Prevalence and types of medication errors among children under five (5) years in 3 primary health care facilities in the Western region of Ghana: a retrospective quantitative study

Abstract

Background

Medication errors (MEs) remain a leading cause of preventable harm in healthcare systems worldwide, contributing to adverse drug events, morbidity, increased healthcare costs, prolonged hospitalization, and mortality. The paediatric population is particularly susceptible to MEs due to the need for individualized dosing based on weight and age, which introduces additional complexity for healthcare providers. Although considerable research has been conducted on this issue, existing studies have predominantly focused on secondary and tertiary healthcare institutions, where specialized personnel and resources are more readily available. In contrast, in certain settings, primary healthcare facilities are staffed by non-specialist healthcare professionals who face unique challenges in paediatric care due to limited training and insufficient clinical support. This disparity highlights a critical gap in the literature and underscores the urgent need for targeted interventions to improve medication safety in primary care settings.

Objective

The study was undertaken with the aim of determining the prevalence and types of medication errors committed by nurses among children under 5 years at three (3) primary healthcare facilities in the western region of Ghana.

Materials and methods

Six (6) months retrospective data was taken to assess the prevalence and types of medication errors. Data was gathered using a checklist from records of paediatric patients aged ≤ 5years who visited the three primary health care facilities from 1st July 2020–31st December 2020 and attended to by a nurse clinician. MEs were categorised into dosage error, frequency error, prescribing error and administration error. Based on the number of errors occurring per treatment, errors were classified as No error, one error and two or multiple errors. Data was analysed using Statistical Package for the Social Sciences (SPSS) version 23.

Results

Out of the 341 paediatric folders examined, the prevalence of MEs among children under 5years from the three primary health facilities was 59.3%. Dosage errors accounted for 47.3% of identified errors, followed by frequency errors (17.8%), prescribing errors of 14.8% and administering errors 6.1%.

Conclusion

The prevalence of medication errors among children under five years across the three primary health care facilities was significantly high (59.3%).

Supplementary Information

The online version contains supplementary material available at 10.1186/s12875-025-03072-w.

Introduction

Medication errors (MEs) are defined as “any preventable event that may cause or lead to inappropriate medication use or patient harm while the medication is in the control of the healthcare professional, patient, or consumer” [1]. MEs can occur at any stage of the medication use process, such as prescribing, transcribing, dispensing, administration or monitoring [2, 3]. Common forms include overdoses, underdoses, incorrect medications, wrong patients, and improper administration routes [4–7].

Globally, MEs are a major contributor to avoidable morbidity and mortality, affecting approximately 1.3 million individuals annually in the United States and costing an estimated USD 42 billion per year [8]. These figures are likely even higher in low- and middle-income countries. Paediatric patients are disproportionately affected due to weight-based dosing requirements and developmental vulnerabilities [9], with estimated error rates ranging from 5% to 27% of medication orders in the U.S [10] and as high as 62.7% of administrations in Ethiopia [11]. MEs are the commonest ways paediatric patients can be harmed and have a much higher risk of causing death in paediatrics than in adults [12].

Inconsistencies in healthcare systems and workforce structure also contribute significantly to variability in ME rates across countries. In the United Kingdom, paediatric care is typically provided by general practitioners (GPs) who receive targeted training in child health as part of their curriculum. In contrast, the United States relies on board-certified paediatricians as primary care providers, with more extensive paediatric-specific training and clinical exposure [6]. These structural differences influence prescribing accuracy and medication safety. For example, UK hospitals report prescribing errors in approximately 1.5% of prescriptions and administration errors in 3–8% of doses [6]. A systematic review found a median prescribing error rate of 6.5%, with dose errors reaching 11.1% among hospitalised children [13]. In comparison, Kaushal et al. reported that 5% to 27% of paediatric prescriptions in the U.S. are affected by medication errors, with administration errors similarly prevalent [14]. More recently, DeCoster et al. documented over 124,000 paediatric medication errors related to ADHD therapies between 2000 and 2021, with a 299% increase in annual frequency and 4.2% resulting in serious medical outcomes [15]. These discrepancies reflect differences in provider training, documentation practices, and the integration of safety technologies such as Computerized Physician Order Entry (CPOE) systems [16].

