Introduction
In 2001, the World Health Organization (WHO) recommended artemisinin-based combination therapy (ACT) including Artemether–lumefantrine (AL) as a first-line treatment for uncomplicated
Plasmodium falciparum (PF) for all countries that experienced resistance to mono-therapies, notably to chloroquine and sulfadoxine-pyrimethamine (SP) [
1]. After three years (2004), Ethiopia adopted AL as a first line treatment following the widespread resistance of SP in the country. In 2017, AL plus single dose of primaquine was recommended as first-line drugs [
2‐
4]. AL is co-formulated tablets of 20 mg of Artemether and 120 mg of Lumefantrine (Coartem®; Novartis) taken twice daily for the subsequent three days; respecting strict eight hour interval between the first and the second dose; also, each dose should be accompanied by fatty meal to maximize absorption. The drug is dosed according to weight or age and supplied in blister packs containing 1–4 tablets. While the artemisinin quickly reduces most of the parasites load, the partner drug clears the remaining ones. If the patient fully adhered, AL has 98% efficacy [
3,
5,
6].
Adherence refers to the extent to which a person takes the prescribed drugs as directed; therefore, adherence to AL is a crucial part of patient care and indispensable for reaching clinical goals. Increasing patient’s adherence to AL may have so far greater impact on the health of the population than any improvement in specific medical treatment. In contrast, not adhering to this drug leads to poor clinical outcomes, increased morbidity and mortality, and unnecessary healthcare expenditure. Eventually, it might lead to the emergence and spread of drug resistant
PF malaria strains [
3,
7,
8]. Patients with uncomplicated
PF malaria have strongly discontinued the full or partial course of AL when they were symptom-free since its symptoms improve rapidly after the treatment was initiated [
9‐
12]. Multiple factors might lead to poor AL adherence, normally classified into five categories: socioeconomic factors [
12‐
16], patient-related factors [
11‐
13,
17‐
20] drug and condition-related factors [
14,
21], and health system-related factors [
13].
Malaria is preventable and treatable, but if the patient does not adhere to treatment, it leads to severe malaria that has a case fatality rate of 10–20% [
22,
23]. The serious consequences of non-adherence to AL implies not only the social and health costs of treatment failure at the patient level but also at national and global level, where resistance to AL has had a significant impact on the cost of malaria control due to the need for a new drug [
24,
25].
Globally, non-adherence to anti-malaria treatment has been identified as a key factor for poor clinical outcomes and is one of the greatest challenges to malaria control efforts today [
26].
Importantly, the mechanism behind the development and spread of SP-resistant
PF strain was a complex one with multiple factors, but non-adherence to SP treatment was the main reason [
27,
28].
Therefore, to keep AL efficacy, patient’s adherence to the drug is crucial. Thus, assessing the magnitude and determinants of optimal AL adherence is urgently needed to develop effective intervention strategies for achieving the national malaria elimination goal in the stated period [
25].
Concerning this specific public health problem in Ethiopia, as per our knowledge, there is little knowledge and available evidence on the raised public health problem in Ethiopia. This study aimed at assessing patients’ adherence with AL and identifying its influencing factors in Asgede Tsimbla district, Tigray, Ethiopia.
The questionnaire tool was adapted from relevant literature [
12,
13,
30]. The study used a structured and pre-tested questionnaire for data collection after translation to the local language (Tigrigna). It also used checklist to inspect the remaining tablet in the blister pack.
An interviewer-administered questionnaire was used to collect the data from March 24 to April 30/2018. On day zero, tracing address was collected, and then patients were visited at their house on the day after AL treatment was supposed to be complete (Day 4). Data were collected by six trained clinical nurse diploma professionals and three BSc. nurses were also assigned as supervisors. Adherence was determined using two different methods: pill count and interview. The availability of blister pack was used to present the visual pill count inspection and to checked out the remaining AL tablet on the pack. The data collectors collected relevant data as per the questionnaire tool under strict supervision. The adherence to AL was determined using pill count only; dose and dose timing interview recall only, and both pill count and interview approaches.
The number of patients recruited each day by data collector was limited to a maximum of five because tracing of the study participants at their household was difficult and time-consuming. Adult patients or caregivers for children patients who were not available in their home on the day of the visit were revisited on the next day and if not found at the second visit, they were considered as non-response.
