Medication errors in malaria management in children: insights from pharmacovigilance data in the Democratic Republic of Congo

Abstract

Introduction

Since 2012, DRC adopted WHO recommendations for the malaria treatment with artemisinin-based drugs. Medication errors are defined as “a failure in the treatment process that results in, or has the potential to result in, harm to the patient”. Medication errors are a major public health problem and one of the leading causes of death in the United States. The impact of medical errors can have severe consequences on children due to physiological features of children.

Objectives

To identify, describe, and propose actionable strategies to address medication errors during malaria treatment in children with adverse effects in the DRC.

Methods

This is a cross-sectional study of the ADR reports of children (< 18 years old) in the DRC recorded in VigiBase^®, the WHO pharmacovigilance database, from 2010 to February 2018. Five treatment process criteria (choice of treatment, dosage, duration, timing and route of administration) were selected to identify medication errors.

Results

Medication errors accounted for 65,9% of the 851 cases retrieved from VigiBase^®. Children aged 2–11 years represented 55.2% of the study population. The choice of treatment, duration and dosage were the main prescription criteria for deviations.

Discussion

The availability of alternative formulations, self-medication and inadequate dosage forms are factors contributing to medication errors. The information available in VigiBase^® did not allow us to evaluate the overall process of malaria management. Pharmacovigilance must be consolidated to raise awareness among consumers and providers and to ensure more effective monitoring.

Conclusion

Non-compliance with national guidelines for the management of malaria is important in DRC. Our study amply demonstrates the need to strengthen the four pillars of the WHO’s third global challenge, “Medication without harm”, to reduce medication errors. This study advocates a significant mobilisation of resources for the training of health professionals and the strengthening of pharmacovigilance. Field studies on the management of malaria in children should be conducted to quantify drug errors.

Introduction

The World Health Organization (WHO) defines a drug-related injury as any adverse reaction to a drug at recommended doses [1]. Medication errors are defined by Ferner R.E. et al. as “a failure in the treatment process that results in, or has the potential to result in, harm to the patient” [2, 3]. Medication errors are a major public health problem and one of the leading causes of death in the United States [4]. The lack of indicators to detect error in developing countries may worsen the situation [5, 6]. Since March 2017, the WHO has launched Medication Without Harm, a challenge to halve the number of avoidable adverse effects caused by medication errors. “Medication Without Harm” acts on 4 axes: (1) patients and the public, (2) health professionals, (3) medicines used as products and (4) drug treatment systems and practices [7, 8].

Malaria is the most common disease in Africa with 94% of 263 million cases according to the 2024 WHO report [9]. Due to the emergence of resistance to the first-generation of anti-malarial molecules, WHO has recommended the use of artemisinin-based combination therapies (ACTs) [10, 11]. In 2023, the National Malaria Control Program (NMCP) of the Democratic Republic of the Congo (DRC) provided 26.2 million ACT doses to healthcare facilities [9]. However, the availability of many ACT formulations may deviate from national recommendations or lead to errors [12, 13]. According to the NMCP, 50% of the 27 million cases of malaria recorded nationwide in 2022 were under-five children [14]. The impact of medical errors can have severe consequences on children due to physiological features of children [15–20].

Over the last 30 years, pharmacovigilance systems have been established across the continent to improve patients’ safety throughout medicines administration and other interventions [21, 22]. However, due to the poor quality of locally available medicines and the weakness of current regulations, evidence of good drug management in low-income countries is limited [23–26]. The pharmacovigilance was gradually implemented in DRC since 2003 to become part of the WHO programme in 2010. But, due to the country size and lack of consistent fundings, the DRC was not able to sustainably implement the pharmacovigilance throughout the country [27]. For example, in 2022 the DRC received 24 million ACTs doses, representing more than 10% of the antimalarial drugs distributed by worldwide [9]. However, the quality of prescription of antimalarial drugs, the widespread of self-medication and drugs misuse may expose important number of individuals to medication errors [13, 28–30]. Additionally, study addressing medication errors in DRC are rare. A 2015 study by Muhindo of malaria control policy and practice in Kinshasa found that only 45.7% of prescriptions aligned with guidelines [31]. More than 30 different antimalarial drugs were prescribed without evidence to treat uncomplicated malaria, according to a study by Nsengi et al. [32]. This study aims to identify, describe, and propose actionable strategies to address medication errors during malaria treatment in children with adverse effects.