The healthcare system in Ghana is tiered into primary, secondary, and tertiary levels. Primary Health Care (PHC) is further stratified into district hospitals, subdistrict health centres, and post and Community-Based Health Planning and Services (CHPS) zones. These facilities, often staffed by non-specialist providers, are the first point of care for most patients, particularly in rural areas [17].

The three PHC facilities in this study are subdistrict facilities operating 24 h services, with registered general nurses as the primary prescribers during afternoons, nights, and weekends. Armah & Kicha (2020) note critical challenges in PHC delivery in Ghana, including inadequate staffing, uneven distribution of personnel, insufficient training, and limited technological infrastructure [17]. These factors, combined with the complex and high-pressure clinical environment, increase the likelihood of MEs, often due to systemic failures [3, 18].

Several studies in Ghana have quantified the financial and clinical burden of medication errors in secondary and tertiary care settings. In the Eastern Region, Vanderpuije et al. estimated that documented medication errors incurred an additional cost of USD 3 432.38 in 2018, based on the National Health Insurance Scheme price list [19]. Of 150 errors reported in the same study, 73% occurred in female patients and 27% in male patients, with the highest error rate (46.5%) among adults aged 20–59 years.

In an adult emergency department of a tertiary hospital in Accra, Acheampong et al. reported a medication administration error rate of 27.2% [20] Paediatric inpatients also experience alarmingly high error rates: Koffuor et al. found a 60.5% administration error rate in one government hospital [21], while a recent study by Wuni et al. documented a 65.2% error rate in another [22].

Despite this evidence, most studies conducted in Ghana focus on adult populations, leaving a significant gap in understanding the prevalence of paediatric MEs especially in PHC settings. Undoubtedly, there is significant evidence that children experience medication errors globally and Ghana is not exempted. Having more studies into paediatric medication errors at all levels of the health delivery system will help discover the magnitude of the problem and also identify cost effective solutions to reduce MEs among paediatric patients.

Materials and methods

Study settings

The study was a hospital-based study and employed quantitative descriptive principles. The study assessed retrospective data of children under 5 years from 1 st July 2020–31st December 2020 who received medication at three (3) primary healthcare facilities under Ghana Armed Forces Medical Service (GAFMS) in the Secondi-Takoradi metropolis in the Western Region of Ghana. The 3 facilities are sub-district primary health care institutions with a total staff strength of 142 health care professionals. The facilities cater for the health needs of troops, their dependents, and as well as the general public.

Facility A is a 32-bed facility with Dental, Out Patient Department (OPD), Maternity, Reproductive and Child Health (RCH), Public health, Laboratory and Pharmacy units. Facility B is an 18-bed facility with a staff strength of 30. Facility C consist of an OPD, Dental, Maternity, Laboratory, Pharmacy and RCH units with a bed capacity of 17. Under the National health Insurance credentialing, facility A is a Clinic, facility B, a health centre with a medical doctor, and facility C, a Health post with a medical doctor. The total patient attendance in the year preceding the study was 35,448. The prescriber to patient ratio of the 3 facilities is 1:5064 as only 3 medical officers and 4 physician assistants were at post at the time of the study.

Being primary healthcare facilities, nurses perform basic consulting, prescribing and administration of medication to most patients including those under the age of 5years.

Sample size

341 paediatric patient folders.

Sample size calculation

Sample size calculation

Sample size was calculated using Yamane’s formula (n) = where.

n = Sample size.

N = Total Out Patient Department (OPD) attendance (35,448) of the three primary facilities in the year 2020.

e = margin of error set at is 0.05.

Sample size (n) =

= = 396.

Total sample size estimated for the study was 396 records of children under 5years.A proportional allocation of sample procedure was undertaken to apportion samples to the various study sites based on the OPD attendance of each facility (Table 1) in the year preceding the study.

