There are still major safety problems for patients who have allergies to antimicrobials, according to Philip Howard (Consultant Antimicrobial Pharmacist, Leeds Teaching Hospitals UK). This was illustrated by the story of a patient, who, in 2001, was treated with Magnapen (flucloxacillin and ampicillin) for an infected insect bite and experienced an anaphylactic reaction that was eventually fatal. Although the patient was known to be allergic to penicillin and was wearing a red ‘allergy band’, the prescribers were unaware of her allergy status. The nurse who gave the drug was aware of the patient’s allergy but did not realise that Magnapen contained penicillins. At the inquest the coroner recorded ‘a catalogue of mistakes’. He recommended that training should be improved and awareness of the issue be raised. However, 11 years later an inspection (by the Care Quality Commission) found two patients in the hospital who were allergic to penicillin who had, once again, been given penicillins.
One problem here is that, although up to 20% of people say that they are allergic to penicillin, when formally tested only 10–15% of these have genuine allergies, and the remainder can tolerate penicillins, explained Mr Howard. In addition, other reactions to penicillins are often recorded as allergies. For example, approximately 90% of patients infected with Epstein-Barr virus (glandular fever) who are given amoxicillin develop a rash, but this is not due to penicillin allergy. Greater exposure to penicillins is associated with a higher frequency of allergy and, on the other hand, allergy to penicillins can be lost over a period of years if the patient is no longer exposed the allergen. Only 20% of penicillin-allergic patients will still be allergic to penicillin after ten years, if not exposed to penicillin, he noted. A 10% cross-reactivity rate with cephalosporins is often quoted, but in fact recent studies suggest that the true rate is 2% or less. Cephalosporins manufactured before 1980 were commonly contaminated with penicillins and so earlier studies are unreliable, he added.
A thorough assessment of ‘penicillin allergic’ patients should always be made. This should include a clear history of previous reactions to antimicrobials and a record of what has been taken (and tolerated) since the reaction. It is important to establish the likelihood of a true allergy because studies show that the clinical outcomes are poorer for patients with antimicrobial allergies and the treatment costs are higher. One reason for poorer outcomes is that patients could be treated with less effective antimicrobial agents, said Mr Howard.
Reports to the National Reporting and Learning System (NRLS) show that each year a small number of patients still receive antimicrobial agents to which they have a documented allergy. A survey conducted at Barts and the London Hospital showed that 25% of clinical staff were unaware that Tazocin contained a penicillin. In addition, 3% of doctors and 15% of nurses believed it was safe to give co-amoxiclav (amoxicillin and clavulanate) to penicillin-allergic patients. At King’s College Hospital a plastic reminder ‘credit card’ that can be worn alongside the hospital identity badge has been developed. The card lists the names of antimicrobials that contain penicillins. Smartphone apps are popular and might be an effective way to provide allergy information, commented Mr Howard. Storing penicillins is a separate cupboard was not usually a successful measure. The use of ‘co’ names such as co-amoxiclav and cotrimoxazole in the UK has only led to confusion because many people do not know what the ingredients are. It would be better to label such products with the two ingredients, for example, ‘Co‑moxiclav 1.2g (amoxicillin-clavulanate)’, he added.
One-to-one feedback to prescribers is helpful in preventing future errors of this type. Another measure that should be considered is giving patients with documented antimicrobial allergies the names of the drugs that they should not be administered so that they can protect themselves.
The risk of errors during reconstitution and dilution of injectable products is very high and therefore measures to avoid or make these processes safer are critical, according to Pascal Bonnabry (Head of Pharmacy, Geneva University Hospitals and President of the Swiss Society of Public Health Administration and Hospital Pharmacists (GSASA)). A simulation study conducted in Geneva showed that anaesthetists had a 6% error rate when preparing common injections. When asked to perform common calculations, 10% of anaesthetists and 27% of nurses made errors. Real-life data were obtained by collecting unused syringes of lidocaine, atracurium, fentanyl and thiopental at the end of an operating session. Of these, 29% had 10% errors in the doses prepared and 4% had errors of 100% or more. There was up to sixfold variation in the fentanyl doses, noted Professor Bonnabry. A further simulation study showed that microbial contamination was present in 13% of prepared bags and 5% of syringes.
