According to World health organization (WHO) 2017 report, globally the cost associated with medication errors has been estimated about 42 billion US dollars annually [
Medication administration error (MAE) is one of the most common errors in the medication error process and occurs when a discrepancy occurs between the drugs received by the patient and the drugs intended by the prescriber [
Voluntary reporting is a critical strategy to detect MAEs. A critical strategy to reduce MAEs is to use error detection, which comprises error recognition and reporting. MAE reporting requires professionals to recognize the occurrence of MAEs and to report them through official channels. MAE reporting is an effective way used to identify the root causes of MAE and to prevent repeating them in future [
A study conducted in North Carolina, Philippines, and Saudi Arabia showed that about 37.9%, 52%, and 28.6% of study participants were reporting MAE, respectively [
A study conducted in Ethiopia indicated that the proportion of MAE reporting among nurses was found to be 57.4% [
There are a lot of studies done on MAE [
The study was conducted in public hospitals of North Shoa Zone, Amhara region, Ethiopia. North Shoa is one of the 10 zones in Amhara region. In North Shoa Zone, there are 9 public hospitals. This study was conducted on three hospitals from March to April. An institutional-based cross-sectional study design was employed.
All nurses working in public hospitals of North Shoa Zone were used as a source population. Nurses that work in selected public hospitals of North Shoa Zone were used as a study population.
All nurses who have a minimum of diploma qualification in nursing and involved in direct patient care, those who have at least six months of work experience, and those who are full time workers were included. Those nurses who were on annual leave, maternal leave, seriously ill, and attending external training courses off-site at the time of the data collection were excluded.
The sample size was determined using single population proportion formula with the assumption of 95% confidence interval with margin of error of 5%, 10% non-response rate and 57.4% of prevalence of medication administration error reporting from a study conducted in Addis Ababa [
To select 230 nurses from the total of nine hospitals, three hospitals were selected by using simple random sampling. Then, the sample size was proportionally allocated to the number of nurses in each hospital. Finally, study participants were selected by using simple random sampling technique.
The instrument used for data collection was a structured self-administered questionnaire. The questionnaire was adapted and modified from a previous study [
The first section deals with the sociodemographic characteristics of the participants; the second section contains work-related aspects of nurses; the third section is regarding knowledge on MAEs; the fourth section is about the reason why MAEs occur; the fifth section is the reason why MAEs are not reported; and the six section deals with the percentage of each type of error reported. The questionnaire that is used in this study is available as a supplementary file.
To assess the validity of the instrument, face validity and content validity were done by five experts. Content validity ratio (CVR) and content validity index (CVI) were measured and were 0.2 and 0.83, respectively, which shows that the instrument is valid. The reliability of the instrument was checked using Cronbach’s alpha and was 0.8 which showed that the instrument was reliable. The questionnaire was pretested on 5% of the sample size at the nearby hospital and appropriate amendment was done on it.
Data were collected by three diploma holder nurses. Training was provided for data collectors about the overall objective of the study, content of the questionnaire, and how to collect the data. The questionnaire was given to the randomly selected participants. Confidentiality of the information was kept by excluding the names of the respondents and names of the hospitals in the questionnaire.
The returned questionnaires were checked for completeness, cleaned and entered into EpiData 3.1, and analyzed using SPSS (Statistical Package for the Social Sciences) version 21. Descriptive analysis was done and presented using tables and texts. Bivariate and multivariate logistic regression analysis was used to identify factors associated with medication administration error reporting. Variables with
This section gives an overview of the sociodemographic characteristics of nurses working in public hospitals of North Shoa Zone, Amhara, Ethiopia, 2019. From the survey, information about sex, age, marital status, educational level, and educational degree attained of the respondents was analyzed.
Out of 230 proposed study participants, 224 nurses participated in this study indicating a response rate of 97.4%. More than half (120) (53.5%) of respondents were females, 117 (52.2%) of them were married, 104 (46.4%) of nurses were in the age group of 25–29 years old, 120 (53.5%) of nurses had Bachelor of Science in Nursing and received their degree from a governmental institution, 195 (87.1%) (See Table
Socio-demographic characteristics of nurses working in public hospitals of North 53 Shoa Zone, Amhara, Ethiopia, 2019 (
Variables | Responses | Frequency ( | Percentage (100%) |
---|---|---|---|
Sex | Male | 104 | 46.5 |
Female | 120 | 53.5 | |
Age | 20–24 | 15 | 6.7 |
25–29 | 104 | 46.4 | |
30–34 | 48 | 21.4 | |
≥35 | 57 | 25.5 | |
Marital status | Single | 96 | 42.9 |
Married | 117 | 52.2 | |
Others | 11 | 4.9 | |
Educational status | Diploma | 93 | 41.5 |
BSc | 120 | 53.5 | |
MSc | 11 | 5.0 | |
Educational degree attained | Government institution | 195 | 87.1 |
Private institution | 29 | 12.9 |
From the participants, 106 (47.3%) had a work experience of ≤4 years, 157 (70.1%) nurses worked in the inpatient department, 143 (63.8%) nurses worked in the day duty shift, and 119 (53.1%) of them worked for 3–6 months on their unit. Regarding average patient care, 81 (36.2%) of the participants provided care for 1–6 patients. The majority of the participants or 189 (84.4%) nurses responded that there is no MAE reporting system in their hospital (see Table
Work-related characteristics of nurses in North Shoa Zone public hospitals 56 Amhara, Ethiopia, 2019.