Table 1.

Sample size of study site based on proportion of total OPD attendance

Study Site	OPD Attendance	% Of Total OPD Attendance	Sample Size N=(%/100)×396
A	11,386	32.1%	127
B	9684	27.3%	108
C	14,378	40.6%	161
TOTAL (N)	35,448	100	396

Source: Field Data, 2021

There was no differentiation for children under 5years attendance at facility B and C thus total OPD attendance was used for the calculation of the sample size.

Subjects & selection method

Data extraction form (checklist) consisting of date, a generated reference number, age in months, weight in kilograms, medical condition, medication prescribed, medication dosage and frequency was used in collecting secondary data from paediatric patients’ folders. This was guided by [5] and [4]. The doses of medications were evaluated using British National Formulary (BNF) for Children 2019–2020 and Ghana standard treatment guidelines 2017 to identify the types of medications errors; dosage error, frequency error, prescribing error and administration error [23, 24].

• Dosing error (DE) - Errors occurring due to underdose, overdose, dosage omission and dosing in millilitres (mls) without a dosage form [3, 6, 25].

• Frequency error (FE) - These are incidences where a medication’s interval is incorrect. This happens when the prescribed frequency is not indicated, frequency too often or not often enough. For example, prescribing a medication to be administered every 12 h instead of every 6 h [4, 5].

• Prescribing Error (PE) - Use of abbreviations and brand names, duplication of the same medication, unspecified medication [3, 26].

• Administration Error (AE), is administering medication at the wrong time, wrong rate, wrong route, and wrong duration of administration [26, 27].

Medication errors were classified as No error: No actual or potential harm occurred and Error: An actual or potential harm occurred using National Coordinating Council for Medication Error Reporting and Prevention’s (NCC MERP) Taxonomy of harm [28].

Inclusion criteria

All medical records of paediatric patients aged 5years and below who visited the 3 primary health care facilities from 1 st July – 31 st December 2020, attended to by a professional nurse and received any form of medication.

Exclusion criteria

Records of paediatric patients (1 day old − 16years) who visited the facilities before and within 1 st July – 31 st December 2020, received no medication therapy and were attended to by a medical doctor or physician assistant.

Statistical analysis

Secondary data obtained from patients’ folders were analysed into frequencies and percentages using Statistical Package for the Social Sciences (SPSS) version 23. Descriptive statistics using tables to show proportions and prevalence of MEs among children under 5 years, using the case definition for medication error adopted from [4, 5, 8]. Diagnoses and medication types were grouped under the top ten and top twenty respectively as occurred in the PHC facilities. To assess whether error distributions varied by facility, a chi-square test of independence was conducted. Expected cell counts satisfied assumptions for the test, and a p-value less than 0.05 was considered statistically significant.

Result

A total of 341 paediatric patients were included in the study, with 1,103 prescriptions reviewed across three primary healthcare facilities. Of these patients, 199 (58.4%) were male and 142 (41.6%) were female. Facility A contributed 89 patients (26.1% of the cohort; 45 males, 44 females) and accounted for 311 prescriptions (28.2%). Facility B enrolled 101 patients (29.6%; 61 males, 40 females) with 304 prescriptions (27.6%), while Facility C had 151 patients (44.3%; 93 males, 58 females) and issued 488 prescriptions (44.2%), as outlined in Table 2.

Table 2.

Gender Distribution, patients folders, and medication error across each facility

Facility	Gender			Medications Prescribed (N,%)	Error Status
	Male	Female	Total (N,%)		No Error (%)	Error (%)
A	45	44	89 (26%)	311 (28.2%)	46.5%	53.5%
B	61	40	101 (29.6%)	304 (27.20%)	36.1%	63.9%
C	93	58	151 (44.3%)	488 (44.24%)	40.2%	59.8%
Total(N)	199	142	341(100%)	1103(100)	40.7%	59.3%

Source: Field Data, 2021

In terms of medication error distribution, Table 2 shows Facility A had a relatively balanced error profile, with 46.5% of medications free of error and 53.5% containing errors. Facility B had the highest medication error rate, with 63.9% of prescriptions involving at least one error. Facility C followed closely, recording a 59.8% error rate. Overall, 59.3% of all prescribed medications across the three facilities contained one or more errors, while 40.7% were error-free.