A global approach to safety is needed, including standardisation of drug concentrations and the use of ready-to-use (RTU) products sourced either from the pharmaceutical industry or from hospital production facilities, said Professor Bonnabry. However, in practice, standardisation can be challenging. In Geneva University Hospitals both the intensive care unit and operating theatres had adopted standardised concentrations but they had chosen different concentrations.
Currently available RTU products include bupivacaine, with and without fentanyl, for epidural use and morphine in bags for patient-controlled analgesia. The list of products for production in the hospital pharmacy was determined by asking customers what was needed. Products were then risk-assessed, the feasibility of preparation (including stability considerations) was investigated and a priority list was compiled. Approximately 15 products are now made in the pharmacy and one or two new products are added each year. Some 30,000 syringes are prepared each year with stabilities of up to 12 months.
It is important to understand and to explain that the preparation of RTU products in the pharmacy represents a changing paradigm – essentially from an artisan process to an industrial process. Many of the steps involved in the industrial process, such as computer-assisted production and the clean room environment, actually suppress the opportunities for error that were previously present, he explained.
The evaluation of medical devices is a key activity of the pharmacy according to Xavier Armoiry (Pharmacist, Pharmacy and Innovation Departments, University of Lyon Hospitals, France). The pharmacy is responsible for both medicines and sterile medical devices, which account for approximately 60% of the overall pharmacy budget, he continued. A medical device is an instrument for use in humans that does not achieve its principal action through pharmacological, immunological or metabolic means, although it may be assisted by such means. For example, a balloon angioplasty device that is coated with paclitaxel is classified as a device because its primary action is through the balloon. The term ‘medical devices’ covers a wide range of products, including those used for diagnosis, prevention, monitoring, treatment or alleviation of disease. Medical devices are also used for replacement or modification of anatomy or of a physiological process.
Innovation departments were created in all French university hospitals in 2007 to evaluate innovative medical devices and determine which offer value for money. The innovation department in Lyon is staffed by pharmacists, researchers and a biomedical engineer. The challenge for teaching hospitals, said Dr Armoiry, is to provide safe and effective care at the forefront of technological development for a reasonable cost. The department aims to provide an informed opinion about the medical and economic benefit of innovative medical devices. The available clinical studies (if any) usually give a very low level of evidence, he noted. The innovation department in Lyon undertakes about 30 evaluations each year.
One example of such an evaluation concerned a percutaneous mitral valve repair device (MitraClip). This is used for treatment of severe mitral regurgitation, (a common complication of myocardial infarction) when medical treatment is insufficient. The device is implanted using a steerable delivery catheter inserted via the femoral vein. This was seen as being a highly innovative, breakthrough technology. However, each unit costs €20,000 and there has been no medico-economic comparison with best medical care. A multicentre, randomised controlled trial with a cost-effectiveness evaluation is now being undertaken.
Another example is magnetically controlled growing rods for severe scoliosis in children. These are designed for use in early-onset scoliosis, a rare condition that causes major deformity of the spine. Conventional treatment involves the implantation of steel rods that are gradually lengthened (during surgical interventions) to straighten the spine. This can involve operations every six months over a period of five-to-ten years. The magnetically controlled rods can be manipulated from the exterior, thereby avoiding the costs and risks of an invasive procedure, explained Dr Armoiry. As each unit costs €26,000, a study has now been set up to evaluate the cost effectiveness of this device.
Understanding why errors occur with injectable medicines is critical to the design of effective interventions, said Darren Ashcroft (Professor of Pharmacoepidemiology, Manchester University, UK). A systematic review of reports of medicines’ administration errors showed that errors occur on more than 20% of occasions, although this falls to 9% if timing errors are removed. Errors with intravenous doses were more common, occurring on 78% of occasions (including timing errors). ‘Wrong time’, ‘wrong preparation’ and wrong administration rate’ were the most common categories for errors with intravenous doses.