Variables | Response | Frequency | Percentage |
---|---|---|---|
Working experience | ≤4 years | 106 | 47.3 |
5–10 years | 71 | 31.7 | |
11–14 years | 27 | 12.1 | |
≥15 years | 20 | 8.9 | |
Working unit | Medical ward | 59 | 26.3 |
Surgical ward | 50 | 22.3 | |
Pediatrics ward | 23 | 10.3 | |
Obstetrics and gynecology ward | 9 | 4.0 | |
Emergency | 28 | 12.5 | |
Intensive care unit | 16 | 7.1 | |
Outpatient department | 18 | 8.0 | |
Others | 21 | 9.4 | |
Duration on present unit | ≤3 months | 15 | 6.7 |
3–6 months | 119 | 53.1 | |
≥6 months | 90 | 40.2 | |
Current duty shift | Day shift | 143 | 63.8 |
Night shift | 61 | 27.2 | |
Alternative shift | 20 | 8.9 | |
Average patient care | 1–6 patients | 81 | 36.2 |
7–10 patients | 76 | 33.9 | |
>10 patients | 67 | 29.9 | |
Presence of MAE reporting system | Yes | 35 | 15.6 |
No | 189 | 84.4 |
Others = NICU, OR, TB.
The proportion of MAE reporting in the last six months that was committed or witnessed among nurses in this study was found to be 85 (37.9%). Out of the reported MAEs (
More than half (132) (58.9%) of the participants perceived that MAEs should be reported as they occur. Out of the total participants (
Magnitude of MAE reporting among nurses working in public hospitals of North Shoa Zone, Amhara, Ethiopia, 2019 (
Variables | Response | Frequency | Percentage |
---|---|---|---|
Report MAE | Yes | 85 | 37.9 |
No | 139 | 62.1 | |
Should medication errors be reported | Yes | 132 | 58.9 |
No | 92 | 41.1 | |
Communicate with another nurse when facing doubt during medication administration | Yes | 201 | 89.7 |
No | 23 | 10.3 | |
Following 6 rights of medication administration avoids MAE | Yes | 205 | 91.5 |
No | 19 | 8.5 |
Among the reasons for MAEs, about 122 (54.5%) of the respondents said that physician orders were not clear/legible, 133 (59.4%) said change of physician orders frequently, 127 (56.7%) said failure of pharmacists to label the medication correctly, 133 (59.4%) of them said the situation in which many patients are on the same or similar medications, 130 (58.0%) of them said the situation in which unit staff do not receive enough service training on new medications, and 136 (60.7%) of them responded that inadequate unit staffing was identified as a reason for MAE.
Regarding the reasons why medication administration errors were not reported, 142 (63.4%) of respondents expressed their disagreement with hospital’s definition of a medication error, 121 (54.0%) sampled nurses did not think the error is important enough to be reported, 143 (63.8%) participants believed the expectation that medications should be given exactly as ordered is unrealistic. Another reason for not reporting MAEs is that about 132 (58.9%) nurses have fear of adverse consequences from reporting medication errors and 145 (64.7%) of respondents believed nursing administration focuses on the individual rather than looking at the system as a potential cause of the error.