Diagnosis identified

The most commonly recorded diagnoses among paediatric patients were upper respiratory tract infection (URTI) (n = 102, 29.9%) and malaria (n = 97, 28.4%). These two conditions accounted for more than half of all diagnoses. Gastroenteritis followed with 66 cases (19.4%), while dermatitis (n = 29, 8.5%) and tonsillitis/pharyngitis (n = 14, 4.1%) were less frequent. Less common conditions included bronchopneumonia and sepsis, each with 12 cases (3.5%), urinary tract infections (n = 8, 2.3%), otitis media (n = 7, 2.1%), and asthma (n = 5, 1.5%). The top ten diagnoses recorded across the study period are summarized below in Table 3.

Table 3.

Top 10 diagnosis identified

SRL	Diagnosis	Total
1.	Upper Respiratory Tract Infection (URTI)	102
2.	Malaria	97
3.	Gastroenteritis	66
4.	Dermatitis	29
5.	Tonsilitis/Pharyngitis	14
6.	Bronchopneumonia	12
7.	Sepsis	12
8.	Urinary tract Infection	8
9.	Otitis Media	7
10.	Asthma	5

Prescribed medication

Among the 1,103 prescriptions reviewed, antibiotics were the most commonly prescribed drug class, accounting for approximately 28% of all medications (Table 4). This trend may reflect the high prevalence of infectious diseases such as URTIs, malaria, and gastroenteritis in the study population. The top 20 prescribed medications and their associated error rates are presented in Table 4.

Table 4.

Top 20 prescribed medications and associated error rates

SRL	Medication	Total Prescriptions 1103(%)	Number of Errors per facility				Error rate
			A	B	C	TOTAL
1.	Amoksiklav Susp	112 (10.1)	7	5	22	34	30.4%
2.	Paracetamol (IV, Syr, Supp.)	92 (8.4)	10	5	32	47	51.1%
3.	Artemether Lumefantrine	78 (7.0)	3	27	6	36	46.2%
4.	Cefuroxime Susp	73 (6.6)	3	19	29	51	69.9%
5.	Cetirizine Syr	67 (6)	20	4	29	53	79.1%
6.	Metronidazole/Flagyl Susp	50 (4.5)	12	18	10	40	80.0%
7.	Cough mixture	44 (4.0)	2	0	7	9	20.5%
8.	Gvither/Artemether Inj.	34 (3.1)	1	25	8	34	100.0%
9.	Zinc Tab	32 (2.9)	6	2	4	12	37.5%
10.	Vit C Syr	32 (2.9)	1	1	11	13	40.6%
11.	Amoxicillin Susp.	29 (6.2)	1	12	7	20	69.0%
12.	Saline nasal drop	27 (2.4)	8	0	16	24	88.9%
13.	Ibuprofen Syr.	24 (2.2)	6	0	11	17	70.8%
14.	Flucloxacillin Susp.	22 (1.9)	1	1	6	8	36.4%
15.	Oral Rehydration Salt (ORS)	22 (1.9)	0	2	0	2	9.1%
16.	Multivitamin Syr	18 (1.6)	5	1	3	9	50.0%
17.	Azithromycin Susp	15 (1.4)	3	0	5	8	53.3%
18.	IV Ceftriaxone	12 (1.1)	5	0	0	5	41.7%
19.	Salbutamol Nebule	7 (0.6)	0	1	0	1	14.3%
20.	IV Artesunate IV	6 (0.5)	0	1	0	1	16.7%

Syr Syrup, Susp Suspension, Supp Suppository, Inj Injection, IV Intravenous

Analysis of medication errors revealed substantial variation in error rates across different drug types. Notably, Gvither/Artemether Injection exhibited a 100% error rate, followed by Saline nasal drops (88.9%), Metronidazole Suspension (80.0%), and Cetirizine Syrup (79.1%). These medications, despite varying prescription volumes, were disproportionately associated with errors, suggesting potential issues in prescribing, dispensing, or administration practices.