Methods

Source

As a WHO’s Drug Safety Program member, the DRC has access to the World Health Organization’s (WHO) international pharmacovigilance database VigiBase^® [33].

VigiBase^® is the unique WHO global database of individual case safety reports (ICSRs). It is the largest pharmacovigilance database in the world, with more than 20 million reports of suspected adverse effects of medicines submitted since 1968 by member countries of the WHO Program for International Drug Monitoring (PIMD). VigiFlow^® is a management system for adverse event reports. It supports the national collection and sharing of ICSR data and provides effective data management [33].

Study design and population

A cross-sectional study was conducted on pharmacovigilance data recorded in the DRC in VigiBase^® from September 2010 to 8 February 2018 (extraction date). The study population consisted of Individual Case Safety Report (ICSR) involving children under 18 years of age and antimalarial drugs selected by their ATC codes. This is a convenience sample based on the availability of safety reports.

Definition of exposure

Antimalarial drugs were selected according to the list provided by VigiBase^®. Antimalarials were coded according to the World Health Organization (WHO) Drug Dictionary and the Anatomical Therapeutic Classification (ATC) [34].

WHO malaria treatment recommendations were published in 2001, with updates in 2006 and again in 2010. Several combinations have been proposed for the treatment of uncomplicated malaria. The first-line treatment used included one of the following ACTs: (1) Artemether + lumefantrine; (2) Artesunate + amodiaquine; (3) Artesunate + mefloquine; (4) Artesunate + sulfadoxine-pyrimethamine; and (5) Dihydroartemisinin + piperaquine. Of note, only severe malaria requires intravenous administration of artesunate [10, 11]. In 2005, the DRC adopted two oral combinations, artesunate + amodiaquine (ASAQ) or artemether-lumefantrine (AL), for the treatment of uncomplicated malaria. Injectable artesunate was introduced in 2012. Posters illustrating the different treatments were displayed in health facilities [35–37].

Outcome

Medication errors can be classified into three categories: (i) A latent error (or risk of error) is an observation that indicates a potential danger for the patient; (ii) The potential of this approach is intercepted before the product is administered to the patient; (iii) Proven error was considered when the wrong drug is administered to the patient, the wrong dose is administered, the wrong route of administration is used, or the wrong therapeutic regimen is administered [38] This type of error can compromise patient safety and cause adverse reactions. One example of this is the administration of the combination Pyrimethamine + Sulfadoxine for the treatment of malaria, despite it only being indicated for prophylaxis in pregnancy [10, 13, 35]. The diagram below shows the relationship between medication errors and adverse reactions (Fig. 1) [3].

Fig. 1.

Relationships between adverse drug reactions and medication errors

Deviation may be related to non-adherence to prescriber recommendations or medication errors (especially when there is a difference in dosage, rhythm, route, or duration). For this study, we had access to reports and narratives extracted through the VigiFlow^® search and statistics function. To identify deviations from best practice, the data were reviewed by two independent reviewers on a case-by-case basis. A reading grid was developed. Based on the information available from the reports, only prescribing and administration errors were assessed. These deviations could correspond to the voluntary use of drugs outside the recommendations or to medication errors (prescription or administration errors). Five criteria were selected to assess medication errors based on the available data:

1. Adherence to national recommendations, applicable according to the year of administration, per the reported indication and the prescribed drug. The term “uncomplicated malaria” included the following ICD-10 terms “Plasmodium falciparum malaria”, “Plasmodium falciparum malaria, unspecified” or “unspecified malaria”, whereas the term “severe malaria”, “complicated malaria”, “Plasmodium falciparum malaria with cerebral complications” or “other severe and complicated Plasmodium falciparum malaria” was used.

2. Dosage: calculation based on the child’s weight, if available;

3. Dosage schedule: adapted to the form and dose of the drug;

4. Route of administration: as prescribed.

5. The duration of treatment was incorrect if it exceeded the recommended duration of treatment for the drug concerned; on the other hand, a shorter duration of treatment than recommended could be explained by the discontinuation of the drug due to adverse effects and was not considered as an error.

Our interpretation is that there was a medication error if the answer to one of these five compliance criteria was negative and that there was a lack of data if no information was available for a decision to be made.