The types of medication administration errors reported among nurses were measured by the frequency of wrong route, wrong time, wrong patient, wrong dose, wrong drug, and medication is omitted; out of the sampled 224 nurses, 155 (69.2%) of them responded wrong route of administration, 126 (56.3%) of the respondents opt wrong time of administration, 167 (74.6%) of respondents responded wrong patient administration, 99 (44.2%) respondents provided wrong dose, 162 (72.3%) of them administered wrong drug, and 121 (54.0%) respondents were not given prescribed medications (see Table
Types of medication administration errors reported among nurses working in public hospitals of North Shoa Zone, Amhara, Ethiopia, 2019 (
Variables | Value | Frequency | Percentage |
---|---|---|---|
Wrong route | 1–20 | 155 | 69.2 |
21–30 | 43 | 19.2 | |
31–40 | 16 | 7.1 | |
41–50 | 7 | 3.1 | |
>50 | 3 | 1.3 | |
Wrong time | 1–20 | 126 | 56.3 |
21–30 | 55 | 24.6 | |
31–40 | 21 | 9.4 | |
41–50 | 12 | 5.4 | |
>50 | 10 | 4.5 | |
Wrong patient | 1–20 | 167 | 74.6 |
21–30 | 32 | 14.3 | |
31–40 | 14 | 6.3 | |
41–50 | 7 | 3.1 | |
>50 | 4 | 1.8 | |
Wrong dose | 1–20 | 99 | 44.2 |
21–30 | 73 | 32.6 | |
31–40 | 25 | 11.2 | |
41–50 | 17 | 7.6 | |
>50 | 10 | 4.5 | |
Wrong drug | 1–20 | 162 | 72.3 |
21–30 | 36 | 16.1 | |
31–40 | 15 | 6.7 | |
41–50 | 8 | 3.6 | |
>50 | 3 | 1.3 | |
Medication is omitted | 1–20 | 121 | 54.0 |
21–30 | 52 | 23.2 | |
31–40 | 23 | 10.3 | |
41–50 | 15 | 6.7 | |
>50 | 13 | 5.8 |
Binary logistic regression analysis was done to identify factors associated with nurses’ MAE reporting. Sex, educational status, educational award, nurses work experiences, average patients care per shift, belief that errors should be reported, the names of many medications being similar/look alike, not agreeing with hospital’s definition of a medication error, and fear of adverse consequences from reporting medication errors had an association with MAE reporting in bivariate logistic regression analysis. All variables that have an association with the outcome variable at
The proportion of MAE reporting was higher among female nurses as compared to male ones. Female nurses were almost three times more likely to report MAEs than male nurses [AOR = 2.91; CI (1.45–5.85)]. Similarly, educational status was an important predictor of MAE reporting. BSc nurses were more than three times more likely to report medication administration errors as compared to those who are diploma nurses [AOR = 3.27; CI (1.61–6.66)]. And MSc nurses were more than six times more likely to report MAEs than Diploma nurses [AOR = 6.4; CI (1.02–40.3)].
Regarding with working experience, participants who worked greater than 15 years were almost four times [ AOR = 3.93; CI (1.11–13.85)] more likely to report MAEs than those who work less than than or equal to four years. Participants who gave care for greater than 10 patients were 0.4 times less likely or 60% times more likely to report than those participants who gave care for less than or equal to 6 patients (AOR = .4; CI (0.16–.96)). Participants who believed an error should not be reported were 0.3 times less likely or 70% times [AOR = .3; CI (.15–.61)] more likely to report MAEs than those participants who believed that errors should be reported (see Table
Bivariate and multivariable logistic regression analysis of factors associated with 278 nurses MAE reporting working in public hospitals of North Shoa zone, Amhara, Ethiopia, 2019.
Variables | Response | Medication administration error reporting | Odds ratio (95% CI) | ||
---|---|---|---|---|---|
Yes | No | Cor | AOR | ||
Sex | Male | 26 (24.8%) | 79 (75.2%) | 1.00 | 1.00 |
Female | 59 (49.6%) | 60 (50.4%) | 2.98 (1.68–5.28) | 2.91 (1.45–5.85) | |
Educational status | Diploma nurse | 21 (21.9%) | 75 (78.1%) | 1,00 | 1.00 |
BSc nurse | 56 (47.5%) | 62 (52.5%) | 3.22 (1.76–5.90) | 3.27 (1.61–6.66) | |
MSc nurse | 8 (80.0%) | 2 (20.0%) | 14.28 (2.81–72.42) | 6.40 (1.02–40.30) | |
Working experience | ≤4 yrs | 38 (35.8%) | 68 (64.2%) | 1.00 | 1.00 |
5–10 yrs | 28 (39.4%) | 43 (60.6%) | 1.35 (0.64–2.81) | 1.17 (0.51–2.70) | |
11–15 yrs | 11 (40.7%) | 16 (59.3%) | 5.17 (2.40–11.13) | 2.11 (0.84–5.30) | |
>15 yrs | 8 (40.0%) | 12 (60.0%) | 9.78 (3.47–27.54) | 3.93 (1.11–13.85) | |
Educational award | Gov’t institution | 81 (41.5%) | 114 (58.5%) | 1.00 | 1.00 |
Private institution | 4 (13.8%) | 25 (86.2%) | 0.22 (0.07–.67) | 0.38 (0.10–1.40) | |
Average patient care | 1–5 | 41 (50.6%) | 40 (49.4%) | 1.00 | 1.00 |
6–10 | 28 (36.8%) | 48 (63.2%) | .56(,30–1.07) | 0.66 (0.30–1.48) | |
≥11 | 16 (23.9%) | 51 (76.1%) | .30(.15–.62) | 0.40 (0.16–.96) | |
Errors should be reported | Yes | 64 (48.5%) | 68 (51.5%) | 1.00 | 1.00 |
No | 21 (22.8%) | 71 (77.2%) | 0.31 (0.17–.57) | 0.30 (0.15–.61) | |
Agree with hospital definition on MAEs | Agree | 29 (59.2%) | 20 (40.8%) | 1.00 | 1.00 |
Disagree | 56 (32.0%) | 119 (68.0%) | 3.08 (1.60–5.90) | 1.40 (0.60–3.23) | |
Fear adverse | Agree | 58 (44.6%) | 72 (55.4%) | 1.00 | 1.00 |
Consequence from MAE reporting | Disagree | 27 (28.7%) | 67 (71.3%) | 0.50 (0.28–.88) | 0.97 (0.47–1.99) |
NB: variables having a
