Drug Errors
Introduction
Administration of medicines is a key element of nursing care. Every day some 7000 doses of medication are administered in a typical NHS hospital (Audit commission 2002).
Drug administration forms a major part of the clinical nurses role. Medicines are prescribed by the doctor and dispensed by the pharmacist but responsibility for correct administration rests with the registered nurse (O'Shea 1999). Each registered nurse is accountable for his/her practice. This practice includes preparing, checking and administering medications, updating knowledge of medications, monitoring the effectiveness of treatment, reporting adverse drug reactions and teaching patients about the drugs that they receive (O'Shea 1999). The patient is expected to receive the correct medication at each drug round but several studies have shown that this is not always the case (Raju et al 1988, Ferner 1995). Medication errors do occur and are a persistent problem associated with nursing practice (O'Shea 1999).
The aim of this paper is to answer the question: Why are nurses still continuing to make drug errors? In order to answer this question this paper shall examine the guidelines that nurses must adhere to when administering drugs, shall provide a definition of a drug errors, look at reasons why drug errors occur and approaches that are aimed at reducing drug errors on the ward.
There are a number of pieces of legislation that relate to prescribing, supply, storage and administration. It is essential that nurses comply with them (NMC 2003).
The 1968 Medicines Act was the first comprehensive legislation on medicines in the United Kingdom. The act provides the legal framework for the manufacture, licensing, prescription, supply and administration of medicines. The act classifies medicines into the following categories: prescription only medicines, which are medicines that may only be supplied or administered to a patient on the instruction of an appropriate practitioner (a doctor or dentist) and from an approved list for a nurse prescriber. The pharmacist is the expert on all aspects of medicine legislation and should be consulted. Pharmacy only medications are only available to buy from a registered primary care pharmacy provided that the sale is supervised by the pharmacist. General sale list medicines do not require either a prescription nor the supervision of a pharmacist and can be obtained from retail outlets (NMC 2002).
In 1992 the UKCC published Standards for administration of medicines and was replaced in 2002 by the NMC's Guidelines for administration of medicines. The guidelines state principles for the administration of medicines. They state that in exercising your professional accountability in the best interests of your patients the registered nurse must know the therapeutic uses of the medicine to be administered, its normal dosage, side effects, precautions and contra-indications, be certain of the identity of the patient to whom the medicine is to be administered, be aware of the patients care plan, check that the prescription, or the label on the medicine dispensed by the pharmacist is clearly written, has considered the dosage, method of administration, route and timing of the administration in the context of the condition of the patient and co-existing therapies, check the expiry date of the medicine to be administered, check that the patient is not allergic to the medicine before administering it, contact the prescriber or other authorised prescriber without delay where contraindications to the prescribed medication are discovered, or where assessment of the patient indicates that the medicine is no longer suitable, make a clear, accurate and immediate record of all medicine administered, intentionally withheld or refused by the patient, ensuring that any written entries and the signature are clear and legible, it is also the responsibility of the nurse to document if the administration of medicines is delegated to another member of the nursing team.
The guidelines also state that when complex calculations are being carried out it may be necessary for a second practitioner to check the calculation in order to minimise the risk of error and that the use of calculators to determine the volume or quantity of medication should not act as a substitute for arithmetical knowledge and skill.
The American Society of Hospital Pharmacists (1982) defined a medication error as a dose of medication that deviates from the physicians order as written in the chart or from standard hospital policy and protocol. From this definition the society have identified 9 categories of medication error. Omission error, unauthorised drug error, wrong rate error, wrong dose error, wrong route error, wrong dosage from error, wrong preparation of a dose and incorrect administration technique. Wolfe (1989) defines a medication error as mistakes associated with drugs and IV solutions that are made during the prescription, transcription, dispensing and administration phase of drug preparation and distribution. Wolfe (1989) further states that errors can be further divided into intentional and unintentional errors. Each category of error ultimately involves failing to give the medication as prescribed by the doctor.
Mathematical proficiency is a pre-requisite to the performance of many nursing functions such as medication calculations, intravenous regulation and fluid balance calculations (Bindler and Bayne 1984). Studies by Bayne and Bindler (1984) and Worrell and Hodson (1989) have identified that medication errors resulting from poor mathematical skills remains an ongoing problem. A descriptive study by Bindler and Bayne (1984) looked at the mathematical skills of 741 student nurses. The study found that between 9% and 38% of student nurses interviewed were unable to pass elements of a mathematical proficiency test. In another study by the same authors, 110 Registered Nurse's mathematical skills were examined and 81% were unable to calculate medication doses at a 90% pass level on a 20 item medication calculation test. The authors concluded from these studies that calculation difficulties continue to exist and have not improved and also imply that student nurses have a lack of mathematical knowledge. This is further supported by Blais and Bath (1992) who tested the drug calculation skills of nursing students. They proposed that the most frequent type of mistake was conceptual errors, mathematical errors and measurement errors. The study also indicated that a large number of student nurses did not possess the basic mathematical skills needed to function as a registered nurse.
