Article Text

Download PDFPDF

The National Patient Safety Agency
  1. T Stephenson
  1. Correspondence to:
    Prof. T Stephenson
    Professor of Child Health, Queen’s Medical Centre, Nottingham NG7 2UH, UK and Clinical Specialty Advisor to the NPSA, 4–8 Maple St, London W1T 5HD, UK;

Statistics from

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.

Assisting the NHS to identify and learn when things go wrong

The National Patient Safety Agency (NPSA) is a Special Health Authority formed in 2001 to improve patient safety in the NHS across England and Wales. Currently, it has a budget of just over £15 million. The NPSA was created following the publication of two key reports by the Chief Medical Officer, Professor Sir Liam Donaldson, An organisation with a memory1 and Building a safer NHS.2An organisation with a memory refers to the death of Wayne Jowett following an inadvertent intrathecal vincristine injection, the 23rd such incident reported worldwide (and the 14th in 15 years in the United Kingdom).3

A central tenet of the NPSA’s creation was that it should assist all those involved in healthcare to identify and learn when things go wrong. When patient safety incidents occurred in the past they may not have been reported locally because they were not seen as important, staff were too busy, patient harm was averted (often known as a prevented patient safety incident or “near miss”), or because of a fear of a blame and punish culture. We need to move from automatically blaming individual front-line staff3 to an understanding of the underlying factors in the system which are more often responsible for incidents.4 For too long, the system has blamed the individual and exhorted staff “to try harder” or “be more careful”. The vast majority of NHS staff work very hard and are very careful, but human error cannot be eradicated by such simplistic calls to arms. In other cases, lessons were learned locally but not disseminated nationally. To overcome this, all clinicians and NHS organisations need to have the safety of patients at the heart of their practice. Safety is not an “add on”, but an integral feature of healthcare systems and processes.

A contrast has been drawn with the aviation industry. In healthcare, errors may not be reported because of fears of disciplinary action. Aircrew face disciplinary action if it transpires that they have not reported a safety incident. It has been estimated that there are 900 000 adverse events per year in the NHS,1 perhaps half in acute hospitals, leading to an additional three million bed days. The total cost to the NHS of these adverse events in extra bed days alone would be around £1 billion a year.5

If we just consider errors arising from prescribing, UK prescribers (mostly doctors) generate 1.5 million GP and 0.5 million hospital prescriptions per day. A single universal solution may not be feasible therefore. This is where comparisons with aviation break down. There are not two million aircraft flights per day in the UK. There may be two million in-flight decisions taken per day, but not by 100 000 different pilots on 100 000 “inpatient flights” and 0.5 million “outpatient flights”. Therefore, the scale of the problem is much bigger than in aviation, although the human cost of a single error is usually much less.


In order to improve patient safety, it is necessary to know the scale and type of problems, so that the NPSA can prioritise its work accordingly. It is also helpful to know if patient safety issues cluster within one region, specialty area, or care setting since this may give clues to the root of the problem. One of the statutory functions of the NPSA is to devise, implement, and monitor a national reporting system for patient safety incidents. Ensuring NHS-wide implementation of the National Reporting and Learning System (NRLS) remains the NPSA’s highest priority. It has been decided that the system should be anonymous, to encourage reporting and respect the confidentiality of those who take the time to report to the Agency, although this limits the ability of the NPSA to seek clarification or updates.

Healthcare staff will be able to report incidents to the NRLS through their organisation’s existing reporting system. The NPSA is aware that not all staff feel comfortable reporting incidents to their local organisations, and in response to these anxieties, the NPSA has developed the eForm, an intuitive electronic reporting form that takes less than 10 minutes to complete. The eForm will allow staff to report independently, confidentially, and anonymously, and is now available on the NPSA’s website, although the Agency encourages staff to share these reports so that local learning and action can take place. The NPSA is also exploring ways to allow patients, their carers, and the public to report problems affecting patient safety.

There may be difficulties analysing what will potentially be a large body of unverifiable reports, but the first aim is to look for clusters and trends which can inform the future research and work of the NPSA.

The NPSA is in the process of rolling out the NRLS across all NHS organisations in England and Wales to enable all staff to report on patient safety incidents, including “near misses” which cause no actual harm. Experience from the aviation industry shows that as reporting rises, the number of serious events begins to decline. Paradoxically, an increase in reporting of patient safety incidents will be a sign that the NPSA has been successful in encouraging an open and fair culture in which we can all learn from the mistakes of others, truly an Organisation with a Memory.


