Article from Nursing Progress: medication errors article

Causes of Medication Errors

Maggie Nicol & Barbara Thompson

Abstract: A study to explore the circumstances surrounding medication errors through a questionnaire asking nurses to recall errors and 'near-misses'.

Contents:
Study design.
Questionnaire.
Results.
Conclusions.
References.

Back to Nursing Progress contents.
Back to Issue 8 contents list.

Figure 1: the route of administration involved in the error.
Figure 2: the grade of the second person and the number of errors.
Figure 3: the duty shift and the number of errors.
Figure 4: the equipment and the number of errors.

Despite numerous studies into the causes and management of medication errors, they continue to occur on a daily basis in most healthcare institutions.(1) It is commonly assumed that errors occur because of a lack of knowledge and additional in-service education is often implemented following an error. However, Baker and Napthine in1994 found that failure to follow the correct procedure was associated with only a small proportion of the errors observed in a large study of over 1000 drug administrations.(2) All nurses receive training on this important aspect of their role and if questioned, most if not all, would be able to recite the correct procedure for checking medicines. But as Cooper points out, "humans will always err, and need assistance and checking procedures to detect mistakes".(3) The literature suggests that other factors such as workload, shift pattern worked, time of day and environmental factors can also contribute to errors.(4),(5)

Although in the vast majority of cases no significant harm befalls the patient, except perhaps to receive sub-therapeutic treatment, making an error can seriously affect the nurse and his/her clinical confidence. As Booth comments, the first feelings of disbelief are rapidly followed by fear for the patient's safety, fear of personal consequences and then feelings of professional failure.(6)

^{Back to Top}

The aim of this study was to explore the circumstances surrounding medication errors. By asking nurses to recall errors and 'near-misses' (i.e. errors narrowly avoided), we hoped to be able to identify the factors that make errors more likely to occur. Examination of the circumstances surrounding near-misses will enable us to develop robust checking procedures that are grounded in the realities of practice and so are likely to find more favour with the practitioners themselves.

The following research questions were addressed:

What conditions, human and environmental, make drug errors more likely to occur?
Which checking procedures are the most effective in preventing errors?

Design of the study

Part 1

An anonymous self-report questionnaire was sent to all wards, asking nurses to recall the details of circumstances surrounding medication errors or near misses in which they had been involved. The questionnaire was developed with reference to the literature and local knowledge of medication errors. It comprised mainly quantitative responses with room for additional qualitative information from two open-ended questions. The quantitative data was collated using SPSS. The qualitative data obtained from the two open-ended questions was analysed manually by the two researchers to identify emerging themes.

Part 2

Participants were also invited to take part in focus groups to discuss issues raised in the questionnaire and the literature on the subject of medication errors.

Ethical approval was obtained from the local research ethics committee before starting the study.

^{Back to Top}

Content of the questionnaire

The questionnaire began by asking respondents to identify whether they had been involved in a medication error or near miss. If the answer was yes, the respondent was asked to give details of the error or near miss. This included the route and type of error, any equipment and/or other professional involved, the time of day, type of shift, how familiar they were with the ward or department, the grade at the time and how many shifts had been worked since the last day off. Respondents were then asked to identify how busy the clinical area was at the time, whether it was well or short staffed, whether the staff were predominantly regular or temporary staff and if, in their opinion, the skill mix was appropriate to the patient dependency.

The final questions asked the nurse if there were any factors which had had contributed to the error, whether they could suggest any ways in which that particular type of error could have been avoided and strategies to prevent errors more generally. If the respondents reported that they had never been involved in an error or near miss, they were directed to the final question: strategies for preventing errors. Copies of the questionnaire are available from the researchers.

^{Back to Top}

Results

Of the 177 who responded, 36 (20%) reported that they had never been involved in a medication error or near miss, 20 (11%) reported having been involved in a near miss, and the remaining 121 all reported having been involved in a medication error. The sample was self-selecting and it is likely that those who had been involved in an error or near miss were likely to be more motivated to respond to such a survey. However, anecdotal evidence suggests that a high proportion of nurses have been involved in medication errors or near misses and so these findings may well be typical.

Figure 1: bar chart of route of administration.

Figure 1: the route of administration involved in the error.

The intravenous (IV) route accounts for the highest proportion 68 (38 %) of all the errors/near misses, with the oral route being the second most common 50 (28%). This is probably explained by the proportion of medicines being administered by these routes. When administering medications by the IV route, nurses are often using stock preparations rather than those dispensed for individual patients, and so the opportunity for error is greater. In addition, IV medications often require mixing with a diluent and may involve a calculation, both of which have been shown to increase the risk of error. Six respondents reported that the error was due to use of the wrong dilutent.

^{Back to Top}

Figure 2 shows the proportion of errors and near misses in which a second person was involved and the grade of that person. A second person was involved in 84 of the 121 errors reported and in almost all of these cases, their involvement included checking the medicine. With the exception of controlled drugs and administration in the Children's Directorate, Trust policy does not require nurses to check medicines with another nurse. However, many nurses comment that they like someone to check their IV medications even though they are happy with single nurse administration for other routes.