A further systematic review to examine the causes of medication administration showed that slips and lapses accounted for the largest number of unsafe acts, followed by deliberate violations and knowledge-based mistakes. Common factors that contributed to error-producing conditions included communication problems, workload and staffing difficulties, medicines supply issues and distractions and interruptions. Product standardisation and the use of prefilled syringes are two measures that could help to reduce administration errors, he concluded.
OPAT (outpatient parenteral antimicrobial therapy) can be an effective way to manage patients who require long-term antibiotic treatment, Mark Gilchrist (Consultant Pharmacist, Infection, Imperial College Healthcare NHS Trust) told the audience. This type of management, first developed for management of children with cystic fibrosis, is now commonly used in the UK. It can involve self-administration, daily visits to an infusion centre or home visits by a nurse. The most common diagnoses treated in this way are cellulitis, osteomyelitis and post-operative wound infections. One of the driving factors for OPAT has been the mounting cost of hospital-acquired infections (HAIs) – studies in the UK show that an episode of HAI increases costs nearly threefold, on average, said Mr Gilchrist. In New Zealand, a randomised trial of OPAT compared with hospital care showed that there were no differences in effectiveness. There are numerous ways that OPAT could go wrong but careful process mapping has allowed the team at Imperial College Healthcare NHS Trust to develop rigorous and systematic procedures. One important aspect is that, when patients are referred to the OPAT team, it is the team that then decides on the most appropriate treatment. Often the team is able to reduce dosing frequency to once daily, thereby making OPAT feasible and tolerable, he explained.
A consensus statement on good practice recommendations for OPAT has been published. This sets out recommendation in five key areas: OPAT team and service structure, patient selection, antimicrobial management, patient monitoring and clinical governance. In addition, the British Society of Antimicrobial Chemotherapy has prepared an OPAT business tool kit to help trusts to establish OPAT services (see Resources).
‘The smaller the child, the more likely an error’ said Steve Tomlin (Consultant Pharmacist, Children’s Services, Evelina Children’s’ Hospital). One study of paediatric medication errors had shown that errors with intravenous medicines were common and 8% involved tenfold errors in dose. Paediatric patients range in weight from 0.5kg to 100kg and this can make dosing complicated. Another factor that increases the risk of errors is the use of adult medicines in children. Nearly one third of doses in the neonatal intensive care unit require less than one tenth of a vial, and about 5% of doses require less than one hundredth and so the risk of accidental overdosage is ever present. Moreover, some adult formulations are unsuitable for children, explained Mr Tomlin. For example, amiodarone and lorazepam injections contain benzyl alcohol as an excipient and this can cause gasping syndrome in neonates, making mechanical ventilation impossible.
Dosage calculations remain a problem in paediatrics and a selection of four or five standardised concentrations for each drug is required. Ideally, these should be terminally sterilised products and they should be barcoded to minimise picking errors, advised Mr Tomlin.
The delivery of home parenteral nutrition (HPN) is a high-risk activity that involves multidisciplinary care but little is known about error rates in this field. Jackie Eastwood (Pharmacy Manager, St Mark’s Hospital, London and Chair, British Pharmaceutical Nutrition Group) described the processes involved in the delivery of HPN and the opportunities for error. A recent report has examined all aspects of home care and made recommendations for effective services (see Resources).
Chapman A et al. Good practice recommendations for outpatient parenteral antimicrobial therapy (OPAT) in adults in the UK: a consensus statement. February 2011. http://e-opat.com/grp/.
OPAT Business Case Toolkit. http://e-opat.com/toolkit/.
Homecare Medicines. ‘Towards a Vision for the Future’. November 2011 http://cmu.dh.gov.uk/files/2011/12/111201-Homecare-Medicines-Towards-a-Vision-for-the-Future2.pdf.
The B Braun Medical Seminar, Changing Practice to Improve Safety, was held at the International Convention Centre, Birmingham on 17 October 2012.