate is weighed while wearing a wet nappy its weight can be greatly increased. Correct identification can be problematic with very young children who, for example, may only be known as “twin Smith 1” and “twin Smith 2”.
Data concerning the true incidence of medication errors in paediatric patients are sparse. One study in Glasgow reported an error rate of 0.15% (or one error per 662 admissions).(1) On the basis of these figures it is estimated that 1,675 medication errors occur in paediatric inpatients in England each year and that 85 of these are likely to be moderate or severe. The real numbers are probably higher but one factor that probably influences low reporting is the extensive use of medicines outside their manufacturers’ licenses in paediatrics. There is still some reluctance to report medication errors where unlicensed products or products used for unlicensed indications are involved. In part, this reluctance is unfounded as it relates to concerns that it may not be regarded as best practice to use medicines outside of their licences – a concern that is not a reality.
Ross and colleagues(1) showed that the error rate was higher in intensive care areas and that errors were more common in younger children – 22% occurred in patients under one month old. IV doses accounted for 56% of all incidents and 8% of incidents were due to 10-fold calculation errors. More than 80% of incidents involved dose errors (including incorrect concentration, rate of administration, missed and extra doses) or administration of the wrong drug. About 10% of incidents concerned incorrect labelling. The medicines most commonly involved were antibiotics or antiviral agents, IV fluids, including parenteral nutrition, and anticancer drugs. These findings clearly suggest that IV treatment in paediatrics presents particular opportunities for error, so risk-reduction measures should be a priority in this area.
The first step is to identify the processes and procedures that are high-risk or error-prone.
Medicines designed for adults
Part of the problem is that most medicines are designed for adults. Often the dose in a single vial or ampoule is the “standard” adult dose. This assumes all adults are more or less the same size. In some ways, dosing children is more logical in that it is based on weight or body surface area, so in children every dose is different and there is no “expected” dose. This immediately throws up problems for those involved in prescribing or checking doses – there is no mental check as to whether a dose is the “expected” one. The situation is further complicated by questions such as: “When does a child become an adult and able to receive the adult dose?” By convention this is at the age of 16, but real-life 16-year-olds vary considerably in size and in practice a 12-year-old is often big enough for an adult dose.
Another question is, “When can doses be rounded up or down?” Unfortunately there is no simple answer to the question, but we know that the use of decimal points only adds to risk. For example, nobody thinks twice about giving an adult 150 mg of ranitidine and their size is never even considered. Paediatric formularies are very specific and say that the dose for children is 2 mg/kg. Ranitidine suspension comes as 15 mg/ml. Would it be wrong to prescribe 15 mg for a child who is 6.5 kg (6.5 x 2 mg = 13 mg)? In reality the child would be fine on 15 mg (1 ml) and yet most prescriptions written up will say 13 mg (0.866 ml). The problem with paediatric prescribing is not having a feel for the limits and it is important to weigh practicalities of administration against the need for safe dosing.
Examples of poor use of decimal points have been reported to my own hospital several times over the past few years. Perhaps the most memorable example is when decimal points are omitted for quantities less than 1. During the 1970s “preceding zeros” were seldom used (ie “.1 mg” was written instead of “0.1 mg”). Errors occurred with aciclovir dosing because the dose was adjusted for body surface area using a value of 6 m2 instead of 0.6 m2. This happened because the decimal point was not noticed on the chart used to calculate surface area. There is nobody in the world who has a body surface area of 6sqm, but once you have a number and multiply it by the dose then at some point it will be administered – and it was!
Another aspect of aciclovir dosing demonstrates that even licensed dosing is not always at all logical and can lead to confusion and potential error. Aciclovir is dosed at 10 mg/kg for those less than three months of age, 500 mg/m2 for those aged three months to 11 years and 10mg/kg for those older than 12 years. The result is that most 11-year-olds get bigger doses than 12-year-olds. It remains unclear what should be done for children who have their 12th birthdays in the middle of a course.
In a recent survey of injection stocks held in NICU at Evelina Children’s Hospital, we compared the quantities in the vials with the doses usually required. We found (in common with other studies) that a third of all injectables on NICU prescriptions call for a dose that amounts to less than one-tenth of the vial contents and nearly 5% require less than one-hundredth of the vial contents. When a doctor or nurse has to open multiple vials to give a dose, mental alarm bells often start to ring, but if one ampoule or vial contains a 100-fold overdose there is no such alarm, and the potential for error is huge.
The pharmaceutical industry must be encouraged to develop paediatric-friendly concentrations and vial or ampoule sizes to overcome this problem. Until things change it would seem very prudent to look at systems such as the use of CIVAS, which have built-in extra checks to avoid these far too easily occurring errors.
In addition to all the usual considerations for safe purchasing of medicines, such as clarity of labelling and ease of use, there are special safety considerations when purchasing injectable medicines for children. Problems can arise due to excipients or additional ingredients in medicines that are primarily designed for adult use. Until recently, licensed preparations (licensed in country of origin) of ibuprofen injection designed for adults, but used for closure of patent ductus arteriosus in neonates, were imported into the UK. Some of these contained lignocaine, which made them completely unsuitable for infusion into premature neonates as there was a risk of cardiac and central nervous system side-effects.
It is essential that paediatric and especially neonatal requirements are borne in mind whenever a new preparation is purchased into a hospital trust or a supplier changes. Labelling, strengths and excipient requirements may all have risk issues in these populations that differ from those in the adult population, so good liaison between purchasers and paediatric pharmacists is crucial to good risk management.
IV administration in neonates is fraught with problems. Compatibility issues are important because IV access is so hard to get and maintain. There is often only one IV line through which all doses must be given – I have known as many as nine infusions, including total parenteral nutrition (TPN), go down one line!