Error Rate (%) is defined as the proportion of prescriptions that resulted in documented medication errors, calculated as: (Total Errors ÷ Total Prescriptions) × 100. This metric reflects the percentage of prescriptions for each medication type that were associated with errors across all facilities.

Medication errors

Types of medication errors identified

Dosing error

Across the reviewed prescriptions, dosing errors were the most prevalent type of medication error, observed in 47.3% of the records. Facility C recorded the highest proportion of dosing errors (50%), with a similar trend seen in Facilities A and B.

Notably, 11.1% of all dosing errors involved overdosing, 7.1% were underdoses, and 5.9% of prescriptions had no dosage stated. A significant observation as seen in Table 5 was the frequent prescription of medications in millilitres (ml) without accompanying standardized dosage guidance, which was noted in 23.2% of dosing errors.

Table 5.

Types of medication errors

Error Category	Subtype	Overall %	Facility A %	Facility B %	Facility C %
Dosing Errors	No dose	5.9%	-	-	-
	Over dose	11.1%	-	-	-
	Under dose	7.1%	-	-	-
	Dose in mils	23.2%	-	-	-
	Overall dosing error rate	47.3%	50.0%	48.2%	43.6%
Frequency Error	No Frequency	8.8%	-	-	-
	Over Frequency	2.4%	-	-	-
	Under Frequency	4.9%	-	-	-
	Overall frequency error rate	17.8%	20.4%	16.4%	16.5%
Prescribing Error	Use of abbreviated name	0.4%	-	-	-
	Use of brand name	8.6%	-	-	-
	Unspecified medication	1.0%	-	-	-
	Inappropriate drug form	2.7%	-	-	-
	No route of administration	0.9%	-	-	-
	Overall prescribing error rate	14.8%	6.4%	22.4%	15.1%
Administration Error	Yes	6.1%	-	-	-
	No	93.9%	-	-	-

Frequency Errors

Taking into consideration frequency error as shown above in Table 5, the numbers were much lesser as compared to dosing errors among the three facilities. There was a proportion of 16.5% frequency errors among patients’ records at facility C whilst facility B had 16.4% and A had the highest proportion of 20.4%. Frequency errors contained 8.8% of unstated or no frequency, 2.4% overstated frequency, and 4.9% understated frequency, with 83.9% of them having no frequency errors.

Prescribing error

Prescribing errors constituted 14.8% of all medication related errors identified in the study. Of these: 0.4% involved the use of non-standard abbreviations such as Pmol for paracetamol and Cef for ceftriaxone. 8.6% of prescriptions were written using brand names, including Amoksiklav, Pizavit, Kofof, Bactroban, and Lonart, rather than their generic equivalents. 1.0% did not specify the medication prescribed, with vague entries such as “syr antimalaria” and “IVF” (intravenous fluid) without indicating the specific formulation. 2.7% involved inappropriate drug formulations, such as prescribing tablets for medications only available as suspensions and 0.9% of prescriptions omitted the route of administration entirely.

Administration Error

There was a total of 6.1% administration errors across the three facilities accounting for the least form of identified ME.

A chi-square test of independence as seen above in Table 6, was performed to examine the relationship between facility and error count category. All expected cell frequencies exceeded five. The test was significant, χ²(4, N = 1073) = 11.64, p = 0.021, indicating that the distribution of error counts differs by facility. Specifically, Facility B showed a higher proportion of single-error prescriptions, whereas Facilities A and C had similar rates of multiple-error prescriptions.

Table 6.