Timing of medication was assessed by considering natural events with local expression, such as the position of the sun, coffee time/cow milking, time from church and time of cattle leaving or entering their shed. We converted these events to approximate hours. Therefore, the time interval for the dose was considered correct if taken ± 2 h. from the expected time when it was supposed to be taken [
12,
13].
Data quality control
The questionnaire was pretested in 5% of participants who were not part of our study to ensure its validity and appropriateness in the local context. Moreover, pre-testing findings such as skip pattern and sequencing of questions were incorporated according to the pre-testing findings. Data collectors and supervisors were trained for two days on how to approach respondents, ethical issues, how to fill the questionnaire using mock exercises and observation forms. Three pharmacy technicians reviewed the questionnaire prior to actual data collection.
At the end of each day, the principal investigator and the supervisors checked out the consistency and completeness of the filled questionnaire. If not, the data collectors went back to the field to complete the questionnaire. The health workers who were responsible for the diagnosis and treatment of malaria did not participate in the study, and the patients/caregivers didn’t know that they would be visited at their household.
Discussion
This study revealed that the magnitude of adherence to AL treatment was low, 53.6% (95% CI 48.4–58.3%). Age group < 5, and being treated in health post were factors associated with AL adherence whereas illiteracy, didn’t know the consequences of AL discontinued, stopped/saved drug when improved before tablets got finished, and had concomitant drug were factors that hindered the AL adherence after adjusting potential confounders. Adherence to AL is a key public health practice in attaining effective implementation of malaria case management strategy and prevention of AL resistance. The study revealed that significant proportion of uncomplicated malaria patients didn’t comply to AL treatment protocol including around one-third of patients didn’t finish their course of AL treatment on day 3 visit, half of patients who experienced vomiting didn’t re-administer and sought replenishment of missed dose, and just waited until the next dose. These findings show that the clinical and public health practices regarding case management and malaria elimination strategy have been implemented traditionally. Healthcare system of developing countries is challenged by a combination factors, poor socio-economic status, reduced availability and accessibility to health services, political issues as well as poor planning and/or poor implementation of health policies and programmes; nearly half of the study participants in the present study were illiterate, three-fourths didn’t get a chance to repeat the prescription instruction and a very small number of patients took their first dose under the direct observation of the dispenser. This indicates that healthcare actors (policy makers, planners, managers, and healthcare professionals) didn’t give more emphasis for provision of standard healthcare practices and minimizing early PF resistance.
This low rate of adherence level to AL treatment could be a big challenge to achieve the malaria elimination goal. The magnitude of adherence to AL treatment is consistent with the studies conducted in Ghana (57.3%), DRC (62%), and Kenya (60%) [
18,
21,
32]. The present study had similar characteristics to cited comparable studies; both study settings were from rural and malaria endemicity with farming as the main source of income and low educational level of the participants. They also used the same outcome measurements (pill count and interview). However, the magnitude of the adherence to AL treatment in the current study was lower compared to studies done in Myanmar (85.7%), Tanzania (74.5% in public health facilities and 69.8% in retailers), and Malawi (65%) [
13,
14,
33]. But it was higher than the study conducted in Tigray, Ethiopia (38.7%) [
12], and in Ghana (36.6%) [
17]. The disparities with the current study finding could be due to different methodological approaches in the previous studies, for instance, in the study of Maynamar, adherence level was classified into three categories: definitely non-adherent, probably non-adherent and probably adherent whereas in the current study, it was classified as adherent and non-adherent. In the study of Malawi, patients were informed that there would be a follow-up visit that could increase patients’ adherence. Moreover, in the study of Tanzania, patients residing within 2.5 km of the dispensary were included whereas in the present study, all patients were included without distance restriction, when the distance from dispensary increases patients’ adherence to AL decreases [
25]. The sample sizes of studies from Ethiopia (n = 195), and Ghana (n = 175) were much smaller than the sample size of the current study (n = 384); this might have caused differences.
In the present study, two out of the five patients didn’t adhere to AL treatment. As the Ethiopian national malaria case management training manual indicates,
PF parasites are only killed when the full course of the treatment is taken [
4]; various methods have been used to measure the level of AL adherence, but none is fully satisfactory; however, adherence rate greater than 95% is mandatory particularly for acute diseases like malaria [
34]. Non-adherence to AL treatment contributes to the recrudescence of malaria cases, affects clinical and parasitological cure rate, increases transmission rate, and eventually leads to the emerging of AL resistant
PF parasite. This could also challenge achievement of the country’s malaria elimination goal on the stated time; therefore, there is a potential need for interventions to improve patient adherence with AL treatment.