However the effectiveness of mathematical questioning can be questioned. Demetrulias and McCubbin (1982) wrote that mathematical tests may measure knowledge at a lower level than needed for actual practice. In addition as stated by Ludwig-Beymer et al (1990) the test's can only establish the theoretical knowledge of the nurse and not his/her actual performance in practice. Although the studies by Bindler and Bayne (1984) identify poor mathematical knowledge as contributing to drug errors, the studies failed to note at what stage of training the student nurses were at and whether the Registered Nurses where asked to calculate drugs that they were familiar or unfamiliar with.
This therefore poses the question 'Is lack of nursing knowledge to blame for drug errors?' Nurses are accountable for the drugs that they administer and therefore require a knowledge of the action, side effects and correct dosage of any drug that they administer (O'Shea 1999). Weiner and Schumacher (1976) looked at the knowledge of 8 groups of healthcare workers in relation to psychotropic drug therapy. The study indicated that pharmacists were competent in this area but many doctors and nurses had an in adequate knowledge of this drug group. However the study can be criticised as it was only testing knowledge of one drug group. Bogg's et al (1988) tested nurses knowledge of 3 commonly administered drugs and sought to determine if a relationship existed between level of knowledge and educational background or experience. Following a 36 item test the researchers concluded that respondents overall had an inadequate level of knowledge. They also found a difference between staff members with different educational backgrounds with nurse managers and educators having a better knowledge of medications than nurses who were responsible for administering them. Bogg's et al (1988) study therefore contradicts the study by Bayne and Bindler (1988) who found that there was no significant difference between the incidence of medication errors and years of experience or educational background of the participants. These studies therefore raises the question whether the level of knowledge that nurses have regarding drugs makes them accountable for their actions (Boggs et al 1988).
Workload factors have been shown to affect the rate of medication errors (Conklin et al 1990, Leape et al 1995). In an exploratory study by Conklin et al (1990) a questionnaire containing open and closed questions was distributed to 175 nurses who were working in acute care settings. 32 of the respondents cited staffing problems, large patient workloads, distractions whilst preparing medications as being responsible for medication errors. The authors therefore recommended that adequate staffing levels are necessary along with a reduced workload are key factors in reducing errors. A larger study conducted by Roseman and Booker (1995) examined 9 workload factors on medication errors. The data was collected from a 140 bed hospital over a 5 year period. The study found that 3 workload factors were found to be significant in medication errors. Errors were found to be increased with the numbers of shifts worked by temporary staff and errors decreased with the amount of overtime worked by permanent nursing staff. The authors concluded that workload and nursing staffing levels do influence medication errors. However in contrast Taunton et al (1994) explored the possible relationships between medication errors and staff sickness, staff sent to other areas and workload. Data was collected over a 6 month period from 4 different hospitals. The study found no relationship between medication errors, staff absenteeism, relief duty or workload. However the study is rather small in comparison to Roseman and Bookers (1995) study although it could be argued that the data is more credible in Taunton et al (1994) study as it collected data from more than one hospital.
Regardless of staffing levels it appears from the literature that although medication policies exist, adherence to these policies is poor (Long and Johnson 1981, Conklin et al 1990). Long and Johnson (1981) conducted a study of medication errors. They found that 72% of medication errors were due to staff not following policies and procedures. Fequa and Stevens (1988) states that written medication policies and procedures are essential for the safe administration of drugs however Pyne (1981) argues that policies and procedures should only act as a guide for registered nurses. Pyne (1981) further argues that many of the medication errors brought before the Nursing and Midwifery Council (NMC) result from a lack of clear medication administration policy.
Another highly influential factor that contributes to medication errors is the quality of written prescriptions. Just as nurses are legally accountable for the drugs that they administer, prescriptions are the legal responsibility of the doctor. Nurses frequently come across poorly written and illegible prescriptions which conflict with policies for the safe administration of medications. Howell (1996) argues that the doctors failure to follow policies puts the nurse responsible for administering medications at risk. If a nurse administers a medication from a poorly written prescription or one that does not meet with legal requirements, the nurse must be prepared to be accountable for administering the medication. Howell (1996) conducted an audit of prescription sheets in the medical unit of a acute hospital. Out of 370 prescriptions 12% had no record of patients allergies which nurses must be aware of prior to administration, and in 15% of cases the drug dosage was not clear. Although the survey was small Howell (1995) concluded that nurses were frequently administering medications in an unsafe manner due to the poor standard of written prescriptions. Although the study highlights problems with the quality of doctors handwriting Williams (1996) points out that if nurses are not satisfied with the quality of written prescriptions then they have a duty not to administer the drug and consult the doctor in order to maintain patients safety and be accountable practicioners.