Distribution of prescription error counts by facility and Chi-Square test of independence

Facility	No Errors n (%)	1 Error n (%)	> 2 Errors n (%)	Total n
A	133 (44.9)	114 (38.5)	49 (16.6)	296
B	116 (38.7)	152 (50.7)	32 (10.7)	300
C	188 (39.4)	211 (44.2)	78 (16.4)	477
Total	437 (40.7)	477 (44.5)	159 (14.8)	1 073

Percentages are within-facility

Discussion

Medication errors identified

This study revealed a high prevalence of medication errors (MEs) in paediatric prescribing across three primary healthcare facilities in Ghana, underscoring systemic vulnerabilities in primary care delivery particularly within GAFMS [29]. The high rate of MEs aligns with global findings from the World Health Organization (2016), which identified primary care settings as especially susceptible to preventable harm due to medication errors, particularly among paediatric populations [2]. Evidence also shows that facilities staffed with paediatric-trained personnel are less likely to report MEs [4], while a meta-analysis by Gates et al. (2019) found that the frequency and nature of medication errors are shaped by facility structure and level of specialization. These findings emphasize the need for policy reform within GAFMS to address infrastructure limitations, staffing gaps, and training deficits in primary healthcare settings.

Among the 341 paediatric records reviewed encompassing 1,103 unique medication regimens, 59.1% included at least one medication error. Antibiotics, antimalarials, and antipyretics were the most commonly prescribed drug classes. Analysis of the top 20 prescribed medications revealed substantial variability in error rates, ranging from 9.1% to 100%, with 13 out of 20 medications exceeding a 40% error rate. Notably, Gvither/Artemether Injection recorded a 100% error rate, indicating that every prescription contained at least one error. This is particularly alarming given its role in treating severe malaria, where dosing precision is critical.

Other medications with exceptionally high error rates included:

• Saline nasal drops (88.9%).

• Metronidazole/Flagyl Suspension (80.0%).

• Cetirizine Syrup (79.1%).

• Ibuprofen Syrup (70.8%).

• Amoxicillin Suspension (69.0%).

• Cefuroxime Suspension (69.9%).

These medications are frequently used in paediatric care, and their high error rates may reflect challenges in weight-based dosing, formulation confusion, and inadequate documentation. For instance, Metronidazole suspensions are available in multiple concentrations, and failure to specify the strength can lead to significant dosing errors a pattern observed in 23.2% of dosing errors across the study. This concern is echoed in a cross-sectional study by Kuitunen et al. (2023), which found that high-alert medications including intravenous antibiotics and analgesics such as Metronidazole and Ibuprofen were disproportionately associated with severe harm and higher error risk classifications in paediatric hospital settings [30].

Conversely, medications such as Oral Rehydration Salt (ORS) and Salbutamol Nebules had relatively low error rates (9.1% and 14.3%, respectively), likely due to their standardized dosing and limited formulation variability. Paracetamol, prescribed in various forms (IV, syrup, suppository), showed a 51.1% error rate, suggesting potential confusion across routes and formulations. Similarly, Artemether Lumefantrine, a first-line antimalarial, had a 46.2% error rate, which could compromise malaria treatment outcomes.

A review by the Royal Pharmaceutical Society highlighted that paediatric prescribing errors affect approximately 13% of prescriptions, with wrong dose and omission errors being among the most frequent, particularly for liquid formulations like antihistamines and analgesics [31].

The distribution of errors across facilities also varied. Facility B consistently recorded higher error counts for medications such as Artemether Lumefantrine and Gvither/Artemether Injection, suggesting possible gaps in prescriber training or supervision during high-volume shifts. These findings highlight the need for targeted safety interventions, including standardized paediatric dosing charts, clear documentation of drug strength and route, regular prescriber training, and integration of decision-support tools to reduce manual calculation errors.

Regionally, Wondmieneh et al. (2020) reported that 68.1% of nurses out of a cohort of 298 randomly selected in Ethiopian tertiary hospitals had committed medication administration errors, citing inadequate training, lack of standardized guidelines, and frequent interruptions as key contributing factors.