Figure 2: bar chart of the grade of the second person.

Figure 2: the grade of the second person and the number of errors.

A second person was involved in half of the near misses reported and so it could be argued that having a second person prevented the error, but in an equal number of near misses only one person was involved. These findings support the literature, which suggests that merely having two nurses involved in administration of medicines will not prevent errors happening. There is a danger that a false sense of security is created in which both nurses subconsciously relies on the other to check thoroughly. In addition there is often a hierarchical difference in the two nurses. If junior nurses act as checkers, they may doubt their own abilities and feel unable to challenge a senior, more experienced colleague. Whatever the reason, it is clear that simply re-instating two-person medication administration would be unlikely to solve the problem.

^{Back to Top}

There were more errors and near misses reported during the night shift (see Figure 3).

Figure 3: the duty shift and the number of errors.

Of the 54 errors or near misses reported to have occurred on night duty, 23 (16%) happened in the evening at the start of the shift, and 31 (22%) in the early morning towards the end of the shift. The high number occurring on night duty may be explained by the fact that on many wards the morning medicines, or at least the IV medicines are administered by the night staff before they leave.

Thus there are two medicine rounds on one shift and so there is more opportunity for errors. Nurses are likely to be tired and less alert at the end of a night shift and again, respondents suggested that tiredness was a common cause of medication errors. However, the shift with the second highest incidence of errors or near misses was the early shift 36 (26%) when nurses should be alert and refreshed after a nights sleep! This may be explained by the fact that there are often two medicine rounds during this shift and a higher proportion of medicines are prescribed to be administered in the morning than at any other time of day.

^{Back to Top}

Figure 4 illustrates the proportion of errors or near misses in which equipment was involved and it can be seen that this was almost exclusively the IV route, involving over half of the 68 intravenous errors. Errors involving the use of IV devices continue to be a source of concern throughout the UK. In the vast majority of cases, human error is found to be the cause. (7)

^{Figure 4: the equipment and the number of errors.
Respondents were asked to indicate how busy the ward was when the error or near miss occurred, and also the staffing levels, whether regular or agency/bank nurses were on duty and the skill mix. Many errors and near misses did occur when the ward was busy and short staffed but this was not always the case, and in most cases, there were regular staff and a reasonable skill mix.
The analysis of the final two questions provided interesting and rich data. Better checking was cited most frequently as a strategy to prevent errors (59 of the 177 respondents) with 14% suggesting that this should take the form of double-checking. The researchers made the assumption that the
respondents who advocated double-checking actually meant two people checking rather than a more careful and second check by the one nurse administering the medication. Staffing issues were the second most frequently cited cause of errors in the opinion of the respondents, 47 (27)% with inappropriate skill mix being highlighted by 16 (9%). However, as noted earlier, the data in the questionnaire did not totally support this view.
Interruptions while undertaking the administration of medicines were also considered to be a significant risk (20%) while loss of concentration because of tiredness, sometimes linked to night duty and long stretches without a day off was also felt to be a factor (15%).
An additional problem, which was considered to be a contributory factor, was a lack of clarity in the prescription charts with 15% having experienced difficulties.
It was interesting to find that 24 (14%) felt that more training around medications would help to prevent errors but only 6 (3%) felt that further assessments would result in a reduction in the number of errors.}

^{Back to Top}

Conclusions

Although the busyness of the ward and number and quality of staff are certainly factors in the incidence of medication errors, interruptions and distractions are also clearly major factors. The next stage of this research is to conduct a number of focus groups with staff from a variety of wards to discuss ways in which these issues can be addressed. The outcomes form the focus groups will be reported in a future issue of Nursing Progress.

On to part 2 of the research

References

1. Gladstone J. Drug administration errors: a study into factors underlying the occurrence and reporting of drug errors in a district general hospital. Journal of Advanced Nursing, 1995: 22: 628-37.

2. Baker H, Napthine R. Medication error: the big stick to beat you with. Australian Nursing Journal, 1994: 2.4: 28-30.

3. Cooper MC. Can a zero defects philosophy be applied to drug errors? Journal of Advanced Nursing, 1995: 21: 487-91.

4. Bechtel GA, Vertees JL, Swartzberg B. A continuous quality improvement approach to medication administration. Journal of Nursing Quality Assurance, 1993: 7.3 : 28-34.

5. Keill P, Johnson T. Shifting gears: improving delivery of medications. Journal of Nursing Quality Assurance,1993: 7.2: 24-33.

6. Booth B. Management of drug errors. Nursing Times, 1994: 90.15: 30-1.

7. Williams C, Lefever J. Reducing the risk of user error with Infusion Pumps. Professional Nurse, 2000; 15. 6; 382-4.

Nursing Progress: Issue 8: May 2000.

Copyright: Nursing Progress, Royal Hospitals NHS Trust

Return to Nursing Progress contents list

Back to Issue 8 content list

^{Back to Top}

Page last updated by DEB on 14/4/02