Nearly eighty percent of non-adherent patients had not taken either 5th or 6th dose. Previous studies showed that doses 5 and 6 contribute most substantially to elevating Lumefantrine day-7 concentrations to levels sufficient to clear all
PF parasites [
6,
27]. Therefore, skipping the last 2 doses could increase the likelihood of recrudescence which in turn could lead to drug resistance.
This study revealed that age group < 5 years were about 60% less likely to experience non-adherence to AL as compared to age group ≥ 18 years. The possible explanation could be due to that children can be monitored by their parents’ to adhere to anti-malarial medication, thereby bring about higher adherence level. This finding is in line with the study done in Garrisa, Kenya in 2015 [
32]. However, this result is inconsistent with the finding of a study from Malawi [
13]. This disparity might be explained by the fact that the current study was carried out after the successful introduction of a specially created pediatric formulation of AL [
35] whereas the former study was conducted before. This could be the reason for the acceptability and adherence to AL in young children in the current study [
3].
Patients who were illiterate were 9 times more likely to experience non-adherence compared to those who attended secondary and above. This is in line with several studies done in sub-Saharan Africa; in Uganda [
11], Kenya, and Tanzania [
19,
33]. Formal education affects the patients’ understanding instructions, the quality of the patients’ relationship with the healthcare provider, and the ability of the patients to interpret the pictorial instructions and those written on the AL pack [
15]. This study showed that patients who were being treated in health posts were about 70% less likely to be non-adherence as compared to those who were being treated in health centers. A study from Malawi indicated that non-adherence rate showed a significant difference across health facilities [
13]. This could be due to that in health posts malaria diagnosis and treatment is given by health extension workers so that they may have more time to spend with each patient to explain details of the treatment schedules, and importance of adhering than the pharmacy technicians who are busy in the health centers. This may improve understanding of AL administration instructions and the importance of adhering in those patients who were being treated in health posts.
In this study, patients who didn’t know the consequence of incomplete treatment were about 4 times more likely to experience non-adherence compared to their counterpart. The present study finding was in agreement with the study done in Ghana [
18]. The possible explanation could be that if patients do not know that they would be cured, and the emergence of drug resistance would be prevented if and only if the full dose is taken then they would adhere less. This appears to indicate that there was a poor understanding of the importance of finishing AL.
In our study, patients who reported to stop/save treatment if improved before tablets got finished were 3 times more likely to experience non-adherence compared to those who reported continuing till tablets got finished. Since
PF malaria symptoms improved rapidly with AL initiation; there is a strong temptation not to complete the three-day course of AL [
24,
36]. Patients who anticipate frequent malaria infections were more likely to be non-adherent, suggestive that the decision could be related to the desire to keep pills for the next malaria episode [
25,
37]. This indicated that there was poor communication between patients and healthcare providers. Those acts, as said earlier, contribute greatly to non-adherence leading to massive drug resistance in future episodes [
11,
20]. Malaria often occurs coincidentally with other diseases which often leads patients to take multiple drugs. Pill burden has been reported as a hindrance to adherence [
36] because many of these drugs have different schedules, and side effects, this could influence the adherence of AL treatment [
4]. However, in this study taking the 1
st dose of AL at dispensary showed no significant association unlike findings from Zambia, Malawi, and Tanzania that reported that it had a significant association [
20,
37,
38]. This disparity could be due to that in our study only fifteen percent of patients received the first dose of AL at the dispensary whereas more than half of patients in the study of Zambia, Malawi, and Tanzania did.
Conclusion
Adherence to Artemether–lumefantrine treatment in the current study was low. Children aged < 5 years, and being treated in health post were determinants of AL adherence whereas illiteracy, didn’t know the consequence of discontinued the drug, and had concomitant treatments were factors that hindered the AL adherence treatment.
Stakeholders should emphasize designing appropriate strategies including educational interventions to avert the AL non-adherence and its consequences on drug resistance.
Specifically, malaria patients should collect additional drugs from nearby health facilities if some doses have been vomited and they should strictly stick to AL dispensing information that is given by healthcare providers. Health facilities workers should provide patient-centered counseling and advice during the process of dispensing of AL. Further research should be conducted to evaluate the drug resistance to AL.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.