Collectively, these studies underscore the urgent need for continuous monitoring, targeted staff education, and system-level reforms to reduce medication error rates in paediatric care settings.

Dosing errors were the most frequent type, comprising 47.3% of all medication errors across the study sites. This mirrors findings from other settings, including Ethiopia and Australia, where similar errors were reported in paediatric wards at Nekemte Referral Hospital, Jimma University Medical Center and a paediatric public tertiary teaching hospital in Australia.

[5, 13, 29]. The prevalence of dosing errors is often attributed to the complexity of paediatric prescribing, which requires individualized, weight-based calculations [16, 32–34]. To address this challenge, Children should be weighed accurately at each visit, and hospital managers should provide standardized paediatric dosing charts to prevent individualised miscalculation and dosing errors [35–37].

Among dosing errors, overdosing (11.1%) was more common than underdosing (7.1%), consistent with prior literature [5, 12, 13, 26]. Both underdosing and overdosing pose significant risks, including treatment failure, drug toxicity and the emergence of drug-resistant strains especially for antibiotics and antimalarial [38–40].

An additional 23.2% of dosing errors involved prescriptions written in millilitres (ml) without reference to drug strength or concentration. This approach is particularly concerning, as many paediatric suspensions such as metronidazole are available in multiple concentrations (e.g., 100 mg/5 ml vs. 200 mg/5 ml), increasing the risk of unintentional overdosing or underdosing when doses are not expressed in milligrams [41]. Best-practice standards recommend expressing paediatric doses in mg not ml to reduce misinterpretation [41].

Frequency errors accounted for 17.8% of all medication errors in this study, making them the second most prevalent error category after dosing errors. The most common subtype observed was the omission of frequency intervals. In children under five, inaccurate dosing frequency can cause subtherapeutic or toxic levels, potentially leading to worsened outcomes especially where accurate timing is critical to maintaining therapeutic drug levels. Inadequate or unclear frequency instructions may lead to missed or duplicated doses, reduced treatment efficacy, or increased adverse effects [42].These errors often stem from incomplete or vague documentation, high patient load, and limited access to paediatric-specific prescribing tools, especially particularly in primary care settings with non-specialist prescribers [4, 5]. WHO (2017) also highlights the role of handwritten prescriptions and inconsistent clinical supervision in such errors. Introducing clinical decision support tools and refresher training could significantly reduce frequency-related mistakes. While frequency errors were fewer than dosing errors, any level of error in paediatric care merits urgent attention.

Prescribing errors, accounting for 14.8% of total MEs, included the use of brand names (8.6%), inappropriate drug formulations (2.7%), unspecified medications (1.0%), omission of administration route (0.9%), and use of non-standard abbreviations (0.4%). Examples included vague entries such as “syr antimalaria” or using abbreviations like “Pmol” for paracetamol and “Cef” for ceftriaxone. Although the rate is lower than studies in a secondary tertiary level public hospital in Western Mexico which had 72% prescribing errors with 50.9% being use of abbreviations [43], it still warrants attention due to the risk of misinterpretation and patient harm.

Administration errors were least frequent (6.1%) and were most common at Facility A (8.6%), which had an inpatient paediatric ward. This may reflect higher in-facility medication handling compared to Facilities B and C, where children were mainly treated on an outpatient basis. This observation aligns with studies indicating that MEs frequently occur at both prescribing and administration stages [2, 4, 27].

A chi-square test of independence revealed a statically significant association between the facilities and the number of prescription errors, χ²(4, N = 1 073) = 11.64, p = 0.021. This suggest that the frequency of medication errors frequencies either none, one or multiple was not uniform across the three primary healthcare facilities.

Facility B showed a higher proportion of single-error prescriptions (50.7), whilst Facilities A and C comparatively each had higher rates of multiple-error prescriptions. These discrepancies may reflect institutional differences in staffing, prescriber experience, supervision or adherence to standard prescribing protocols.