However, some patient safety issues are already obvious. The NPSA has tackled some of these pending the implementation of the NRLS. Two examples illustrate how longstanding systems problems created error prone situations which led to recurrent patient safety incidents.

Safety of infusion pumps was highlighted by incidents when batteries went flat during patient transfer and when patients were unintentionally administered excessive and life threatening bolus doses of drugs. A pilot study involving six acute Trusts by the NPSA revealed that, on average, each Trust owned 1095 infusion pumps at a total cost of £1.6 million, each Trust had 31 different types of pump, and 65% were not utilised at any one time. Annual maintenance costs for each pump varied from £20 to £200. For England and Wales, this represents approximately 200 000 pumps costing £300 million. This situation arose because each directorate purchased its own devices, instead of a Trust wide policy, and therefore purchased the number to meet a “worst case” scenario for that directorate, irrespective of unused stocks elsewhere in the hospital. A 10% reduction in the stockpile would save the NHS £33 million in new purchases (when the current devices expire) and £400 000–£4 million per annum recurrently in maintenance costs saved.

Practice points

  • Safety is not an “add on”

  • In the NHS each year, including adult and paediatric patients, there are 624 million prescriptions, 300 million GP visits, 13 million OPD visits, 5.3 million admissions, 2.9 million “999” calls, and an estimated 900 000 adverse events per year in the NHS

  • We need an open and fair culture in which staff can report errors

  • Human error cannot be totally eradicated

  • Therefore, systems must be mistake proofed against individual human error, wherever feasible

Two new infusion devices initiatives will improve safety for patients. Firstly, a one-stop information website has been established at which helps staff to purchase, manage, and use infusion devices safely and efficiently. Secondly, Trusts are encouraged to establish “equipment libraries” with the minimum number of different devices compatible with the Trust’s services. What we can learn from this example is that this approach promotes centralisation (avoiding wasteful under-utilisation), standardisation (easier staff training), and ensures devices are properly stored, labelled, cleaned, maintained, and charged. These benefits are likely to accrue if the same strategy is adopted in relation to other expensive and complex equipment.

Deaths have occurred when potassium chloride has been injected intravenously instead of sodium chloride. Any design engineer, looking at ampoules of the two solutions, would stare in disbelief. Two colourless liquids with no odour, both in identically sized clear glass ampoules with tiny black typescript. The crucial difference lies between “K” and “Na”. A subtle distinction in an emergency situation.

In July 2002, the NPSA issued its first patient safety alert. It advised on preventing further accidental overdoses with intravenous potassium chloride concentrate that have been misidentified with sodium chloride 0.9% and water for injections. The alert identified safe medication practice recommendations concerning the prescribing, distribution, storage, and preparation of potassium chloride solutions in hospitals. They recommended the withdrawal of concentrated potassium solutions from ward stock and replacement of them with ready to use infusion products. Before publication, formal written safety controls for potassium chloride concentrate were only present in 25% of the trusts taking part in the initiative. After publication, 68% of these trusts had implemented all the recommended safe medication practices, and this number is expected to continue to rise. There have been significant reductions in the number of general ward areas and other clinical areas outside of intensive care and high dependency areas holding stocks of potassium chloride concentrate.6 In this example, the authority of a national patient safety alert empowered local pharmacists to act to withdraw concentrated KCl ampoules from ward areas. Prior to the alert, medical and nursing staff may have felt that pharmacists were intruding on their “clinical freedom” to administer whatever electrolyte solutions they thought appropriate.


Some of the NPSA’s safety solutions are generic and apply to all age groups, for example, the “cleanyourhands” campaign and the alert on confusion over once weekly methotrexate.7 In addition, the NPSA has formed an External Reference Group, including lawyers, nurses, paediatricians, and paediatric pharmacists as well as patient representation, to advise specifically on the safety of children in the NHS. For children, one of the obvious priority areas for patient safety is in medication errors (see below). We hope to work with the National Programme for IT, which is spearheading the world’s largest ever IT programme including the electronic patient record, to develop innovative solutions which will reduce the risk of serious drug errors in children. One obvious example is computerised dose calculation and electronic prescribing (see below). The NPSA is committed to reducing by 40% the number of serious errors in the use of prescribed drugs by the end of 2005.