Data collection and analysis

The data were extracted from VigiBase^® by a National Pharmacovigilance Centre member and saved in an Excel^® 2016 file. The statistical processing and data analysis were performed with SAS^® University Edition, SAS Version 9.4 software (SAS Institute, Inc., North Carolina, USA).

The statistical analysis included the frequency and nature of the prescription and administration medication errors. Patients were also stratified according to the severity and frequency of adverse events by patient age and sex. The data were presented in tables and graphs, and the qualitative variables were summarized in the form of proportions and percentages. Quantitative variables were synthesized as the mean, median, and mode by central tendency indices and as standard deviations and quartiles for dispersion indices. The chi-square test was used to compare proportions (α = 0.05).

Regulatory aspects

As a member of the WHO Program for International Drug Monitoring, the DRC is a permanent access code holder for VigiBase^® for data extraction for research purposes in the public health context.

Results

Study population

As of 8 February 2018, a total of 851, representing 6.3%, of the DRC’s Individual Security Reports (ICSRs) had been retained and constituted the study population. (Fig. 2).

Fig. 2.

Population study inclusion flow diagram

When extracted, 16 420 047 reports were found in VigiBase, of which 13 439 were from the DRC. A total of 851 safety reports were extracted after applying the inclusion criteria of age under 17 years and presence of at least one antimalarial drug.

The sex ratio was approximately 1.01:1, indicating a nearly equal representation of males and females. The median age was 5 years, and the overall mean age was 6.8 years (95% CI 6.4, 7.2). The average ages for both sexes were 6.9 years for females and 6.8 years for males. The largest proportion of ICSRs were collected from children aged 2 to 11 years (55.2%), followed by adolescents (24.4%), infants (19.7%) and five neonates (0.6%). Six hundred and ninety-nine (699; 82.1%) ICSRs contained at least two suspected or concomitant drugs, and 80.7% reported at least two adverse events, more than one-third (39.2%) of which were serious, according to the reporters. Artemisinin-based combination therapies (ACTs) were the most common drugs of recovered in the data (61.8%). Uncomplicated malaria was the most common indication (88.8%). Two-thirds of the notifiers were “other health professionals” (67.9%), such as nurses, midwives, and students; 28.3% were physicians; 10 were pharmacists; and two lawyers were also found among the reporters.

Characteristics of medication errors

Prescription and administration medication errors were identified in 65.9% (561) of the ICSRs. We missed data for 11.16% of the patients.

The results of the overall analysis of medication errors are presented in Table 1. After stratifying the patients according to the ICSRs, with or without errors, no statistically significant differences were found between the different age groups for age, sex or number of medications. However, the indication, number and evolution of adverse effects were significantly associated with prescription and administration medication errors. Uncomplicated malaria was the main indication among the notifications, with differences of 55.4% versus 22.6% (p < 0.001). Medication errors were associated with reports containing more than two drugs (22.4%) compared with reports containing ‘1–2’ drugs (61%), (p < 0.001). And all ICSRs reporting deaths (3; 0.4%) contained medication errors.

Table 1.

ICSR characteristics by deviation from recommendations

Variables	Medication errors, n (%)			Total N = 851	p
	Yes n = 561 (65.9%)	No n = 195 (22.9%)	Missing data n = 95 (11.2%)
Age groups
Neonate	4 (0.5)	1 (0.1)	0	5(0.6)
Infant	110 (12.9)	39 (4.7)	19 (2.2)	168(19.8)
Child	303 (35.6)	113 (13.2)	54 (6.4)	470(55.2)
Adolescent	144 (16.9)	42 (4.9)	22 (2.6)	208(24.4)
Sex
Male	287 (33.7)	87 (10.2)	50 (5.9)	424(49.8)
Female	270 (31.7)	107 (12.6)	44 (5.2)	421(49.5)
Not specified	4 (0.5)	1 (0.1)	1 (0.1)	6(0.7)
Indication					< 0.001
Uncomplicated malaria	471 (55.4)	192 (22.6)	93 (10.9)	756(88.8)
Severe malaria	89 (10.5)	3 (0.4)	2 (0.2)	94(11.1)
Others	1 (0.1)	0	0	1(0.1)
Number of drugs					P = 0.01
1–2	456 (53.7)	172 (20.2)	71 (8.3)	699(82.2)
> 2	105 (12.3)	23 (2.7)	24 (2.8)	152(17.8)
Number of ADR					< 0.001
1–2	451 (53.0)	170 (20.0)	66 (7.7)	687>(80.7)
3–4	101 (11.9)	25 (2.9)	29 (3.4)	155(18.2)
5–6	9 (1.06)	0	0	9(1.06)
Seriousness of ADR
yes	267 (31.4)	41 (4.8)	26 (3.0)	334(39.2)
No				517(60.8)
ADR evolution					< 0.001
Favorable	519 (61.0)	191 (22.4)	89 (10.5)	799(93.9)
With Sequellae	2 (0.2)	0	0	2(0.2)
Ongoing	16 (1.8)	4 (0.5)	4 (0.5)	24(2.8)
Death	3 (0.4)	0	0	3(0.4)
Unknown	10 (1.2)	0	0	10(1.2)
Missing data	11 (1.3)	0	2 (0.2)	13(1.5)