Reduction of Errors
The fact that most drug errors occur without causing the patient harm or otherwise being detected by the practitioner makes assessing the effectiveness of inititives concerned with drug administration difficult (Bates et al 1995). The belief that nurses are to blame for medication errors has been well documented in the literature. These articles also offer checklists and test's to help nurses combat error (McGovern 1986, Cohen 1990). Whilst Hunt (1988) acknowledges that these types of checklists can be helpful in terms of safety they are infact ineffective in isolation and are unable to entirely stop errors. Hunt (1988) further states that published nursing checklists addresses only the nurses behaviour and do not take into account factors in the environment which impact the nurses ability to administer correctly.
Double checking has also been proposed as a way to stop error. However any kind of checking alone cannot be relied on to significantly reduce errors in the long term (Anderson et al 2001). Kohn et al (1990) believes that double checking can become ritualistic with the second nurse simply repeating what the first nurse has said.
A systems approach has been widely discussed in the literature as a means of reducing drug errors on the wards. Perrow (1984) writes that systems are made of many parts which can be grouped under the broad categories of design, equipment, procedures, operators, supplies and environment (DEPOSE). Factors such as good design, effective safety measures, tolerance limits and recommended operating procedures can all be seen as a systems defence against accident or fault (Perrow 1984, Sagan 1993). Defects in design, violations such as not following correct procedures, a lack of proper equipment maintenance, or simple human error can be thought of as gaps in the system defences. The size of the gaps depends on the problem. Having fewer defence systems increases will coincide (Reason 1990). The systems approach recognises that no defence or safety mechanism is perfect in every circumstances. Latent defects are inevitable and nature of these defects will change over time with changes in the supplies procedures and other system factors.
A successful incident reporting scheme is central to the implementation of an effective systems approach to drug error reduction (Runciman et al 1993). However many current incident reporting schemes are unsuccessful due to low levels of response (Cullin et al 1995). Anderson et al (2001) writes that attitudes towards error must change. Reporting an error must not be seen immediately as admitting incompetence. The strong desire to blame must be resisted and the many other environmental factors which contribute to error must also be recognised as significant. An incident reporting scheme which is not anonymous as seen as maintaining the blame culture. As Runciman et al (1993) points out that this is the reason that most successful incident reporting schemes are anonymous. The advantages of such schemes are that they are cheap, universally applicable, medico-legally safe and able to supply a large volume of relevant specific information. The value of anonymous incident reporting has recently been recognised in reports from the NHS (DOH 2000). Anonymity encourages good rates of reporting and avoids individuals being singled out. Anderson et al (2001) points out that as near-misses are typically more common than accidents, systems based incident reporting that includes near missed and system approaches has the potential to identify and correct problems before and accident occurs but only if reporting levels are high. Reporting near-misses also helps to overcome the perception that anyone filling out and incident form has just made and error.
Cooper (1995) proposes the so called zero defect philosophy for application in nursing which is similar to the systems approach. It involves a wide ranging analysis of all aspects of a human system interface to find ways to error proof problem areas by using devices which prevent dangerous or incorrect alternatives. These systems attempt to remove the potential for human error out of the drug administration system, however as Anderson et al (2001) points out such systems are only effective once it is understood why the system is going wrong.
The NMC (2002) states in its Guidelines for administration of medications that it is important that an open culture exists in order to encourage the immediate reporting of errors or incidents in the administration of medicines. They believe that all errors and incidents require a thorough and careful investigation at a local level, taking full account of the context and circumstances and the position of the practitioner involved. Such incidents require sensitive management and a comprehensive assessment of all the circumstances before a professional and managerial decision is reached on the appropriate way to proceed.
Conclusion
This paper has highlighted that drug errors are still a persistent problem that is associated with nursing practice. It has identified that nurses mathematical ability and knowledge of drugs are key factors that contribute to drug errors. The paper has also highlighted how the problem of drug errors can be tackled in a modern NHS hospital using the systems approach. From the research the author has noted that there appears to be a lack of research into the education that student nurses receive regarding mathematical drug calculations and drug knowledge whilst at university. However it is clear that there has to be a continuous long term investment into the management of drug errors in NHS hospitals.