Children’s medication errors (includes prescribing, dispensing, and administering)

  • Medication errors are the single largest group of errors by doctors8

  • Drug errors account for 25% of medical negligence claims against GPs9

  • 6% of medication prescriptions may include an error; three studies suggest that 2.3–7.5% of paediatric patients may suffer an adverse drug event as a result10–,12

  • It is estimated that there is one medication error for every 660 admissions to a UK children’s hospital13

  • An estimate of 1675 avoidable medication errors occur per year in paediatric inpatients in England, of which 85 are moderate/severe reactions14

  • Children vary in weight from 300 g to over 90 kg (300-fold range; compared to adults, 40 kg to 120 kg, a 3-fold range), and there have been several highly publicised deaths from 10-fold errors in drug doses following miscalculations15,16

  • Risk factors for medication errors in a paediatric emergency department include: patients seen between 4 am and 8 am, patients with severe disease, medication ordered by a trainee (especially at the beginning a training year), and patients seen during weekends17

  • Computerised prescribing with computerised decision support decreased serious medication errors by 55%18


The Department of Health (DoH) is undergoing a radical transformation, reducing in size by over one third. The DoH is supported by 38 “arm’s length bodies” (ALBs), including the NPSA. Other ALBs include bodies as diverse as the National Blood Authority, the National Radiation Protection Board, the National Institute of Clinical Excellence (NICE), and NHS Direct. The DoH has undertaken a review of all ALBs, which currently spend £4.8 billion per year and employ 25 000 staff, and the number will reduce to 20 with projected savings of £0.5 billion and a 25% reduction in staff. As part of this reorganisation, and in recognition of the NPSA’s lead role in safety and protection of patients, the NPSA will take on the role of the National Clinical Assessment Authority (NCAA); hospital food, cleanliness, and safe hospital design from NHS Estates; the three National Confidential Enquiries (the National Confidential Inquiry into Suicide and Homicide (NCISH), the Confidential Enquiry into Maternal and Child Health (CEMACH), and the National Confidential Enquiry into Patient Outcome and Death (NCEPOD)); and the Central Office of Research Ethics Committees (COREC).

Case scenario: now and in the future

Imagine a distressed ventilated newborn weighing 1 kg is prescribed morphine as sedation. This seems appropriate, although the drug is not licensed for this purpose. The infant is prescribed 15 micrograms/kg. However, the barely legible and inappropriate abbreviation “mcg” is interpreted by both of the nurses involved in the administration of the drug as “mg” (milligrams). Unfortunately, the infant can easily be given 15 milligrams because no “tailor-made” paediatric formulation exists for this unlicensed drug and therefore commercially available 2 ml adult vials are stocked on the neonatal intensive care unit containing 10 mg/ml. The infant has a respiratory arrest and, although resuscitated, suffers permanent neurological damage leading to lifelong learning difficulties, cerebral palsy, and incontinence. The litigation costs the NHS over £3 million.

Staff in the future can anticipate computerised order entry. Even if the doctor inadvertently prescribes 15 milligrams, the software will absolutely bar this prescription being printed out for use in a 1 kg neonate. Imagine that, coincidentally, the doctor also mistakenly entered the child’s weight as 100 kg (a weight for which 15 milligrams would be an appropriate dose for intramuscular or subcutaneous injection). The software will alert the prescriber that, given the patient was only born three days ago (from the date of birth entered at the same time), 100 kg is an implausible weight. Alternatively, for less extreme examples, the electronic prescriber would be linked by the infant’s identity bracelet barcode to the baby’s own electronic record (the old fashioned and often illegible “notes”) which would contain the most recent weight measurement. An appropriate dose for the child’s age and weight will be suggested on-screen, and once the prescription is consistent with this, it will be printed out with MICROGRAMS in capitals accompanied by the reminder NOTE: THIS DOSE IS IN MICROGRAMS, NOT MILLIGRAMS. The computer will never resort to the abbreviation “mcg”.


Of the top 20 risk factors which account for nearly three quarters of all deaths annually, adverse in-hospital healthcare events come eleventh. This is above air pollution, alcohol, drugs, violence, and road traffic accidents, and approaching one third of the deaths due to tobacco.19 The NPSA was established to be the driving force to improve patient safety. The NRLS is the most ambitious national confidential reporting system in the world. This will inform future research and the development of systems solutions rather than a knee-jerk reaction of individual blame. An important lesson from other industries is that an open and fair culture where safety problems are acknowledged is more important than attributing blame or issuing punishments and sanctions. The NHS spends millions annually on litigation and yet few of these cases lead to national changes in practice. The NPSA aims to change that.



  • Competing interests: The author is Clinical Specialty Advisor in paediatrics to the National Patient Safety Agency