Another common problem is extravasation as neonatal veins are so fragile. By the time a problem is seen a whole limb may be affected – visualise a needle strapped to an arm only a few inches long. The best way to tackle these problems is to work with a limited number of drugs and clear policies.
An example of an unexpected mishap due to a change in infusion handling has been seen with enoximone administered via a scalp vein. Enoximone is known to be very unstable; we have seen cases of correct delivery, but due to the slow circulation of the scalp vein the enoximone actually crystallised along the length of the vein, leaving a solid white precipitate. A thorough understanding of medicines and the physiology of neonates and children is an essential part of the risk management of use of medicines in this age group.
Issues of stability and concentration are also important. Insulin presents a problem because it sticks to plastic tubing and syringes and the dose is effectively reduced. This happens in all patients, but due to the very low concentrations used in neonates the infusion loses virtually all of its effectiveness over the time of administration. Many paediatric units have now resorted to pretreating syringes with insulin or adding albumin to minimise this effect.
The stability of phenytoin injection remains a challenge. The dose can be as small as 0.02 ml and this has to be given slowly over 20 minutes. If the injection is diluted then the stability is altered. Thus, it is now normal practice to put a filter on all diluted phenytoin infusions to ensure that the potentially crystallised phenytoin does not enter the child.
Total parenteral nutrition is risky business in neonates who are the most vulnerable subpopulation of paediatric patients. Whilst the amount of TPN used in neonates is large across the UK there is no licensed, standardised product that can be used and most bags are specially made and not end-product tested to ensure that the contents are accurate. There are great similarities between formulations when the bigger picture is taken into account and experience suggests that it might be possible to use standardised solutions for 70–80%. Some joined-up thinking and changes to the way we work are needed here to reduce this unacceptable risk.
Unfortunately there are some well-established practices in paediatrics that contribute to the risks of IV treatment. The so-called “rule of six” is a case in point (see Box 1). Over the years people have adapted the rule and learned local versions. For example, if the infant’s weight is multiplied by three and the resulting number of milligrams of drug added to 50 ml of carrier solution, a flow rate of 1 ml/hour will still deliver 1 μg/kg/min. Another variation is to multiply the infant’s weight by 0.75 and add the resulting number of milligrams of drug to a volume of 25 ml. This produces a concentration of 30 μg/ml and so when given at the rate of 1 ml/hr, delivers 30 μg/kg/hr (0.5 μg/kg/min).
These variations are both arithmetically correct, but are potentially confusing because they have lost the obvious link with the number six. Moreover, the differences in volumes and administration rates make them harder to remember. Errors have been reported in connection with use of the rule of six and several bodies have recommended the technique’s abandonment in favour of standard concentrations.
In the USA, the Joint Commission National Patient Safety Goal related to improving safety of high-alert medications requires the use of standardised drug concentrations and therefore the Joint Commission has openly discouraged use of the rule of six.(2)
In the UK, stopping this practice and moving over to standard concentrations would be a huge change to paediatric practice and one that would need to involve industry, manufacturing units, storage facilities and training programmes before it could be thought that changing practice could decrease rather than increase the risk to children.
IV fluids of themselves present further risks for neonates. If injections are diluted there can be a risk of fluid overload as pre-term neonates can have a circulating volume of as little as 80 ml. Saline flushes can quickly deliver large amounts of sodium. For example, infusion of 1 ml/hr of sodium chloride 0.9% to a 0.5 kg baby would deliver more than 6 mmol/kg/day sodium giving them more than their total daily sodium requirement and a fluid volume of 24 ml – a quarter of their total circulating volume.
The risks associated with the use of hypotonic fluids in children were highlighted in a recent NPSA Alert.(3) This has raised many questions. Not only does it lead us to question our total view of fluid management and fluid requirements, but raises the problem that for total compliance we have at present got to use unlicensed infusion bags and potentially need to add strong potassium to these. A thorough local review is required, but probably a coordinated national approach is the real way forward
The risk management measures for IV treatment in children and neonates are shown in Box 2.
Dose-banding and premixed injections might be feasible for young children but would be far more difficult for neonates as weight changes can alter the required dose significantly (there is 500% difference between a 1 kg and 5 kg neonate). Standardised procedures are required and robust systems should be used to ensure products are made up accurately and are sterile. Amazingly, Parshuram (2003) showed that 65% of all morphine syringes prepared at ward level involved concentration errors greater than 10%.4 This is unacceptable and mass production or CIVAS must be considered preferable to this.
A “second check” is good provided it is a genuine check and not just a matter of showing someone else what you have done. Quiet areas for working, electronic prescribing and smart pumps can also contribute to effective risk management.
In summary, making use of IV medicines in children safer depends on having good knowledge of the medicines themselves and of the processes and procedures involved in the care of children and neonates. Some issues in paediatrics are the same as in adult medicine but some are specific to this group of patients and this makes the paediatric pharmacist’s job particularly challenging. ■
This article is based on a lecture given as part of a one-day conference on delivering safer IV medication sponsored by B Braun Medical Ltd
1. Ross LM, Wallace J, Paton JY. Medication errors in a paediatric teaching hospital in the UK: five years’ operational experience. Arch Dis Child 2000;83:492-7.
2. ISMP. Avoid preparing infusions using the rule of 6 or Broselow tape. Institute for Safe Medication Practices (ISMP) Medication Safety Alert 2005;10(4). Available online at: http://www.medscape.com/viewarticle/501053
3. NPSA. Reducing the risk of hyponatraemia when administering intravenous infusions to children. Patient safety alert 22. London: NPSA; 2007.
4. Parshuram CS, Ng GY, Ho TK, et al. Discrepancies between ordered and delivered concentrations of opiate infusions in critical care. Crit Care Med 2003;31:2483-7.