Overall, the majority (59.3%) of paediatric prescriptions in the three PHC facilities contained at least one error. While not all errors result in harm, the potential for adverse outcomes particularly in a vulnerable population such as children under 5 years emphasizes the need for enhanced pharmacovigilance. Future research should assess the clinical impact of these errors, including harm severity, to inform context-specific interventions aimed at improving medication safety in primary care.

Recommendation

1. Standardized Paediatric Treatment Guidelines: Collaborate with the Paediatric Society of Ghana to develop and disseminate national paediatric protocols. In the interim, provide every PHC facility with WHO’s Pocket Book of Hospital Care for Children: Guidelines for the Management of Common Illnesses with Limited Resources [44].

2. Technological Interventions: Implement Computerized Physician Order Entry (CPOE) systems integrated with clinical decision support and paediatric drug dosing applications to minimize manual calculation errors.

3. Training, Documentation, and Oversight: Conduct regular training programs on weight-based prescribing and error reporting to help eliminate ambiguous instructions.

4. Equitable Deployment of Paediatric Nurses and Nurse Specialists: Ensure primary care teams include paediatric nurses and nurse specialists, as nurses are the 1 st point of contact in primary healthcare settings. This would help improve paediatric clinical outcomes in primary healthcare facilities.

Conclusion

This study reveals a high prevalence of paediatric medication errors in primary healthcare facilities in Ghana, with dosing errors accounting for nearly half of all identified incidents. The absence of standardized paediatric dosing tools, reliance on non-specialist prescribers, and inconsistent prescription practices including frequent use of brand names and vague instructions highlight systemic vulnerabilities in primary healthcare delivery. Although administration errors were relatively infrequent, the overall error burden underscores the urgent need for targeted interventions. Strengthening paediatric medication safety in Ghana’s primary healthcare settings will require a multifaceted approach:

Abbreviations

AAP

American Academy of Pediatrics

AE

Administration Error

AMR

Antimicrobial Resistance

CDC

Center for Disease Control and Prevention

CHPS

Community-based Health Prevention Service

CPOE

Computerized Physician Order Entry

DE

Dosing Error

EHR

Electronic Health Record

FE

Frequency Error

FACILITY A

2nd Medical Reception Station(2MRS)

FACILITY B

Air Force Medical Centre(AFMC)

FACILITY C

Western Naval Command Medical Centre(WNCMC)

GAFMS

Ghana Armed Forces Medical Service

ME

Medication Error

NCCMERP

National Coordinating Council for Medication Error Reporting and Prevention

OPD

Out Patient Department

PHC

Primary Health Care

PE

Prescribing Error

SDG

Sustainable Development Goal

WHO

World Health Organization

Authors' contributions

BBA: I declare that the work presented is the results of my own original which was in partial fulfilment of my paediatric membership programme at Ghana College of Nurses and Midwives. This has neither in part nor in whole been presented to any journal for publishing. Work done by other authors which served as useful sources of information have been duly acknowledged by making references to them.VES: As the principal supervisor for this research, she contributed in the design, analysis and interpretation of data.VA: Contributed in the development of the checklist and assisted in pre-testing the checklist at 37 military hospital PEU prior to it being used at the three primary healthcare facilities. She also Proof read the research and manuscript.

Funding

This research was funded by the corresponding author.

Data availability

Data that supports the findings are available at 2Garrison medical facilities, but access is restricted due to ethical considerations. Data sharing requires approval from GAFMS.

Declarations

Ethics approval and consent to participate

This was granted by Ghana Armed Forces Medical Service (GAFMS) IRB at 37 Military Hospital with reference 37/MH-IRB/FLP/IPN/479/2021.

The authors declare that they obtained the authorisation from the GAFMS IRB to collect secondary data for this research. The authors assure that data collected will be kept confidential in accordance with research ethics and applicable regulations.

Consent for publication

The authors declare that they have obtained all the necessary consent to publish this manuscript from the GAFMS IRB.

Competing interests

The authors declare no competing interests.