This study was carried out with the aim of determining the magnitude of MAE reporting and the associated factors. In this study, the proportion of MAE reporting was low. This was in line with the finding in North Carolina which indicated that 37.9% of the participants reported all types of medication errors that occurred on their unit [
The finding of this study was slightly high as compared to studies in Saudi Arabia and University of Gondar Referral Hospital in which 28.6% and 29.1% of MAEs were reported, respectively [
In this study, the proportion of female nurses who reported medication errors was higher than the male nurses and was statistically significant. Female nurses were almost three times more likely to report MAEs than male nurses. The result was consistent with that of a study from Addis Ababa [
Educational status was an important predictor of MAE reporting. BSc nurses are more than three times more likely to report MAEs as compared to those who are Diploma nurses. MSc nurses were more than six times more likely to report than Diploma nurses. The result was consistent with that of the study from Addis Ababa, Gondar [
From the participants, 58.9% perceived that errors should be reported as they occur for the safety of patients and this is lower than the study from Addis Ababa. The possible difference may be due to lack of a readily available practice system of MAE reporting [
Participants who say medication administration errors should not be reported were 70.1% times less likely to report MAEs than those who say medication administration errors should be reported.
This result is lower than the previous study conducted in Addis Ababa [
Pertinent to work experience, participants who worked greater than fifteen years were almost four times more likely to report medication administration errors than those who worked less than or equal to four years. This result is consistent with the study conducted in Saudi Arabia [
Participants who gave care for greater than 10 patients were 60.4% times less likely to report medication administration errors than those who gave care for less than or equal to 6 patients. This result is contradicted with a study conducted in Saudi Arabia [
The result of this study shows that medication administration error reporting among nurses was low. This implies that there is a problem in nursing practice. So, each hospital should create and apply a reporting system and nurses should practice the documentation and reporting of errors through the reporting system.
As a limitation, since the study was done by cross-sectional study design, it does not determine cause effect relationship. The number of the participants might have contributed to the absence of a significant association between some of the factors and MAE reporting, as well as to the generalizability of the findings.
The magnitude of MAER among nurses was found to be low. Being female, belief that errors should be reported, working experience, educational status, and average patient care were factors significantly associated with nurses’ medication administration error reporting. To increase medication administration, error reporting efforts should be made to change the attitude of nurses on the belief that errors should be reported, retaining staffs that have longer experience, upgrading staff educational status, and limiting the number of patients cared by a single nurse.
Debre Berhan Referral Hospital
Debre Berhan University
Ethiopia Nursing Association
Food and Drug Agency
Intensive care unit
Intravenous
Medication administration errors
Medication administration error reporting
Ministry of Health
Master of Science
Neonatal intensive care unit
Outpatient department
Operation room
Statistical Package of Social Sciences
Tuberculosis
University of Gondar Referral Hospital
United States
World Health Organization.
All the data are available from the corresponding author on reasonable request.
Before starting the data collection process, the study protocol was approved by Debre Birhan University Institute of Medicine and Health Science with approval number IHRERC-650/2019. Official letter of co-operation was written to all hospitals and concerned bodies in the region to obtain their co-operation in facilitating the study.
Written informed consent was obtained from each study participant. Confidentiality of information was ensured by excluding names and identification from the questionnaire.
The authors declare that they have no conflicts of interest.
KK was involved in title selection, critical review of the design, literature, analysis, and report writing. DD was involved in literature search and review, data collection and analysis, data interpretation, and report writing. KK was also involved in the preparation of the draft manuscript. WN was involved in reviewing the full document and the draft manuscript. KK, DD, and WN were involved in critically reviewing the manuscript and edited and approved the final manuscript. All authors read and approved the final manuscript.
The authors would like to express a profound sense of gratitude and sincere thanks to Debre Birhan University, the selected hospitals, data collectors, and study participants. This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.