Values in bold are those with a significant p-value

The results of the analyses according to the evaluation criteria are presented in the table below.

Of the 561 patients whose data deviated from the recommendations, 55.4% were not compliant with the national guidelines for malaria management. Incorrect dosages were found in 209 (37.2%) notifications, 52.6% of which were higher than national recommendations, represented mainly by 75.5% of the quinine prescriptions. The artesunate + amodiaquine combination (ASAQ) was mostly prescribed (67.7%) at 47.4% at lower dosages. An inadequate administration frequency was found in 29.1% of the patients. The route of drug administration was inappropriate for 12 ICSRs (2.1%). Treatment duration was excessive in 52.9% of the patients.

According to the number of negative criteria, of 561 notifications with errors, 284 (33.4%) had only one criterion assigned, and in two ICSRs, all criteria were assigned.

For other drugs found in VigiBase^®-derived reports, analgesics were the most common used (302, 49.6%) among concomitant medications in reports with medication errors, followed by antimicrobials (26.4%) and vitamins (5.1%).

Discussion

According to the inclusion criteria, 851 ICSRs of antimalarial drugs prescribed to children were extracted from VigiBase^® over 8 years. The study population consisted mainly of children aged between 2 and 11 years. The average number of adverse reactions reported in the ICSRs was 1.8.

The age distribution of our study population is consistent with findings from studies conducted in Nigeria and Ghana [39, 40]. This distribution underscores the vulnerability of children to malaria, particularly those under 5 years old who have not yet developed immunity through regular exposure to the Plasmodium parasite [41, 42]. In European studies, the mortality rate of 0.4% ranged from 0.1 to 2.2%, which is lower than the 3.4–13% observed in North America (), where medication errors were more prevalent [43–46]. The three recorded deaths involved children aged 1 to 5 years who were treated for uncomplicated malaria with drugs not recommended by national guidelines such as sulfadoxine-pyrimethamine (SP), which is reserved for malaria prophylaxis during pregnancy in DRC [13]. Nevertheless, the following questions remain: were these deaths due to medication errors, the severity of adverse events, or a under-diagnosed severe malaria? [44–47].

The proportion of prescription and administration errors in our study is similar to that found in Muhindo’s 2015 study [31]. Nevertheless, our findings are lower than the 75.1% reported in the Ethiopian study by Dedefo and colleagues [48]. But are within the range of 15.2-88.6% previously reported in South Asian countries (Singapore, Malaysia, Thailand, Vietnam, the Philippines and Indonesia) [49]. In Thailand, Sangtawesin et al. found a frequency of 35.4% prescription errors due to incorrect dosing in a pediatric hospital [50]. The fact that our study only examined errors related to adverse events and the high number of missing data reports may have led to an underestimation of the true burden. The criteria most influenced by guideline adherence and dosage are similar to those found in several studies reporting pediatric medication errors in Denmark, Egypt and Ethiopia [6, 48, 51]. Childhood is a period of rapid weight gain, and certain drug formulations are not necessarily adapted to individual differences. In some remote areas without appropriate equipment, community workers may administer drugs based on age. Additionally, dosages may mislead prescribers if they do not consider factors such as malnutrition and obesity [12, 19, 20, 52–54].

In our study, 57% of reports included a first-line ACT recommended by the NMCP: artesunate and amodiaquine (ASAQ). In the Muhindo study, the two recommended ACTs accounted for 50% of the total ACTs prescribed [31]. Our rate is much higher than in Dodoo’s 2009 study (43.0%) [12]. However, this comparison is limited because our study only covered a later time period (2010–2017), and this compliance would demonstrate commitment to fighting antimalarial misuse. Although the NMCP has a two-year transition plan, six years after the adoption of the WHO recommendations, non-conforming formulations still exist in the DRC [37].

The availability of several non-recommended medicines is thought to encourage these deviations. According to a survey conducted by ACTwatch in 2015, non-ACT combinations accounted for almost half of the antimalarials available on the market. This is due to the private sector’s dominance and price differences [13, 31, 55]. At the same time, the national program suspected that harmful behaviors among the population were at the root of non-compliance with national guidelines, such as self-medication and rejection of ASAQ due to fear of adverse effects. Moreover, these new treatments are not affordable for low incomes [10, 13]. This highlights the need for public education and training of health professionals [56–58].

We considered the difference in duration of treatment in relation to recommendations to extend treatment, using information available in the International Drug Monitoring Database. Drug discontinuation, with or without substitution, was included in the scope of drug use studies. Therefore, it was not considered in this study.

Adverse reactions to antimalarials derived from VigiBase^® could also had been underestimated, given the size of the pediatric population and the prevalence of malaria in the country. However, the DRC could face under-reporting of the pharmacovigilance phenomena as observed worldwide. Mandatory reporting of adverse events related to healthcare products, as required by pharmacovigilance guidelines, has not yet been integrated into daily practice [59]. Furthermore, other limiting factors could have delayed the implementation of pharmacovigilance measures, such as the weakness of health systems and the lack of financial resources [21, 22, 26, 60]. A useful step towards curbing this scourge would be training and disseminating understandable messaging to health providers, emphasising the safety of healthcare products. Training in electronic reporting and mobile application use could overcome accessibility issues in some regions. Another step would be raising community awareness of the dangers of self-medication and misuse of health products [56–58].

Our study used restricted access SAS University software at the time of paper writing, and it is no longer possible to update pharmacovigilance data using the same methodology. Furthermore, the lack of multivariate adjustments restricts the imputation of our results. Another limitation is that because VigiBase^® is the basis for reporting adverse drugs reactions, only prescription and administration errors were analysed. This approach does not provide a comprehensive overview of all medication errors at different stages of malaria treatment, including dispensing, preparation, and administration. Reports of adverse events, mainly from health districts supported by international agencies involved in malaria control, are likely to be subject to selection bias to minimize the proportion of non-compliant antimalarial drugs [21, 22, 31, 60]. Moreover, the lack of relevant information meaning that many notifications could have not be definitively concluded.

This study is the first to be conducted on medication errors in the DRC. It is a step forward in antimalarial pharmacovigilance. It also provides insight into the antimalarials prescribed to the paediatric population, which is the largest consumer of antimalarials and most at risk of medication errors.

Conclusion

Non-compliance with national guidelines for the management of malaria is important. This constitutes medication errors that can lead to potentially serious side-effects in children undergoing treatment. The results of our study amply demonstrate the need to strengthen the four pillars of the WHO’s third global challenge, “Medication without harm”, to achieve the ambitious goal of reducing the adverse effects of medication errors. It is the responsibility of political and regulatory authorities to ensure that national guidelines are implemented and monitored at all levels of the health pyramid. They should also encourage healthcare providers to identify and report medication errors, and raise public awareness of the dangers of self-medication and abuse. Furthermore, authorities should ensure that medicines supplied are of good quality. Pharmaceutical companies should enhance the pharmacovigilance of their products, but also consider the development of formulas adapted to the weight of each child. This study advocates for the mobilization of meaningful resources to strengthen pharmacovigilance through the application of regulations, infrastructure development, and capacity building for health professionals. This could be a starting point for field studies assessing other stages in the process of managing childhood malaria in the DRC.

Author contributions

M.N.A., G.M., and A.F. conceived the work, designed the work, analyzed and interpreted the data, and wrote the main manuscript. A.E. acquired and interpreted the data. E.K.N, A.E.B, N.N., G.M.K, J.B.M, C.N.N, and G.T.L critically reviewed the work for important intellectual content and gave final approval to the version to be published. All authors reviewed the manuscript.

Funding

This research received no external funding.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.